Disertación/Tesis

Clique aqui para acessar os arquivos diretamente da Biblioteca Digital de Teses e Dissertações da UFRN

2024
Disertaciones
1
  • BEATRIZ SOARES DE SOUZA
  • Application of Graph Theory in the Analysis of Query Similarity and Complexity in Relational Databases

  • Líder : LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • MIEMBROS DE LA BANCA :
  • DANIEL GOUVEIA COSTA
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • Data: 31-ene-2024
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Information Systems constant changes and refactoring processes eventually result in design debts, one of them being related to the database management. Redundancy and complexity are often found in databases and eventually affect the overall system performance. In this study, the analysis conducted was based on graph analysis, which is an essential technique in several fields demanding data management. The investigation involved the study of relationships and connections in a real-world financial organization database, between SQL queries represented as nodes and edges in a graph structure. By analyzing the graph structure and properties, it was possible to identify important nodes, detect clusters of related data, and uncover hidden relationships and redundancy. The results indicate that 50% of the database queries had medium to high similarity in subgraphs, which allows the organization to gain valuable insights into their data, make informed decisions, and optimize their database performance.

2
  • DANIEL SILVA DE MORAIS
  • Identification, Control and Prediction Techniques applied to Differential Wheeled Robots

  • Líder : PABLO JAVIER ALSINA
  • MIEMBROS DE LA BANCA :
  • ADELARDO ADELINO DANTAS DE MEDEIROS
  • CARLOS EDUARDO TRABUCO DOREA
  • FILIPE CAMPOS DE ALCANTARA LINS
  • PABLO JAVIER ALSINA
  • Data: 16-feb-2024
    Ata de defesa assinada:


  • Resumen Espectáculo
  • This work presents the identification, control, filtering and prediction of a two-wheeled differentially-driven mobile robot. For this, the variable substitution methodology will be used in the mathematical model of the robot, with its position variables, represented in the Cartesian coordinate system, replaced by its linear displacement. As the model becomes essentially linear, without the use of linearization techniques, classic control approaches, such as PID, can be used, as well as the use of predictors and filters. For the identification of the model, real experiments were performed for data collection and the least squares method was used for different structures considering the order and delay time. Using Akaike information criterion and real robot reaction experiments, it was possible to choose a model that best represents it. Decoupling the system from the robot is proposed to reduce the unwanted cross relationship that exists between inputs and outputs. It is also proposed to use the Smith Predictor and Kalman Filter to minimize the delay time and estimate the filtered values of linear displacement and orientation, respectively. Finally, some identified models and the execution schedule for the completion of this work will be demonstrated.

3
  • MARIA DA GUIA TORRES DA SILVA
  • Educational Robotics as a Pedagogical Resource for Students with Learning Difficulties

  • Líder : LUIZ MARCOS GARCIA GONCALVES
  • MIEMBROS DE LA BANCA :
  • CARLA DA COSTA FERNANDES CURVELO
  • LUCIANE TERRA DOS SANTOS GARCIA
  • LUIZ MARCOS GARCIA GONCALVES
  • SARAH THOMAZ DE SA ROSSITER
  • Data: 19-feb-2024
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The use of Educational Robotics as a support tool in the teaching-learning process for children with learning difficulties is the core of our investigation. In this work, we try not to distinguish between (normal) difficulties caused by lack of interest or another solvable problem in the classroom or those caused by well-defined disorders in medicine. In an initial systematic review, in which we did a general search of articles and dissertations in conferences and journals using a well defined set of keywords that covered the most well-known learning problems and that also met the keyword "educational robotics", we identified that there are very few works on this topic, around 14 contributions. Given the novelty of the topic, we carried out a more in-depth analysis of these and other works, including several issues, not only related to learning difficulties. Our main question to be answered, from then on, was to identify how much and how educational robotics can be an adjunct in the teaching-learning process, for these students with difficulties. With this in mind, we defined the quali-quanti research methodology to identify the remaining points, where we adopted questionnaires as a data collection instrument. We provide forms with questions for students, teachers, and monitors from selected schools, among those that have educational robotics activities (all public schools) in the city of Natal, RN, Brazil. Through an analysis of the data obtained from the forms, we observed an increase in the use of robotics in education, in general. However, we mainly identified that, although students with learning difficulties or disorders are found in educational institutions, they are not completely included in robotics activities. Nonetheless, after compiling and analyzing the data, we reached positive results, which verify the possibility of application to improve learning in these children. That is, the majority of students and a good number of teachers and monitors agree that robotics has managed to improve the ability of students with difficulties in learning, especially transversal content. Regarding "how", we noticed that some factors are important. The first factor is the ease of understanding the robotics itself, brought about by new software and hardware tools developed by this pretty new scientific community, which greatly facilitate its use as a support tool. Furthermore, we note that the interest of teachers is another factor in the "how" that can collaborate in trying to improve the issue of learning difficulties. We noticed that interest improves with the use of these new tools, demystifying robotics, which is no longer a big deal for lay teachers. Furthermore, the existence of complete support systems for its application, free of charge and freely available, involving content and tools on the Web, in addition to an effective continuing education policy for teachers, have contributed to the tool becoming increasingly more effective, mainly, as highlighted, in content transversal to robotics.

4
  • MATEUS ARNAUD SANTOS DE SOUSA GOLDBARG
  • Aware Compression of Deep Neural Network Models Based on Pruning Followed by Quantization.

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • MARCELO AUGUSTO COSTA FERNANDES
  • ALLAN DE MEDEIROS MARTINS
  • SERGIO NATAN SILVA
  • FLAVIA MARISTELA SANTOS NASCIMENTO
  • MARIA GRACIELLY FERNANDES COUTINHO
  • Data: 20-feb-2024
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Deep learning techniques, particularly deep neural networks (DNNs), have been successfully employed in numerous problems. However, these types of algorithms demand substantial computational effort due to the large number of parameters and mathematical operations involved, which can pose challenges for applications with limited computational resources, low latency requirements, or low energy consumption. Thus, this work proposes the application of a novel training strategy for the conscious compression of DNN models based on pruning, quantization, and pruning followed by quantization, capable of reducing processing time and memory footprint. The compression strategy was applied in two domains: the first for automatic modulation classification, where the model size was reduced by 13 times, maintaining an accuracy only 1.8% lower than the uncompressed model. In the second domain, focused on image classification in microservices environments, the same compression strategy was applied. In this context, a 7.6-fold reduction in model size was observed, with accuracy close to the uncompressed model. Furthermore, the implementation of these techniques reduced prediction latency by 1.7 times and significantly decreased the time required for the deployment of microservices containing these models. These results underscore the effectiveness of the proposed approach, indicating its potential positive impact in scenarios that require computational efficiency and resource conservation.

5
  • EDUARDO NUNES VELLOSO
  • End-to-End Optimization of Multiuser MIMO Systems Using Autoencoders with Bidirectional Channel Estimation

  • Líder : LUIZ FELIPE DE QUEIROZ SILVEIRA
  • MIEMBROS DE LA BANCA :
  • ALLAN DE MEDEIROS MARTINS
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • MARCELO SAMPAIO DE ALENCAR
  • Waslon Terllizzie Araujo Lopes.
  • Data: 21-mar-2024
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The spectral efficiency gains introduced by multiuser MIMO systems render them relevant schemes for current and upcoming generations of mobile communication networks. Due to the intrinsic complexity of the mathematical models of these systems under realistic conditions and the interdependence between the processing steps of the transmitters and receivers, machine learning is an option that allows the complete system to be designed by training a noisy autoencoder. This paper proposes a neural network architecture for end-to-end optimization of a multiuser MIMO system. The performance of the system, measured in terms of symbol error rate, was compared to an M-PSK baseline with zero-forcing equalization and least-squares channel estimation. Simulations were performed using a Rayleigh fading channel model and the realistic 3GPP TR 38.901 model. A bidirectional channel estimator, based on the interpolation of sparse pilots, is proposed, reducing the control signaling to less than 3% in exchange for a fixed 10 ms delay. The results show that signicant gains can be achieved by applying the proposed model, but those vary with respect to the estimation errors during the pilot transmission times.

6
  • ANTONIO ALCIR DE FREITAS JUNIOR
  • Detection of Distributed Denial of Service Attacks using Convolutional Neural Network.

  • Líder : LUIZ FELIPE DE QUEIROZ SILVEIRA
  • MIEMBROS DE LA BANCA :
  • AGOSTINHO DE MEDEIROS BRITO JUNIOR
  • FRANCISCO SALES DE LIMA FILHO
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • Data: 16-abr-2024


  • Resumen Espectáculo
  • With the expansion of the Internet, combined with the growing number of Internet of Things (IoT) devices, denial of service attacks (Denial of Service - DoS), as well as its distributed variant (Distributed Denial of Service - DDoS), have become more widespread, making it a significant problem for the availability of services operating on the Internet. In recent years, the number of research in academia and industry on the detection and mitigation of these attacks has been growing, but without a definitive solution yet. Techniques involving machine learning are being widely used to detect and mitigate these attacks. Although efficient, the proposed techniques present a high computational cost, which may make them unfeasible in network scenarios with intense data flows, due to the temporal restrictions imposed by real-time processing of the data flow. Inspired by these works in the literature, but seeking to reduce computational complexity, this work proposes the use of a low-complexity convolutional neural network to detect DDoS attacks. The complexity reduction of the proposed convolutional network is based on the use of descriptors obtained from a set of metrics calculated on sampled network traffic header data. The developed method has a high success rate, low false positive rate, and relative simplicity of implementation, making it suitable for the task of detecting DDoS attacks in network scenarios with high throughput.

7
  • JORDAO PAULINO CASSIANO DA SILVA
  • A Proposal for Pothole Detection with Machine Learning at the Edge

  • Líder : IVANOVITCH MEDEIROS DANTAS DA SILVA
  • MIEMBROS DE LA BANCA :
  • DANIEL GOUVEIA COSTA
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • Data: 30-abr-2024


  • Resumen Espectáculo
  • Potholes in urban roads represent a significant issue, impacting both user safety and vehicle durability. This study introduces an innovative approach integrating machine learning on edge devices, with an emphasis on YOLOv8 and FOMO models in the context of TinyML. We utilize a specialized dataset, containing annotated images, to effectively train these models for accurate pothole detection. The focus of the research is on optimizing the performance of these models for devices with limited computational resources, aiming for real-time efficiency and reduced energy consumption. This work not only provides effectively trained models but also introduces an adaptable framework for pothole detection, ensuring practical and efficient implementation. Furthermore, we present a complete pipeline for pothole detection at the end, validating the models' accuracy and efficiency. This approach proposes a robust and energy-efficient solution for the automatic recognition of potholes, significantly contributing to improvements in urban infrastructure maintenance.

Tesis
1
  • MAURÍCIO RABELLO SILVA
  • Data Transport Protocol for FANET in Rocket Impact Area Scanning over the Sea by Unmanned Aerial Vehicles

  • Líder : PABLO JAVIER ALSINA
  • MIEMBROS DE LA BANCA :
  • CARLOS MANUEL DIAS VIEGAS
  • FELIPE DENIS MENDONCA DE OLIVEIRA
  • FRANCISCO ARY ALVES DE SOUZA
  • MARCELO BORGES NOGUEIRA
  • PABLO JAVIER ALSINA
  • Data: 23-feb-2024
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Systems involving multiple Unmanned Aerial Vehicles (Multi-UAVs) performing area scanning procedures have garnered attention in recent years. Developing these systems faces significant challenges, with the robustness of the data communication network being a crucial aspect for efficient collaboration among UAVs. This work is dedicated to specifying a data transport protocol for a communication network within a squadron of Unmanned Aerial Vehicles, focusing on scanning the rocket impact area over the sea in the SPACEVANT 2 project.

    The research encompasses the key characteristics of communication networks for multi-UAV systems, considering application-specific aspects such as path and trajectory planning and sensor data collection. Strategies are proposed for the rocket impact area scanning, analyzing characteristics of networks best suited for these strategies, and defining a suitable protocol for the application. Utilizing Xbee Pro 900HP S3B modules in a network architecture integrated with a computer-controlled embedded hardware platform, a test plan is devised and implemented to assess the protocol's performance. Performance evaluations include measurements of bandwidth (throughput), latency, and indicators on communication links among network nodes, such as retransmission, packet loss, flow control, and congestion. This study demonstrates the robustness and reliability of the protocol in the proposed network architecture for multi-UAV systems.

2
  • RUTE SOUZA DE ABREU
  • Enhancing Fault Prediction in Redundancy Models: A novel approach using Generalized Stochastic Petri Networks and Spiking Neural Networks

  • Líder : IVANOVITCH MEDEIROS DANTAS DA SILVA
  • MIEMBROS DE LA BANCA :
  • DANIEL GOUVEIA COSTA
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • JUAN MOISES MAURICIO VILLANUEVA
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • RENAN CIPRIANO MOIOLI
  • Data: 29-feb-2024
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Fault prediction plays an important role across several sectors such as industry, technology, medical sector, among others. This task can help in the reducing of equipment maintenance costs, prevention of accidents and disasters, and improvement of system dependability since it can increase availability by reducing system downtime. This work presents a methodology for fault prediction in redundancy models designed using the formality of Generalized Stochastic Petri Networks. The approach comprehends the steps of modeling and simulation of systems with active and passive redundancy under different fault scenarios, such as non-perfect switches, standby failures, and common cause failures, as well as fault datasets generation and the implementation of a machine learning model for performing the fault prediction. For forecasting, this research utilizes Spiking Neural Networks (SNNs), which have been recognized as the third generation of Artificial Neural Networks. Just like typical artificial neural networks, SNNs draw inspiration from the biological dynamics of the brain, incorporating the interconnected topology of neurons into their architecture. However, while conventional neural networks rely on error minimization by weight adjustment, SNNs aim to replicate the learning process by simulating neuron behavior by taking into account elements of the biological process such as synapse, energy accumulation, electric impulse firing, and refractory periods between emissions. Due to the ability to capture temporal aspects from data, SNNs are vastly used in problems with time dynamics. Additionally, literature has shown these networks to be task and energy-efficient serving as a low-cost alternative compared to conventional ANNs.

3
  • DUNFREY PIRES ARAGÃO
  • A Set of Independent Variables for Time Series Regression Tasks of Pandemic Scenarios Based on COVID-19

  • Líder : LUIZ MARCOS GARCIA GONCALVES
  • MIEMBROS DE LA BANCA :
  • ADRIANO MONDINI
  • ANDOUGLAS GONÇALVES DA SILVA JÚNIOR
  • DAVI HENRIQUE DOS SANTOS
  • LUIZ MARCOS GARCIA GONCALVES
  • TIAGO PEREIRA DO NASCIMENTO
  • Data: 10-may-2024


  • Resumen Espectáculo
  • We propose the creation of a set of independent variables, derived from the analysis of phenomena, government interventions and events that may influence the spread of the SARS-CoV-2 virus, whether positively or negatively in relation to the trend of the curve of cases and deaths. This approach aims to identify relevant characteristics and parameters applicable to these scenarios, using regression techniques. To achieve this, we carried out studies using machine learning methodologies to determine which variables should be selected and applied to the regression models. Our strategy included collecting and cleaning data, evaluating the generalization of models, and applying machine learning techniques, such as regression using Seasonal ARIMA, clustering, dimensionality reduction through principal component analysis and neural networks, in addition to creating informative data visualizations. As a result, we compiled and proposed the use of a set of independent variables to predict the number of deaths from COVID-19, aiming to increase accuracy and reduce the standard deviation in relation to real values, and avoid underspecification of the problem. The main contributions of this work include the investigation of the causes and possible relationships between phenomena, government measures and events in the spread of the virus in urban areas, such as cities and countries. The results of this study can serve as a complementary resource to assist governments and authorities in making decisions in the face of possible future pandemic scenarios, such as the one faced during the COVID-19 crisis, which, although still present, has a lower mortality rate.

4
  • YURI THOMAS PINHEIRO NUNES
  • Concept Drift Detection Heuristic Based on TEDA

  • Líder : LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • MIEMBROS DE LA BANCA :
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • MARCELO AUGUSTO COSTA FERNANDES
  • IGNACIO SANCHEZ GENDRIZ
  • JUAN MOISES MAURICIO VILLANUEVA
  • Data: 21-may-2024


  • Resumen Espectáculo
  • The enormous amount of machine learning applications and data produced presents several challenges today. In many contexts, data may have temporal relevance, for example, seasonality and trends, resulting in non-stationary behavior. This characteristic present in several systems makes it difficult to apply machine learning models, which, in general, assume that the data is stationary. In this scenario, data sources can be considered as data streams: ordered and unlimited sources of non-stationary data. These sources feed machine learning applications unreliably because they violate stationarity. When the data stream presents a significant variation that could lead to performance degradation, it is said that a concept drift has occurred. A data stream that presents concept drift is considered to represent an evolutionary system (evolving system). A system that evolves, presenting changes in its internal concepts, for example, emergence of new concepts, extinction of concepts, division and fusion of concepts, etc. In this context, machine learning techniques must be adapted to the context of data streams. An example would be a classifier for data stream samples (data stream classifier). This type of model needs to consider real-time retraining, robustness to non-stationarity, data unavailability, and limited data set, among others. To implement these different characteristics, it is essential to use concept drift (CDD) detectors. CDDs are not models capable of identifying when one or more concepts in the data stream have changed significantly. The literature is rich in works on concept drift detection distributed into three groups: supervised, semi-supervised, and unsupervised. Supervised methods have access to the true classes of data stream samples at the time of detection, while semi-supervised methods have limited access. Semi-supervised methods can have access to the true classes during training, during offline steps, or even to a subset of samples at the time of detection. Unsupervised methods do not access the true classes of the samples, being theoretically more limited than other approaches. However, unsupervised methods allow for shorter detection delays in real applications, as it is reasonable not to have access to the true class at the time of detection. Examples of unsupervised methods are ADWIN, KSWIN, and PageHinkley. This work presents a new concept drift detection method, TEDA-CDD. This detector is composed of two models to represent concepts based on TEDA: the reference model and the evolutionary model. The reference model represents the concept known to the machine learning model, while the reference model is free to adapt to any new model that emerges from the data stream. The models are compared heuristically using the Jaccard index to indicate similarity. When the index indicates low similarity between the models, the detector indicates a concept drift. To compare the proposed method with other methods present in the literature, initially, a realistic approach for data stream classifiers is proposed. This approach makes it possible to apply several classifiers and detectors to the data stream classification task and estimate performance metrics specific to the data streams context. In the experiments, the proposed method is compared to other methods present in the literature using synthetic and real benchmarks. The proposed method has comparable performance in terms of accuracy compared to methods consolidated in the literature, while it is the most efficient in terms of memory consumption.

2023
Disertaciones
1
  • IRLETE PEREIRA MOTA ALVES
  • Optimization of the Speed Vector Control Drive of a Three-Phase Induction Motor without Split Winding Bearings

  • Líder : ANDRES ORTIZ SALAZAR
  • MIEMBROS DE LA BANCA :
  • ANDRES ORTIZ SALAZAR
  • JOSSANA MARIA DE SOUZA FERREIRA
  • JOSÉ ÁLVARO DE PAIVA
  • RICARDO FERREIRA PINHEIRO FILHO
  • Data: 03-mar-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  •  
    The objective of vector control is to guarantee a fast speed response even with high torque for induction machines, both in transient and steady state. To implement a speed control in a three-phase induction motor without bearings with split winding, three phases are normally used to control the currents. This standard procedure requires a complete drive structure, containing the three arms, with various power electronics.

    The strategy presented in this work consists of controlling only two of the three available currents, obtaining the third through the sum of the two previous ones. With this change, only two inverter arms are needed, considerably reducing the electronic components used. For this implementation, an induction bearing motor working in the vertical position was used.
2
  • JULIANA OLIVEIRA DE MEDEIROS
  • Use of the Particle Swarm algorithm to estimate the error introduced by TIs in calculating the fault location distance using fundamental frequency phasors

  • Líder : MANOEL FIRMINO DE MEDEIROS JUNIOR
  • MIEMBROS DE LA BANCA :
  • MANOEL FIRMINO DE MEDEIROS JUNIOR
  • FLAVIO BEZERRA COSTA
  • JOSE TAVARES DE OLIVEIRA
  • MELINDA CESIANARA SILVA DA CRUZ
  • ROANA D' ÁVILA SOUZA MONTEIRO
  • Data: 10-mar-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • An Electric Power System (EPS) has the basic function of providing consumers with a continuous and quality product. For this, performance standards specified by ANEEL (National Electric Energy Agency) must be complied with by the concessionaire companies, in such a way as to guarantee the reliability of the energy supply and the conformity of voltage levels when disturbances occur and during operation. normal on a permanent basis. Thus, it ensures that the energy system will be able to keep working for a longer time, inhibiting failures due to equipment wear and avoiding the occurrence of financial losses. In order to meet reliability and quality requirements, the electrical system operation centers have data servers with oscillographic records from relays, oscillographs and qualimeters installed in the system's substations. These files are normally used for various functions, such as protection or power quality analysis. With regard to power quality, oscillographic records are normally used to classify disturbances in the electrical network.
    From a protection point of view, the files are often used for post-operation analysis of the proper functioning of protective relays or for fault location algorithms. Therefore, it is essential that the records have a reliable time base, so that they can be considered synchronized, when more than one record is needed in a specific algorithm. Even considering that efficient algorithms for synchronizing records and locating faults are available, it is not possible to find in the literature an algorithm to estimate the error produced in the process of calculating the fault distance, a calculation performed according to the methodology that adopts phasors from the fundamental frequency.
    It is noteworthy here that the oscillographic records available for calculating the phasors come from sampling signals, collected by means of TPs and CTs (TIs), which inherently have relation and phase errors, which depend on the accuracy class of these instruments. In addition to those related to TIs, other errors can be introduced by the sampling process or composition of the representative phasors of the sampled signals. Thus, it is necessary to develop an algorithm to estimate the error produced in determining the fault distance, even if this distance has been determined from reliable algorithms for synchronizing the records and locating the fault. In this Dissertation, we propose an algorithm based on Particle Swarm (PSO) to meet this purpose.

3
  • EVELYN CRISTINA DE OLIVEIRA LIMA
  • PROGRAMMABLE BANDPASS SIGMA-DELTA MODULATORS USING N-PATH FILTERS

  • Líder : DIOMADSON RODRIGUES BELFORT
  • MIEMBROS DE LA BANCA :
  • DIOMADSON RODRIGUES BELFORT
  • FRANCISCO DE ASSIS BRITO FILHO
  • SEBASTIAN YURI CAVALCANTI CATUNDA
  • Data: 15-mar-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • In Software Defined Radio (SDR) most analog functions such as downconversion and channel selection are moved into the digital domain. With these features, an SDR can be easily programmed and adapted to new standards. However, its implementation using a Nyquist rate analog-to-digital converter (ADC) has been challenging and very power-hungry, which makes it unsuitable for mobile applications. SDM (Sigma-Delta Modulator) using LC resonators are well suited for this purpose, as they can convert a narrowband around a radio frequency (RF) center frequency. In the literature, high-order LC-based SDM uses nodes between LC tanks for feedforward or feedback coefficients, to obtain the desired noise transfer function (NTF). These nodes are often an additional source of noise, non-linearity, and power consumption, affecting overall SDM performance. In this project, we aim to explore the functionality of N-Path filters to implement a programmable Sigma-Delta Bandpass Modulator that meets various communication standards. Therefore, it was necessary to carry out a bibliographic and technological research on the state of the art of N-path filters, continuous-time and discrete-time Sigma-Delta modulators. There was also the development of high-level models for different topology categories of Sigma-Delta Modulators implemented with LC and discrete-time filters to serve as comparison models with the topology design methodology using the N-path filter proposed in this work. A high-level model of the N-Path filter was developed, and the work is currently in the stage of integrating the N-path filter and SDM. The results obtained from the SDM implementation with LC filters show that the modulator has a good agreement with its DT equivalent. In initial experiments, the implementation of SDM with an RLC circuit equivalent to the N-path filter proved to be promising, presenting results similar to those obtained with the discrete-time topology.

4
  • FRANCINALDO DE ALMEIDA PEREIRA
  • Data Lake and Analytics platform for COVID-19 Data Analysis

  • Líder : LUIZ MARCOS GARCIA GONCALVES
  • MIEMBROS DE LA BANCA :
  • ALYSSON NASCIMENTO DE LUCENA
  • DAVI HENRIQUE DOS SANTOS
  • LUIZ MARCOS GARCIA GONCALVES
  • SARAH THOMAZ DE SA ROSSITER
  • Data: 05-abr-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • We propose to develop a platform consisting of a data lake, to be implemented as a web-based service. The main idea is that it can be used by data scientists working on COVID-19-related projects in order to access as much data as possible in one repository and be able not only to analyze that data but also to manage and contribute to new data. Through this platform, it will be possible to dynamically aggregate different data repositories related to the COVID-19 pandemic, in order to provide users, through a web interface, tools for use, transformation, and collaboration of data, as well as analysis and visualization tools integrated to geographic information systems.

5
  • RENAN ARAUJO DE LIMA
  • SOFTWARE TOOLS FOR BEHAVIORAL BOX FOR THE STUDY OF TACTILE DISCRIMINATION TASKS IN RODENTS

  • Líder : LUIZ MARCOS GARCIA GONCALVES
  • MIEMBROS DE LA BANCA :
  • ALMIR KIMURA JUNIOR
  • ANDOUGLAS GONÇALVES DA SILVA JÚNIOR
  • LUIZ MARCOS GARCIA GONCALVES
  • Data: 12-abr-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The area of Neurophysiology of Social Behaviors in Rodents has the general objective of investigating the electrophysiological correlates of social behaviors in rodents. Examples of studies in the area include the investigation of neural processing disorders and synaptic plasticity in animal models of autism and changes in synaptic plasticity in animal models of epilepsy. This area uses sophisticated and expensive equipment, and some of the existing platforms are proprietary brands, obsolete, and using old technology. There is a prominent need to improve this equipment to aid research, seek cheaper alternatives, with open construction, and use more up-to-date components. With that in mind, in this work, we propose improvements in the software platform that receives data and controls the necessary operations in the experiments using the behavioral box. Specifically, we contribute on three fronts. At first, we developed software aimed at automating a behavioral box in a tactile discrimination task, which, in its new version, uses the Arduino microcontroller. Also, we developed the software part of the Supervisory System, in Python, which communicates with Arduino, for visualization and definition of the tasks to be carried out in the behavioral box. Finally, we developed basic computer vision routines for locating the rodent within the behavioral apparatus, indicating its position and orientation. These developments collaborate so that the box can be automated, reducing manual interaction with the experimenter, and thus contributing to improving the process of studying tactile discrimination in rodents.

6
  • AZIZ OLOROUN - SHOLA BISSIRIOU
  • Contributions to Energy Management of Single Phase AC Microgrids Used in Isolated Communities.

  • Líder : RICARDO LUCIO DE ARAUJO RIBEIRO
  • MIEMBROS DE LA BANCA :
  • ALEXANDRE CUNHA OLIVEIRA
  • FLAVIO BEZERRA COSTA
  • RICARDO LUCIO DE ARAUJO RIBEIRO
  • THIAGO DE OLIVEIRA ALVES ROCHA
  • Data: 26-abr-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Isolated communities require assistance to maintain a reliable power system that can provide an efficient power flow to their consumers, as they are often not connected to a bulk transmission systems. To meet local energy needs, small power systems, called microgrids, are created. Typically, these microgrids use Renewable Energy Resources (RESs) and Energy Storage Systems (ESSs) to control power flow and maintain energy quality within specified limits. However, due to the intermittency of renewable sources, particularly wind and solar, finding the appropriate topology and defining a power management system for these microgrids has been challenging. This dissertation presents a methodology to determine a suitable topology for a single-phase AC microgrid to provide power to the Adjarra/Benin community. Additionally, it proposes an energy management scheme to ensure a reliable energy flow, taking into account the minimum deployment and operation costs of the microgrid. Finally, the viability of microgrid operation under different scenarios is evaluated through simulation experiments.

7
  • PEDRO VICTOR ANDRADE ALVES
  • Proposal for a Real-Time Testing Platform Applied to Tactile E-commerce

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • MARCELO AUGUSTO COSTA FERNANDES
  • ALLAN DE MEDEIROS MARTINS
  • LUCILEIDE MEDEIROS DANTAS DA SILVA
  • VICENTE ANGELO DE SOUSA JUNIOR
  • PATRICIA DELLA MÉA PLENTZ
  • Data: 31-jul-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • This work aimed to develop a real-time test platform for systems associated with the tactile internet area. The proposal comprises a master device, a communication channel and a slave device. The master device is a tactile glove (wearable technology) that works as a tactile interface based on vibratory feedback, responsible for enabling interaction with virtual or remote elements. Additionally, the communication channel introduces a bidirectional connection of variable latency. In turn, the default slave device is a robotic Phantom Omni manipulator integrated with the Matlab/Simulink environment and the robotics toolbox. The platform can generate tactile sensations such as coarse roughness, fine roughness, smoothness, dripping and softness. Furthermore, the results of the proposal have been adequate and allow testing various algorithms and methods related to tactile internet.  Therefore, the developed system has the capacity to serve as a basis for a wide variety of use cases related to tactile internet, providing a diverse opportunity for exploration and application. In this sense, the proposed platform serves as a foundation for a use case focused on tactile exploration of product textures on an e-commerce website, enabling interaction with virtual clothing through online shopping capable of providing tactile stimuli. In the scenario of this use case, the slave device is a cursor that functions as an element controlled by the master device and a mouse, which allows interaction with different clothing textures, generating specific tactile sensations on the tactile glove. In order to enhance the platform performance, Virtual Network Functions (VNFs) were introduced. VNFs are virtualized processes implemented and executed in cloud services, capable of enhancing the scalability and agility of a network, while also enabling more efficient utilization of infrastructure resources. In Fifth Generation (5G) mobile networks, tactile internet connections are associated with ultra-reliable low-latency communications (URLLC) services. This category poses challenges due to its stringent requirements of low latency in a few milliseconds, low packet loss probability, and high network availability. Thus, the application of VNFs during tactile internet connections in 5G networks allows for adjustments of connection parameters to maintain good connectivity, ensuring the low-latency criterion and a high degree of realism in tactile interactions. Consequently, the use of VNFs in URLLC communications aims to fulfill the requirements of tactile internet systems while maintaining an acceptable level of Quality of Experience (QoE), which is a metric defined by the Telecommunication Standardization Sector as the degree of satisfaction or annoyance of users with an application or service.

8
  • JESSIKA CRISTINA DA SILVA
  • Performance Evaluation of Data Rate Adaptation Mechanisms for LoRaWAN networks for Scenarios of Livestock in Semi-Confinement

     

  • Líder : VICENTE ANGELO DE SOUSA JUNIOR
  • MIEMBROS DE LA BANCA :
  • VALDEMIR PRAXEDES DA SILVA NETO
  • VICENTE ANGELO DE SOUSA JUNIOR
  • ÁLVARO AUGUSTO MACHADO DE MEDEIROS
  • Data: 01-ago-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • This work aims to investigate Adaptive Data Rate (ADR) mechanisms in LoRaWAN networks as a solution for dynamic IoT scenarios and also propose a new solution based on latter investigations. The standard ADR technique, defined in the LoRaWAN network protocol, is a simple technique that allows an adjustment of transmission rate by reading the SNR (Signal-to-Noise Ratio) value. Due to the multiplicity and dynamics of IoT scenarios, it is necessary to investigate ADR techniques that establish a good compromise between coverage and capacity. This thesis aims to investigate IoT scenarios of livestock in semi-confinement, especially in time-varying scenarios (emergence of concentrated traffic demand, network with mobile sensors, for example). Preliminary results using the ns-3 simulator demonstrate the need to dynamically adapt the ADR parameters, as each scenario requires different ADR strategies (or different parameterization of pre-existing strategies). Finally, we propose an adaptation of classic ADR algorithms to promote flexibility between coverage and capacity in such scenarios.

9
  • CASIMIRO WAETE AGOSTINHO
  • Automatic Orientation of Photovoltaic Solar Panels for Monitoring the Seasonal Movement of the Sun Using Intelligent Control

  • Líder : FABIO MENEGHETTI UGULINO DE ARAUJO
  • MIEMBROS DE LA BANCA :
  • ANDRES ORTIZ SALAZAR
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • JOÃO TEIXEIRA DE CARVALHO NETO
  • ROMÊNIA GURGEL VIEIRA
  • Data: 09-ago-2023


  • Resumen Espectáculo
  • Solar trackers, besides serving as support for photovoltaic modules, follow the move- ment of the sun throughout the day and try to keep the plane of the photovoltaic modules perpendicular to the sun’s rays, thereby increasing the performance of the photovoltaic module. With the aim of demonstrating the effectiveness of using the automatic orien- tation method of photovoltaic solar modules in following the seasonal movement of the sun as a way of improving energy production, this work analyzes two systems: a solar tracker and a fixed-axis system. The first part simulates two photovoltaic plants with solar tracking and fixed-axis systems using the software PVsyst, consisting of a total of twenty modules. The results show that the solar tracker improved the efficiency of electricity production, with a gain of 25.97% during the day and an annual gain of 25.6% more than a fixed system. The second part presents a proposal for an automatic solar tracking system (with a controller) and a fixed-axis system (without a controller) based on the equations of the sun’s movement angles. This was modeled using the software Matlab/Simulink using Fuzzy logic as the controller, with the aim of evaluating the use of solar radiation throughout the day and to build a real system with the same characteristics as the pro- posed simulation. The gain of this system was approximately 17% to 35% during some hours of the day for the system with a controller. Both software simulations showed that photovoltaic systems that track the movement of the sun during the day and year are more advantageous compared to fixed-axis systems because they take better advantage of solar radiation.

10
  • ALEXANDRE HENRIQUE SOARES DIAS
  • Promoting Research Relevance: A Natural Language Processing-based Model for Identifying SDG-aligned Scholarly Publications

  • Líder : IVANOVITCH MEDEIROS DANTAS DA SILVA
  • MIEMBROS DE LA BANCA :
  • LIVY MARIA REAL COELHO
  • ALLAN DE MEDEIROS MARTINS
  • EDUARDO ALMEIDA SOARES
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • OLIVIA MORAIS DE MEDEIROS NETA
  • Data: 31-ago-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • In 2015, the United Nations established the 17 Sustainable Development Goals (SDGs) to promote environmental stewardship, economic advancement, and social equity. Within this framework, scientific research plays a pivotal role in addressing the challenges encompassed by the SDGs. An exemplary tool, SciVal, facilitates the correlation of scientific outputs with the SDGs through expert analyses. However, in order to mitigate the reliance on specialized expertise and offer a self-reliant solution, this endeavor proposes a natural language processing-based, deep machine learning-powered, multi-target classification model bolstered by interpretability techniques and good practices for the development and analysis of data streams.. The objective is to effectively map academic publications to the SDGs. By employing this proposed model, the vast potential of scholarly research can be harnessed, directly aligning it with the global agenda for sustainable development. Researchers, policymakers, and organizations can adeptly navigate the extensive landscape of research papers and identify those that harmonize with their specific areas of interest within the SDG framework. Over one million scientific publications were utilized to train and evaluate the model. The corpus encompassed publication titles extracted from the Scopus database, accessed via the SciVal tool, and annotated with respect to 16 of the 17 SDGs. To substantiate the efficacy of the proposed model, it was applied to associate publications from the Brazilian Automation Congress (CBA 2020) with the SDGs, thereby measuring the contribution of scientific endeavors in automation towards the attainment of the SDGs. The outcomes within the context of CBA 2020 revealed prevalent themes affiliated with SDGs 7 and 9, relating to clean energy and industrial innovation, respectively. Given the extensive training data and the comprehensive range of SDGs addressed, the model can confidently be deployed to correlate academic output from diverse domains with the SDGs.

11
  • JÚLIA DA LUZ ANDRADE SILVA
  • Measurement and Evaluation of Exposure to Non-Ionizing Radiation in Indoor Environments

  • Líder : VICENTE ANGELO DE SOUSA JUNIOR
  • MIEMBROS DE LA BANCA :
  • VICENTE ANGELO DE SOUSA JUNIOR
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • HUMBERTO DIONISIO DE ANDRADE
  • Data: 22-sep-2023


  • Resumen Espectáculo
  • It is essential to monitor the levels of Non-Ionizing Radiation (NIR) that the general population may be exposed to and compare them with the limits defined in current regulations, bearing in mind the rapid rise of telecommunication services and the perspectives of an extremely connected. Indoor environments (\textit{indoor}), such as residences and shopping malls, are places that meet the interests of measurements, mainly due to the presence of different NIR sources and the location of these sources about end users. This work presents NIR measurements in a \textit{shopping center} and residences in Natal, RN, Brazil. At shopping, a set of six measurement points was proposed, following two criteria: places with a large flow of people and the presence of one or more Distributed Antenna System (DAS), co-located or not with Wi-Fi access points. The results are presented and discussed in terms of distance from the DAS (conditions: near and far) and flow density of people in the mall (scenarios: low and high number of people). The highest mean and highest peak electric fields measured were 1.96 and 3.26~V/m, corresponding respectively to 5\% and 8\% of the limits defined by the International Commission on Non-Ionizing Radiation Protection (ICNIRP) and Agency National Telecommunications Agency (ANATEL). In the homes, measurements were taken at 40 Wi-Fi access points where the highest exposure level was 4.66 V/m (7.64\% of the limit), verified for the network load on the network of 2.4 GHz (Situation 1), at a value of 4.66 V/m (7.64\% of the limit). The results are discussed regarding the measurement situation, brand and time of use of the equipment. Measurements were also carried out in 51 microwave ovens. Only one exceeded 50~W/m² (limit determined in current regulations), caused by rust that compromised the door's structure. Thus, two types of repairs were carried out: the use of epoxy putty to fill the opening in the door (Repair 1) and the replacement of the outer surface of the oven (Repair 2). Only Repair 1 solved the leakage problem with an average of 0.1841 W/m² and a peak of 0.4222 W/m². As for the other 50 microwave ovens, the relation between power density and factors such as worst point, brand, time of use and state of conservation of the device were analyzed.

12
  • ARTHUR DINIZ FLOR TORQUATO FERNANDES
  • Machine Learning Technics for the prediction of extreme overirradiance events

  • Líder : ADRIAO DUARTE DORIA NETO
  • MIEMBROS DE LA BANCA :
  • ADRIAO DUARTE DORIA NETO
  • DANIEL LOPES MARTINS
  • IGNACIO SANCHEZ GENDRIZ
  • SAMIRA DE AZEVEDO SANTOS EMILIAVACA
  • Data: 22-sep-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The current dissertation is centered on the forecasting of Overirradiance within intervals of up to five minutes, achieved through the utilization of machine learning methodologies. Overirradiance, a phenomenon characterized by solar irradiance surpassing anticipated values under clear-sky conditions at the Earth's surface, has generated scholarly interest within the sphere of solar energy research and its implications for photovoltaic power generation systems. To date, no dedicated studies investigating the application of Machine Learning techniques for forecasting this phenomenon have been identified.

    In pursuit of this aim, the performance of four distinct machine learning algorithms has been meticulously examined: Random Forest, Support Vector Machines (SVM), Multilayer Perceptron (MLP), and Long Short-Term Memory (LSTM) neural networks. The present study endeavors to bridge a lacuna in research by scrutinizing the feasibility and efficacy of these algorithms in predicting Overirradiance events, thereby augmenting the comprehension and pragmatic application of this phenomenon within solar energy systems.

13
  • TAYNÁ ARRUDA CÂMARA DA SILVA SALVIANO
  • An Approach for Generating and Visualizing Association Rules for Access to News Portal Content
  • Líder : LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • MIEMBROS DE LA BANCA :
  • DANIEL GOUVEIA COSTA
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • Data: 05-dic-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • This work aims to propose and validate an approach for the generation and visualization of association rules and sequence rules obtained from the content access history data of a Brazilian journal. The proposed approach is composed of four phases: exploratory data analysis (EDA), data preprocessing, generation of association and sequence rules, and visualization of results. The algorithmsAprioriand FP-Growth were used to generate the association rules. To generate sequence rules, the algorithm used was SPADE. Parallel coordinate graphs were used to visualize the association rules and graphs for visualization of sequence rules. An outstanding aspect of the proposed approach is the visualization of the rules obtained using graphic resources to enhance the analysis of the results in support of business decisions and contribute to mapping the users’ access profile. The proposal was validated by using data from user access to a digital news portal.

14
  • RAQUEL NUNES PEREIRA

  • Design of DGS Sensors for Monitoring Humidity in Civil Construction Materials

  • Líder : VALDEMIR PRAXEDES DA SILVA NETO
  • MIEMBROS DE LA BANCA :
  • ADAILDO GOMES D ASSUNCAO
  • CRISTHIANNE DE FATIMA LINHARES DE VASCONCELOS
  • FRANCISCO DE ASSIS BRITO FILHO
  • KLEBER CAVALCANTI CABRAL
  • VALDEMIR PRAXEDES DA SILVA NETO
  • Data: 20-dic-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Humidity monitoring is essential in the construction industry to ensure the durability of materials and prevent structural problems. In this study, the analysis and development of planar structures using the defect ground structure (DGS) technique and two-port resonator, for sensing applications, are performed. The operating principle consists of varying the resonant frequencies of the studied elements when the relative permittivity of the material under test (MUT) is changed. The effectiveness of the sensor is proven through the analysis of materials with previously known dielectric constants that validate its applicability. In this project, its application is directed to the determination of the percentage of water present in two samples of dry sand and to the monitoring of water loss in three specimens made of mortar. The prototype was manufactured and the preliminary results obtained in the measurements were compared with the results of other works, certifying the effectiveness of the presented sensor.

Tesis
1
  • LUCILEIDE MEDEIROS DANTAS DA SILVA
  • Hardware Proposal of Evolutionary Algorithm for Outlier Detection in Streaming Applications

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • IGNACIO SANCHEZ GENDRIZ
  • LEONARDO ALVES DIAS
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • MARCELO AUGUSTO COSTA FERNANDES
  • SERGIO NATAN SILVA
  • Data: 12-ene-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The amount of data in real-time, such as time series and streaming data, continues to grow. Analysing this data the moment it arrives can bring immense added value. However, it also requires much computational effort and new acceleration techniques. As a possible solution to this problem, we propose a hardware architecture for Typicality and Eccentricity Data Analytic (TEDA) algorithm implemented on Field Programmable Gate Arrays (FPGA) for data streaming. TEDA is based on a new approach to outlier detection in the data stream context. The suggested design has a full parallel input of N elements and a 3-stage pipelined architecture to reduce the critical path and thus optimise the throughput. To validate the proposals, results of the occupation, throughput and power efficiency of the proposed hardware are presented. Compared to other software platforms, the design reached a speed of up to 693x, with a throughput of up to 10.96 MSPs (Mega Sample Per second) with a dynamic power of 16mW. Besides, the bit-accurate simulation results are also presented for different application scenarios with multiple sensors ranging from applications in Industry 4.0 environments to the Internet of Medical Things (IoMT). This work is a pioneer in the hardware implementation of the TEDA technique in specialised hardware.

2
  • ELVIS MEDEIROS DE MELO
  • Learning Analytics and Online Evaluations: A Graph Data Science Methodology

  • Líder : IVANOVITCH MEDEIROS DANTAS DA SILVA
  • MIEMBROS DE LA BANCA :
  • BETANIA LEITE RAMALHO
  • DANIEL GOUVEIA COSTA
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • RAYMUNDO CARLOS MACHADO FERREIRA FILHO
  • Data: 24-ene-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Evaluation plays an extremely important role as a systematic instrument for correcting failures and promoting successes in the learning process. It is one of the tripods of the students' schooling, along with the curriculum and the teaching and learning process. In view of the new reality with online education (e-learning) spread on a larger scale due to the Covid-19 pandemic, the Federal University of Rio Grande do Norte (UFRN) institutionalized the Multiprova platform to support online evaluation processes in the institution. With the computerization of evaluation processes, Learning Analytics (LA) challenges and the need to use new techniques, such as graph data science. Thus, the task of understanding how students behave, identifying student profiles, and gaining insights through online evaluation resolution records is a field of LA research that can be optimized with graph data science methodology. With the use of LA techniques such as Machine Learning (ML), there is a need to transform interpretable models in education. For this, data visualization and eXplainable Artificial Intelligence (XAI) techniques need to be considered. Based on this reality, the thesis hypothesis arises: Is it possible to use data from online evaluation resolution logs to obtain insights into the learning process and student profiles using LA techniques such as graph modeling and ML? To this end, the theoretical framework about the topics that make up the object of study is presented, such as graphs, LA and online evaluations. Among the results, a systematic literature review pointed to 40 papers involving LA and online evaluations, but no papers used graph metrics with LA techniques such as ML to analyze student performance. Thus, two case studies were modeled according to the proposed graph data science methodology. We realized the importance of using graph features in LA techniques in identifying insights about student learning considering their journey in online evaluation, as well as graph metrics and XAI for the interpretation of the results.

3
  • JONATHA REVOREDO LEITE DA FONSÊCA
  • APPLICATION OF ADAPTIVE PROTECTION IN CORRECTIVE SWITCHING FOR ELIMINATION OF OVERLOADS CONSIDERING PROTECTION RESTRICTION.

  • Líder : MANOEL FIRMINO DE MEDEIROS JUNIOR
  • MIEMBROS DE LA BANCA :
  • ARRHENIUS VINICIUS DA COSTA OLIVEIRA
  • JOSE JULIO DE ALMEIDA LINS LEITAO
  • MANOEL FIRMINO DE MEDEIROS JUNIOR
  • MARCOS ANTONIO DIAS DE ALMEIDA
  • MELINDA CESIANARA SILVA DA CRUZ
  • RENATO MACHADO MONARO
  • Data: 08-feb-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The use of Corrective Switching technique to eliminate overloads in electrical systems, addressed in several studies in the last decades, has proved to be a very effective load flow control technique. Some of these studies are based on electrical circuit theories or Relief Function methodologies. Network switching causes changes to the system configuration and consequently the short-circuit levels of the buses are changed. These changes can lead to a network insecure operation when the protection parameters are not modified, since the overcurrent protection functions are based on the values of the rated currents and the short-circuit currents, which are previously defined in the protection studies. Adaptive Protection techniques makes it possible to recalculate the network protection parameters in real time, for any system configuration, enabling network operation with optimized protection settings. The purpose of this workis to implement the Corrective Switching technique, using, as a restriction condition in the choice of switching variants, the adjustments of the network protection parameters. If the network operation is not safe for the configuration proposed by the relief variants, new network protection parameters are calculated, as well as coordination and selectivity, is adopted and the Adaptative Protection technique is implemented. The network is automatically reconfigured through the Self Healing technique. 

4
  • DANIEL RODRIGUES DE LUNA
  • Adaptive Bandwidth Partitioning in 5G NR Systems Using Machine Learning Solution

  • Líder : VICENTE ANGELO DE SOUSA JUNIOR
  • MIEMBROS DE LA BANCA :
  • ANDRÉ MENDES CAVALCANTE
  • MARCELO AUGUSTO COSTA FERNANDES
  • MARCIO EDUARDO DA COSTA RODRIGUES
  • VICENTE ANGELO DE SOUSA JUNIOR
  • WALTER DA CRUZ FREITAS JÚNIOR
  • Data: 13-feb-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The fifth generation (5G NR) of the 3GPP communication network proposes a variety of use cases, ranging from improved mobile broadband communications (eMBB) to ultra-reliable low latency communications (URLLC), in addition to massive communications between machines (mMTC). Introducing flexibility in bandwidth utilization is one of the critical requirements of 5G technologies. As such, the concept of bandwidth parts is introduced to give flexibility, fast-changing configurations, and energy saving to UEs that do not need the whole bandwidth available by using just a part of it. The use case that can benefit from this feature is mMTC, characterized by its massive number of devices and the need for small power consumption. This work proposes a reinforcement learning solution of bandwidth part adaptation in mMTC scenarios to save energy and improve system statistics. Firstly, the subject and a complete section of related works with the more recent papers are introduced, highlighting the gap in this area of research. In addition, a brief theoretical approach to 5G is presented as part of the basis of this work. Then, the system model and main system parameters are described, along with the simulation tool used, the ns-3 NR 5G-LENA, in which calibration campaigns are carried out to validate its use. Afterward, we detail the investigation scenario in which we can apply the reinforcement learning (RL) solution. After decentralized and centralized solutions are proposed, new campaigns are created using both proposed RL-based bandwidth part adaptation solutions. The final results campaign shows the gains attained compared to a traditional fixed approach. Finally, the papers published and the main discussions of this thesis are gathered in the end.

5
  • RAFAEL LUCAS DA SILVA FRANCA
  • One-Terminal TravelingWave-Based Protection for HVAC and HVDC Transmission Lines

  • Líder : FLAVIO BEZERRA COSTA
  • MIEMBROS DE LA BANCA :
  • DIRK VAN HERTEM
  • FELIPE VIGOLVINO LOPES
  • FLAVIO BEZERRA COSTA
  • KAI STRUNZ
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • WASHINGTON LUIZ ARAUJO NEVES
  • Data: 28-feb-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • This work proposes the development of one-terminal traveling wave-based transmission line
    protection methods for HVAC (High Voltage Alternating Current) and HVDC (High Voltage
    Direct Current) systems. The effect of the sampling frequency on the protection is considered,
    which until now has not yet been investigated for one-terminal methods. Inaccuracies on the
    wave velocity estimations, which classically yields to problems on methods based on traveling
    waves, are addressed and solutions to such problem are presented. The proposed methods have
    been evaluated by means of computer simulations. The proposed earth fault distance protection
    for HVAC transmission lines was evaluated using a real commercial relay with time domain
    based protection functions. The results related to the proposed earth fault distance protection
    for HVAC transmission lines show that the proposed function, when associated with other existing
    protection functions, can achieve a quite remarkable enhancement for the dependability and
    velocity of the transmission line protection. The results concerning the method based on traveling
    wave reflections show that it is possible to protect most point-to-point transmission lines in
    a fast way, without the need for communication, and independent of the estimation of the wave
    speed. Finally, the results concerning distance protection for meshed HVDC systems demonstrate
    the applicability of the proposed method for such systems. The method showed operating
    time below 2 ms for a line of 500 km length. This operation time meets the likely requirements
    that HVDC meshed systems will present. In addition, the method presented selectivity for a
    4-terminal system interconnected by 5 transmission lines.

6
  • JOSE WANDERSON OLIVEIRA SILVA
  • Improved Behavioral Box for Analysis of Tactile Discrimination Tasks in Rodents

  • Líder : LUIZ MARCOS GARCIA GONCALVES
  • MIEMBROS DE LA BANCA :
  • ANDERSON ABNER DE SANTANA SOUZA
  • ANDOUGLAS GONÇALVES DA SILVA JÚNIOR
  • BRUNO MARQUES FERREIRA DA SILVA
  • JULIO CESAR PAULINO DE MELO
  • LUIZ MARCOS GARCIA GONCALVES
  • Data: 10-mar-2023


  • Resumen Espectáculo
  • Among the research works carried out with rodents in the area of Neuroscience, behavioral studies stand out. In order to carry out these studies, closed and controlled platforms are generally used in which the rodent is inserted, and tasks are elaborated, depending on the study, which the rodent must carry out. To study the learning of behaviors from tactile discrimination, for example, tasks are usually accompanied by a reward, normally following the reinforcement learning model. These platforms are commonly known as behavioral boxes. In this proposal, we aim to improve the methodologies and techniques used in this research area, developing new technologies for this type of study, including the automation of some procedures with the behavioral box aimed at tactile discrimination tasks. We did some initial tests of a behavioral box, using a computational model of physical simulation, from which we designed a new structure for the platform, with the best-elaborated spaces and materials, and using only components that can be bought in the market or easily developed. Our demonstrated hypothesis is that this new structure for the behavioral box improves the study of models related to the sensorimotor system in a more refined way. For example, the platform design enables more precise control of the opening of the discrimination bars, which is currently not done satisfactorily. The final model also allows for the design of more complex decision-making experiments using the camera and sensor system, allowing a better evaluation of rodent performance. This includes improvements in determining the number of correct answers in the tasks performed in the studies in question. Therefore, as a main practical contribution, we believe that the present study provides that laboratories that work with this type of research can enjoy a low-cost tool that is easy to develop. All material and documents developed are available on a multi-user collaborative platform.

7
  • DIEGO DA SILVA PEREIRA
  • Softwarized and Resilient Multi-UAV Network Applied to Rocket Impact Area Scanning

  • Líder : PABLO JAVIER ALSINA
  • MIEMBROS DE LA BANCA :
  • AUGUSTO JOSE VENANCIO NETO
  • CARLOS MANUEL DIAS VIEGAS
  • HELBER WAGNER DA SILVA
  • PABLO JAVIER ALSINA
  • TADEU FERREIRA OLIVEIRA
  • Data: 14-abr-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The scanning area procedure of the probable impact region is a mandatory step in the pre-launch 

    phase of any rocket. The objective is restrict the presence of unauthorized vessels, avoiding 

    possible accidents and, consequently, preserv human lives. Considering the high operational cost 

    of this procedure, normally performed by a P-95 aircraft, the SpaceVANT II project proposed the 

    execution of the area scan by a fleet of Unmanned Aerial Vehicles (UAVs) that, acting in a 

    cooperative way, have the ability to execute the mission with lower cost and operational complexity. 

    In this context, it is essential to create a communication network able to ensure the exchange of 

    information between the aircraft that make up the system. One of the most used alternatives to 

    provide this communication network is mobile ad hoc network (MANET) protocols adapted for 

    flying ad hoc networks (FANET). However, managing the dynamic behavior of this type of network 

    in order to minimize disconnections and enable the necessary conditions for the operation of 

    applications on board aircraft is not simple and has found limitations in the solutions adopted. In 

    this perspective, this thesis proposed a management based Software Defined Networking (SDN) 

    called SD-FANET. Design on a distributed and hierarchical architecture, the SD-FANET control 

    plane allows centralized management and assigns the controller the ability to program the 

    switches, changing the data forwarding from a holistic view of the network. The SDN approach 

    allows the use of different path selection algorithms, which adds more flexibility and improvement 

    to the forwarding process. To validate the proposed solution, experiments were carried out in the 

    emulated environment provided by Mininet-Wifi and a proof of concept based on a prototype using 

    a communication protocol developed to enable transmission of images in ZigBee communication 

    links embedded in DJI Phantom 3 Standard drones. The results showed that the throughput 

    obtained by SD-FANET was 22.2% better than MANET protocols and 5.4% in relation to the local 

    SDN controller. The average recovery time provides by resilience mechanisms was about 1.5 s.


8
  • SAMUEL BELARMINO DE PAIVA
  • Simulation and Design of MIMO Antennas for Applications in UWB, 5G and Smartphone Technologies

  • Líder : ADAILDO GOMES D ASSUNCAO
  • MIEMBROS DE LA BANCA :
  • ADAILDO GOMES D ASSUNCAO
  • ADAILDO GOMES D ASSUNCAO JUNIOR
  • ALEXANDRE JEAN RENE SERRES
  • CUSTÓDIO JOSÉ DE OLIVEIRA PEIXEIRO
  • LAERCIO MARTINS DE MENDONCA
  • VALDEMIR PRAXEDES DA SILVA NETO
  • Data: 23-may-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • This work presents three MIMO (multiple-input multiple-output) antennas for applications in modern wireless communication systems. The first proposed MIMO antenna has compact dimensions, high isolation between ports and can be used for wireless local area network (WLAN) applications; the second proposed MIMO antenna is a miniaturized MIMO antenna, with high isolation between the ports and which can be used in ultra-wideband (UWB) applications; the proposed third MIMO antenna can be used for smartphone applications and covers the lower and upper WLAN bands, as well as 5G band. Among the advantages that 5G offers, we can highlight ultra-fast speeds, low latency, large channel capacity, high spectral efficiency and massive connection density. Then, MIMO antennas have been extensively studied for applications in this technology, because they can considerably increase the data rate and the channel capacity, due to the channel capacity being proportional to the number of antennas used in transmission and reception, as well they can also improve the link reliability of wireless communications systems through multi-path data transmission and reception. The proposed MIMO antennas were designed using ANSYS HFSS software for numerical characterization, and the prototypes were manufactured with FR-4 dielectric and measured, showing good agreement between simulated and measured results. A key performance parameter of MIMO antennas, the envelope correlation coefficient (ECC), is analyzed and presents good results for the three proposed MIMO antennas. Furthermore, for the proposed antenna for smartphone applications, the effects caused by the user's hands are investigated, which is an extremely important performance parameter.

9
  • JOÃO MARIA ARAÚJO DO NASCIMENTO
  • Diagnosis of Operating Conditions and Sensor Failures in Wells Operating by Mechanical Pumping Using Machime Learning

     
     
  • Líder : ANDRE LAURINDO MAITELLI
  • MIEMBROS DE LA BANCA :
  • ANDERSON LUIZ DE OLIVEIRA CAVALCANTI
  • ANDRE LAURINDO MAITELLI
  • CARLA WILZA SOUZA DE PAULA MAITELLI
  • FABIO SOARES DE LIMA
  • OSCAR GABRIEL FILHO
  • RUTACIO DE OLIVEIRA COSTA
  • Data: 16-jun-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • In oil fields with many wells operating by Sucker-rod pumping, due to the lack of early diagnosis of operating conditions or sensor faults, several problems can go unnoticed. These problems can generate large losses, such as increased downtime, increased OPEX (Operational Expenditure), decreased efficiency and lost production. In practice, the identification and diagnosis of operating conditions are carried out from surface and downhole dynamometric cards, using pre-established standards, with human visual effort in the operation centers. This task requires a lot of time and work, in addition to requiring experience, as it can be influenced by subjective factors. However, in recent years, with the facilities inherent to Machine Learning (ML) algorithms, several researches on the subject have achieved good results in the diagnosis of operating conditions, showing that ML can be used for this purpose. However, it is still common to have doubts about the difficulty level of the dynamometer card classification task, the best algorithm, the best shape descriptor, the best metrics and what is the impact of the imbalanced datasets. In the search for answers to these questions, this work used real data from 38 wells operating by sucker-rod pumping in the region of Mossoró, RN, Brazil. More than 50,000 cards have been classified by specialists and distributed in eight modes of operation and two common sensor faults in this field. Sixty tests were performed and divided into seven groups. Three algorithms were tested (with and without hyperparameter tuning): Decision Tree, Random Forest and XgBoost, in addition to pipelines provided by Automated Machine Learning (AutoML). The descriptors used were: Fourier descriptors and Wavelet descriptors, in addition to the load values of the downhole dynamometric card. Balanced and imbalanced datasets were also tested. The results confirm the feasibility of applying ML for diagnosis of operating conditions and sensor faults in sucker-rod pumping systems, since 75% of the tests reached accuracy above 92% and a maximum accuracy was 99.84%.

     
     
10
  • JOSE KLEITON EWERTON DA COSTA MARTINS
  • A new method for design of nonlinear PID controllers under constraints using Fuzzy Takagi-Sugeno models

  • Líder : CARLOS EDUARDO TRABUCO DOREA
  • MIEMBROS DE LA BANCA :
  • ANDRÉ FELIPE OLIVEIRA DE AZEVEDO DANTAS
  • CARLOS EDUARDO TRABUCO DOREA
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • LEANDRO LUTTIANE DA SILVA LINHARES
  • WALLACE MOREIRA BESSA
  • Data: 26-jun-2023


  • Resumen Espectáculo
  • Most of the existing controllers in industries are from the PID family (Proportional-Integral-Derivative) developed for linear systems, although the processes are non-linear. Therefore, they present a significant loss of performance when operating outside the operating point for which they were designed. An alternative controller is the so-called PDC (Distributed Parallel Compensator) which, based on a Fuzzy Takagi-Sugeno model, results in a non-linear controller.

    The Fuzzy Takagi-Sugeno model is intrinsically non-linear, due to the membership functions that make up its structure. Thus, this model can represent nonlinear systems in such a way that, from the physical model of the system, it is possible to obtain a Fuzzy Takagi-Sugeno model that represents it exactly, which contributes to obtaining the PDC. The Fuzzy Takagi-Sugeno model can be represented by the composition of local models. In the PDC controller methodology, local controllers are designed, which can be a PID, for each local model, thus presenting the same rules structure as the Fuzzy Takagi-Sugeno model, that is, the controller shares the same membership functions.

    Normally, the PDC stability guarantee considers a quadratic Lyapunov function as a candidate, thus allowing it to be tested for linear matrix inequalities (LMIs) that are solvable using efficient convex optimization algorithms. However, polyhedral Lyapunov functions have shown advantages over the quadratic function, especially with regard to satisfying constraints on state, input and output variables, showing their potential to be used to guarantee stability and performance of a PDC-controlled system. In view of this, this work proposes an optimization-based method for tuning Proportional-Integral (PI) controllers for nonlinear systems subject to constraints on the controlled variable and actuator saturation. Conditions are presented for a polyhedron contained in the set of constraints to be invariant with respect to a closed-loop system with a controller subject to saturation. These conditions are used in the formulation of a bilinear programming problem whose solution provides the controller parameters that satisfy the constraints and an associated invariant set. Throughout the development of the work, several variations of the main technique were explored and their advantages and disadvantages were shown through the results.

    The results are presented in numerical examples. Starting with linear systems and testing I-P controllers and a traditional PI controller with a filter for the reference, we explore case studies with: saturation allowance, state constraints, time-delay, high order systems, servo and regulator problem. Then we explore nonlinear systems testing systems with and without the affine term and controllers with and without the feedfoward term, exploring case studies with: saturation permission, state constraints and servo problem. The results show that the developed method obtains a tuning for the controller that does not violate any state constraint and allows saturation, thus illustrating its efficiency.

11
  • MAGNO MEDEIROS DE ARAÚJO
  • The Design of a Glycerol Concentration Sensor Based on a Photonic Biosensor for Point-of-Care Application

  • Líder : JOSE PATROCINIO DA SILVA
  • MIEMBROS DE LA BANCA :
  • JOSE PATROCINIO DA SILVA
  • LAERCIO MARTINS DE MENDONCA
  • VALDEMIR PRAXEDES DA SILVA NETO
  • IGUATEMI EDUARDO DA FONSECA
  • VITALY FÉLIX RODRÍGUEZ ESQUERRE
  • Data: 27-jul-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The manufacturing processes advancement in silicon wafers has enabled integration between electronics and photonics. Data will no longer be processed with electrons for the technology sectors but with photons, providing higher processing and transmission rates. In this context, the health and well-being sectors, for example, will have new equipment, allowing remote and individual monitoring and diagnosis, enhancing the prevention and treatment of diseases. Thus, this research proposes a photonic device with graphene applied and a surface plasmon resonance structure for sensing applications. Initially, they are concepts related to photonic devices and their different applications. Photonic crystals, plasmonic devices, graphene, and their use in optical guides are also presented. Then, the methodology used in the research development is explained, dealing with modeling and simulation. The results are presented and analyzed as the structure’s spectral behavior for different glycerol concentrations, which reinforces the potential for point-of-care applications considering its compactness and simplicity of construction.

12
  • FILIPE CAMPOS DE ALCANTARA LINS
  • DPO: Direct Planar Odometry with Stereo Camera
  • Líder : PABLO JAVIER ALSINA
  • MIEMBROS DE LA BANCA :
  • VALDIR GRASSI JUNIOR
  • ADELARDO ADELINO DANTAS DE MEDEIROS
  • ANDERSON ABNER DE SANTANA SOUZA
  • MARCELO BORGES NOGUEIRA
  • PABLO JAVIER ALSINA
  • Data: 28-jul-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Contemporary Visual Odometry (VO) methodologies generally are created upon point-based approaches to estimate the camera's pose and a representation of the environment explored. Direct Sparse Odometry (DSO) is the most popular point-based technique of a class of approaches called direct methods. Many works presented improvements in the point-based VO area using DSO characteristics as the fundamentals of these approaches. However, only recently, two new monocular plane-based DSO have been presented. The first approach utilizes a learning-based plane estimator to generate initial plane estimates, which may lead to optimization issues. The second approach restricts plane detection to horizontal and vertical orientations, making it more suitable for structured environments.

    This thesis presents a stereo plane-based VO technique - Direct Planar Odometry (DPO) - that employs planes as features in a Sliding window optimization framework and utilizes unit dual quaternion as the pose parameter. Our experiments show that the proposed methods achieved comparable results to the Stereo DSO point-based approach.

13
  • MICHEL SANTANA DE DEUS
  • PROGRAMMABLE PULSE INTEGRATOR AMPLIFIER FOR SYSTEMS ON CHIP

  • Líder : SEBASTIAN YURI CAVALCANTI CATUNDA
  • MIEMBROS DE LA BANCA :
  • VALNER JOÃO BRUSAMARELLO
  • DIOMADSON RODRIGUES BELFORT
  • FERNANDO RANGEL DE SOUSA
  • RAIMUNDO CARLOS SILVÉRIO FREIRE
  • SEBASTIAN YURI CAVALCANTI CATUNDA
  • Data: 28-jul-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • This work is an extension of previous studies and presents the research and development of a pulse width programmable gain integrating amplifier for chip systems. One of the main advantages of the proposed PGA architecture is that it uses few components and the number of gain values is independent of the circuit area. In addition, the architecture can be configured to operate with single-ended or differential signals.

     

    Before starting the Integrated Circuit (IC) development, equations defining the uncertainties of the circuit were obtained and a discrete circuit was designed, simulated, and developed, which went through a series of experiments to validate the proposed architecture.

     

    The implementation of the architecture in an IC aims to achieve higher operating frequencies, allowing its use in a wider range of applications. The integrated circuit was designed using the TSMC 0.18 µm CMOS technology, presenting an average power dissipation of 123.2 µW and occupying an area of approximately 0.065 mm2.

     

    Transient simulations were performed with and without noise, DC, AC, PSS, PAC, Corner, and Monte Carlo, evaluating means, uncertainties, relative and absolute errors related to various parameters. This suggests that the circuit was carefully designed and tested, and the presented results provide a solid basis for the application of the proposed PGA in a wide range of analog conditioning applications.

14
  • ÁLVARO PINTO FERNANDES DE NEGREIROS
  • High-level subsumption-based control architecture for sail-powered autonomous surface vehicles

  • Líder : LUIZ MARCOS GARCIA GONCALVES
  • MIEMBROS DE LA BANCA :
  • EDUARDO CHARLES VASCONCELLOS
  • DAVI HENRIQUE DOS SANTOS
  • ESTEBAN WALTER GONZALEZ CLUA
  • JOAO MORENO VILAS BOAS DE SOUZA SILVA
  • LUIZ MARCOS GARCIA GONCALVES
  • PABLO JAVIER ALSINA
  • Data: 04-ago-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • We propose a high-level control architecture for an Autonomous Surface Vessel (ASV) designed to overcome the challenges associated with executing missions under varying weather conditions and with energy autonomy. The project is built on the understanding that while a number of low-level control techniques are readily available as open source, there remains a need for a high-level control architecture. This architecture would facilitate the creation of a resilient, versatile sailing vessel capable of handling any mission without requiring the user to master navigation specifics, naval procedures, or corner cases. The proposed solution involves developing a control architecture inspired by subsumption, and centered on hierarchical behaviors. This structure incorporates a variety of specialized behaviors adapted to different contexts, each of which is established using reinforcement learning techniques (PPO). A combination of Gazebo simulation environment with the ROS framework for training was used to validate the proposed architecture. This simulation enables the digital replication of the vessel's behaviors, which simplifies the implementation process and mitigates the challenges and costs tied to real-world sailing operations. The results of this study indicate that the high-level control architecture of the virtual sailing vessel was successful in passing both perimeter scanning and long-distance tests. This suggests that the ASV is equipped to navigate the several situations it might encounter in real-world missions.

15
  • ISAAC DANTAS ISIDÓRIO
  • Observer-Based Output Feedback Control Using Invariant Polyhedral Sets for Fuzzy T-S Models Under Constraints

     

  • Líder : CARLOS EDUARDO TRABUCO DOREA
  • MIEMBROS DE LA BANCA :
  • CARLOS EDUARDO TRABUCO DOREA
  • EUGENIO DE BONA CASTELAN NETO
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • KURIOS IURI PINHEIRO DE MELO QUEIROZ
  • VILMA ALVES DE OLIVEIRA
  • Data: 07-ago-2023


  • Resumen Espectáculo
  • In this work, a numerical method for the computation of observer-based output feedback controllers is proposed for fuzzy Takagi-Sugeno (T-S) systems subject to constraints, based on set-invariance theory. Positively Invariant (PI) polyhedral sets are used to ensure that state and control constraints are satisfied at all times. Sufficient conditions are established for a polyhedron defined in the augmented state space (state + estimation error) to be PI. From the invariance conditions, a bilinear optimization problem is formulated to simultaneously calculate the controller and observer gains and the positively invariant polyhedron that guarantee the satisfaction of the constraints. The two types of observers found in the literature of fuzzy T-S systems are considered: the first considers the membership functions dependent only on the system output, while the second refers to the general case, where these functions can be associated with any state variable. In the simplest case, although the membership functions depend only on the output, the estimated state feedback results, in general, in controllers with better performance and with larger sets of admissible states associated with them than the output static feedback control. For the general case, as membership functions depend on non-accessible states, an estimation mechanism is needed to calculate these variables. In both cases, this role is played by the fuzzy observer T-S. The problem of tracking a constant reference signal is also considered, for which the concept of robust positive invariance is used in conjunction with an Integral-Proportional (I-P) controller. Sufficient conditions are established for a polyhedron defined in the augmented state space (state + estimation error + tracking error integral) to be PI in the presence of a constant reference signal. Several numerical experiments illustrate the effectiveness of the proposed approach.

16
  • NELSON JOSÉ BONFIM DANTAS
  • Resonant and Antiresonant control of Multiple-Input Second-Order Systems with Time Delay using Receptance

  • Líder : CARLOS EDUARDO TRABUCO DOREA
  • MIEMBROS DE LA BANCA :
  • CARLOS EDUARDO TRABUCO DOREA
  • FERNANDO DE OLIVEIRA SOUZA
  • JOSÉ MÁRIO ARAÚJO
  • PERICLES REZENDE BARROS
  • WALLACE MOREIRA BESSA
  • Data: 22-ago-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • A variety of problems in control engineering field deal with mathematical models under the form of systems of second-order differential equations, resulting from first-principles analysis using discretization (i.e., finite element method) to obtain finite-dimensional models. These models require, in general, an accurate knowledge of the parameters involved, such as, for example, masses, damping and stiffness coefficients in mechanical systems, for the design of controllers with high performance. The need to consider delays in measuring variables or actuation in these types of systems makes the problem more challenging, in view of the transcendent nature of the resulting transfer functions. In this context, the receptance concept emerges as an alternative, enabling the modeling of these systems entirely from experimental data, dispensing with the use of simplifying hypotheses and approximations, as is recurrently done in mathematical models with discrete parameters. Another virtue is highlighted in the waiver of approximations for the exponential term of the delay, which can result in undesirable solutions or very high order transfer functions. Since their introduction, receptance-based models have gained prominence in several works, with emphasis on the active control of vibrations by allocating zeros (anti-resonant frequency) and/or poles (resonant frequency). This work proposes a method for controlling second-order systems with multiple inputs and delays using receptance. The structure of the controllers uses state feedback, and the control objectives include partial placement of zeros and poles with guaranteed stability and robustness, based on the maximum peak of the sensitivity function. Stability is guaranteed from the extended concept of the Nyquist criterion for systems with multiple inputs. The synthesis of the controller is conducted from an optimization problem and the solutions are obtained using an algorithm based on evolutionary theory, where population evaluations are defined according to functions related to the frequency response of the systems. Numerical examples are presented to illustrate and discuss the performance of the solutions using the proposed method.

17
  • BRENO SANTANA SANTOS
  • A methodology oriented to unstructured data of scientific production for the temporal evaluation of research groups

  • Líder : IVANOVITCH MEDEIROS DANTAS DA SILVA
  • MIEMBROS DE LA BANCA :
  • RICARDO BARROS SAMPAIO
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • JUAN MOISES MAURICIO VILLANUEVA
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • MARCELO AUGUSTO COSTA FERNANDES
  • Data: 29-ago-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Funding agencies and research institutions often employ quantitative methods and scientometric techniques to evaluate scientific groups. These evaluations typically rely on a single type of metric, whether it is based on counts (such as the h-index) or derived from Complex Network Analysis. However, the use of multiple measurement approaches and the proper exploration of the temporal dimension of academic production is still a recurrent issue. Particularly, an underexplored approach involves combining these indicators with Machine Learning and Graph Embedding techniques, which could enhance the evaluation process of research groups. In this context, this study proposes a Graph Data Science-oriented methodology to analyze scientific teams over time. Through a case study, the results suggest the feasibility and suitability of the proposed method for quantitatively assessing research groups. The presented approach has the potential to provide strategic and proactive insights for scientific teams, contributing to a better understanding of their dynamics and limitations.

18
  • INGRIDY MARINA PIERRE BARBALHO
  • RevELA Platform: A digital health solution for the management of care and surveillance of patients with Amyotrophic Lateral Sclerosis in Brazil

  • Líder : RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • MIEMBROS DE LA BANCA :
  • ANTONIO HIGOR FREIRE DE MORAIS
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • CRISTINE MARTINS GOMES DE GUSMÃO
  • DANILO ALVES PINTO NAGEM
  • JAILTON CARLOS DE PAIVA
  • KARILANY DANTAS COUTINHO
  • RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • Data: 30-ago-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Amyotrophic Lateral Sclerosis (ALS) is a rare neurodegenerative disease considered complex due to its heterogeneity. Despite being known for many years, few countries have accurate information about its epidemiology and the characteristics of individuals diagnosed with ALS, such as data related to the diagnosis and progression of the disease. In Brazil, the lack of information about ALS limits the use of data for advancing research and developing public policies that favor people affected by this health condition. In this sense, this work aims to present the development of a digital health solution for managing the care and monitoring of the ALS disease in Brazil to provide health surveillance with valuable and timely information to support the formulation of public policies and, at the same time, time to qualify and improve the follow-up of patients with ALS through an electronic health record. The proposed solution is composed of two tools: the Brazilian National Registry of ALS, responsible for collecting epidemiological data from patients with ALS throughout Brazil in a structured way; and the Electronic Health Record for ALS patients (PEP ELA, acronym in Portuguese), responsible for assisting in the multidisciplinary follow-up of patients with ALS, safely and adequately, based on the ALS Clinical Protocols and Therapeutic Guidelines (PCDT, acronym in Portuguese). This work was guided by the action-research methodology, and all phases of development of the platform were carried out sequentially in several cycles through the application of the iterative and incremental model using the SCRUM framework. It is essential to highlight that the development of this solution had the support of the Brazilian Ministry of Health and was conceived in response to the needs identified in the epidemiological scenario related to the disease, as well as the limitations reported by health professionals specialized in the care of patients with ALS. To promote security in the sharing of patient data, a blockchain network was modeled to receive smart contracts related to transactions carried out on the platform, thus ensuring the highest level of privacy and security. The developed platform has great potential for strengthening public policies since the data stored on the platform can be analyzed by teams of care specialists and public health managers, both in the individual and collective monitoring of people living with ALS in Brazil. The integration of surveillance and monitoring data can have significant benefits for the development of public policies and the planning of strategies at all levels of health care, in addition to providing a realistic view of the situation in the country about cases. Therefore, this digital health solution can be used for research, intervention, monitoring, and strengthening the response to ALS in the Brazilian Health System.

19
  • FELIPE RICARDO DOS SANTOS FERNANDES
  • DIGITAL HEALTH SOLUTION FOR ALTERNATIVE COMMUNICATION FOR PEOPLE WITH AMYOTROPHIC LATERAL SCLEROSIS

     

  • Líder : RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • MIEMBROS DE LA BANCA :
  • ANTONIO HIGOR FREIRE DE MORAIS
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • CRISTINE MARTINS GOMES DE GUSMÃO
  • DANILO ALVES PINTO NAGEM
  • ERNANO ARRAIS JUNIOR
  • KARILANY DANTAS COUTINHO
  • RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • Data: 31-ago-2023


  • Resumen Espectáculo
  • Amyotrophic Lateral Sclerosis (ALS) is a neurodegenerative disease that irreversibly impairs an individual's motor system and functional abilities, even causing the progressive loss of communication skills and autonomy. Technological resources based on Augmentative and Alternative Communication, Computer Vision, and Machine Learning are essential for developing digital health solutions to enable the communicative process and autonomy that, consequently, promote improvements in the quality of life and survival of people with ALS. Focused on a Human-Computer Interaction (HCI) approach based on images of the eyes from a simple camera not mounted on the body, this work presents an assistive technology resource for Augmentative and Alternative Communication for people with ALS. The approach proposed in this work consists of an algorithmic model capable of recognizing the state of the eye (open or closed) in real-time and interoperating with Autonomus, a digital health solution designed by the Laboratory of Technological Innovation in Health at the Federal University of Rio Grande do Norte (LAIS/UFRN) for the communication of people with ALS. The model consists of four methodological processes: (i) image acquisition; (ii) Face detection; (iii) eye detection; and (iv) classifying the state of the eye, which is the foremost step for Human-Computer Interaction. An algorithmic study with a control group was conducted to evaluate the model's overall performance and the Convolutional Neural Network (CNN) classification ability. The results related to the proposed model for classifying the state of the eye in real-time are promising and reach significant values of accuracy and f1-score above 92%. The results also point to the viability of developing low-cost assistive technology resources that guarantee universal access, health promotion, well-being, and reduced inequalities, which go beyond improvements in the communicative process of people with ALS. Therefore, the object of study of this work is also to enable and promotes the exercise of rights, citizenship, fundamental freedoms, and health care for people with ALS.



20
  • JOSÉ RAIMUNDO LIMA JÚNIOR
  • Two-Terminal Traveling-Wave-Based Non-Homogeneous Transmission-Line Protection

  • Líder : FLAVIO BEZERRA COSTA
  • MIEMBROS DE LA BANCA :
  • FELIPE VIGOLVINO LOPES
  • FLAVIO BEZERRA COSTA
  • KLEBER MELO E SILVA
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • WASHINGTON ARAUJO NEVES
  • Data: 04-sep-2023


  • Resumen Espectáculo
  • This work presents a two-terminal traveling-wave-based protection algorithm applied to non-homogeneous transmission lines comprising any number of sections with different topologies and considering the effect of the sampling rate. Existing two-terminal traveling wave protection functions cannot protect the line under close-in faults and present limitations in non-homogeneous transmission lines. However, the effects of the sampling rate, considered in the proposed method, result in well-defined protection and unprotected zones, essential for protection security and development of solutions to deal with the issue of close-in faults in non-homogeneous transmission lines. Furthermore, the algorithm can accurately detect the faulted section, allowing its use in advanced protection functions such as adaptive automatic auto-reclosing and high-speed protection schemes. A protection device was modeled considering a sampling frequency equal to 1 MHz, including functions to detect traveling waves via wavelet transform, and the proposed protection algorithm to discriminate line internal from external faults, and to select the faulted section. The algorithm was evaluated using a large number of Alternative Transients Program (ATP) fault simulations. The results show that the algorithm is robust and reliable for protection devices installed in non-homogeneous lines.

21
  • GABRIELA DE ARAUJO ALBUQUERQUE LEMOS
  • Machine Learning Aplicado a Triagem de Osteoporose: modelo baseado em atenuação de ondas eletromagnéticas

  • Líder : RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • MIEMBROS DE LA BANCA :
  • ANTONIO HIGOR FREIRE DE MORAIS
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • CRISTINE MARTINS GOMES DE GUSMÃO
  • GUILHERME MEDEIROS MACHADO
  • JOÃO PAULO QUEIROZ DOS SANTOS
  • LORENA ITATI PETRELLA
  • RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • Data: 21-sep-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Osteoporosis is a silent and still underdiagnosed condition, with a mortality rate higher than several types of cancer, especially when patients suffer fractures. The gold standard equipment for the diagnosis of densitometry, Dual-energy X-ray absorptiometry (DXA, or DEXA), which requires an invasive and costly procedure, is scarce in countries considered middle or low-income, thus hindering timely access to diagnosis. In this context, a portable device, known in the literature as Osseus, was developed for the screening of patients who need the densitometry exam, i.e., to qualify the referrals of exams to the DEXA equipment. The thesis aimed to validate the Osseus device using machine learning techniques. For this, the planning and data collection of 505 patients who underwent the DEXA exam and the test on the Osseus device were carried out, of which 21.8% were healthy and 78.2% were diseased (they had low bone mineral density or osteoporosis). Therefore, to implement the studies and develop the research, the database was separated into 80% for training and validation (5-fold cross-validation) and 20% for testing. The performance obtained in the test base with the best model (Random Forest) corresponded to sensitivity=0.853, specificity=0.871, and F1(harmonic average of precision and recall rate)=0.859. The results showed that the most relevant variables to indicate the individual health status were age, body mass index (BMI), and the attenuation measured in the Osseus. When compared to the results of DEXA scans, the model has proven to be effective and consistent in screening individuals with osteoporosis and facilitating early diagnosis of the disease, which consequently entails improved productivity and reduced costs for surgery, treatment, and hospitalization. This way, by qualifying the referral of patients from primary care to the specialized network, the Osseus can improve the osteoporosis care network and consolidate itself as a resource of easy access in primary care, also impacting the reduction of waiting lines in the specialized network of the BrazilianNational Health System (SUS).


22
  • ANDRE HENRIQUE MATIAS PIRES
  • Fuzzy Controllers Optimization by Multiobjective Genetic Algorithms in the Wavelet Domain

  • Líder : FABIO MENEGHETTI UGULINO DE ARAUJO
  • MIEMBROS DE LA BANCA :
  • CARLOS EDUARDO TRABUCO DOREA
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • JEAN MARIO MOREIRA DE LIMA
  • LEANDRO LUTTIANE DA SILVA LINHARES
  • MÁRCIO EMANUEL UGULINO DE ARAÚJO JÚNIOR
  • Data: 02-oct-2023


  • Resumen Espectáculo
  • Due to the increasing competitiveness in the industry, it is imperative to use more efficient tuning techniques that can in fact find controllers with the desired performance. With this proposal, optimization techniques can be used to obtain the controller parameters according to an evaluation criterion, which should encode how good a particular controller is, properly expressing the desired specifications, so that the algorithm employed can find the controller. wanted. The methods traditionally used in tuning present the difficulty in expressing the desired specifications. The difficulty is due to the fact that the traditionally used criteria, in general, only use the total error information, through indices such as the Integral Absolute Error (IAE) or the Integral Square Error (ISE), which do not describe aspects of system behavior, such as if the response is very aggressive and oscillatory, steady state error, rise time and stabilization time, as a human designer would do. Some of these impressions are not well defined for references other than the step, lacking generality. Thus the optimization algorithm responsible for obtaining the controller parameters according to an evaluation function, which must actually be able to encode how good a given controller is, adequately expressing the desired specifications, so that the optimization algorithm employed can find the controller that best satisfies such a function. In view of this, a generic methodology for using wavelet analysis will be presented along with multiobjective optimization techniques to express more closely and closely related to the human behavior of the controlled system, allowing a more accurate optimization. In the proposed methodology, wavelet analysis, very present in the literature, focused on other applications, especially in the analysis of signals, sounds and images, is used to obtain descriptors that describe aspects of system behavior, such as its steady state behavior, behavior In the transient, no amplification of noise and rejection of disturbances, these descriptors become objectives that will be optimized by multiobjective techniques. The study carried out used Multiobjective Genetic Algorithm (MOGAs) techniques for optimization, due to their being widely used in the literature and known for their simplicity and efficiency.

23
  • MARIA GRACIELLY FERNANDES COUTINHO
  • Stacked Sparse Autoencoder applied to SARS-CoV-2 virus classification based on image representations of genome sequences

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • ADRIANA TAKAHASHI
  • ADRIAO DUARTE DORIA NETO
  • LEONARDO ALVES DIAS
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • MARCELO AUGUSTO COSTA FERNANDES
  • Data: 06-oct-2023


  • Resumen Espectáculo
  • Since December 2019, the COVID-19 pandemic caused by the SARS-CoV-2 virus has intensely affected the world. In the case of a novel virus identification, the early elucidation of taxonomic classification and origin of the virus genomic sequence is essential for strategic planning, containment, and treatments. Deep learning techniques have been successfully used in many viral classification problems associated with viral infection diagnosis, metagenomics, phylogenetics, and analysis. Considering that motivation, this work proposed an efficient viral genome classifier for SARS-CoV-2 using the deep neural network based on the stacked sparse autoencoder (SSAE). For the best performance of the model, we explored the utilization of image representations of the complete genome sequences as the SSAE input to provide a classification of the SARS-CoV-2. For that, two datasets were explored: based on k-mers image representation and based on CGR image representation. The dataset based on k-mers image representation was applied in the experiments of different levels of taxonomic classification of the SARS-CoV-2 virus, and the dataset based on CGR images was applied to the experiments of classification of SARS-CoV-2 variants of concern (VOC). For the experiments of taxonomy classification, the SSAE technique provided great performance results, achieving classification accuracy between 92% and 100% for the validation set and between 98.9% and 100% when the SARS-CoV-2 samples were applied for the test set. These results indicate that our model can be adapted to classify other emerging viruses. For the experiments of SARS-CoV-2 variants classification using CGR images, the SSAE technique provided even better results, achieving classification accuracy of 99.9% for the validation set and 99.8% for the test set. Finally, the results indicated the applicability of this deep learning technique in genome classification problems.

24
  • YURI PEDRO DOS SANTOS
  • Adaptive clustering of NOMA users in the power domain

  • Líder : LUIZ FELIPE DE QUEIROZ SILVEIRA
  • MIEMBROS DE LA BANCA :
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • VICENTE ANGELO DE SOUSA JUNIOR
  • TIAGO TAVARES LEITE BARROS
  • FABRÍCIO BRAGA SOARES DE CARVALHO
  • PEDRO THIAGO VALERIO DE SOUZA
  • Data: 19-oct-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The technique of Non-Orthogonal Multiple Access (NOMA, Non-Orthogonal Multiple Access) is intended to enable the transmission of two or more users sharing the same resources of time, frequency, and code, and thus significantly improve the spectral efficiency of wireless communication networks in future generations. Signal multiplexing can be achieved in the Power Domain NOMA (PD-NOMA), where superposed signals are transmitted with sufficiently different power levels. The efficiency of this method fundamentally depends on two previous processing steps: an adequate clustering of users (transmission candidates) with different channel gains and the choice of power levels that will be used to transmit each signal. The solutions presented in the literature to solve the user clustering problem do not consider the dynamics of the communication systems, that is, the temporal variation of the number of users and the channel conditions. To consider these dynamic characteristics when grouping users in NOMA systems, this work proposes a new clustering technique based on a modification of the evolutionary algorithm DenStream, chosen for its evolutionary capacity, robustness to noise, and online processing. Results show that the proposed clustering technique follows the system's dynamics, clustering all users and favoring the uniformity of the transmission rate between the clusters.

25
  • TARCIANA CABRAL DE BRITO GUERRA
  • Multi-Connectivity Solutions for LTE and NR Networks

  • Líder : VICENTE ANGELO DE SOUSA JUNIOR
  • MIEMBROS DE LA BANCA :
  • FUAD MOUSSE ABINADER JUNIOR
  • LEONARDO AGUAYO
  • MARCIO EDUARDO DA COSTA RODRIGUES
  • VALDEMIR PRAXEDES DA SILVA NETO
  • VICENTE ANGELO DE SOUSA JUNIOR
  • WALTER DA CRUZ FREITAS JÚNIOR
  • Data: 27-oct-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Dual connectivity (DC) technology (or, more generally, multiple connectivity (MC)) is an important feature in the initial journey of New Radio (NR), the access network of 3GPP to 5G, being the basis for the first ways of implementing this generation of communication systems. With MC, users can be simultaneously connected to legacy technologies (4G-LTE and Wi-Fi) and the new 5G-NR technology. However, such technology poses additional challenges for the network, especially with regard to energy management and the numerous connectivity configuration options. This work aims to conceive strategies based on machine learning capable of efficiently exploring the performance of MC towards 5G eMBB (enhanced Mobile Broadband) and 5G URLLC (Ultra Reliable Low Latency Communications) in heterogeneous networks. The idea is to dynamically configure the best set of MC parameters to provide increased data throughput (eMBB target) and a more robust connection (URLLC target), being relevant in scenarios with the presence of small cells using mmWaves, for example, both considered essential to meet 5G specifications.

26
  • ALCEMY GABRIEL VITOR SEVERINO
  • Representation Based on Stacked Autoenconder Optimized by Particle Swarm

  • Líder : FABIO MENEGHETTI UGULINO DE ARAUJO
  • MIEMBROS DE LA BANCA :
  • CARLOS EDUARDO TRABUCO DOREA
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • JEAN MARIO MOREIRA DE LIMA
  • LEANDRO LUTTIANE DA SILVA LINHARES
  • SERGIO NATAN SILVA
  • Data: 20-nov-2023


  • Resumen Espectáculo
  • Market competitiveness drives organizations to pursue technological development in order to improve product quality and reduce production costs while meeting the socio- environmental demands of consumers. However, industrial processes can pose challenges in real-time monitoring and control of critical variables. One solution to this problem is the use of soft sensors, which are algorithms capable of estimating difficult-to-measure variables based on easily measurable secondary variables. A common challenge in soft sensor projects is the lack of labeled data, making semi-supervised methods more promi- sing than traditional methods. In this context, the Stacked Autoencoder neural network architecture has been widely employed. This architecture is trained in an unsupervised manner and subsequently fine-tuned in a supervised manner. However, appropriately de- fining the hyperparameters of the Stacked Autoencoder, such as batch size, learning rate, and number of hidden features, presents a challenge. Traditional methods like Grid Se- arch and Random Search are computationally intensive and may not quickly find the best combination of hyperparameters. A more efficient alternative is the use of metaheu- ristic algorithms, such as Particle Swarm Optimization. These algorithms intelligently explore the search space and are more effective in high-dimensional spaces. A promising approach is to incorporate Mutual Information into the evaluation function of Particle Swarm Optimization, along with the Mean Squared Error. Mutual Information captures nonlinear relationships between the outputs of the Stacked Autoencoder and the actual system outputs, while the Mean Squared Error measures the difference between these outputs. In this context, the present thesis proposes the Representation Based on Particle Swarm Optimized Stacked AutoEncoder method, which utilizes Particle Swarm Optimi- zation with a modified evaluation function to optimize the hyperparameters of a Stacked Autoencoder-based soft sensor. It is expected that this approach will improve the accu- racy and representation capacity of the Stacked Autoencoder compared to conventional approaches that only utilize the Mean Squared Error. In order to evaluate the performance of the models generated by the proposed method, two widely used nonlinear processes in the industry were selected. These processes were chosen due to their relevance in virtual sensor implementation and are frequently employed in comparative analyses. The results demonstrate that the incorporation of Mutual Information in the evaluation function al- lows for a more efficient and balanced search, resulting in a Stacked Autoencoder with improved performance and representation capacity.

27
  • GLAUBER RODRIGUES LEITE
  • Uncalibrated visual servoing in the presence of non-gaussian feature tracking noise

  • Líder : ALLAN DE MEDEIROS MARTINS
  • MIEMBROS DE LA BANCA :
  • ALLAN DE MEDEIROS MARTINS
  • JOÃO PAULO FERREIRA GUIMARÃES
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • PABLO JAVIER ALSINA
  • ÍCARO BEZERRA QUEIROZ DE ARAÚJO
  • Data: 08-dic-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Visual servoing is a control strategy that uses visual feedback from cameras to control the motion of a robot or a system. Image-based visual servoing relies on image processing and computer vision algorithms to detect and track image features, incorporating them directly in the control loop. That approach considers that there is a map, also known as interaction jacobian, between feature motion and camera pose based on the camera's intrinsic and extrinsic parameters. Although there are calibration techniques to compute the camera's parameters, they can become error-prone or need online changes, especially in unstructured scenarios. Some examples that could happen are when a task requires image zoom, the camera presents lens distortion, or its sensor has temperature sensitivity. Uncalibrated visual servoing studies aim to estimate the interaction jacobian using environment information and the measurement of features displacement, generally at run-time, with the help of an estimator, such as the Kalman filter. While most studies approximate the estimation uncertainty to a gaussian distribution, the environment in which the robot actuate could present more challenging characteristics. In that case, target occlusion, reflection, or similar appearance to other image objects can confuse the computer vision algorithm leading to outliers in the feature extraction. If not treated correctly, these errors may poor the performance of the visual servoing controller, or even affect its convergence. The maximum correntropy criterion can take advantage of the statistical properties of non-gaussian random variables. Thus, this work proposes a thesis theme study on dealing with non-gaussian feature tracking noise, preserving its statistical properties through the maximum correntropy criterion applied to the Kalman filter.
28
  • JOSE ILTON SARMENTO SILVEIRA JUNIOR
  • Reference Tracking via Output Feedback for Constrained Uncertain Linear Systems

  • Líder : CARLOS EDUARDO TRABUCO DOREA
  • MIEMBROS DE LA BANCA :
  • CARLOS EDUARDO TRABUCO DOREA
  • KURIOS IURI PINHEIRO DE MELO QUEIROZ
  • JOSÉ MÁRIO ARAÚJO
  • TIAGO ALVES DE ALMEIDA
  • TITO LUÍS MAIA SANTOS
  • Data: 08-dic-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • This thesis presents a robust output feedback control approach designed to address constant reference tracking in linear systems affected by uncertainties, disturbances, and measurement noise. Our research contributions revolve around three primary aspects. Firstly, we establish the conditions for achieving the Output Feedback Controlled Invariance (OFCI) property within the context of linear systems that incorporate polytopic uncertainties. This OFCI property guarantees robust constraint satisfaction through output feedback even in the presence of uncertainties. Secondly, we introduce a dynamic output feedback controller tailored to uncertain models, which employs set-membership observers. These observers effectively reduce the set of feasible states, enhancing tracking performance by minimizing errors. Thirdly, to further reduce tracking errors, we propose a model update procedure that adjusts the nominal model used for tracking based on output measurements. We conduct extensive numerical experiments to assess the controller's effectiveness, demonstrating its capability to achieve reference tracking with significantly reduced errors for the uncertain systems under consideration. Our research offers valuable insights into addressing constant reference tracking in linear systems with output feedback, constrained control and state, while accounting for uncertainties, disturbances, and noise. The potential applications of our approach hold promise for advancing control strategies in practical systems operating under uncertain conditions.

29
  • MÁRIO SÉRGIO FREITAS FERREIRA CAVALCANTE
  • Modified Type-2 Neuro-Fuzzy Structure for Identification and Behavior Prediction of Nonlinear Systems

  • Líder : FABIO MENEGHETTI UGULINO DE ARAUJO
  • MIEMBROS DE LA BANCA :
  • THIAGO DE SOUZA ROCHA
  • ALLAN DE MEDEIROS MARTINS
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • JEAN MARIO MOREIRA DE LIMA
  • ÍCARO BEZERRA QUEIROZ DE ARAÚJO
  • Data: 11-dic-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • System identification is a crucial sphere of engineering dedicated to finding econo- mical yet accurate models for fully understanding how systems behave. In effectuating this aim, these models predict future behavior while enabling simulations for optimization purposes inclusive of parameter adjustments where necessary for enhanced performance levels.

    However, what makes identifying systems challenging is the selection process regar- ding model structure choice and the estimation method used when making predictions concerning non-linearities present in complex phenomena affecting multiple variables. Nonetheless, experts have devised viable options toward precise modelling solutions by employing sophisticated techniques such as artificial intelligence algorithms or polyno- mial multi-model frameworks.

    he proposed thesis offers an approach that fuses interval type-2 fuzzy logic together with neural network training skills towards producing a generalized structure that enables both local model selection combined modeling which permits approximating or forecas- ting the behavior of any given system.

    The results were obtained using three case studies: the chaotic Mackey-Glass time equation, a furnace system, and a multisection tank system. The results of the pro- posed network for the approximation and prediction of these systems were compared with techniques from the literature, and the modified type-2 neuro-fuzzy interval network (MIT2FNN) showed lower mean squared error (MSE) values than the other techniques.

30
  • MYCHAEL JALES DUARTE
  • Analysis and Design of Ultraminiaturized FSS Structures for Applications in Temperature Sensing Systems
  • Líder : VALDEMIR PRAXEDES DA SILVA NETO
  • MIEMBROS DE LA BANCA :
  • ADAILDO GOMES D ASSUNCAO
  • ADAILDO GOMES D ASSUNCAO JUNIOR
  • ALFREDO GOMES NETO
  • LAERCIO MARTINS DE MENDONCA
  • MARCIO EDUARDO DA COSTA RODRIGUES
  • VALDEMIR PRAXEDES DA SILVA NETO
  • Data: 21-dic-2023
    Ata de defesa assinada:


  • Resumen Espectáculo
  • This paper presents a study on ultra-miniaturized Frequency Selective Surfaces (FSS) for applications in temperature sensing systems. Two 2.5D FSS are proposed with elements inspired by convoluted metal lines printed on an FR-4 dielectric substrate. The inserted vias contribute to the capacitive and inductive effects in the structure, providing ultraminiaturization of the unit cell dimensions. The FSS have unit cell sizes equal to 4.68 % and 6.80 % of the wavelengths in free space for the frequencies of 2.34 GHz and 2.72 GHz (resonance frequencies), respectively, with the second structure exhibiting dual-band behavior. Two equivalent circuit models are proposed to better understand the operating principle of the FSS. The application of these structures for use in temperature sensing is studied. According to the literature, the electrical permittivity of the dielectric material used changes with the change in the material's temperature. With this in mind, simulations were carried out and the FSS showed a linear relationship between resonance frequency and temperature. The two proposed sensors were built and measured from room temperature up to 120 °C, and both showed excellent sensitivity to temperature changes. The numerical results simulated for the prototypes designed were obtained using the ANSYS HFSS software and the equivalent circuit model. The prototypes were built and the experimental characterization of the transmission coefficients, bandwidth and resonance frequency was carried out. The values obtained in the experiments were compared and discussed with the simulation results, which showed good agreement.

31
  • MARCOS TULIO ANTUNES BEZERRA SEGUNDO
  • The Application of Quasi-Periodic Sequence Models in Optical Waveguide

  • Líder : JOSE PATROCINIO DA SILVA
  • MIEMBROS DE LA BANCA :
  • HUMBERTO DIONISIO DE ANDRADE
  • JOSE PATROCINIO DA SILVA
  • LAERCIO MARTINS DE MENDONCA
  • MAGNO MEDEIROS DE ARAÚJO
  • VICENTE ANGELO DE SOUSA JUNIOR
  • Data: 22-dic-2023


  • Resumen Espectáculo
  • In this work, waveguides based on directing light through media with sequentially refractive indices in the direction of propagation are proposed. The core guiding model of these structures is based on quasi-periodic sequencing of the refractive index that forms the signal propagation region. In this context, well-known optical fiber models are modified to adapt to the segmentation structures of optical signal guiding regions.

    These modifications enable the development of new optical device models, such as lasers, filters, and optical sensors. One of the most traditional segmentation models applied in optical waveguides is the Bragg Grating, also known as Fiber Bragg Grating (FBG), where the sequencing in the propagation region follows a periodic pattern. More recently, some quasi-periodic sequence models have been applied in photonic crystals, primarily considering the cross-sectional area of the waveguide. In this context, a new segmentation strategy is proposed, based on quasi-periodic sequencing of the refractive index along the direction of propagation.

    Structures of core fractionation in waveguides will be investigated by applying quasi-periodic models, such as Fibonacci sequences, Thue Morse, Period Doubling, and Octonacci, with the aim of comparing their performance with waveguides that use Bragg Grating as a segmentation strategy. For this study, a mathematical formulation based on the finite element method, in conjunction with the optical beam propagation method, will be used as a simulation tool. Parameters of transmission and reflection at the interfaces between the different media composing the applied sequences will be analyzed.

2022
Disertaciones
1
  • MATHEUS NUNES ORSANO AIRES
  • A Wavelet-based Restricted Earth-fault Power Transformer Differential Protection

  • Líder : FLAVIO BEZERRA COSTA
  • MIEMBROS DE LA BANCA :
  • FLAVIO BEZERRA COSTA
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • BRUCE A. MORK
  • RODRIGO PRADO DE MEDEIROS
  • WASHINGTON LUIZ ARAUJO NEVES
  • Data: 20-ene-2022


  • Resumen Espectáculo
  • Power transformer differential protection schemes have been widely studied in recent years to improve their accuracy and speed. Researches have been developed by using the most varied digital signal processing techniques. Internal faults may cause critical damages to the power transformer, and they need special attention for this kind of disturbance, mainly in faults on the winding of the power transformer that is not too sensitive to the principal differential protections. This work aims to present a restricted earth fault protection based on the wavelet transform (REFW) for detecting ground faults close to the transformer neutral point (turn-to-ground faults) and supporting the conventional phase differential protection, which presents limitations in this kind of faults. The proposed REFW protection uses only high-frequency components instead of traditional low-frequency components obtained by means of phasor estimation methods, speeding up the detection of turn-to-ground faults.

2
  • LUCAS SOLANO CADENGUE
  • Intelligent Control of Omnidirectional Robots using Recurrent Neural Networks

  • Líder : WALLACE MOREIRA BESSA
  • MIEMBROS DE LA BANCA :
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • ORIVALDO VIEIRA DE SANTANA JUNIOR
  • PHILIPPE EDUARDO DE MEDEIROS
  • WALLACE MOREIRA BESSA
  • Data: 11-mar-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Due to their great efficiency, security and flexibility, mobile robots are being increasingly used in industry. However, their positioning control is a great challenge due to the non-linear nature of this plant and the difficulty of estimating certain parameters, for example, the friction effects. In this work, non-linear controllers are applied to the trajectory control of an omnidirectional robot under the effect of unmodeled dynamics. The control approaches used in this work were both non-linear control strategies, Feedback Linearization (FBL) and Sliding Modes (SMC) both incorporated with an intelligent compensator utilizing Recurrent Neural Networks in order to assist the control by estimating uncertainties. The chosen architecture of the neural network was based in the need to compensate more complex dynamics and at the same time the restriction of computational complexity so that it could be embedded in the hardware of a mobile robot. The stability properties were proven by the principle of assintotic stability proposed by Lyapunov and the performance of the strategies were verifed through both simulations and experiments using Robotino, an omnidirectional mobile robot produced by Festo Didatics and a performance gain was observed when compared with the neural network without the recurrence.

3
  • REINALDO AGOSTINHO DE SOUZA FILHO
  • An OpenMP Implementation for the Nanvix Operating System

  • Líder : SAMUEL XAVIER DE SOUZA
  • MIEMBROS DE LA BANCA :
  • HENRIQUE COTA DE FREITAS
  • KAYO GONCALVES E SILVA
  • SAMUEL XAVIER DE SOUZA
  • TIAGO TAVARES LEITE BARROS
  • Data: 26-abr-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Multicore programming is not a trivial task. In some cases, multicore systems have no parallel programming interfaces, which requires operating system support to be made available. This work implements a version of the development interface of parallel programming Open Multi-Processing (OpenMP) for the Nanvix operating system. OpenMP is a parallel application development interface that allows for the intuitive abstraction of parallelism and the division of workload across processes. It is common to have memory limitations in platforms that embed codes in their projects. These memory limitations could make unavailable the use of an operating system with programming-friendly interfaces. The system software of embedded processing platforms is limited in programmability, requiring more knowledge from the programmer about the process and the hardware architecture used. The Nanvix is an operating system embedded in multicore platforms, being light enough to fit in memory restriction environments; however, it has standards that ease the programming. The Nanvix includes native parallel programming interfaces inspired on the Portable Operating System Interface (POSIX) adopted to implement some versions of OpenMP, that hopes to build an OpenMP version to the Nanvix it is going to result in easy coding for the operating system. In this work, the version of OpenMP uses Nanvix’s compilator to translate the compilation directives, and it assembles a library to the running routines. It was tested in an emulator of the RISC-V architecture. Some applications were built in the OpenMP version and their equivalent on the native Nanvix library of parallel programming to validate the library. The results show that, in some cases, OpenMP has more parallel efficiency compared to Nanvix’s native API, but in other cases, it has more performance. The programmability of Nanvix is improved with OpenMP, lowering the development time of parallel applications without a considerable performance loss, and in some cases, it is improving efficiency.

4
  • MÁRCIO LUIZ BEZERRA LOPES JÚNIOR
  • Stratification of Preterm Birth Risk in Brazil Through Unsupervised Learning Methods and Socioeconomic Data

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • ALEXANDRE DIAS PORTO CHIAVEGATTO FILHO
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • LEONARDO ALVES DIAS
  • MARCELO AUGUSTO COSTA FERNANDES
  • RAQUEL DE MELO BARBOSA
  • Data: 29-abr-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Preterm birth (PTB) is a phenomenon that brings risks and challenges for the survival of the newborn child. Despite many advances in research, not all the causes of PTB are yet clear. It is currently understood that PTB risk is multi-factorial and may also be associated with socioeconomic factors. In order to analyse this possible relationship, this work seeks to stratify PTB risk in Brazil using only socioeconomic data, extracting and analysing those clusters that present relevant PTB divergence, all of which will be found by automatic clustering processes using a series of unsupervised machine learning methods. Through the use of datasets made publicly available by the Federal Government of Brazil, a new dataset was generated with municipality-level socioeconomic data and a PTB occurrence rate. This dataset was processed using two separate clustering methods, both built by assembling unsupervised learning techniques, such as $k$-means, principal component analysis (PCA), density-based spatial clustering of applications with noise (DBSCAN), self-organising maps (SOM) and hierarchical clustering. The methods discovered clusters of municipalities with both high levels and low levels of PTB occurrence. The clusters with high PTB were comprised mostly of municipalities with lower levels of education, worse quality of public services -- such as basic sanitation and garbage collection -- and a less white population. The regional distribution of the clusters was also observed, with clusters of high PTB located mostly in the North and Northeast regions of Brazil. The results indicate a positive influence of the quality of life and the offer of public services on the reduction of PTB risk.

5
  • PAULO RANNIER COSTA DA SILVA
  • Measurement of Oxygen Uptake Rate using Sigma Delta Modulator in the Biological Domain in Activated Sludge Systems

  • Líder : SEBASTIAN YURI CAVALCANTI CATUNDA
  • MIEMBROS DE LA BANCA :
  • ANTONIO WALLACE ANTUNES SOARES
  • BRUNO AUGUSTO FERREIRA VITORINO
  • DIOMADSON RODRIGUES BELFORT
  • SEBASTIAN YURI CAVALCANTI CATUNDA
  • Data: 03-may-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The activated sludge system is a wastewater treatment method, capable of removing water pollutants using aerobic bacteria metabolism. The oxygen uptake rate (OUR) is an important parameter to the evaluation and monitoring of the activated sludge system, since it allows evaluating the biodegradation of polluting organic matter. In this work,
    we propose a method to measure OUR based on the architecture of the first order sigma delta modulator in the biological domain. In this method, the operational functions of the sigma delta modulator will be performed by biological processes and electronic circuits. The proposed method has some advantages over existing OUR measurement methods. When compared to the classical method, it improves the dynamic response, decreases and
    fixes the sampling time; when compared to the PWM method, it introduces the benefits of noise modeling and oversampling. The performance of the proposed model is demonstrated through simulations.

6
  • ARTHUR DA COSTA FERNANDES
  • Real-time studies about the technique of Corrective Switching

  • Líder : MANOEL FIRMINO DE MEDEIROS JUNIOR
  • MIEMBROS DE LA BANCA :
  • ARRHENIUS VINICIUS DA COSTA OLIVEIRA
  • AYLANNA RAQUEL DA COSTA OLIVEIRA
  • EDNARDO PEREIRA DA ROCHA
  • MANOEL FIRMINO DE MEDEIROS JUNIOR
  • MAX CHIANCA PIMENTEL FILHO
  • Data: 26-jul-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • With the growth of technologies for the automation of electrical systems, it is important to create ways to maintain the supply of electrical energy even in abnormal situations, with fast and low-cost resolutions, keeping load restriction as a last resort. The Corrective Switching technique is a tool capable of controlling the power flow of meshed networks by changing the network topology. The main advantage of this control methodology is that it does not involve additional costs, since its implementation depends on maneuvers by elements already present in the network. In previous works, several techniques and methodologies were developed to reduce the computational time needed to implement the technique, envisioning its application in the operation of systems in real-time. The present work aims to develop a computer program capable of identifying overloads in branches of the system and reconfiguring the network in real-time through the Corrective Switching technique. The simulations of electrical networks in real-time will be carried out through RTS. The computer program with the techniques and methodologies used to choose the switching variant will be developed using Scilab software, which will exchange information with the simulation in real-time in RTS. To improve the performance of the technique in the simulations, linearization techniques were also used to gain computational time. Finally, analysis of the results obtained from simulating the corrective switching technique in real time is presented.

7
  • EMANOEL LUCAS RODRIGUES COSTA
  • Self-Organizing Maps Applied to the Analysis of Atmospheric Pollutants

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • ADRIAO DUARTE DORIA NETO
  • CRISTIANO HORA DE OLIVEIRA FONTES
  • MARCELO AUGUSTO COSTA FERNANDES
  • ÉDLER LINS DE ALBUQUERQUE
  • Data: 27-jul-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Air pollution is a problem that is increasingly present in our society due to the growing development of countries. In the study of air pollutants, multivariate statistical methods are commonly used, however, machine learning has proved to be a great alternative, presenting techniques capable of dealing with highly complex problems, such as air pollution. In this work, the machine learning technique, Self-Organizing Maps (SOM), was used to explore and analyze data on atmospheric pollutants and meteorological parameters from an air quality monitoring network, with stations located in the city of Salvador - Bahia. SOM offers several resources capable of making the study of data more comprehensive, which were used for the development of an individual and mutual analysis on the stations, being also briefly compared with a principal component analysis. From the visualization of the component planes, patterns between the air quality variables could be identified, as well as the observation of the present correlations, which were more specifically described by a hierarchical analysis of similarity, allowing to raise assumptions about their influence, formations and possible sources of emission, with a better description of the results. In addition, based on the arrangement of the neurons on the map, a study regarding data clusters could be carried out, enabling a balance on the samples and formation of clusters, characterizing in this way information related to the concentrations of pollutants, with their specificities and how they can be related to each monitoring station according to the division and arrangement of neurons.

8
  • LUISA CHRISTINA DE SOUZA
  • New proposal for viral genome representation applied to the classification of SARS-CoV-2 with deep learning

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • MARCELO AUGUSTO COSTA FERNANDES
  • LUIZ MARCOS GARCIA GONCALVES
  • LEONARDO ALVES DIAS
  • Data: 19-ago-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • In December of 2019, the first case of COVID-19 was found in Wuhan, China, and in April of 2021, there were already 136 million confirmed cases. Due to the virus fast propagation, the scientific community has been making efforts to develop viral classifications techniques for the SARS-CoV-2. In this work, was developed, using a set of techniques from Genomic Signal Processing, a new proposal of genomic data representation of six viruses from the Coronaviridae family, which the SARS-CoV-2 belongs to. Then, the accomplished mapping was applied in a deep learning architecture for the samples' viral classification, obtaining accuracy of 94% e 91% for the sequences resized for the sizes of 64 and 128, respectively, also obtaining sensibility of 100% for the vectors with size 64. Lastly, given the mutation rate of the RNA virus, new variants emerged, and with them the possibility of an increase in cases. It was then, using the developed technique, carried out an analysis of the evolution of four variants of concern in three viral classification procedures, the results obtained aided comprehending the phylogenetic relationships between the variants.

9
  • KAROLAYNE SANTOS DE AZEVEDO
  • Deep Learning Applied to Classification and behavior analysis SARS-CoV-2 virus

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • MARCELO AUGUSTO COSTA FERNANDES
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • SERGIO NATAN SILVA
  • Data: 19-ago-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The new Beta Coronavirus, officially named SARS-CoV-2 (Severe Acute Respiratory Syndrome Coronavirus - 2 - SARS-CoV-2) is the virus causing COVID-19 disease. A member of the Coronaviridae family of viruses, SARS-CoV-2 is a positive-sense, single-stranded RNA enveloped virus that contains nearly 30,000 base-pair pairs - bp).RNA viruses tend to undergo more modifications than DNA viruses. Thus, when a virus is circulating widely in a population and causing many infections, the probability of its genome undergoing modifications increases, which may negatively affect some of its properties, becoming more transmissible and/or even more lethal. Within this context, this work proposes a tool, based on machine learning, which makes use of a deep one-dimensional (1D) convolutional neural network (CNN), intended for the classification and comparison of viral genomes of the new SARS-CoV - 2. As input, complete genomic cDNA samples (complementary DNA) were used, whose size varies between 26342 and 31029 base pairs (base-pair - bp) in length. Contrary to most approaches presented in the literature, the results obtained by this tool involving the classification of viruses, from the same family, reveal high values for the performance metrics, proving to be more reliable when compared to the works discussed in the state of the art. The proposed model was also used to verify possible changes in the genomic sequences of the main concern variants (alpha, beta, gamma), over a period of time, through their accuracy values, obtained through the classification between the variants.For this experiment, genomic samples from GISAID (Global Initiative on Sharing All Influenza Data - GISAID) were used, which also hosts epidemiological and clinical data referring to all variants related to SARS-CoV-2. The results obtained in this experiment indicate that the model can be used not only for classifying the virus of the Coronaviridae family, but also for predicting the behavior of SARS-CoV-2 variants over time.

10
  • ARTHUR ANDRADE BEZERRA
  • Indoor devices for automatic data collection of particulate matter in the air in the prevention of COVID-19 and other severe acute respiratory syndrome endemics

  • Líder : LUIZ MARCOS GARCIA GONCALVES
  • MIEMBROS DE LA BANCA :
  • ANDOUGLAS GONÇALVES DA SILVA JÚNIOR
  • DAVI HENRIQUE DOS SANTOS
  • JULIO CESAR PAULINO DE MELO
  • LUIZ MARCOS GARCIA GONCALVES
  • Data: 26-ago-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Este trabalho ou projeto de detecção de qualidade relevante, detectou e detectou ativamente um dispositivo de previsão de previsão e comparação com índices de dados de medição de previsão/final de qualidade com um microcontrole capaz de enviar como informações apresentadas para um banco de todos os dados. Por ser um projeto que necessita de grande escala para que seja possível capturar índices de qualidade do maior número de pontos possível para obter dados com uma granularidade muito alta, o desenvolvimento está sendo pensado semper visualizando o custo-benefício dos componentes para que replicar, e também o desenvolvimento faz parte de um projeto maior possível, que deve ser disponibilizado à comunidade uma plataforma completa, capaz de fornecer dados de qualidade do ar em tempo real.

11
  • FÁBIO HENRIQUE DE CARVALHO FERRAZ
  • An Optmization-Based MIMO PID Controller Tuning Method for Linear Systems Under Constraints

  • Líder : CARLOS EDUARDO TRABUCO DOREA
  • MIEMBROS DE LA BANCA :
  • ANDRE LAURINDO MAITELLI
  • CARLOS EDUARDO TRABUCO DOREA
  • TIAGO ALVES DE ALMEIDA
  • Data: 07-oct-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • In this work we propose a methodbased on optimization for tuning PID controllers, for linear Multiple-Input Multiple-Output (MIMO) discrete-time systems subject to constraints on the output variables and saturation of the actuators. From a polytopic modeling of saturation, conditions are presented for a polyhedron contained in the set of constraints to be positively invariant with respect to a closed-loop system with a MIMO PID controller. Such conditions are used in the formulation of a nonlinear optimization problem, whose solution provides the controller parameters that satisfy the constraints, guarantee local stability and optimize the performance of the closed loop system. Both a centralized and a decentralized controller are analyzed. The efficiency of the proposed method is evaluated from numerical simulation results.

12
  • ALAN RODRIGUES DE SOUSA
  • Contributions to the Sliding Mode Control Strategy Applied to Grid-Forming Converters.

  • Líder : RICARDO LUCIO DE ARAUJO RIBEIRO
  • MIEMBROS DE LA BANCA :
  • ALEXANDRE CUNHA OLIVEIRA
  • FLAVIO BEZERRA COSTA
  • RICARDO LUCIO DE ARAUJO RIBEIRO
  • THIAGO DE OLIVEIRA ALVES ROCHA
  • Data: 09-nov-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Distributed generation systems, which use renewable energy sources, are increasingly present
    due to the need to meet energy demand and environmental restrictions. Generally, such
    generation systems are connected to the electrical grid by means of power converters and,
    their integration into it, can produce instability in the electrical system. With the advancement
    in power electronics and the wide range of applications involving single-phase voltage
    converters, it is possible, with the appropriate control strategy, to dodge instability problems
    and also to offer improvements in power quality. For this, several control techniques have
    been studied for this purpose, however, the classical techniques fail for not having robustness
    to parametric uncertainties and are dependent on the system model. Thus, the control by
    sliding modes has gained space in the control of converters because it is a robust technique to
    parametric uncertainties and is independent of the system model, but the presence of
    chattering in the controlled variable in steady state is its main disadvantage. This master
    thesis presents a sliding mode control strategy applied to single-phase grid-forming converters.
    The proposed strategy employs linear systems analytical tools to determine the of closed-loop
    poles' suitable location when constrained to the splip surface. Furthermore, it is using a
    simplified structure for the internal current control. Simulation results obtained from a single-
    phase VSI connected to the grid operating under different operational conditions demonstrate
    the effectiveness of the proposed solution.

13
  • AMANDA KARINE MIRANDA DE OLIVEIRA

  • CHARACTERIZATION OF THE ATTENUATION OF RADIO FREQUENCY SIGNALS BY DIFFERENT MATERIALS USED IN CIVIL CONSTRUCTION

  • Líder : VALDEMIR PRAXEDES DA SILVA NETO
  • MIEMBROS DE LA BANCA :
  • ADAILDO GOMES D ASSUNCAO
  • ERICA NATASCHE DE MEDEIROS GURGEL PINTO
  • FRED SIZENANDO ROSSITER PINHEIRO
  • VALDEMIR PRAXEDES DA SILVA NETO
  • Data: 27-dic-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • In recent years, technological development and advances in the studies of communication systems based on radio-frequency (RF) have fostered several researches, with emphasis on the propagation of RF signals in the indoor environment. However, there is still an unfilled space with regard to research on the influence of construction materials on the propagation performance of these radio signals. This work aims to study the behavior of RF signals when reaching obstacles of building elements based on ceramic blocks, gypsum blocks and concrete blocks, and to experimentally determine the attenuation caused by these materials, which are currently most used in the construction civil. A measurement setup is proposed in this work and samples of these materials were subjected to RF signals in the frequency range between 700 MHz to 2500 MHz.They were analyzed in 03 samples of each type of material and the measured results for the attenuation of the signals were statistically treated and compared. The results indicate that the attenuation of the RF signal depends on the type of material used and its composition, therefore, it is necessary to develop studies in an attempt to propose new materials that meet the strength requirements demanded by civil construction and minimize the propagation losses of these signs in the indoor environment.

Tesis
1
  • BRUNO SÁTIRO DA SILVA
  • Equivalent Circuit Method Applied in Ring Geometries for Design of Selective Surfaces in Complementary Frequency

  • Líder : ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • MIEMBROS DE LA BANCA :
  • ALFREDO GOMES NETO
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • JOSE PATROCINIO DA SILVA
  • LAERCIO MARTINS DE MENDONCA
  • MAURICIO WEBER BENJÓ DA SILVA
  • Data: 16-feb-2022


  • Resumen Espectáculo
  • The benefits of frequency-selective control of electromagnetic wave flow make frequency selective surface (FSS) a continuous subject of study in applied electromagnetism areas. With a wide range of applications in various engineering fields, FSS can be used, for example, as an antenna efficiency improvement tool, used for selective electromagnetic shielding, and even for fault detection in concrete structures such as bridge pillars. One particular class of FSS is the Complementary Frequency Selective Surface (CFSS) which has very interesting features together such as high angular stability, multiple transmission and/or reflection bands, and unit cell miniaturization. The analysis and design of this type of structure are commonly performed using commercial software that implements the so-called full-wave methods (FWM) which, despite having reliable responses, requires a license to use and demands high computational effort, impacting a longer time to obtain results. The Equivalent Circuit Model (ECM) provides approximate equations that, by calculating impedances related to unit cell geometry, model the behavior of frequency-selective surfaces in the frequency domain faster due to simplified formulas, which can be implemented in free software in many different programming languages, showing good agreement when compared to full-wave methods. This research aims to optimize the ECM when applied to complementary frequency selective surfaces formed by rings, uniting the interesting characteristics of both the cascade structure and the geometry. The study is paralleled with output techniques from transmission line theory and genetic algorithms for greater effectiveness of the method.

2
  • ELIZABETH VIVIANA CABRERA AVILA
  • Windowed Optimization for Stereo Visual Odometry Fusion

  • Líder : LUIZ MARCOS GARCIA GONCALVES
  • MIEMBROS DE LA BANCA :
  • ANDERSON ABNER DE SANTANA SOUZA
  • BRUNO MARQUES FERREIRA DA SILVA
  • JOAO MORENO VILAS BOAS DE SOUZA SILVA
  • LUIZ MARCOS GARCIA GONCALVES
  • PABLO JAVIER ALSINA
  • Data: 18-feb-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Accurate motion estimation is an essential task accomplished in autonomous mobile agents' navigation (aerial or ground) and in robots navigation. In this thesis, we propose to reduce the error of stereo visual odometry when using 6 degrees of freedom poses through a graph optimization-based visual odometry fusion approach using the redundancy of captured information from the environment. Our approach uses two stereo images sets of a public dataset captured with a moving platform mounted on the top of a robot Pioneer 3AT to compute independent stereo odometry employing the LIBVISO algorithm and later fuse them. Our results are compared against two recognized SLAM frameworks ORB-SLAM2 and UCOSLAM and with the stereo odometry algorithm input. The relative pose error of the fused poses decreases by up to 94\% in relation to the error of stereo odometry and by up to 91\% compared with the results of UCOSLAM. Our implementations are open source and use public libraries.

3
  • ROSANA CIBELY BATISTA REGO
  • Lyapunov-based Intelligent Control

  • Líder : FABIO MENEGHETTI UGULINO DE ARAUJO
  • MIEMBROS DE LA BANCA :
  • ADEMAR GONÇALVES DA COSTA JÚNIOR
  • CARLOS EDUARDO TRABUCO DOREA
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • MARCUS VINICUS SILVÉRIO COSTA
  • TAKASHI YONEYAMA
  • Data: 23-feb-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Nonlinear dynamical systems play a crucial role in control systems because, in prac- tice, most of the plants are nonlinear. However, achieving control for nonlinear systems is not simple though many methods have been developed. There are still some problems to be solved, as robust control balance in humanoid robots and the modeling inaccuracies of the autonomous underwater vehicle, which has a small-pitch-angle. Usually, a Lyapunov function is used to perform a control and stability analysis of a nonlinear system. The procedure for obtaining a Lyapunov function is not a simple task. There have been many efforts and numerical methods in the literature on how to estimate Lyapunov functions for several kinds of systems. An artificial neural network is a useful tool for generat- ing functions. Motivated by this, we investigated the capability of a neural network to compute Lyapunov functions and provide a deep neural network to compute a control Lyapunov function without any linear approximation for nonlinear systems. Moreover, we examined the equilibrium point stability and obtained an estimation of its region of attraction contained in the set. Numerical examples and experimental simulations using some nonlinear systems, such as the inverted pendulum and the rotary inverted pendulum, are performed and compared with some conventional control techniques.

4
  • MARIA IZABEL DA SILVA GUERRA
  • Study of Intelligent Controllers for Tracking the Maximum Power Point of a Photovoltaic System

  • Líder : FABIO MENEGHETTI UGULINO DE ARAUJO
  • MIEMBROS DE LA BANCA :
  • ANDRE LAURINDO MAITELLI
  • ANDRES ORTIZ SALAZAR
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • JOÃO TEIXEIRA DE CARVALHO NETO
  • MARCELO ROBERTO BASTOS GUERRA VALE
  • Data: 07-abr-2022


  • Resumen Espectáculo
  • Photovoltaic (PV) systems have shown growth in the world's electrical matrix. However, the non-linear nature of PV array and their strong dependence on ambient conditions decrease the maximum power they can produce and, consequently, reduce their performance and commercial attractiveness. Maximum Power Point Tracking (MPPT) techniques have been studied over the years to minimize these problems. Among the various control techniques used for spot tracking and maximum power, those that use intelligent algorithms to control the switching of DC-DC converters have shown a high potential for use. Therefore, the present work proposes to develop MPPT techniques based on Artificial Neural Network (ANN), fuzzy and Adaptive Neuro-Fuzzy Inference System (ANFIS) intelligent algorithms, to be applied to PV systems that have the buck-boost as a CC-CC converter. Three proposed architectures were developed for each algorithm. They were compared with each other and with the classic Perturb and Observe (P&O) algorithm. The proposals are distinguished by the input variables used, namely: irradiance and ambient temperature, for Proposal 1, with purely environmental parameters as input variables; irradiance and instantaneous output power of the PV array, for Proposal 2, with input variables that mixed environmental and electrical parameters; and instantaneous and previous instantaneous output power of the PV array, for Proposal 3, with purely electrical parameters as input variables. To assist in the study of the performance of the intelligent algorithms, two scenarios of PV systems, composed of PV array, buck-boost converter, MPPT and load, were modeled, identified Scenario 1 and Scenario 2. The scenarios were differentiated by the total power of the system. At the end of the analyses, it was noticed that the intelligent algorithms had a high tracking speed and were more stable than the P&O algorithms. The PV systems controlled by the intelligent algorithms of Proposal 1 showed the highest efficiency in reaching the maximum power point. The ANFIS and ANN algorithms were more prominent. In power generation, ANN recovered up to 12.05% of the energy lost when using P&O. In the Proposal 2 study, the PV systems also performed well, but lower than the Proposal 1 algorithms. The highest power generated was also achieved by the ANN. It generated 12.01% more power than the P&O. In Proposal 3, the intelligent algorithms had their efficiency compromised because the overlapping of some database values. Anyway, under Random condition, the intelligent algorithms still proved to be superior to P&O in tracking the maximum power point, recovering 8.27% of the generated power. Therefore, intelligent algorithms, especially ANN and ANFIS, have shown the relevance of their use in photovoltaic applications, especially in regions with random environmental conditions. Furthermore, the proposed intelligent algorithms are more attractive as the power of the PV system to be used is high.

5
  • VICTOR RAMON FRANÇA BEZERRA DE SOUZA
  • CONTRIBUTION FOR IMPROVEMENT OF LOW VOLTAGE RIDE THROUGH CAPABILITY ON DOUBLE FED INDUCTION GENERATOR WITH A BACK-TO-BACK MODULAR MULTILEVEL CONVERTER TOPOLOGY

  • Líder : LUCIANO SALES BARROS
  • MIEMBROS DE LA BANCA :
  • KLEBER CARNEIRO DE OLIVEIRA
  • ANDRES ORTIZ SALAZAR
  • FLAVIO BEZERRA COSTA
  • FRANCISCO KLEBER DE ARAÚJO LIMA
  • LUCIANO SALES BARROS
  • Data: 02-may-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The advancements in power electronics have supported the widespread penetration of wind energy conversion systems (WECS) in electric grids. In this sense, the system based on the Double Fed Induction Generator (DFIG) has a large share of installed wind farms power generation technology, in this configuration the stator is directly connected to the grid and the rotor is interfaced by means of two back-to-back power converters, which have a crucial functionality in the machine control by means of voltages applied to the rotor terminals in order to ensure that the stator voltages and currents have the same frequency as the electrical grid, besides controlling the active and reactive power and are directly related to the harmonic distortion, thermal and mechanical stress levels in the WECS. Currently, several power converters topologies have been employed in order to improve the performance and increase the WECS processed power capability to support the network demand, in this sense, the multi-level converter topologies stand out and can be appropriately used to fulfill these requirements. On the other hand, one of the main features of the DFIG-based WECS is its high sensitivity to electrical disturbances, especially voltage sags, since the stator is directly connected to the grid, this feature highlights a fundamental challenge which is to provide fault support and in particular voltage sags to DFIG. Considering the relevance of this problem, this thesis presents a proposal focused on the improvement of DFIG supportability to a voltage sag in the grid by means of a recent multilevel converter topology, the Modular Multilevel Converter (MMC). The proposal is based on the use of the specific characteristics of the MMC through the presence of internal impedances, in order to provide significant damping of the DFIG transient components of the rotor and stator fault currents during a voltage sag in the grid, effectively contributing to maintaining the DFIG connection, while keeping the controllability and avoiding the protection activation, without the need of structural modifications or additional control loops. Moreover, the proposed structure using the MMC directly contributes to the improvement of the system performance by reducing the harmonic distortion rate and electromagnetic torque oscillation, reducing the heating and internal losses in DFIG windings and increasing the useful life of the gearbox, in addition to providing greater robustness, efficiency and reliability to the set of back-to-back power converters. Results obtained by means of simulations indicate the MMC effective damping capability of DFIG currents and the increase in supportability during a grid fault.

6
  • WERBET LUIZ ALMEIDA DA SILVA
  • Implementation of the Active Disturbance Rejection Control (ADRC) technique in the Radial Position Control of a Split-winding Bearingless Induction Machine.

  • Líder : ANDRES ORTIZ SALAZAR
  • MIEMBROS DE LA BANCA :
  • ANDRES ORTIZ SALAZAR
  • ANDRE LAURINDO MAITELLI
  • JOSSANA MARIA DE SOUZA FERREIRA
  • JOSE ALVARO DE PAIVA
  • JOSE ANDRES SANTISTEBAN LARREA
  • Data: 19-may-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • This work aims to realize the rotor radial position control of a three-phase bearingless induction machine, with split-winding and optimized drive structure, using the Active Disturbance Rejection Control (ADRC) technique.  The development of vector control techniques, along with microcontrollers and power electronics components have increased the use of induction motors and turned it the most used machine topology in the world nowadays. Researches point that mechanical bearings are main cause of faults in electrical machines. Aiming to reduce maintenance, noise and vibration, many researches have been realized to develop and optimize bearingless machines structures. This special kind of machine has an operation strongly dependent on closed-loop control systems. As a system with multi-variable, non-linear, time varying and with coupled variables, it demands advanced control strategies to an efficient operation with a good dynamic performance. The ADRC controller considers the total disturbance (composed by non-modeled dynamics, non-linearity, uncertainty and changes in load) as a new system state, to be estimated in real-time throughout an extended state observer. On this way, these disturbances can be compensated in real-time, eliminating regime errors and with a suitable response to general disturbances. This technique will be applied as a position controller and the results will be compared with previous works for system operation with and without load applied to the machine shaft.

7
  • MARIANNE BATISTA DINIZ DA SILVA
  • A data stream-driven methodology for driver behavior modeling.

  • Líder : IVANOVITCH MEDEIROS DANTAS DA SILVA
  • MIEMBROS DE LA BANCA :
  • MAX MAURO DIAS SANTOS
  • EDUARDO ALMEIDA SOARES
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • JUAN MOISES MAURICIO VILLANUEVA
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • Data: 26-may-2022


  • Resumen Espectáculo
  • The Internet of Things (IoT) is a growing network of objects — sensors, devices, systems, and more — that capture, transfer data, and communicate with each other through communication protocols. In addition, such objects have the ability to produce potentially unlimited sequences of data, called data streams. In this sense, it is clear that the IoT is creating new opportunities for various sectors, as can be seen in an unconventional way in the automotive sector. It is known that as a consequence of the advancement of architecture, vehicles are becoming increasingly equipped with various sensors and computational power. And, from available interfaces, it becomes possible to capture and extract, in an automated way, information through sensors and communication protocols present in vehicles and enabling a scenario known as the Internet of Intelligent Vehicles (IoIV). One of the benefits of IoIV is the creation of diagnostic applications, such as characterizing the behavior of drivers. This type of diagnosis is an essential requirement since the way you drive can impact different contexts, such as traffic safety, fuel consumption, emissions, and maintenance, among others. Furthermore, solutions generally available in the literature for analyzing drivers’ behavior focus on supervised offline learning models, fed with the entire dataset for training and testing. On the other hand, such solutions do not handle data streams suitable for online learning, that is, without knowledge of subsequent data. In face of this reality, the objective of this work is to identify patterns in the behavior of drivers, from a methodology oriented to data streams and unsupervised online algorithms. The methodology is adaptable and flexible, and considers the historical-temporal relationship between the samples, adapting in an autonomous and evolutionary way, without the need for a supervised training phase. In order to validate the proposed methodology, a case study was carried out in a real scenario with different conditions, which allowed the identification of daily driving operations. The results indicated the feasibility of the proposal regarding the identification of event detection and indicators of driver behavior. Therefore, the methodology can contribute to several applications, such as industry 4.0 — customized maintenance, fault detection — smart cities and urban mobility — improvement of pavements, increase in the number of speed reducers and crosswalks, decrease
    in the maximum speed of roads, among others.

8
  • MANOEL DO BONFIM LINS DE AQUINO
  • Circular Correntropy: Definition, properties and applications

  • Líder : ALLAN DE MEDEIROS MARTINS
  • MIEMBROS DE LA BANCA :
  • ALLAN DE MEDEIROS MARTINS
  • ALUISIO IGOR REGO FONTES
  • GUILHERME DE ALENCAR BARRETO
  • JOILSON BATISTA DE ALMEIDA REGO
  • JOÃO PAULO FERREIRA GUIMARÃES
  • Data: 30-jun-2022


  • Resumen Espectáculo
  • Circular statistics has been applied to several areas of knowledge in which the input data is circular. Noisy measurements are still a problem in circular data applications and, like non-circular data, second-order statistics have some limitations to deal with non- Gaussian noise. Recently, a similarity function called correntropy has been successfully employed in applications involving impulsive noise for being capable of extracting more information than second-order methods. However, correntropy has not been studied from the perspective of circular data so far. This thesis defines a novel statistical measure called circular correntropy (CC). It uses the von Mises density function in order to redefine correntropy in this domain. In particular, it is proved analytically that the CC contains information regarding second-order and higher-order moments, being a generalization of the circular correlation measure. Its properties are studied as well as a new recursive solution for the Maximum Circular Correntropy Criterion (MCCC). The performance of this new similarity measure is evaluated as a cost function in nonlinear regression and time series prediction problems, where signals are contaminated with additive impulsive noise. Simulations demonstrate that CC is more robust than second-order circular statistics in impulsive noise environments.

9
  • MARCELLA ANDRADE DA ROCHA
  • Text Mining Applied to Analysis of Public Health Policy Interventions: the case of the Syphilis Epidemic in Brazil


  • Líder : RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • MIEMBROS DE LA BANCA :
  • RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • CIPRIANO MAIA DE VASCONCELOS
  • DANIELE MONTENEGRO DA SILVA BARROS
  • MARQUIONY MARQUES DOS SANTOS
  • ANGELICA ESPINOSA BARBOSA MIRANDA
  • THAISA GOIS FARIAS DE MOURA SANTOS LIMA
  • ANTONIO HIGOR FREIRE DE MORAIS
  • Data: 15-jul-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Syphilis is a chronic infectious disease that remains a serious public health problem in much of the world. In Brazil, the rapid response to syphilis project, No Syphilis!, was created in 2018. This project has several strategies to combat syphilis, including the creation of a group of field researchers who worked in priority municipalities and produced thousands of text reports and added them to the platform. The objective of the work is to use the textual productions of the platform of field researchers of the No Syphilis project, LUES, for Text Mining to understand the impact of syphilis in the territory. The texts will be analyzed using some Data Mining (DM) algorithms. In addition, the answers from questionnaires from the "No Syphilis!" project will be used. with the purpose of finding relationships between the productions carried out and the questionnaires in the contribution of the reduction of syphilis. Associations of the data extracted from the questionnaires and reports with indicators of syphilis and its impact of the epidemic in the territory are tested. With this, it is possible to understand the most effective processes for reducing syphilis in the territories from the support strategy and the LUES platform.


10
  • JOSÉ JAIME GUIMARÃES PEIXOTO NETO
  • Frequency Selective Surfaces of the Type Absorb/Transmit Broadband

  • Líder : ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • MIEMBROS DE LA BANCA :
  • ALFREDO GOMES NETO
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • JOSE PATROCINIO DA SILVA
  • MARCIO EDUARDO DA COSTA RODRIGUES
  • MAURICIO WEBER BENJÓ DA SILVA
  • Data: 20-jul-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Frequency Selective Surfaces (FSS) are increasingly being used in telecommunications systems due to the numerous advantages presented by this type of structure, among them the low cost, ease of fabrication and low profile. As well as frequency selective surfaces, electromagnetic wave absorbers have also been widely used, mainly with the objective of solving the problem of multiple paths. Absorbers are structures that aim to absorb electromagnetic waves in a certain frequency range, while allowing the passage of these waves outside it. Thus, this work proposes to investigate the use of FSS, for the design of multilayer absorbers, to obtain broadband absorption. The study consists of the application of resistive-type FSS combined with conductive-type FSS, which use square turns as unit cell geometry, to design an absorb/transmit structure with broadband-type frequency response. The proposed structure should operate in the frequency range between 2 GHz and 6 GHz. An extensive parametric analysis was performed to optimize the desired response. At the end of this analysis, two resistive and two conductive FSS were designed and cascaded to obtain the desired bandwidth. The analyzes showed that the broadband absorber has polarization independence and angular stability. The results obtained show that the absorber presents angular stability up to 30◦. Furthermore, experimental results show that the proposed structure can absorb signals in a frequency range from 2.4 to 6.13 GHz. Absorption above 80% occurs over the entire proposed frequency range. The structure can absorb signals for the entire ISM band (2.4 – 2.4835 GHz), 5G at 3.5 GHz and UNII (5 – 6 GHz), without blocking the other frequencies, avoiding multiple paths in the vicinity where the absorber it's installed.

11
  • VICTOR CARVALHO GALVÃO DE FREITAS
  • Velocity prediction of a pipeline inspection gauge (PIG) based on differential pressure and acceleration with artificial neural networks

  • Líder : ANDRES ORTIZ SALAZAR
  • MIEMBROS DE LA BANCA :
  • ANDRES ORTIZ SALAZAR
  • ADRIAO DUARTE DORIA NETO
  • ANDRE LAURINDO MAITELLI
  • GUSTAVO FERNANDES DE LIMA
  • JUAN MOISES MAURICIO VILLANUEVA
  • Data: 28-jul-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • In the oil and gas industry, a device known as the Pipeline Inspection Gauge (PIG) runs through oil and gas pipelines performing various maintenance operations. The efficiency of these operations can be increased by employing a closed loop speed control system. To get the speed, it is usually resorted to the use of odometers. Although such a method is relatively simple, it can cause certain measurement problems resulting from slippage between the odometer wheel and the duct. In order to contribute to the solution of these problems, the objective of this work is to develop a soft sensor (virtual sensor) to measure the velocity of PIGs from the pressure difference to which the device is submitted in the duct. A soft sensor is basically made up of two elements: a mathematical model of the system and sensors that measure the physical variables required by the model. To obtain the model, it is intended to use artificial neural networks. This model will be shipped in a Raspberry Pi to be installed in the PIG, which will also be responsible for obtaining the sensor data. The SIP testing pipeline from the Petroleum Assessment and Measurement Laboratory (LAMP / UFRN) will be used to evaluate the results. The proposed system is expected to be able to complement the use of odometers, thereby increasing the reliability of speed measurement.

12
  • JUAN RAFAEL FILGUEIRA GUERRA
  • Monitoring Relative Humidity in Concrete Blocks Using Frequency Selective Surface Sensor

  • Líder : ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • MIEMBROS DE LA BANCA :
  • ALFREDO GOMES NETO
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • FRED SIZENANDO ROSSITER PINHEIRO
  • GUTEMBERGUE SOARES DA SILVA
  • HUMBERTO DIONISIO DE ANDRADE
  • Data: 28-jul-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Due to the development of civil construction, there is a greater need for methods to indicate failures and damages of various kinds that may compromise the functionality of buildings. Within this aspect, the concept of Structural Health Monitoring (SHM) is used to define the use of technologies in order to establish assessments that enable the diagnosis of the conditions of a structure. There are technologies available for this purpose, such as optical, resistive and mechanical wave based sensors, however, they present difficulties regarding complexity and cost. Thus, the main objective of this work is to present a preliminary study of the use of Frequency Selective Surface (FSS) in the development of a transmission coefficient - based relative humidity sensor system, applied to concrete structures. Thus enabling a measurement in free space (noninvasive) and with less complexity than current methods. For this, it is necessary to understand the behavior of the electrical permittivity of the medium and the relationship between this and the resonant frequency of the FSS. Therefore, the electrical allowances of various materials used in civil construction were analyzed. Next, two distinct FSSs were designed and their frequency responses measured for various relative humidity levels of commercial concrete blocks. The results are in good agreement with the behavior discussed in the literature regarding central frequency displacement, and it is verified that it is possible to determine a direct correlation between the frequency variation and the relative humidity level for the chosen study object.

13
  • ALYSSON NASCIMENTO DE LUCENA
  • Micro-Brosh: Unmanned Micro-Aerial Vehicle with Low Operation Risk

  • Líder : LUIZ MARCOS GARCIA GONCALVES
  • MIEMBROS DE LA BANCA :
  • ANDOUGLAS GONÇALVES DA SILVA JÚNIOR
  • BRUNO MARQUES FERREIRA DA SILVA
  • LUIZ MARCOS GARCIA GONCALVES
  • PABLO JAVIER ALSINA
  • PAULO FERNANDO FERREIRA ROSA
  • RAIMUNDO CARLOS SILVERIO FREIRE JUNIOR
  • Data: 04-ago-2022
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The risk involving the use of Unmanned Aerial Vehicles (UAV) has grown directly with its popularization. It is a global concern of regulatory agencies to reduce these risks. In Brazil the National Civil Aviation Agency (ANAC) waives most requirements for UAVs up to 0.250 kg. In this perspective, the present study aims to propose the design and development of a micro UAV flying wing model with remote or autonomous piloting capability with low risk of collision with people and facilities (Micro-Brosh). With a maximum wingspan of 0.30 m and a final weight of 0.160 Kg, the Micro-Brosh UAV's main feature is its easy and safe operation, being able to be launched manually and with the realization of landing on grass cover, without causing damage to the aircraft. We introduce the proposed architectural design, with aerodynamic and performance analyses, demonstrating parameters such as lift, thrust, drag, stall speed, and flight time. In our tests, the prototype proved to be superior to almost all other UAVs in the literature studied, except for one that has similar size and flight time characteristics, but with greater weight and size than the Micro-Brosh. In all, 6 prototypes were built and analyzed in more detail, resulting in the final model that was validated through flight tests and bench tests. In addition to the safety advantages, another important aspect is that the costs associated with the construction and maintenance of the Micro-Brosh is below US$ 500, a value considered low in relation to the traditional UAVs surveyed.

     

    O risco envolvendo o uso de Veículos Aéreos não Tripulados (UAV) tem crescido diretamente com a sua popularização. É uma preocupação mundial das agências reguladoras diminuir esses riscos. No Brasil a Agência Nacional de Aviação Civil (ANAC) dispensa a maioria das exigências para UAVs com até 0,250 Kg. Nesta perspectiva, o presente estudo tem como objetivo propor o projeto e desenvolvimento de um micro UAV modelo asa voadora com capacidade de pilotagem remota ou autônoma de baixo risco de colisão às pessoas e instalações (Micro-Brosh). Com uma envergadura máxima de 0,30 m e peso final de 0,160 Kg, o UAV Micro-Brosh tem como principal característica a operação fácil e segura, podendo ser lançado manualmente e com a realização do pouso em capo gramado, sem causar danos à aeronave. Introduzimos o projeto arquitetônico proposto, com análises aerodinâmica e de desempenho, demonstrando parâmetros como sustentação, empuxo, arrasto, velocidade de estol e tempo de voo. Em nossos testes, o protótipo demonstrou ser superior a quase todos os outros UAV da literatura estudados, menos um que possui características de tamanho e tempo de voo similars, porém com peso e tamanho maiores que o Micro-Brosh. Ao todo, foram construídos e analisados mais detalhadamente 6 protótipos, resultando no modelo final que foi validado através de testes de voo e testes de bancada. Além das vantagens de segurança, outro aspecto importante é que os custos associados com a construção e manutenção do Micro-Brosh fica abaixo de \$500, um valor considerando baixo em relação aos UAVs tradicionais pesquisados.

     
14
  • THALES QUEIROZ FONSECA
  • Virtual Impedance-Based Control Strategies for Voltage Supporting in LCL-VSC Converters

  • Líder : RICARDO LUCIO DE ARAUJO RIBEIRO
  • MIEMBROS DE LA BANCA :
  • RICARDO LUCIO DE ARAUJO RIBEIRO
  • FLAVIO BEZERRA COSTA
  • ANTONIO MARCUS NOGUEIRA LIMA
  • JOSEP MARIA GUERRERO ZAPATA
  • MAURICIO AREDES
  • Data: 05-ago-2022


  • Resumen Espectáculo
  • The advances in power electronics and environment restrictions have driven the insertion, as distributed generators (DG), of renewable energy based systems on the power grid, the main one being those who use the photovoltaic and wind systems as primary energy sources. The interconnection of these sources with the power grid is generally obtainedbyVSCconvertersandLCorLCLfilters,beingcalledasLC-VSCorLCL-VSC systems. Such systems are highly variable in power generation, which may cause voltage fluctuation and frequency deviation in the electric grid. LC-VSC systems are controlled to operate as a controlled voltage source connected to the grid and they can regulate the amplitude and frequency of the voltage of the point of common coupling (PCC). LCLVSC are controlled to operate as controlled current source connected to the grid and they are still the most common system employed currently. Therefore, in this doctoral qualification, are proposed virtual impedance-based control strategies applied in LCL-VSC systemsforPCCvoltageregulation. Thevirtualimpedanceisusedforshapingtheequivalent DG system impedance connected to the grid and regulate the power flow of these systems for achieving the required PCC voltage regulatio

15
  • ANAMARIA SENA MAIA
  • Design of dual band microwave absorbers using frequency selective surfaces

  • Líder : ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • MIEMBROS DE LA BANCA :
  • ALFREDO GOMES NETO
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • JOSE PATROCINIO DA SILVA
  • LAERCIO MARTINS DE MENDONCA
  • MAURICIO WEBER BENJÓ DA SILVA
  • Data: 08-nov-2022


  • Resumen Espectáculo
  • Currently designs developed in wireless network systems are on signal confinement. We observe the need to increase security and restriction of services, in which electromagnetic signals must remain only in a specific room, without external transmittance. Consequently, several researches are studying the use of Frequency Selective Surfaces (FSS) as a way to reduction electromagnetic interference (EMI) and to maintain the systems' compliance. With characteristic and distinct performance of the conventional passive absorbers of Salisbury and Jaumann, that present good absorption in the band of stop, but, low transmission in the external bands. The absorbers with FSS propose to act with transparency to the electromagnetic waves of the incident plane in the passing band, while absorbing the waves in the wanted frequency bands. Therefore, the objective of this work is to design a microwave absorber structure integrated with frequency selective surfaces with multiband performance. Studies described in the literature are analyzed to obtain such desired absorption characteristics, later structures are projected with variation of their respective parameters. Demonstrating mastery in the use of the commercial software used in the simulations, and understanding the effects of the parametric variations of the devices examined.

16
  • VERIVAN SANTOS LIMA
  • Thermal profiles in water injection wells: reduction of the systematic error in the flow measurement during the
    transitional regime.

  • Líder : ANDRES ORTIZ SALAZAR
  • MIEMBROS DE LA BANCA :
  • ANDRE LAURINDO MAITELLI
  • ANDRES ORTIZ SALAZAR
  • CARLA WILZA SOUZA DE PAULA MAITELLI
  • DIEGO ANTONIO DE MOURA FONSECA
  • GUSTAVO FERNANDES DE LIMA
  • WERBET LUIZ ALMEIDA DA SILVA
  • Data: 12-dic-2022


  • Resumen Espectáculo
  • This work is a contribution to flow measurement techniques in water injection wells in oil
    production fields, focusing on the first moments of the operation. The technique developed by
    Ramey (RAMEY JR et al., 1962) was chosen, which, although it is aimed at calculating the
    temperature in the injection fluid, it has been adapted to calculate the flow. In this technique,
    the calculation is based on the relationship between the heat flow established in the well and the
    temperature difference between the injection fluid and the geothermal temperature, naturally
    established in the reservoir. Due to the mathematical complexity involved in the heat and mass
    transfer mechanisms, many simplifications were adopted in the development of the theory, limiting its application in the first moments of the operating cycle. In the considered times, the
    neglect of heat transfer in completion constitutes the main source of systematic error inherent to
    Ramey’s methodology. With lesser intensity, but still significant, the failure to observe the temporal variation of the injection fluid also results in a systematic error that needs to be addressed.
    The reduction of the systematic errors listed is the main product of this thesis.


17
  • RUANN VÍCTOR DE ANDRADE LIRA
  • Elliptical UHF Sensor for Partial Discharge Detection

  • Líder : ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • MIEMBROS DE LA BANCA :
  • ALEXANDRE JEAN RENE SERRES
  • ALFREDO GOMES NETO
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • LAERCIO MARTINS DE MENDONCA
  • VALDEMIR PRAXEDES DA SILVA NETO
  • Data: 15-dic-2022


  • Resumen Espectáculo
  • In this work, simplified geometries for microstrip antennas were evaluated and optimized, to select one appropriate geometry, aiming for its application of partial discharge (PD) detection, as a UHF sensor. To improve the performance of the proposed antenna, cascaded frequency selective surfaces (FSS), using double square loops as a unit cell element, were designed and evaluated. Antenna reflection coefficient measurements without and with the integrated FSS were performed and compared with simulated results. The results show that the insertion of the FSSs did not impact the operational bandwidth of the antenna, which comprises 100$\%$ of the characteristic frequency range, of the PD activity (300 – 1500 MHz). The insertion of the FSSs resulted in an average gain of 4 dBi, concerning the isolated antenna, increasing the sensitivity for PD detection, within the proposed frequency range. We evaluated the PD detection of the structure through measurements in three devices, a bar of a hydro generator, an oil tank with electrodes in the flat-tip configuration, and a potential transformer (PT). The proposed UHF sensor for PD detection could detect PD activity in all three scenarios and for levels lower than 10 mV. 

18
  • EVANDSON CLAUDE SEABRA DANTAS
  • ELECTRICAL EQUIVALENCE PYRANOMETER WITH AMBIENT TEMPERATURE ANALOG COMPENSATION

  • Líder : SEBASTIAN YURI CAVALCANTI CATUNDA
  • MIEMBROS DE LA BANCA :
  • ANTONIO WALLACE ANTUNES SOARES
  • DIOMADSON RODRIGUES BELFORT
  • FERNANDO RANGEL DE SOUSA
  • RAIMUNDO CARLOS SILVÉRIO FREIRE
  • SEBASTIAN YURI CAVALCANTI CATUNDA
  • Data: 20-dic-2022


  • Resumen Espectáculo
  • Pyranometers are devices used to measure incident radiation per unit area. This type of device is found in many applications in the field of solar energy, UV treatment, atmospheric research, etc. Some of these devices are based on the principle of electrical equivalence, in which radiant power is equated with electrical power in a thermoelectric balance. The balance of this balance usually occurs by compensating the radiant power by an electrical power to maintain a resistance thermometer at constant temperature. This configuration, however, has limitations when the ambient temperature varies, varying parameters such as sensitivity, useful output voltage, power consumption, among others. In this work, it is proposed to replace the configuration that keeps the sensor temperature constant for the configuration of constant temperature difference. The proposal is formulated by modifying the architecture of the feedback Wheatstone bridge so that it works with a constant temperature difference. The implemented
    proposal is then validated through computer simulations in a SPICE environment and through field experiments. In this work, a prototype was developed that was tested in 3 different climatic conditions, namely: clear weather, partially cloudy weather and cloudy wheater. Results of this work point to an improvement in the useful voltage range by 5 times, a reduction in the influence of ambient temperature variation on the output voltage, a reduction in electrical consumption and an increase in the sensitivity to incident radiation.

2021
Disertaciones
1
  • MAILSON RIBEIRO SANTOS
  • A Methodology Based on Evolving Systems for Fault Detection and Identification of Dynamic Systems

  • Líder : LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • MIEMBROS DE LA BANCA :
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • CLAUBER GOMES BEZERRA
  • Data: 07-ene-2021


  • Resumen Espectáculo
  • This work proposes a methodology for the detection and identification of failures in dynamic systems, through an online and evolutionary approach. The proposal is divided into three stages, in which data pre-processing and post-processing are carried out to increase the robustness of the methodology in the presence of outliers and noise, in the pre-processing the selection of characteristics, normalization is carried out of data, filtering and adding regressors, in post-processing time filtering is performed. In the processing stage, an adaptive and unsupervised approach is applied, through the Auto-Cloud algorithm, which performs grouping and classification of data streams. To validate this proposal, different evaluation metrics were used, such as Adjusted Rand Index (ARI), homogeneity, completeness, precision, f1_score, recall, and satisfactory results were obtained. Finally, the conclusion of this work is presented, in addition to proposals for future work.

2
  • VINÍCIUS SOUSA DE OLIVEIRA
  • Study and application of cryptographic algorithms for wireless sensor networks in a software-defined radio environment

  • Líder : ANDRES ORTIZ SALAZAR
  • MIEMBROS DE LA BANCA :
  • ANDRES ORTIZ SALAZAR
  • DIEGO ANTONIO DE MOURA FONSECA
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • RODRIGO SOARES SEMENTE
  • Data: 29-ene-2021


  • Resumen Espectáculo
  • In this work is proposed a comparative study of cryptographic algorithms to be applied in Wireless Sensor Networks. Having as motivation the lack of researches with practical results, the main goal of this work is to verify the viability of the insertion of studied algorithms in a real environment based on a Software-Defined Radio system. Due to the internal limitations of each node, efficiency is an aspect as important as cryptographic quality, in order to ensure a good performance of the security technique. For this, it will be analyzed data such as execution time, memory usage, processing cost and power consumption. Thereby, it is expected to increase the security in Wireless Sensor Networks, having in consideration its performance, memory and energy constraints.

3
  • FELIPE FERNANDES LOPES
  • Fully parallel implementation of an SVM with SGD-based training on FPGA

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • MARCELO AUGUSTO COSTA FERNANDES
  • ORIVALDO VIEIRA DE SANTANA JUNIOR
  • VICENTE IDALBERTO BECERRA SABLON
  • Data: 29-ene-2021


  • Resumen Espectáculo
  • Artificial intelligence, machine learning, and deep learning have proven to be powerful techniques for solving natural language processing problems, computer vision, and others. However, its computational and statistical performance depends on other factors of the algorithms used for training, the computing platform used, and even the numerical precision used to represent the data. One of the main machine learning algorithms is the Support Vector Machine (SVM), most commonly trained through minimal sequential optimization, contributing to low computational performance. As an alternative, algorithms based on Stochastic Gradient Descent (SGD) have better scalability and can be used as good options for training machine learning algorithms. However, even with algorithms based on the stochastic gradient, the training and inference times can become long depending on the computing platform employed. Thus, accelerators based on Field Programmable Gate Arrays (FPGAs) can be used to improve performance in terms of processing speed. This work proposes a fully parallel FPGA implementation of an SVM with SGD-based training. Results associated with hardware occupation, processing speed (or throughput), and accuracy for training and inference in various quantization formats are presented.

4
  • ANDRESSA STÉFANY SILVA DE OLIVEIRA
  • Macro SOStream: An evolving algorithm to self organizing density-based clustering

  • Líder : LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • MIEMBROS DE LA BANCA :
  • DANIEL FURTADO LEITE
  • DANIEL ALOISE
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • MARCELO AUGUSTO COSTA FERNANDES
  • Data: 19-mar-2021


  • Resumen Espectáculo
  • Situations that generate a continuous data stream, such as TCP / IP traffic, e-commerce, and industrial monitoring, can make the usability of algorithms that have machine learning completely offline unviable. That is due to the need for data storage, the infinite growth of data generation, and limited memory restrictions. With that, the algorithms that have the learning totally or partially on-line appeared. Among them, there are the evolving algorithms, which have been of interest because they can develop and update in unknown environments and detect concepts drift and evolution in the input data over time. Because of these algorithms' broad applicability in real problems, this work proposes a new evolving algorithm named Macro SOStream. This algorithm has on-line learning and is based on self-organizing density for data stream clustering. The Macro SOStream is based on the SOStream algorithm, but we incorporated macroclusters composed of the microclusters. While microclusters have spherical shapes, macroclusters can have arbitrary shapes. Besides, the Macro SOStream's performance was compared to SOStream and DenStream algorithms' performance using the datasets and the ARI performance metric to validate our proposal.
5
  • VÍTOR SARAIVA RAMOS
  • Real-Time Highlight Removal from a Single Image

  • Líder : LUIZ FELIPE DE QUEIROZ SILVEIRA
  • MIEMBROS DE LA BANCA :
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • LUIZ GONZAGA DE QUEIROZ SILVEIRA JUNIOR
  • RAFAEL BESERRA GOMES
  • FRANCISCO MADEIRO BERNARDINO JUNIOR
  • Data: 31-mar-2021


  • Resumen Espectáculo
  • The problem of highlight removal from image data refers to an open problem in computer vision concerning the estimation of specular reflection components and the removal thereof. In recent applications, highlight removal methods have been employed for the reproduction of specular highlights on high dynamic range (HDR) displays; to increase glossiness of images in specular reflection control technologies; to improve image quality in display systems such as TVs; and to enhance the dynamic range of low dynamic range (LDR) images. However, the underlying processing required by state-of-the-art methods is computationally expensive and does not meet real-time operational requirements in image processing pipelines found in consumer electronics applications. In addition, these applications may require that methods work with a single frame in imaging or video streams. Consequently, this work has the objective of proposing a novel method for the real-time removal of specular highlights from a single image. The essence of the proposed method consists in matching the histogram of the luminance component of a pseudo-specular-free representation using as reference the luminance component of the input image. The operations performed by the proposed method have, at most, linear time complexity. In experimental evaluations, the proposed method is capable of matching or improving upon state-of-the-art results on the task of diffuse reflection component estimation from a single image, while being 5X faster than the method with the best computational time and 1500X faster than the method with the best results. The proposed method has high industrial applicability, and targeted use cases can take advantage of contributions of this work by incorporating the proposed method as a building block in image processing pipelines.

6
  • VICTOR RAMON FIRMO MOREIRA
  • Intelligent control of an omnidirectional mobile robot with reinforcement learning for decision making

  • Líder : WALLACE MOREIRA BESSA
  • MIEMBROS DE LA BANCA :
  • WALLACE MOREIRA BESSA
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • ORIVALDO VIEIRA DE SANTANA JUNIOR
  • ESTHER LUNA COLOMBINI
  • Data: 12-abr-2021


  • Resumen Espectáculo
  • The evolution of robotic systems has become evident over time. Due to the advances in mechanical manufacturing and the new algorithms used, mobile robots have become increasingly independent in their actions. Regarding machine learning strategies, special attention is given to reinforcement learning algorithms, because of its similarities with the biological learning process. This work proposes the development of an autonomous agent, combining intelligent control strategies with decision-making algorithms. For the implementation of the proposed strategy, the Robotino omnidirectional mobile robot will be used. Simulations of the robot's performance were performed to explore space in an environment, for which a specific mathematical model is applied. For system control, the Linearization by Feedback strategy was combined with a compensator based on Artificial Neural Networks to deal with uncertainties and possible external disturbances. The epsilon greedy algorithm, in turn, was chosen to enable the robot in the decision-making process. The results show that the intelligent control strategy was efficient and the proposed intelligent agent was able to explore the environment effectively, obtaining a high average reward. The perspective is that the strategy is still implemented experimentally in Robotino.

7
  • ERIKA AKEMI YANAGUIBASHI ALBUQUERQUE
  • EDUCATIONAL ROBOTICS ON THE PREVENTION OF MENTAL DISORDERS IN THE SCHOOL ENVIRONMENT

  • Líder : LUIZ MARCOS GARCIA GONCALVES
  • MIEMBROS DE LA BANCA :
  • ANDERSON ABNER DE SANTANA SOUZA
  • AQUILES MEDEIROS FILGUEIRA BURLAMAQUI
  • LUIZ MARCOS GARCIA GONCALVES
  • Data: 26-abr-2021


  • Resumen Espectáculo
  • In this work, we propose to study and verify the potential of using Educational Robotics in the classroom to improve cognitive functions and at the same time act in the prevention of the occurrence of mental disorders, a problem that has hindered the socio-cultural development of many young people today. It is worth mentioning that there are studies verifying the effectiveness of Educational Robotics with children and adolescents in order to assist in the diagnosis and treatment of the autistic spectrum and in visual impairment, but we do not find in the literature a study similar to the one proposed in this work. Furthermore, it is already demonstrated in the literature that Educational Robotics improves the child's cognitive functions and also their and teachers' self-esteem. During the work, from the analysis of collected data, we were able to verify strong indications that children and adolescents who take educational robotics classes are less susceptible to major problems of depression, stress, among other types of mental disorders. To reach this inference, we developed and applied questionnaires for teachers and students, from which variables were determined to be analyzed, such as psychomotor, cognitive and affective. The questionnaires were applied in the practical phase of the Brazilian Robotics Olympiad that took place at the Federal Institute of Rio Grande do Norte, Natal, Brazil. These data were submitted to an analysis and later verification through qualitative and quantitative methodologies on the educational robotics tool used.

     
8
  • RODRIGO DE ANDRADE TEIXEIRA
  • Power Quality Analysis Applied to a Three-Phase PWM Rectifier System Using the One-Cycle Control Technique.

  • Líder : ANDRES ORTIZ SALAZAR
  • MIEMBROS DE LA BANCA :
  • ANDRES ORTIZ SALAZAR
  • ELMER ROLANDO LLANOS VILLARREAL
  • JOÃO TEIXEIRA DE CARVALHO NETO
  • RICARDO FERREIRA PINHEIRO
  • Data: 27-abr-2021


  • Resumen Espectáculo
  • This work proposes the use of the one-cycle control technique to control a three-phase rectifier based on boost topology, without the use of voltage sensors at the coupling point of the system. Furthermore, unlike most of the works developed with the OCC technique, a digital implementation using Texas Instruments' DSP TMS 320F28335 was used. The objective of the proposed implementation is to obtain a sinusoidal current at the rectifier input with a power factor as close as possible to unity, while the rectifier output is controlled at a fixed voltage value. To evaluate the proposed system, simulations were performed in Matlab and PSIM software and the practical results were verified using a laboratory prototype.

9
  • PAULO VICTOR QUEIROZ CORREIA
  • DETECTION AND CLASSIFICATION PERFORMANCE ANALYSIS OF FAILURES IN INDUSTRIAL PROCESSES USING LSTM NEURAL NETWORKS WITH DATA COMPRESSION TECHNIQUES

     
  • Líder : LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • MIEMBROS DE LA BANCA :
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • MARCELO AUGUSTO COSTA FERNANDES
  • CELSO JOSÉ MUNARO
  • Data: 30-jun-2021


  • Resumen Espectáculo
  • Industry 4.0 set a paradigm shift in industrial process monitoring and control. It installed several sensors in different parts of the plant, connecting these industrial processes with the Internet of Things and Cloud Computing. Although, the data generation growth demanded engineers build proper environments to process and store the absurd amount of data. This growth caused an increasing energy consumption, computational complexity and environmental degradation. Therefore to address these demands, this dissertation proposes efficient approaches to perform Fault Detection and Identification in industrial processes. The first approach consists of using Symbolic Aggregate Approximation (SAX) to compress process variables to reduce the load on data warehouses. Then, we train a Long Short-Term Memory (LSTM) neural network with those compressed inputs to perform fault detection. Finally, the second approach addresses efficient edge computing systems, performing LSTM neural network compression with pruning technique. The compression reduces the memory usage and number of operations of these networks, saving energy and accelerating inference speed in edge computation. To assess the performance of both approaches, we use the Tennessee Eastman Process (TEP) as the benchmark with classification metrics of accuracy, precision, recall and F1-Score. We are also going to analyze the compression efficiency of both approaches, studying their viability and parameter reduction in LSTM networks.

     
10
  • MAXWEL DA SILVA SANTOS
  • Improved Margin Voltage Control Strategy for Multi-terminal High Voltage Direct Current Systems based on the Modular Multilevel Converter for Robustness Towards Disturbances

  • Líder : LUCIANO SALES BARROS
  • MIEMBROS DE LA BANCA :
  • LUCIANO SALES BARROS
  • CARLOS EDUARDO TRABUCO DOREA
  • FLAVIO BEZERRA COSTA
  • RODRIGO ANDRADE RAMOS
  • Data: 02-jul-2021


  • Resumen Espectáculo
  • In modern systems, multi-terminal high-voltage Direct Current (MTDC) transmission technology has been considered the key technology for long-distant bulk electric power transmission, asynchronous grid interconnections, and offshore wind energy converter systems (WECS). The main purpose of an MTDC grid is to share power among the AC grids. However, this technology presents challenges in its operation and control. Therefore, for its control is necessary the use of appropriate control system strategies, such as master-slave, margin voltage, voltage droop, and their combinations. In this work, it is proposed a modification in the DC voltage control loop of the margin voltage control strategy aiming to avoid degradation of current control and to increase the supportability to disturbances in the power grid. This modification is based on closed-loop control of the power and feed-forward compensation of the DC grid power. The performance of the proposed method was assessed in a four-terminal meshed MTDC system topology based on the multilevel modular converter (MMC) under abnormal operating conditions and a comparison with the margin voltage and voltage droop control strategies was accomplished. The operation scenarios consider challenging and common failures for MTDC network control systems such as power flow variation in a converter station, open-circuit of a DC transmission line, and failure of a converter station. The proposed control presented the best performance, obtaining a new balance of power flow with less power and voltage oscillations.

11
  • ISMAEL ALVES DE AZEVEDO
  • Comparison of Control Strategies for Squirrel-Cage Induction Generator-Based Wind Energy Conversion Systems

  • Líder : LUCIANO SALES BARROS
  • MIEMBROS DE LA BANCA :
  • LUCIANO SALES BARROS
  • ANDRES ORTIZ SALAZAR
  • DANIEL BARBOSA
  • Data: 05-jul-2021


  • Resumen Espectáculo
  • The squirrel cage induction generator is a robust and low cost alternative to variable speed wind energy conversion systems. For this type of generator, direct torque control, direct stator-field-oriented control, direct rotor-field-oriented control, stator-flux slip control and rotor-flux slip control can be used for maximum power point tracking. Stator-flux slip control and direct torque control have simple schemes with fast dynamic responses, since they do not have internal current controllers. However, stator-flux slip control presents high inrush-current and a poor dynamic response and direct torque control presents high torque ripples and poor performance at low speeds. On the other hand, rotor-flux slip control, direct stator-field-oriented control and direct rotor-field-oriented control have fast dynamic response, low inrush current and low error at maximum power point tracking, but have current controllers and offer a larger complexity in the control design. The objective of this work is to define the most suitable control strategy by evaluating the following performance indexes simultaneously: maximum power point tracking, total harmonic distortion, inrush current and dynamic response. Simulations are performed in order to analyze the performance of the control strategies. Tests were carried out for the five control strategies and the direct rotor-field-oriented control presented the best trade off considering maximum power point tracking, total harmonic distortion and dynamic performance, despite it does not have the best inrush current.

12
  • DANIEL DE LUCENA FLOR
  • Acoustic Noise Evaluation inside Vehicles under Different Traffic Conditions

  • Líder : VICENTE ANGELO DE SOUSA JUNIOR
  • MIEMBROS DE LA BANCA :
  • EMANUEL BEZERRA RODRIGUES
  • ALLAN DE MEDEIROS MARTINS
  • DANILO DE SANTANA PENA
  • VICENTE ANGELO DE SOUSA JUNIOR
  • Data: 23-jul-2021


  • Resumen Espectáculo
  • This work aims to analyze the interior acoustic noise in a vehicle under several traffic conditions by comparing different statistical models used for evaluating noise impulsiveness. Initially, we present the importance of studying the sound quality for the development of multimedia and control applications and of smart vehicles. Next, we discuss previous works on the acoustic channel inside vehicles, as well as the approaches to statistically model this type of channel. The measurement setup we used and the analyses of the collected data are presented in order to highlight which measurements conditions contribute the most to the interior noise levels and to impulsivity.

13
  • GABRIEL LUCAS ALBUQUERQUE MAIA SIGNORETTI
  • An online evolving algorithm for automatic data compression in IoT scenarios

  • Líder : IVANOVITCH MEDEIROS DANTAS DA SILVA
  • MIEMBROS DE LA BANCA :
  • GUSTAVO BEZERRA PAZ LEITAO
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • JUAN MOISES MAURICIO VILLANUEVA
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • Data: 23-jul-2021


  • Resumen Espectáculo
  • With the advancement and mass adoption of solutions in the fields of Internet of Things (IoT) and connected cities, the number of devices and sensors connected to the network tends to grow exponentially. In this scenario, the transmission and storage of the growing volume of data bring new challenges. When devices transmit potentially irrelevant or redundant data, there is increased energy and processing waste, as well as unnecessary use of the communication channel. Thus, local data compression solutions on the IoT devices themselves become increasingly attractive, enabling the elimination of samples that would have little or no contribution to the application, in order to significantly reduce the volume of data needed to represent the information. However, such devices present on the market today have serious storage and processing power limitations. In order to circumvent these limitations, the TinyML field emerges, which seeks ways to implement machine learning models in low-power devices. Given this context, one of the sectors that can benefit most from these new technologies is the automobile industry, as currently all cars produced must be instrumented with a series of sensors. In this way, by connecting an intelligent device to the vehicle, it is possible to process the data locally and transmit it to a remote server later. In this context, the present work proposes the development of a new online, unsupervised, and automatically adaptable data compression algorithm for IoT applications. The proposed approach is called Tiny Anomaly Compressor (TAC) and is based on data eccentricity and does not require pre-established mathematical models or any assumptions about data distribution. To test the effectiveness of the solution and validate it, two tests were carried out with different objectives. First, a comparative analysis on two real-world datasets was developed with two other algorithms from the literature, the Swing Door Trending (SDT) and the Discrete Cosine Transform (DCT). Finally, the proposal was embedded in an IoT device based on an Arduino and connected to a car to verify the impact of the algorithm on the processing time of the system's primary operations. Preliminary results show that it is possible to achieve significant compression rates without significant impacts on the generated error and system processing.

14
  • ANA THERESA FERNANDES DE OLIVEIRA MANCINI
  • Design of Dynamic Output Feedback Controllers for Linear Systemas Under Constraints

  • Líder : CARLOS EDUARDO TRABUCO DOREA
  • MIEMBROS DE LA BANCA :
  • AMANDA DANIELLE OLIVEIRA DA SILVA DANTAS
  • ANDERSON LUIZ DE OLIVEIRA CAVALCANTI
  • CARLOS EDUARDO TRABUCO DOREA
  • Data: 04-ago-2021


  • Resumen Espectáculo
  • In this work an improvement in the design of output feedback controllers using invariant sets is proposed. Controlled invariant sets have been widely used to solve constrained systems problems. Despite having been well studied in state feedback control, the use of controlled invariant sets for output feedback is still little explored. A state observer is incorporated into the compensator structure in order to obtain a dynamic compensator. The proposed output feedback controlled invariant set is constructed from a conditioned invariant set and a controlled invariant set. The uncertainty of states is reduced using the contraction of the conditioned invariant set. An output feedback control strategy is to minimize the admissible states consistent with the measurements one step ahead. Here we propose the optimization of this strategy by using the result of the linear programming problem as an additional information in the calculation of the next control action in order to accelerate convergence. Results obtained from the optimization strategy using the conditioned invariant set as a target for the optimization of the distance to the origin are also analyzed. First, the theory of invariant sets and its application in state feedback control is presented. Next, the strategies for static and dynamic output feedback are presented without the use of additional information in the calculation of the control action. Finally, the design of output dynamic and static feedback controllers using the optimization strategies with additional information is presented and the results obtained with both strategies are compared.

15
  • MILLENA MICHELY DE MEDEIROS CAMPOS
  • RF Signal Based Classification of Number of People in an Environment: A Machine Learning Approach

  • Líder : VICENTE ANGELO DE SOUSA JUNIOR
  • MIEMBROS DE LA BANCA :
  • EDUARDO RODRIGUES DE LIMA
  • LEONARDO HENRIQUE GONSIOROSKI FURTADO DA SILVA
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • VICENTE ANGELO DE SOUSA JUNIOR
  • ÁLVARO AUGUSTO MACHADO DE MEDEIROS
  • Data: 26-ago-2021


  • Resumen Espectáculo
  • This work proposes a technique for counting people in an already populated environment. Initially, a survey is made of the technologies and solutions designed for this purpose. As a proof of concept, a counting solution is analyzed for a small number of people, applying machine learning to the descriptive statistics of an RF signal. Finally, the classification results are presented for a more realistic scenario, with up to 350 people in the environment, using a software-defined radio measurement system for data collection. The results show significant precision in counting the number of people by classification in groups of individuals.

16
  • RAFFAEL SADITE CORDOVILLE GOMES DE LIMA
  • A parallel software-defined ultra-low-power receiver for a satellite message forwarding system

  • Líder : SAMUEL XAVIER DE SOUZA
  • MIEMBROS DE LA BANCA :
  • SAMUEL XAVIER DE SOUZA
  • KAYO GONCALVES E SILVA
  • TIAGO TAVARES LEITE BARROS
  • HENRIQUE COTA DE FREITAS
  • JOSÉ MARCELO LIMA DUARTE
  • Data: 17-sep-2021


  • Resumen Espectáculo
  • In the beginning, satellite communications faced a major challenge of putting complex on-board processing, due to the high cost, energy consumption, weight, and volume of the equipment. With the advancement of microelectronics, today we have miniaturized devices with low cost and low power consumption. The CubeSat standard is an alternative for replacing large satellites in specific applications, due to its size, there are restrictions on energy consumption. In this paper, we present a parallel implementation of pi/3-Phase Shift Keying multi-user receiver in the GAP8 Parallel Ultra-Low Power processor for low earth orbit (LEO) nanosatellite communication system. The near-threshold operation will guarantee high energy efficiency and parallelism with high-efficiency processing. The system consists of pi/3-PSK signals transmitted by terrestrial platforms, satellite links, noise, and receiver. The noise channel that was used in the system is Additive White Gaussian Noise (AWGN), loss of power due to the satellite communication link is considered. The Channel coding that was used in Manchester. A MatLab implementation of the receiver will be used as a reference model for validating the receiver implementation on GAP8. The receiver validation in GAP8 will be done by comparing its output with that of the reference model for the same input.

17
  • IAN DA SILVA VIGANÓ
  • Fuzzy Interval Theory Applied in a Magnetic Levitation System Control

  • Líder : FABIO MENEGHETTI UGULINO DE ARAUJO
  • MIEMBROS DE LA BANCA :
  • ANDRE LAURINDO MAITELLI
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • MÁRCIA LISSANDRA MACHADO PRADO
  • OSCAR GABRIEL FILHO
  • Data: 18-oct-2021


  • Resumen Espectáculo
  • Magnetic levitation teory attracts a lot of interest from the academic community currently, as it has a great potential in present promising technologies to the market, mainly in the passenger and cargo transportation sector and in the industrial sector, providing a great positive impact for society and contributing to the your well-being. Under this premise this work proposes to analyze the application of type-2 fuzzy control as a better alternative for control design for systems of this class, analyzing the positive characteristics of the application of this control technique in the face of type-1 fuzzy. The controllers were designed and tuned manually, in order to observe if the extra degree of freedom provided by the type-2 fuzzy technique presents a superior performance. The experimental results show that the type-2 fuzzy controller provides a superior performance to its type-1 counterpart in different aspects. With this methodology, the best characteristics of the different classes of control systems can be combined to achieve a more efficient and intelligent control scheme.

18
  • WYSTERLÂNYA KYURY PEREIRA BARROS
  •  Hardware Implementation of the Otsu’s Method Applied to Real-Time Worm Segmentation

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • MARCELO AUGUSTO COSTA FERNANDES
  • RAFAEL BESERRA GOMES
  • AGOSTINHO DE MEDEIROS BRITO JUNIOR
  • MARCO ANTONIO GARCIA DE CARVALHO
  • Data: 17-nov-2021


  • Resumen Espectáculo
  • Behavioral genomic studies employing the worm Caenorhabditis elegans have aided the discovery of new gene-behavioral associations and the screening of new drugs. High-resolution cameras record experiments with this worm, generating videos that computational solutions will later process for automated behavioral analysis. Because of the large volume of data to be processed, these analyses usually have to be performed offline. However, it is desired to develop a high-throughput implementation capable of operating in real-time, seeking to reduce the memory occupation by storing videos and allow the realization of new kinds of experiments. One way to speed up the algorithms employed is through the use of reconfigurable computing. Therefore, this work proposes the hardware development of the Otsu method for worm segmentation in real-time. The proposed implementation was developed in Field Programmable Gate Array (FPGA) using a fully parallel strategy with fixed-point representation. Architecture details are presented, as well as synthesis results related to the hardware area occupation, throughput, and dynamic power consumption. Results about validation of the implementation using images of the worms are also provided. The data show that the proposed architecture can achieve high speedups compared to similar work presented in the literature, besides allowing the segmentation of worms in real-time

19
  • IGOR MACEDO SILVA
  • CEVERO: an open hardware processor for aerospace missions

  • Líder : SAMUEL XAVIER DE SOUZA
  • MIEMBROS DE LA BANCA :
  • FERNANDA GUSMÃO DE LIMA KASTENSMINDT
  • Frank Kagan Gürkaynak
  • SAMUEL XAVIER DE SOUZA
  • TIAGO TAVARES LEITE BARROS
  • Data: 03-dic-2021
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The space environment is a harsh environment for digital circuits. The effects of space radiation on the circuit might result in malfunction and, as a consequence, jeopardize the security and execution of space missions, manned or not. Therefore, there was a need to develop digital circuits which can withstand radioactive particle’s strikes and, nowadays, there are several processor technologies that can indeed work safely under such conditions. However, these technologies generally involve proprietary transistor fabrication steps or proprietary cell libraries which inhibit usage and development. There is, nonetheless, another set of architectural methods which operate on a higher abstraction level and are arguably as reliable as the mentioned techniques for a set of space missions. This work uses these techniques to develop the CEVERO, a general-purpose processor base on the existing PULP Platform, which uses architectural modifications such as redundancy and other methods to detect and correct faults, for deploying on aerospace missions.

20
  • CASSIANO PERIN DE CARVALHO
  • Deep Learning Architecture for Automatic Modulation Classification in Time-Varying Fading and Impulsive Noise Channels.

  • Líder : LUIZ MARCOS GARCIA GONCALVES
  • MIEMBROS DE LA BANCA :
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • LUIZ MARCOS GARCIA GONCALVES
  • PEDRO THIAGO VALERIO DE SOUZA
  • TALES VINÍCIUS RODRIGUES DE OLIVEIRA CÂMARA
  • Data: 10-dic-2021


  • Resumen Espectáculo
  • The automatic modulation classification (AMC) allows identifying the kind of modulation of the received signal, being a key part of the development of cognitive radio devices that adapt the type of modulation to the characteristics of the communication environment. Several types of research on AMC have been done based on the analysis of the modulation signals and using its parameters for developing powerful feature descriptors to be used on this automatic classification. Recently, a new trend appears related to the use of architectures based on deep learning for this classification. Hence, in this work, we propose to use methods based on deep learning to classify the modulation type of a signal in an environment with doppler fading and impulsive noise. We studied and propose a model based on CNN that has shown to be comparable to the state-of-the-art methods.

21
  • JANAILSON MACIEL DE QUEIROZ
  • Study of DFIG Differential Protection Including Analysis of Interturn Faults

  • Líder : LUCIANO SALES BARROS
  • MIEMBROS DE LA BANCA :
  • DANIEL BARBOSA
  • FLAVIO BEZERRA COSTA
  • LUCIANO SALES BARROS
  • RICARDO AUGUSTO SOUZA FERNANDES
  • Data: 15-dic-2021


  • Resumen Espectáculo
  • With the increase in interest in renewable energies in the world, the importance of research in wind energy, methods, and associated technologies has grown. The increase in wind turbines connected to the electrical systems also brought with it new challenges related to the stability of the system and protection of its components. Differential protection is used in many electrical system components and for DFIG-based wind turbines this protection has been proposed recently. Therefore, this work proposes criteria in the decision-making of the differential protection based on the Park transform, which in addition to the restriction current also uses the voltage level and the direction of the currents at the stator terminals. In a simulation with DFIG operating at 1.51 MW, tests were carried out for internal and external faults to the protection zone, of the single-phase and threephase type with different levels of voltage drops. Inter-turn faults on the machine’s rotor and stator have also been tested with varying levels of defects. Thus, with the proper def inition of the operating logic, it was possible to show the effectiveness of the protection, with detection for all single-phase and three-phase faults within the protection zone, detection of faults in windings in the stator with 10% defect or more and detection of faults in the rotor windings with a defect of 20% or more.

22
  • THIAGO FIGUEIREDO DO NASCIMENTO
  • Modeling and Controllers Design for an Electromagnetic Frequency Regulator Applied to Wind Systems

  • Líder : ANDRES ORTIZ SALAZAR
  • MIEMBROS DE LA BANCA :
  • ANDRES ORTIZ SALAZAR
  • EVANDRO AILSON DE FREITAS NUNES
  • PAULO VITOR SILVA
  • RICARDO FERREIRA PINHEIRO
  • Data: 17-dic-2021


  • Resumen Espectáculo
  • With the high penetration of distributed generation (DG) systems based on wind energy sources, it is necessary to develop technologies for improving the efficiency of system operation. Among the solutions proposed in the literature, the Electromagnetic Frequency Regulator (EFR) device has proven to be an attractive solution for improving the performance of the wind systems. This work proposes a dynamic modeling and control design for EFR applied to a wind systems. The system controllers parameters are designed by using root-locus method (RLM) in order to realize desired dynamic performance. Besides, a stability analysis for closed-loop system is assessed to obtain the operating limits. Finally, simulation results demonstrate the effectiveness of the proposed dynamic modeling and system controllers performance.
Tesis
1
  • THIAGO HENRIQUE FREIRE DE OLIVEIRA
  • Reinforcement Learning Algorithms for Multiobjective Optimization Problems

  • Líder : ADRIAO DUARTE DORIA NETO
  • MIEMBROS DE LA BANCA :
  • ADRIAO DUARTE DORIA NETO
  • ALUIZIO FAUSTO RIBEIRO ARAÚJO
  • DANIEL SABINO AMORIM DE ARAUJO
  • FRANCISCO CHAGAS DE LIMA JUNIOR
  • JORGE DANTAS DE MELO
  • MARCELO AUGUSTO COSTA FERNANDES
  • Data: 11-ene-2021


  • Resumen Espectáculo
  • Multi-objective optimization problems depict real situations and therefore, this class of problems is extremely important. However, even though it has been studied for decades, this class of problems continues to provide challenging situations, especially by the continuing lack of effective techniques. Among all the difficulties that we can find in the optimization of multiple objectives simultaneously, whether conflicting or not, one of the main difficulties found by the algorithms and existing approaches is the need for a priori knowledge of the problem, causing a predefined importance for each of the objectives. When dealing with this class of problems through reinforcement learning, two approaches are predominant: single policy (single-policy) and multiple policies (multi-policy). Algorithms and techniques that use the first approach suffer from the need for prior knowledge of the problem, an inherent characteristic of multi-objective problems. The second approach has other difficulties, such as: limiting the set of solutions and high computational cost. Given this presented context, the work proposes two hybrid algorithms, called Q-Managed with reset and Q-Managed without reset. Both are a hybridization of the Q-learning algorithm and the ε−constraint approach, respectively belonging to reinforcement learning and multi-objective optimization. In summary, the proposed algorithms work as follows: Q- Learning is used for environment exploration, while the ε−constraint approach is used for the environment dynamic delimitation, allowing to keep intact the essence of how the algorithm Q-Learning works. This delimitation has the following purpose: to impose the learning agent can learn other solutions by blocking actions that lead to solutions already learned and without improving them, that is, solutions to which the learning agent has already converged. This blocking actions feature is performed by the figure of a manager, where it is responsible for observing everything that occurs in the environment. Regarding the difference between the proposed algorithms, basically it is the choice of whether or not to take advantage of the knowledge already acquired of the environment after a solution is considered to be learned, that is, the learning agent has converged to a particular solution. As a way of testing the effectiveness of Q-Managed two versions, traditional benchmarks were used, which were also adopted in other works, thus allowing a fairer comparison. Thus, two comparative approaches were adopted, the first of which was through the implementation of third-party algorithms for direct comparison, while the second was done through a common metric to everyone who used the same benchmarks. In all possible tests, the algorithms proposed here proved to be effective, always finding the entire Pareto Front.

2
  • SERGIO NATAN SILVA
  • Reconfigurable computing applied to reduce latency in control and prediction systems associated with tactile internet

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • JOSÉ CLÁUDIO VIEIRA E SILVA JUNIOR
  • LEONARDO ALVES DIAS
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • MARCELO AUGUSTO COSTA FERNANDES
  • Data: 28-ene-2021


  • Resumen Espectáculo
  • Tactile internet is the current technological advance for the Internet. The devices associated with this new internet paradigm will be responsible for man-machine applications with the sending of touch information and the other stimuli already sent. Thus, it is necessary to guarantee an extremely low latency between the devices that make up the tactile interaction. This latency is associated with propagating information through the communication channel, processing power of local devices, and complexity of the techniques being executed, among others. Accordingly, this work proposes using dedicated hardware-based reconfigurable computing to reduce latency in control and prediction systems applied to tactile Internet. Two approaches are proposed to address the problem of latency. In the first approach, reconfigurable hardware is proposed for an intelligent control system based on Fuzzy logic. The system is a Takagi - Sugeno Fuzzy-PI type controller that aims to reduce the latency associated with processing data using a tactile tool. The implementation uses a fully parallel strategy associated with a hybrid bit format scheme (fixed-point and floating-point). In the second approach, the implementation in reconfigurable hardware of linear and nonlinear prediction techniques is proposed. In the nonlinear case, a technique based on multilayer Perceptron artificial neural networks is used. In this approach, prediction techniques are used to minimize the impacts caused by delays and loss of information associated with tactile Internet. The proposals were tested for a field-programmable gate array (FPGA) on the Virtex 6 xc6vlx240t-1ff1156 platform. Data related to hardware occupation and throughput associated with the target platform are presented, and a comparison between results through simulation and implementations in dedicated hardware. The results are superior to those presented in other studies in the literature.

3
  • AUGUSTO CÉSAR REBOUÇAS DE BRITO
  • BOW-TIE ANTENNA INTEGRATED TO A REFLECTIVE FSS FOR 5G SYSTEM APPLICATIONS

  • Líder : ADAILDO GOMES D ASSUNCAO
  • MIEMBROS DE LA BANCA :
  • ADAILDO GOMES D ASSUNCAO
  • LAERCIO MARTINS DE MENDONCA
  • CRISTHIANNE DE FATIMA LINHARES DE VASCONCELOS
  • ADAILDO GOMES D ASSUNCAO JUNIOR
  • CUSTÓDIO JOSÉ OLIVEIRA PEIXEIRO
  • JOSE DE RIBAMAR SILVA OLIVEIRA
  • Data: 02-feb-2021


  • Resumen Espectáculo
  • In recent years, advances in wireless communication systems and the growth in the number of applications have made it attractive to use compact multiband and broadband antennas for both commercial and military communications. The most common examples are found in the variety of portable wireless communication devices, such as: smartphones, handsets, palmtops and laptops, among others. The development of different wireless technologies, such as GSM (Global System for Mobile Communications), RFID (Radio Frequency Identification), Bluetooth, Wi-Fi and WiMAX, has served as a motivation to boost searches in search of lighter devices , compact and low cost, as is the case of planar microstrip antennas. The development of antennas with these characteristics is still a challenge to current research. This work aims at the study, design, manufacture and measurement of new planar antenna geometries, with openings in the ground plane and the use of FSS (Frequency Selective Surface), aiming to reduce its size, volume and weight as well as gain of antennas, to meet the demands of the services offered by the Mobile Systems 5G, for example. The insertion of air layers between the antenna and the EBG structure was also used aiming at a greater degree of freedom to adjust the resonant frequencies. Computational simulations with the objective of predicting the values of the parameters that characterize the proposed structures were performed through the CST and HFSS 13.0 softwares. Prototypes were manufactured and measured. A good agreement was observed between the simulated and measured results.

4
  • ANDOUGLAS GONÇALVES DA SILVA JÚNIOR
  • Holographic Projection with Deep Learning for Microparticles Detection from Water Samples

  • Líder : LUIZ MARCOS GARCIA GONCALVES
  • MIEMBROS DE LA BANCA :
  • VITTORIO BIANCO
  • COSIMO DISTANTE
  • ESTEBAN WALTER GONZALEZ CLUA
  • LUIZ MARCOS GARCIA GONCALVES
  • PABLO JAVIER ALSINA
  • Data: 05-feb-2021


  • Resumen Espectáculo
  • This thesis proposes a complete holographic system to be applied in scientific research and monitoring, which is able to detect microparticles from the holographic projection of water samples, using a deep learning approach. The system proposed in this thesis uses digital holography techniques to acquire holograms from these particles (a device was built for this purpose), reconstruct them numerically by obtaining phase and intensity information, and classify them using machine learning models. In addition, we have developed an application on the web capable of performing all stages of the hologram reconstruction and the classification process using trained models, which is also available. The need for studies on particles that are invisible to the naked eye and that can be dangerous to the health of living beings is an increasingly important research topic and there are many concerns about it. An example is the various types of microplastics found on a large scale in different parts of the planet, even within the human body. Another particle that can help identify microplastics and that can be used to calculate bioindicators of water quality are diatoms. The detection of microplastics and diatoms is subject to difficult studies due to their size, in the order of the micrometer.

5
  • ALEX FABIANO DE ARAÚJO FURTUNATO
  • Analytical modeling for performance prediction of parallel applications in symmetric architectures considering variations in the data access delay

  • Líder : SAMUEL XAVIER DE SOUZA
  • MIEMBROS DE LA BANCA :
  • SAMUEL XAVIER DE SOUZA
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • Arthur Francisco Lorenzon
  • EDSON BORIN
  • Data: 09-feb-2021


  • Resumen Espectáculo
  • Several analytical models created since Amdahl's pioneering work have explored aspects such as variation in the size of the problem, memory size, communication overhead, and synchronization overhead, but delays in accessing data are considered constant. However, such delays can vary, for example, according to the number of cores used, the relationship between processor and memory frequencies, and the problem's size. Given different problem sizes and the large number of possible configurations of operational frequency and number of cores than current architectures can offer, speedup models suitable for describing such variations among these configurations are quite desirable for offline or online scheduling decisions. A new analytical performance model that considers variations in the average data access delay to describe the limiting effect of the memory wall in parallel applications in homogeneous shared memory architectures is presented in this thesis. The experimental results indicate that the proposed modeling captures the application's behavior well. Besides, we show that considering parameters that reflect the applications' intrinsic characteristics, the proposal presented in this work has significant advantages over statistical models such as those based on machine learning. Our experiments also show that conventional machine learning modeling may need about an order of magnitude more measurements than the proposed model to achieve the same level of precision achieved by the proposed model.

6
  • FELIPE FERREIRA DE ARAUJO
  • Wireless communications, Microstrip antennas, Metasurfaces

  • Líder : ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • MIEMBROS DE LA BANCA :
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • ADAILDO GOMES D ASSUNCAO
  • VALDEMIR PRAXEDES DA SILVA NETO
  • ALFREDO GOMES NETO
  • JEFFERSON COSTA E SILVA
  • Data: 18-feb-2021


  • Resumen Espectáculo
  • The microstrip antennas are the object of study in several research fields due to their advantages such as: small thickness, reduced weight and easy integration of electronic circuits. Similarly, the evolution observed in the development of microstrip antennas has also been noted in relation to the metasurfaces, a two-dimensional version of a metamaterial. Metasurfaces are artificial materials that are not found in nature and have unconventional electromagnetic characteristics, such as the index of refraction and coefficients of electrical permittivity and negative magnetic permeability, in addition, they have the ability to be used in planar structures, such as microstrip, without interfering with its traditional geometry. The objective of this work is to increase the bandwidth of a microstrip antenna using a new metasurface model. This new geometry is used to generate high order modes in the antenna and with rotation of the metasurface in relation to the antenna we can overlap the modes and obtain a large bandwidth. The proposed antenna presents bandwidth from 5.1 GHz to 8.0 GHz and can be applied in Wi-Fi 5 and 6. Numerical results were chosen with the Ansys HFSS software. A prototype was built and built for S11 and gains were made. The numerical and experimental results are in agreement.

7
  • PHILIPPI SEDIR GRILO DE MORAIS
  • Salus: A Digital Health Architecture Applied to Syphilis Case Management


  • Líder : RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • MIEMBROS DE LA BANCA :
  • GUILHERME MEDEIROS MACHADO
  • ANGELICA ESPINOSA BARBOSA MIRANDA
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • ION GARCIA MASCARENHAS DE ANDRADE
  • JAILTON CARLOS DE PAIVA
  • KARILANY DANTAS COUTINHO
  • LYANE RAMALHO CORTEZ
  • RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • Data: 26-mar-2021


  • Resumen Espectáculo
  • Syphilis in Brazil grew by over 5000% between 2012 and 2017, this drew the attention of the Ministry of Health, which in 2017 declared a syphilis epidemic Brazil. The hardest side of syphilis is in vertical transmission, that is, when a pregnant woman transmits Treponema Pallidum to her baby. The Federal Court of Accounts (TCU - Brazil), through a serious and in-depth audit regarding the increase in syphilis cases in Brazil, made several notes and recommendations, among them the need to develop tools and technologies that make it possible to monitor more efficiently the evolution of syphilis countrywide. However, when it comes to monitoring it is important to look beyond epidemiological issues, diagnosis, care, treatment and cure. These factors are important, for example, to eliminate vertical transmission of syphilis and, consequently, congenital syphilis, the latter being one of the greatest challenges for Brazil. In this context, it is necessary, therefore, that the actions to face syphilis in Brazil are coordinated between Health Surveillance and Health Care. These two areas, despite being conceived by the Brazilian Health System (SUS) as essentially articulated with each other, operate in a large part of the country in a dichotomous way, including in the Ministry of Health itself. This sometimes occurs due to the lack of integration between the teams that they work in different areas of health, or because the technologies made available by DATASUS are far from understanding the need to mediate a logic of integration between these strategic areas of SUS. The present doctoral thesis is, therefore, situated in this dimension, it brings in its discussion the development of a technological architecture that in fact is a digital health solution that mediates the integration between Health Surveillance and Health Care through technology. both, it incorporates in its flow the management of cases of Syphilis in the territory, that is, in the municipality, places where health is closest to the population. At the same time, it produces epidemiological and care indicators on syphilis. The construction and validation of this architecture was carried out in the city of Natal-RN and the entire discussion surrounding the development and application is presented in this doctoral thesis.


8
  • EVANDRO AILSON DE FREITAS NUNES
  • Contributions to Speed Control Strategy Applied to Speed Multiplication of Frequency Electromagnetic Regulator.

  • Líder : ANDRES ORTIZ SALAZAR
  • MIEMBROS DE LA BANCA :
  • ANDRES ORTIZ SALAZAR
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • RICARDO FERREIRA PINHEIRO
  • DARLAN ALEXANDRIA FERNANDES
  • PAULO VITOR SILVA
  • Data: 08-abr-2021


  • Resumen Espectáculo
  • Currently, incentives for clean energy generation have been increasing considerably due to the limitation of fossil fuels, until then used in abundance. Among these alternative sources, wind energy has been receiving special attention from the energy sector and scientific community and, in Brazil, which in 2017 took the 8th position in the world ranking of countries with the largest installed capacity in wind farms, the participation of this natural resource in the energy matrix shows an exponential growth.
    Speed multiplication is one of the most important steps in wind power generation systems and it is traditionally performed by means of mechanical gearboxes. However, due to its reduced service life, high failure ratio and because it is considered an intermittent noise source, the search for new speed multiplication devices for gearboxes replacement has become an important research topic to increase the efficiency of wind power systems. Electromagnetic Frequency Regulator (REF) is an example of an alternative application that allows a high speed gain while making use of the main advantages of a squirrel cage rotor induction machine. The current control system of the REF consists of the conventional PID strategy, which has several limitations widely discussed in the literature. This strategy also limits REF performance at different points of operation. In this work, a control strategy based on artificial intelligence (AI) of fuzzy logic will be adopted for REF speed regulation, with the aim of eliminate the dependence of the control system in relation to the physical parameters of the prototype. Simulation and experimental results will be presented for the validation of the efficiency of the proposed control strategy.

9
  • BRUNO VICENTE ALVES DE LIMA
  • Semi-supervised Learning by Deep Learning Techniques and Information Theory

  • Líder : ADRIAO DUARTE DORIA NETO
  • MIEMBROS DE LA BANCA :
  • ADRIAO DUARTE DORIA NETO
  • DANIEL SABINO AMORIM DE ARAUJO
  • IVAN NUNES DA SILVA
  • JORGE DANTAS DE MELO
  • VINICIUS PONTE MACHADO
  • Data: 09-jun-2021


  • Resumen Espectáculo
  • The expressive growth of modern data sets, combined with the difficulty of obtaining information about labels, has made semi-supervised learning one of the problems of practical importance in modern data analysis. In most cases, obtaining a set of data with enough examples to induce a classifier, can be costly, as it is necessary to carry out data labeling by a specialist. Unlabeled data is easier to obtain but more difficult to analyze and interpret. In of semi-supervised learning problem, there is a database formed by a small part labeled and a larger part not labeled, being possible two aspects: semi-supervised classification and semi-supervised clustering. With this, this work aims to apply models that use Deep Learning techniques in semi-supervised learning, where a neural network is trained, in this case, an autoencoder using unlabeled data. Then, an additional layer is embedded in the encoder. This new layer has its weights initialized by the K-means ++ algorithm and adjusted through the backpropagation algorithm using information theory learning. The labeled data is assigned to the clusters generated by the encoder, influencing the unlabeled, cluster by cluster, thus labeling the non-labeled data that was previously clustered. With the experiments carried out, it was noted that the satisfactory performanceof the proposed model when compared with other semi-supervised algorithms, both the classics such as self-training and co-training, as well as other more recent works, showing the proposed model feasibility for the learning semi-supervised problem.

10
  • FELIPE OLIVEIRA SIMÕES GAMA
  • Development of Wavelet Coding and IEEE 802.15.4-based Communications Systems for Industrial Wireless Sensor Networks

  • Líder : ANDRES ORTIZ SALAZAR
  • MIEMBROS DE LA BANCA :
  • ANDRES ORTIZ SALAZAR
  • ELMER ROLANDO LLANOS VILLARREAL
  • JEFFERSON DOOLAN FERNANDES
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • RODRIGO SOARES SEMENTE
  • VICENTE ANGELO DE SOUSA JUNIOR
  • Data: 24-jun-2021


  • Resumen Espectáculo
  • With the advent of Industry 4.0 wireless communication has become a trend for signal transmission in industrial environments, sparing features and enabling new applications when compared to wired communications systems. In this context, the transmission of data between sensor and controller within control loop using wireless interconnected nodes is susceptible to the degenerative effects produced by the multiple propagation paths. Nevertheless, most industrial wireless sensor network standards provide no error-control mechanism, which would be useful for improving communication reliability and efficiency, as they are based on the IEEE 802.15.4 protocol. In order to minimize these destructive effects characteristic of wireless channels, several techniques have been investigated recently. Among them, the wavelet coding is presented as a good alternative for the nodes interconnected by wireless communication due to its robustness to the effects of the wireless channel and its low computational complexity. This work is based on an approach that explores wavelet matrices for the design of finite impulse response (FIR) filters used in the coding of the complex signals generated by the physical layer of the IEEE 802.15.4 protocol. A new transmission system is proposed that is still based on the IEEE 802.15.4 protocol, which allows to maintain the interoperability with legacy systems and to still achieve superior performance to the requirements for the establishment of control systems by networks wireless. Specifically, it is expected that the proposed system will increase the reliability of the communication and thereby decrease the retransmission events caused by the wireless channel over the transmitted data packets. In particular, the communication delays derived from these retransmission events seriously interfere with the control system, as it does not guarantee the time constraints of the control signal. In addition, an important point regarding the performance of the proposed system is the possibility to improve the reliability of the communication without diminishing the spectral efficiency. The proposed system also allows for an adaptive performance in terms of the coding rate and parameters of the wavelet matrix, without a significant increase in the implementation complexity. The proposed communication system will be implemented using the software concept radio concept in the GNU Radio development environment. The performance of the communication system will be evaluated in terms of bit error rate (BER) versus Eb /N0 and spectral efficiency, considering a time-variant channel model with flat Rayleigh fading. In addition, the effects of the transmission system will be considered in the controllability analysis of a control system.

11
  • SAMANTA MESQUITA DE HOLANDA
  • Study of Textile Metamaterial for Applications in Planar Antenna Substrates for WBAN Technology
  • Líder : JOSE PATROCINIO DA SILVA
  • MIEMBROS DE LA BANCA :
  • JOSE PATROCINIO DA SILVA
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • MARCOS SILVA DE AQUINO
  • HUMBERTO DIONISIO DE ANDRADE
  • IDALMIR DE SOUZA QUEIROZ JÚNIOR
  • Data: 30-jun-2021


  • Resumen Espectáculo
  • The incessant development of implanted or promoted biometric sensors has led to the growth of Wireless Body Area Network (WBAN) technology. A WBAN can be used for many applications, such as monitoring of physiological signals and industrial communication applications. In this context, the textile antenna is an important element in wireless communication in smart fabrics, being the subject of recurrent research in industrial, military and medical applications. These antennas are flexible and, therefore, become convenient in applications where antenna rigidity is an obstacle This work presents the development of planar antennas with a textile metamaterial substrate in the 5G frequency range for WBAN applications, showing all process from making the substrates to the manufacture and testing of the antennas. Three types of textile substrate with different structure and composition were selected, and in each of them a conductive thread was inserted through different techniques (manual insertion, knitting and automatic sewing). The characterization of the materials was performed using simulations in the HFSS and the coaxial probe method, to obtain the electrical parameters. The antennas were designed and simulated to operate at 3.5 GHz, the best results being manufactured and tested in a Vector Network Analyzer (VNA). The results obtained in the simulations and tests were analyzed and the presence of a metamaterial characteristic was found in the loop and straight stitch geometries. Antennas with a textile metamaterial substrate had a lower resonance frequency than those with a dielectric textile substrate, indicating the possibility of reducing the dimensions of the device.

12
  • LEIDIANE CAROLINA MARTINS DE MOURA FONTOURA
  • A Novel Synthesis Method of Multiband FSS Based on Machine Learning for Wireless Communication Systems

  • Líder : ADAILDO GOMES D ASSUNCAO
  • MIEMBROS DE LA BANCA :
  • ADAILDO GOMES D ASSUNCAO
  • ADAILDO GOMES D ASSUNCAO JUNIOR
  • ALFREDO GOMES NETO
  • CUSTÓDIO JOSÉ OLIVEIRA PEIXEIRO
  • HERTZ WILTON DE CASTRO LINS
  • LAERCIO MARTINS DE MENDONCA
  • VALDEMIR PRAXEDES DA SILVA NETO
  • Data: 09-jul-2021


  • Resumen Espectáculo
  • Frequency selective surfaces, or simply FSS, play a fundamental role in optimizing telecommunications systems by reducing undesirable signals, among other applications. Combining the dimensions and arrangement of the elements and defining the physical characteristics of these devices, such as thickness and permittivity of the substrate, can cause conflict of objectives and make more complex the analysis and synthesis of the FSS. In this context, the present work is a study on the application of supervised machine learning with the decision tree algorithm in the synthesis of frequency selective surfaces. For this, the sunflower (Helianthus annuus) was used as a base element, being an original and simplified geometry, with frequency response characteristics similar to those of fractal structures. The thesis work is thus divided in two parts: the proposed element characterization and synthesis of the multiband FSS. Initially, the evolution of geometry and design equations are presented. The intermediate and the proposed structures are numerically characterized using the commercial software Ansoft Designer, manufactured, and experimentally characterized, with good agreement between the simulated and measured results. In the second step, the sunflower geometry is partially modified to define parameterization variables. The Ansoft Designer numerically characterizes the value of each variable of the new geometry, and it generates the frequency responses without repetition. The decision tree algorithm performs the dataset classification and evaluation, and the random forest algorithm validates and confirms the results. This process and the synthesis of the FSS using the decision tree algorithm occur in less than 10 seconds, with accuracy greater than 90\%, meeting the desirable criteria, under two different scenarios. Based on these scenarios, two FSS are manufactured and experimentally characterized, obtaining results with good agreement. Moreover, it is observed that the agility and precision of this classification algorithm make the synthesis of the structures particularly interesting. Intuitive implementation, simplicity in training and validation, and an efficient data analysis model are highlighted.

13
  • DEMÉTRIOS ARAÚJO MAGALHÃES COUTINHO
  • Performance-Energy Trade-offs Prediction and Runtime Selection for Parallel Applications on Heterogeneous Multi-Processing Systems

  • Líder : SAMUEL XAVIER DE SOUZA
  • MIEMBROS DE LA BANCA :
  • DANIELE DE SENSI
  • ANTONIO CARLOS SCHNEIDER BECK FILHO
  • CARLOS AVELINO DE BARROS
  • HENRIQUE COTA DE FREITAS
  • SAMUEL XAVIER DE SOUZA
  • Data: 15-jul-2021


  • Resumen Espectáculo
  • In the multi-core era, the size of the software operation space, i.e. hardware configurations that provide different software performance and energy consumption, is significantly larger. It becomes even more complex to choose a configuration that optimizes heterogeneous processors' performance and energy consumption. Heterogeneous multicore architectures offer flexibility in different core types and voltage and frequency pairings, defining a vast design space to explore. Furthermore, energy efficiency solutions are crucial on smaller devices as they can lead to longer battery life and a better user experience, including more complex applications. This thesis proposes a methodology to find performance-energy trade-offs for single parallel applications with dynamically balanced workloads running on HMP systems with a single instruction-set architecture (ISA). Our method devises novel analytical models for performance and power consumption whose parameters can be fitted using only a few strategically sampled offline measurements. These models are then used to estimate an application's performance and energy consumption for the whole configuration space. In turn, these offline predictions define the choice of estimated Pareto-optimal configurations of the model, which are used to inform the configuration that the application should execute. The methodology was validated on an ODROID-XU3 board for eight programs from the PARSEC Benchmark, Phoronix Test Suite and Rodinia applications. Energy savings of up to 59.77%, 61.38% and 17.7% were observed compared to the performance, ondemand and powersave Linux governors, respectively, with higher or similar performance. This method aims to provide an optimal start point for a runtime energy manager to make better decisions according to the given application's performance and energy consumption requirements and running system. Therefore, this thesis also proposes a strategy using the Pareto-optimal configuration selected by our models as an appropriate start point for a runtime support framework called Nornir.  This framework performs a  local search dynamically for a more desirable configuration of cores and frequency adapting to workload fluctuations and external interference.  Also, we extend our power model to predict the whole device's consumption, i.e. the sum of all internal components' consumption. This hybrid approach was employed on an ODROID-XU3 board on two multi-thread applications. Nornir starting with Pareto configuration can achieve up to 50% of energy savings compared to random starting configurations. We also observed that the performance and interactive Linux governors consumed up to 1.62X more energy than Nornir using Pareto.

14
  • TADEU FERREIRA OLIVEIRA
  • Use of Parallel and Distributed Processing in the Control Plan of Software Defined Networks to Increase Energy Efficiency in Data Center Networks

     
  • Líder : LUIZ FELIPE DE QUEIROZ SILVEIRA
  • MIEMBROS DE LA BANCA :
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • SAMUEL XAVIER DE SOUZA
  • AGOSTINHO DE MEDEIROS BRITO JUNIOR
  • ANDREY ELÍSIO MONTEIRO BRITO
  • CARLOS AVELINO DE BARROS
  • Data: 30-jul-2021


  • Resumen Espectáculo
  • The main feature of the software-defined networks is the separation of the role of decision-making, known as control-plane, and the role of routing of the packages, known as data-plane. This separation allowed the introduction of the concept of network programmability, with which new applications could be implemented to interact directly with the operation of the networks. Today, these applications enable data center environments to match demand elastically, enabling cloud computing services. In this scenario, the datacenters became the prominent service providers, and one of its main costs is the consumption of energy in the infrastructure of servers and network equipment. Many papers indicate that software-defined network in datacenter networks allows for better energy efficiency, especially in the data plane. In this work, we present a strategy to – using parallel processing and distributed processing with lower operating frequency on the processing elements – reduce the energy consumption of the controllers on a software-defined network. The implementation of parallel and distributed versions of an SDN controller offers a fault-tolerant energy-aware solution to the presented problem.

     
15
  • LUIS ENRIQUE ORTIZ FERNANDEZ
  • Method to Measure, Model, and Predict Depth and Positioning Errors of RGB-D Cameras in Function of Distance, Velocity, and Vibration

  • Líder : RAFAEL BESERRA GOMES
  • MIEMBROS DE LA BANCA :
  • BRUNO MARQUES FERREIRA DA SILVA
  • COSIMO DISTANTE
  • ESTEBAN WALTER GONZALEZ CLUA
  • LUIZ MARCOS GARCIA GONCALVES
  • RAFAEL BESERRA GOMES
  • Data: 02-ago-2021


  • Resumen Espectáculo
  • This thesis proposes a versatile methodology for measuring, modeling, and predicting errors as the RMSE in the depth and the RPE in the camera positioning using data captured from an RGB-D camera mounted on the top of a low-cost mobile robot platform. The proposed method is divided in three stages, with the very first one consisting on the creation of ground truth data for both 3d points (mapping) and camera poses (localization). The next stage is the acquisition of a data set for RMSE and RPE errors calculation using the mobile platform with the RGB-D camera. Finally, the third step is to model and predict the errors in the measurements of depth and positioning of the camera as a function of distance, speed, and vibration. For this modeling and prediction stage, a simple approach based on MLP neural networks is used. This results in two networks, the N rmse Z for the depth error prediction and the N RPE for the prediction of camera positioning error. Experiments show that the N rmse Z and N RPE have an accuracy of ± 1% and ± 2.5%, respectively. The proposed methodology can be straight used in techniques that require an estimation of the error dynamics, as for example probabilistic robotics for localization and mapping, with RGB-D cameras mounted on UAVs, UGVs, and also USVs (including sailboats). Tasks that use RGB-D sensors, such as environmental monitoring, maintenance of engineering works, and public security, could rely on this approach to obtain the error information associated to the camera measurements (depth and positioning).

16
  • LUÍS BRUNO PEREIRA DO NASCIMENTO
  • Smooth and Safe Path Planning based on Probabilistic Foam for Autonomous Robotic Systems

  • Líder : PABLO JAVIER ALSINA
  • MIEMBROS DE LA BANCA :
  • ADELARDO ADELINO DANTAS DE MEDEIROS
  • ALLAN DE MEDEIROS MARTINS
  • DENNIS BARRIOS ARANIBAR
  • EDUARDO OLIVEIRA FREIRE
  • PABLO JAVIER ALSINA
  • Data: 30-ago-2021


  • Resumen Espectáculo
  • Planning a path for a robot to navigate between two points in a given environment and avoiding colliding with obstacles is one of the main issues for autonomous robotics. The search for short paths and reduced search time are aspects of most planning methods, but for applications where the robots interact directly with human beings, such as assistive robotics, ensuring a greater degree of safety in movements is a fundamental requirement. In this context, this Ph.D. thesis presents a set of new strategies for the planning of safe paths for autonomous robots. The methods developed are fundamentally based on the concepts of the Probabilistic Foam Method (PFM). PFM is a sampling-based path planning method capable of generating paths bounded by a set of connected bubbles, which guarantees a volumetric region in the free space for safe maneuverability. In order to compute the bubbles, PFM requires an explicit representation of the obstacles region in the configuration space, which is computationally impracticable considering its application for most problems. Thus, we applied a new strategy to compute bubbles without representing these obstacles regions. Besides, we present an analysis to reduce the number of PFM parameters. To improve the quality of the paths, two optimization procedures were implemented in order to reduce the path length and increase the path smoothness, maintaining the high clearance. New variants of PFM were developed to explore different mechanisms for the propagation of the foam, to ensure the planning of shorter paths, with short searching time, and guaranteeing paths with high clearance.  In order to demonstrate the main contributions of this Ph.D. thesis, some simulated experiments were performed, considering the path planning for two assistive robots: The first one is a lower limbs exoskeleton, with the tasks of overcoming obstacles, walking up and down a stair, resulting in smooth movements, with a more anthropomorphic pattern.  These results illustrate the ability of PFM to plan safe and smooth paths for open kinematic-chain robots without the explicit representation of the obstacles region in configuration space. The Smart Walker was the second robot considered in this work. It was possible to illustrate the planning of safe paths for a mobile robot with differential drive in addition to showing some aspects of the paths planned by the new methods developed.

17
  • JEAN MARIO MOREIRA DE LIMA
  • Representative Feature Extraction for Industrial Virtual Sensors Development: An Approach Based on Deep Learning

  • Líder : FABIO MENEGHETTI UGULINO DE ARAUJO
  • MIEMBROS DE LA BANCA :
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • LEANDRO LUTTIANE DA SILVA LINHARES
  • MARCELO AUGUSTO COSTA FERNANDES
  • ROBERTO KAWAKAMI HARROP GALVAO
  • WALLACE MOREIRA BESSA
  • Data: 03-sep-2021


  • Resumen Espectáculo
  • Deep learning is growing in popularity in virtual sensor modeling problems - the soft sensors - applied to industrial processes of accentuated nonlinearity. Virtual sensors can generate estimates of process variables, which are associated with quality indexes in real- time. Thus, such sensors are a viable alternative when the variables of interest are difficult to measure due to some limiting factor: unavailability of hardware sensors or large measu- rement intervals. Traditional machine learning strategies show difficulties to model such sensors. Typically, industrial processes are highly nonlinear, and the amount of available labeled data is scarce. Due to that, the extraction of representative features present in the abundant amount of unlabeled data has become an area of interest in the development of virtual sensors. From the aforementioned premises, a new virtual sensor modeling tech- nique based on deep learning and representation, which integrates stacked autoencoders (SAE), mutual information (MI), long short-term memory (LSTM), and aggregation bo- otstrap , is proposed. First, in the unsupervised stage, the SAE structure is hierarchically trained layer-by-layer. After a layer’s training, MI analysis is conducted between the tar- get outputs of the model and the representations of the current layer to assess the learned characteristics. The proposed method removes irrelevant information and weights the re- tained ones. The given weights being proportional to the relevance of the representation. Therefore, this approach can extract deep representative information. In the supervised step, called fine-tuning, an LSTM structure is coupled to the tail of the SAE structure to address the intrinsic dynamic behavior of the evaluated industrial systems. Further, a ensemble strategy, called bootstrap aggregation, combines the models obtained in the supervised training phase to improve the performance and credibility of the virtual sen- sor. The proposal uses two industrial nonlinear processes, widely used as benchmarks, to evaluate the performance of the models generated by the proposed technique in the implementation of soft sensors. The results show that the proposed virtual sensors ob- tained better prediction performance than traditional methods and several state-of-the-art methods.

18
  • JULIANO COSTA LEAL DA SILVA
  • Modeling and Harmonic Impact Analysis of a Squirrel Cage Induction Generator Interconnected to the Power Network and Driven by an Electromagnetic Frequency Regulator

  • Líder : MANOEL FIRMINO DE MEDEIROS JUNIOR
  • MIEMBROS DE LA BANCA :
  • DAMÁSIO FERNANDES JUNIOR
  • JOSE TAVARES DE OLIVEIRA
  • MANOEL FIRMINO DE MEDEIROS JUNIOR
  • RICARDO FERREIRA PINHEIRO
  • THALES AUGUSTO DE OLIVEIRA RAMOS
  • Data: 30-sep-2021
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The analysis of electricity quality is an essential issue in the integration of distributed energy sources. Due to the increased penetration of such sources connected to the electricity grid, using power converters, new harmonic disturbance limits are regulated by technical standards. . This system is based on a Squirrel Cage Induction Generator (SCIG), and driven by an Electromagnetic Frequency Regulator (EFR). The steady state harmonic model of the REF is developed from the stationary frame, according to the conventional induction machine modeling. Thus, it is possible to analyze the harmonic disturbances occurring in electrical and mechanical quantities due to the EFR armature voltage inverter. The electrical connection between the EFR and the SCIG is non-existent, and the results demonstrate that the system's inherent inertia contributes to the mitigation of the harmonic content on the network side, arising from the converter switching. In addition to steady state results, Total Nominal Distortion (TND), which includes harmonic and interharmonic components, was calculated and performed better compared to the IEEE 1547-2018 standard and real data extracted from a single Generator Powered Induction System (SCIG). Finally, the harmonic performance of the proposed system was evaluated taking into account the impact of the Thevenin equivalent impedance of the electrical network on the Point of Common Coupling (PCC).

19
  • ROGER ROMMEL FERREIRA DE ARAÚJO
  • Boosting Memory Access Locality of the Spectral Element Method with Hilbert Space-Filling Curves

  • Líder : SAMUEL XAVIER DE SOUZA
  • MIEMBROS DE LA BANCA :
  • SAMUEL XAVIER DE SOUZA
  • JOAO MEDEIROS DE ARAUJO
  • HIROSHI OKUDA
  • Lucia Catabriga
  • LUTZ GROSS
  • Data: 07-oct-2021


  • Resumen Espectáculo
  • The wave equation is pervasive in mathematical physics and engineering, and we need to solve it repeatedly to simulate wave propagations in software. The spectral element method, one of several approaches for the numerical solution of the wave equation, discretizes the underlying domain in a mesh made of elements and nodes, and traverses every element and every node at each time step as it marches the target equation through time. We propose a memory reordering algorithm, meant to be used with the spectral element method, that rearranges mesh-related data to reduce the number of cache misses and boost locality of data reference, thereby improving the execution speed of the mesh traversal process. We devise a spectral element method formulation for 2D waves over unstructured meshes made of triangles, and we pair it to our memory reordering algorithm to construct an acoustic wave propagation simulator. Our experiments show that the reordering technique based on Hilbert space-filling curves performs well in meshes of different granularities, and also when the variation in element sizes is either small or large. In addition, we compare the proposed approach with three other memory reordering schemes, and find that our algorithm runs between 9% and 25% faster than the alternatives we tested. We recommend this memory reordering algorithm to any application that requires successive traversals across domains.

20
  • GUILHERME PENHA DA SILVA JUNIOR
  • Implementation of a Synchronverter Virtual Synchronous Machine for Double Fed Induction Generator Connected to a Microgrid

  • Líder : LUCIANO SALES BARROS
  • MIEMBROS DE LA BANCA :
  • DANIEL BARBOSA
  • FLAVIO BEZERRA COSTA
  • FRANCISCO KLEBER DE ARAÚJO LIMA
  • LUCIANO SALES BARROS
  • RODRIGO ANDRADE RAMOS
  • Data: 08-oct-2021


  • Resumen Espectáculo
  • The high integration of Distributed Generation (GD) to the conventional electric system
    brings many challenges to the operating sector, given the growing concerns about the
    reliability of the system and the quality of the energy generation. The DGs have no control
    over the generated energy, although they have very fast power dynamics, with little or no
    inertia. In order to mitigate the impacts caused by the integration between the GD and the
    existing system, it is essential to use the concept of a virtual synchronous machine, whose
    fundament search to emulate the behavior of a real synchronous machine by VSI (Voltage
    Source Inverter) control. In this context, this work proposes the development of a control
    strategy aimed at controlling the DFIG (Doubly Fed Induction Generator) aerogenerator
    so that all necessary requirements can be met to maintain the stability of the grid. The
    purpose will be achieved starting from the assumption that the VSI connected to the grid
    will have similar behavior to a synchronous generator, since the control technique will
    be used Synchronverter. Thus, in addition to greater control over the active and reactive
    powers generated, it is also possible to obtain a contribution with the ancillary services. In
    addition, the DC bus of the back-to-back converter must be connected to a battery bank,
    which will enable greater control over the power dispatch. Therefore, it is possible to
    charge the batteries in the horaires of greater generation and lower energy consumption
    of the grid, as well as to discharge the batteries in the moment of demand of support to
    the frequency of the grid. Preliminary results confirmed the functioning reported in the
    literature, proving the effectiveness of the technique, with potential for improvements that
    the work aims. Finally, with the use of the Synchronverter, it will be possible to operate
    the DFIG autonomously in a microgrid in the occurrence of a Blackout in the system and,
    after ceasing the collapse in the system, synchronize the VSI to the grid without the need
    to use a PLL (Phase Locked Loop).

21
  • AGUINALDO BEZERRA BATISTA JÚNIOR
  • A knowledge graph data-driven approach for analyzing industrial alarm and event records 

  • Líder : IVANOVITCH MEDEIROS DANTAS DA SILVA
  • MIEMBROS DE LA BANCA :
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • DIEGO RODRIGO CABRAL SILVA
  • GUSTAVO BEZERRA PAZ LEITAO
  • JUAN MOISES MAURICIO VILLANUEVA
  • PLACIDO ANTONIO DE SOUZA NETO
  • Data: 22-oct-2021


  • Resumen Espectáculo
  • Alarm and event logs make up a voluminous and dormant historical repository of tabular-like data, commonly undervalued or overlooked in manufacturing. Although they are a potentially rich source of relevant information about the monitored plant or process, these records are taken for analysis only as a last resort, mainly due to the difficulties imposed by the large volume and low expressiveness of those databases. Such oversight is no longer acceptable in the contemporary data-oriented scenario, already ubiquitous in several productive sectors and gaining prominence in traditional manufacturing, especially due to the advent of the Industry 4.0 paradigm. Therefore, it is proposed to transpose these bases to a more expressive and flexible representation domain, allowing a more proactive exploration of the episodes reported in the records and, consequently, entailing more agile incident, anomaly, compliance, and performance analysis tasks. For such, from the recognition of an ontology, entities, attributes, and associations virtually immersed in the operational context described in the records are mapped into a Knowledge Graph (KG). The approach uses Exploratory Data Analysis, Natural Language Processing, Network Analysis, Multivariate Analysis, and Composite Indicators techniques to derive a myriad of aspects, properties, and relations from data, which are incorporated as hierarchical, temporal, and similarity relationships (edges) between identified entities (nodes). The visualization of the KG is dynamic and interactive, with different visualization modes and levels of detail. Evaluation scenarios are designed to demonstrate the effectiveness of the approach.


22
  • DIEGO ROCHA LIMA
  • Visual attractiveness in vehicle routing through bi-objective optimization

  • Líder : DANIEL ALOISE
  • MIEMBROS DE LA BANCA :
  • ANAND SUBRAMANIAN
  • BRUNO JEFFERSON DE SOUSA PESSOA
  • DANIEL ALOISE
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • LUCIANO FERREIRA
  • Data: 03-dic-2021


  • Resumen Espectáculo
  • In this thesis we approach a vehicle routing problem where the route distribution system must consider both its effective cost and its visual attractiveness. Clustering methods are in principle not designed for the Vehicle Routing Problem, but when used, they can provide visually attractive and possibly cost-effective solutions. So, our proposal is to work in an integrated way in a bi-objective method, which are the route cost minimization and the optimization of a grouping criterion, thus making customers better partitioned in different routes. For this we use a multi-objective evolutionary algorithm based on non-dominance ordering, in order to approximate its Pareto Frontier. We show through computational experiments that our model is capable of generating solutions for vehicle routing that have a low cost and at the same time are visually attractive according to the metrics proposed in the literature. Furthermore, the model was tested with a group of instances based on data from a real road network.

23
  • JOSÉ DE ARIMATÉIA PINTO MAGNO
  • ANALYSIS AND DESIGN OF MICROSTRIP ANTENNA ON CERAMIC SUBSTRATES FOR IEEE 802.11ax SYSTEM APPLICATIONS
  • Líder : VALDEMIR PRAXEDES DA SILVA NETO
  • MIEMBROS DE LA BANCA :
  • ADAILDO GOMES D ASSUNCAO
  • IDALMIR DE SOUZA QUEIROZ JÚNIOR
  • JOAO BOSCO LUCENA DE OLIVEIRA
  • JOSE PATROCINIO DA SILVA
  • SAMANTA MESQUITA DE HOLANDA
  • VALDEMIR PRAXEDES DA SILVA NETO
  • Data: 10-dic-2021


  • Resumen Espectáculo
  • The needs of new wireless communication systems demand that radiant devices are increasingly stable and reliable. In recent years, several proposed solutions are observed in the literature to improve the performance of planar antennas applied in modern wireless communication systems. This work presents the use of ceriferous substrates as a biodegradable solution, of low cost, electrically stable and with many application possibilities to improve the operation of planar antennas in wireless communication systems. Although the use of organic substrates is not new, a variety of manufacturers have lines of boards for planar circuits based on organic compounds. This work proposes in an innovative way the use of ceriferous substrates for the development of miniaturized microstrip antennas. Planar antennas were developed for applications in IEEE 802.11 ax protocol services. Throughout this work, the process of characterizing the proposed material, making the substrates, designing and manufacturing the proposed antennas, numerical analysis and experimental characterization will be presented. Four types of ceriferous substrates based on carnauba wax were selected, and the main characteristics were determined by analyzing the material's electrical properties, X-ray diffraction and thermogravimetry. With results from the characterization of the substrates, the radiating devices were designed and numerically analyzed by the Finite Element method, using the Ansys HFSS® software. The antennas were experimentally characterized in the frequency range from 1GHz to 8.5 GHz, where it was possible to verify that the material used as substrate presented stability and good properties for applications in planar circuits, as in the case of microstrip antennas. There is also the possibility of miniaturization of the radiant elements using carnauba wax as a substrate, when compared to the same structure designed on a standard commercial substrate such as FR4, a miniaturization factor of 44.5% was achieved in the volume of the miniaturized antenna on wax substrate.  

24
  • BRUNO DE MELO PINHEIRO
  • Design of internal antennas and electromagnetic analysis of Osseus, a diagnostic and patient screening device for osteoporosis

  • Líder : ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • MIEMBROS DE LA BANCA :
  • AGNALDO SOUZA CRUZ
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • FRANCISCO CARLOS GURGEL DA SILVA SEGUNDO
  • RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • ROBSON HEBRAICO CIPRIANO MANICOBA
  • Data: 10-dic-2021


  • Resumen Espectáculo
  • Osteoporosis is a systemic osteometabolic disease that affects an increasing portion of the Brazilian population, which generates significant costs for the Brazilian Unified Health System (SUS). However, with a previous diagnosis, it is possible to implement preventive measures, which can prevent the occurrence of fractures and reduce SUS costs. In order to enable the necessary exams to identify the pathology and observe its evolution during treatment, maintaining agility and low costs, the development of Osseus was proposed, an instrument that combines techniques and concepts from different areas such as: software engineering, electrical, electronics, computing and biomedical. In addition, the equipment was proposed to be developed at low cost, to be easily accessible to the population and non-invasive, being developed at the Laboratory of Health Innovations and Technologies (LAIS). The latest version of Osseus, version 3.0, had some problems that prevent it from being implemented on an industrial scale. These problems are related to some operational instabilities observed in the part of radio frequency (RF) antennas, which are planar microstrip antennas. Therefore, this work proposes the implementation of improvements in Osseus, aiming to eliminate these instabilities and propose the use of new computational intelligence algorithms, to be used in the equipment. It is hoped, with this, to improve the equipment and provide a low-cost device, helping LAIS fulfill its mission, which is to make science an instrument of love for others.

25
  • ROMÊNIA GURGEL VIEIRA
  • Application of Artificial Intelligence Techniques for Fault Identification in Photovoltaic Modules

  • Líder : FABIO MENEGHETTI UGULINO DE ARAUJO
  • MIEMBROS DE LA BANCA :
  • ANDRE LAURINDO MAITELLI
  • ANDRES ORTIZ SALAZAR
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • JOÃO TEIXEIRA DE CARVALHO NETO
  • MARCELO ROBERTO BASTOS GUERRA VALE
  • Data: 14-dic-2021


  • Resumen Espectáculo
  • Photovoltaic solar energy has proven to be a viable alternative that contributes to sustainable development and ensures energy supply around the world. However, the exponential growth of installed capacity in recent years has highlighted the need to ensure the safe operation and reliability of photovoltaic systems. In this context, faults in such systems are a crucial issue, since they can significantly impact the generated power, reduce the useful life, and cause potential risks in operation. Thus, this research applied artificial intelligence techniques to detect and diagnose faults in photovoltaic modules. The faults identified by the proposed methods are short-circuit modules, string disconnection and partial shading. In addition, multilayer perceptron neural network algorithms, probabilistic neural networks, and a neuro-fuzzy method were developed, combining a neural network with fuzzy logic. All trained algorithms were used from simulated and tested experimental data from three different photovoltaic systems. Moreover, training situations in which the dataset is contaminated by random noise were also considered. The results indicated maximum accuracy of 99.1% for the lack of short-circuited modules, 100% for string disconnection and 82.2% for the lack of partial shading. Furthermore, the analyzes allowed to reaffirm the robustness of the multi-layer perceptron network for fault detection in photovoltaic systems, even with the presence of noise in the training data.

26
  • WELLINGTON GUILHERME DA SILVA
  • Miniaturization of Microstrip Antennas for Applications in Cubesats

  • Líder : ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • MIEMBROS DE LA BANCA :
  • ALFREDO GOMES NETO
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • HUMBERTO DIONISIO DE ANDRADE
  • LAERCIO MARTINS DE MENDONCA
  • MARCIO EDUARDO DA COSTA RODRIGUES
  • Data: 15-dic-2021


  • Resumen Espectáculo
  • The objective of this work was to develop a compact microstrip antenna for use in CubeSats, a type of miniaturized satellite mainly used for space exploration and research, used in low earth orbit. The small size of satellites presents a great challenge to the project. One of the main components being a satellite board, an antenna determines the performance of all wireless systems, including: telemetry, tracking, and control (TT&C); downlink of high-speed data; navigation; communications between satellites; radars; and sensors. For the TT&C function, expandable wire antennas are often employed in V-UHF. However, the use of the release mechanism of these antennas can fail and compromise a mission. Therefore, a compact microstrip antenna is desirable. The work presents a literature review covering some concepts about the CubeSat pattern, the fundamentals of antenna theory, such as the main techniques for the miniaturization of antennas, the fundamentals of electrically small antennas, and the main works already developed on the subject. Further investigation was carried out on Wheeler’s method to measure the radiation efficiency of compact antennas, this method is applied to measure the radio efficiency of the built prototype. The work methodology consists of designing, simulating, building, and characterizing a microstrip antenna operating at two frequencies (401,6 MHz and 435,0 MHz) based on the fundamentals of the literature review. These frequencies were chosen to meet the requirements of most CubeSats in use, but in particular, they meet two of the frequencies planned for CONASAT, a project of the Northeast Regional Center (CRN) of the National Institute for Space Research (INPE), located in Natal, in partnership with the Federal University of Rio Grande do Norte (UFRN). The contribution of this work lies in the ability of the developed prototype to operate at two frequencies with a single power supply point, thus reducing the weight, complexity, and risk of failures in the CubeSats communication systems.

27
  • ROANA D' ÁVILA SOUZA MONTEIRO
  • Parameters-Free Non-Iterative Two-terminal Measurements Synchronization and Fault Location Algorithms

  • Líder : MANOEL FIRMINO DE MEDEIROS JUNIOR
  • MIEMBROS DE LA BANCA :
  • FELIPE VIGOLVINO LOPES
  • JOSE TAVARES DE OLIVEIRA
  • MANOEL FIRMINO DE MEDEIROS JUNIOR
  • MELINDA CESIANARA SILVA DA CRUZ
  • WASHINGTON LUIZ ARAUJO NEVES
  • Data: 16-dic-2021


  • Resumen Espectáculo
  • In this work, angle synchronization method and fault location method were proposed, based on the fundamental frequency phasor approach, independent of the electrical parameters of the transmission line and of iterative processes. In the synchronization algorithm proposed the line capacitive effect is taken into account and only the analysis of steady-state voltage and current phasors is required, eliminating the need for settings based on line electrical parameters. In addition, it identifies angles in any quadrant and is insensitive to operation conditions of the system. Real operating conditions of the Brazilian system are considered to assess the algorithm performance under different scenarios, including light, moderate and heavy line loading cases, and various power factors and line lengths.  The obtained results show that the synchronization proposed solution is reliable, straightforward, accurate, being useful for practical two-terminal phasor-based applications when traditional common time reference sources are not available. The tool developed for fault location requires pre and post-fault current and voltage phasors, work for all types of faults (symmetrical and asymmetrical), and for transposed and untransposed transmission lines. The evaluation of the fault location algorithms is performed through fault simulations by the ATPDraw software. The obtained results prove the efficiency of the method in finding fault location with good precision.

2020
Disertaciones
1
  • JURGEN KLINSMANN AZEVEDO NOGUEIRA
  •  Analysis of Planar Antennas for Medical Applications at 435 MHz, 915 MHz and 2.45 GHz ISM Bands

  • Líder : ADAILDO GOMES D ASSUNCAO
  • MIEMBROS DE LA BANCA :
  • ADAILDO GOMES D ASSUNCAO
  • LAERCIO MARTINS DE MENDONCA
  • VALDEMIR PRAXEDES DA SILVA NETO
  • HERTZ WILTON DE CASTRO LINS
  • JOAO BOSCO LUCENA DE OLIVEIRA
  • JOSE DE RIBAMAR SILVA OLIVEIRA
  • Data: 10-ene-2020


  • Resumen Espectáculo
  • Wireless devices are presented in the latest research works as alternatives for monitoring several medical parameters, with applications in the frequency bands designated for Medical Device Radiocommunications Service - MedRadio (401 - 405 MHz), Medical Implant Communication Services - MICS (402 - 405 MHz) and Industrial, Scientific and Medical - ISM (433 - 435.8 MHz; 2.4 - 2.5 GHz). This master thesis presents a frequency response analysis of planar microstrip antennas with medical applications in the ISM band for the three frequency bands (433 - 435.8 MHz; 902 - 928 MHz; 2.4 - 2.5 GHz). Three rectangular slotted antennas are designed with the slots placed near the edges of the patch. Simulation results were obtained by Ansoft HFSS software for resonance frequency, bandwidth and reflection coefficient. Measurements were performed with solid samples of pork. The simulated and measured results showed excellent agreement.

2
  • CARLA DOS SANTOS SANTANA
  • Workload scheduling analysis in geophysical numerical methods

  • Líder : SAMUEL XAVIER DE SOUZA
  • MIEMBROS DE LA BANCA :
  • Angelo Amâncio Duarte
  • CALEBE DE PAULA BIANCHINI
  • HERVÉ CHAURIS
  • SAMUEL XAVIER DE SOUZA
  • TIAGO TAVARES LEITE BARROS
  • Data: 13-ene-2020


  • Resumen Espectáculo
  • The tasks organization among computational nodes affects the performance of the program. In computationally expensive applications such as geophysical problems, the impact is more significant. The imbalance caused by inefficient task scheduling can generate an application non-scalable. Therefore this work investigates the workload scheduling in geophysical methods.

     Three types of workload scheduling will be compared: centralized dynamic (CD), decentralized static (DS), and decentralized dynamic (DD).  The CD was implemented with the technique master-slave, where the master node is responsible for the distribution of the tasks to other nodes called slaves. The DS with an equal division of tasks. The DD implements the work-stealing method proposed by Assis et al. 2019, where an idle node can steal the tasks of an overloaded node.

    The principal geophysics method used was Full waveform inversion (FWI) 2D with the acoustic wave. To analyze the performance of workload scheduling methods, we employ a synthetic velocity model and present speedup, efficiency, and load distribution plots generated with different model sizes and different quantity of nodes. The FWI code and the workload scheduling methods were implemented in C to distributed memory parallelization and using the message passing interface (MPI) library.

    With the results of the workload scheduling methods in FWI, it was applied the work-stealing (because this technique presented the more effective performance) in another geophysics problem: Least-squares migration (LSM).  We used the LSM with DS implemented by Chauris et al. 2017 to compare with the LSM with work-stealing. The LSM code and DS used in this problem were implemented in Fortran and the work-stealing in C. The communication between the nodes was implemented using MPI.  To analyze the performance of workload scheduling methods in LSM, we used the marmousi velocity model.

3
  • JOÃO BATISTA FERNANDES
  • Auto-tuning of load scheduling of granularity for multi-core processors applied to reverse time migration

  • Líder : SAMUEL XAVIER DE SOUZA
  • MIEMBROS DE LA BANCA :
  • HUGO ALEXANDRE DANTAS DO NASCIMENTO
  • IDALMIS MILIAN SARDINA MARTINS
  • SAMUEL XAVIER DE SOUZA
  • TIAGO TAVARES LEITE BARROS
  • Data: 14-ene-2020


  • Resumen Espectáculo
  • Reverse time migration (RTM) is an algorithm widely used in the oil and gas industry to process seismic data. It is a computationally intensive task that suits well in parallel computers. Because of it being massive and regular, this type of task is often equally and statically distributed among the available parallel workers. However, this strategy is often not optimal. When the workers are heterogeneous, and even when most have similar processing power, many of them might still have to wait idly for the slower workers. In this paper, we show that even small performance differences between homogeneous cores can considerably affect the overall performance of a 3D RTM application. We show that dynamic load distribution has a significant advantage over the conventional static distribution, and others defaults OpenMP scheduling, auto and guided. However, the granularity of the dynamically distributed chunks of work plays a key role in harvesting this advantage. In order to find the optimal granularity, we propose a coupled simulated annealing (CSA) based auto-tuning strategy that adjusts the chunk size of work that OpenMP parallel loops assign dynamically to worker threads during the initialization of a 3D RTM application. Experiments performed on computational systems with different processor and RAM specifications and for different sizes of input show that the proposed method is consistently better than the default OpenMP loop schedulers.

4
  • Vitor Hugo Mickus Rodrigues
  • GPU Support for Automatic Generation of Finite-Differences Stencil Kernels

  • Líder : SAMUEL XAVIER DE SOUZA
  • MIEMBROS DE LA BANCA :
  • SAMUEL XAVIER DE SOUZA
  • LUCAS COSTA PEREIRA CAVALCANTE
  • Pedro da Silva Peixoto
  • CALEBE DE PAULA BIANCHINI
  • Gerard John Gorman
  • Data: 16-ene-2020


  • Resumen Espectáculo
  • The growth of data to be processed in the Oil & Gas industry matches the requirements imposed by evolving algorithms based on stencil computations, such as Full Waveform Inversion and Reverse Time Migration. Graphical processing units (GPUs) are an attractive architectural target for stencil computations because of its high degree of data parallelism. However, the rapid architectural and technological progression makes it difficult for even the most proficient programmers to remain up-to-date with the technological advances at a micro-architectural level. In this work, we present an extension for an open source compiler designed to produce highly optimized finite difference kernels for use in inversion methods named Devito c . We embed it with the Oxford Parallel Domain Specific Language (OP-DSL) in order to enable automatic code generation for GPU architectures from a high-level representation. We aim to enable users coding in a symbolic representation level to effortlessly get their implementations leveraged by the processing capacities of GPU architectures. The implemented backend is evaluated on a NVIDIA R GTX Titan Z, and on a NVIDIA R Tesla V100. in terms of operational intensity through the roofline model. Experimented with a 3D acoustic isotropic wave propagation stencil kernels for varying space-order discretization levels. It achieves approximately 63% of V100’s peak performance and 24% of Titan Z’s peak performance for stencil kernels over grids with 256 3 points. This study reveals that improving memory usage should be the most efficient strategy for leveraging the performance of the implemented solution on the evaluated architectures.

5
  • GISLIANY LILLIAN ALVES DE OLIVEIRA
  • A data-driven approach for the generation of a liveability index based on the UBER API

  • Líder : IVANOVITCH MEDEIROS DANTAS DA SILVA
  • MIEMBROS DE LA BANCA :
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • LUCIANA CONCEICAO DE LIMA
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • PATRICIA TAKAKO ENDO
  • Data: 24-ene-2020


  • Resumen Espectáculo
  • One of the global dilemmas concerns the accelerated urban transition in the last decades. Therefore, promoting sustainable urban development to accommodate population growth is extremely important. Under those circumstances, the concept of liveability arises, being defined as a principle that combines economic, social and environmental attributes to promote quality of life and human well-being, and it is widely discussed in the New Urban Agenda (NUA) adopted by the United Nations in 2016. NUA defines policies to promote the consolidation of  Sustainable Development Goals (SDGs), with its Goal 11 focusing on a pro-urban future. To supervise SDGs implementation and ensure that these goals are met, it is recommended the use of indicators and the liveability concept can then be associated with an indicator for this purpose. However, there are known issues related to data unavailability, poor quality and aggregation, that make the SDGs monitoring difficult. Considering the described scenario, this work proposes a liveability indicator that combines traditional census data with alternative data sources, such as data  from Uber, a popular ride-sharing service. Assuming that Uber service behavior can act as a proxy to liveability, a data science approach based on exploratory and spatial data analysis was conducted using Uber’s Estimated Time of Arrival (ETA) data sourced for the Brazilian city of Natal (RN). This approach aims to build a composite indicator which can portray at some level the liveability for that city. The proposed methodology was applied at two different spatial aggregation levels: neighborhoods and Human Development Unities (HDUs). Results showed how the Uber service oscillates spatially and how it reacts to weather variations, festivals, and other events, as well as its relations with existing social and infrastructural indicators. It was also observed that different spatial aggregation levels affect the Uber ETA and its relations with socioeconomic variables. Finally, the proposed indicator was created at HDU scale to be applied in sustainable development monitoring. Furthermore, it was concluded that West and North administrative regions of Natal predominantly have localities with the worst liveability indicators. 




6
  • TALINE DOS SANTOS NÓBREGA
  • Statistical analysis of electroencephalographic data in human applications

  • Líder : ALLAN DE MEDEIROS MARTINS
  • MIEMBROS DE LA BANCA :
  • ALLAN DE MEDEIROS MARTINS
  • ANDRÉ MENDES CAVALCANTE
  • EDGARD MORYA
  • HELIANA BEZERRA SOARES
  • VICENTE ANGELO DE SOUSA JUNIOR
  • Data: 31-ene-2020


  • Resumen Espectáculo
  • The motor activities of the human body, as well as those related to decision making and emotional and psychic issues, can be understood through the analysis of electrical signals from the brain, also known as electroencephalogram (EEG) signals. The study and application of electroencephalographic data has been growing within the scientific community. It is known that the use of these signals forms the basis of the development of the Computer Brain Interface (ICC), and that these represent the future of assistive technologies, especially those aimed at people who do not have motor control. However, the extraction of characteristics and patterns of these signals is still a complex process. Surveys involving ICC and EEG signals usually implement event related potentials (ERP) analyzes; being the main ones: the static potentials evoked visually (SSVEP) and the potential P300. In general, they are responses to external stimuli (visual, auditory, tactile), and are widely used to recognize patterns in EEG signals associated with changes in brain activity. The purpose of this paper is to analyze the signs of the neural activity of individuals who are exposed to external stimuli using the identification of potential SSVEP and P300. The project uses a low-cost, non-invasive EEG signal sensor with wireless, wireless technology. It is expected to extract EEG data to the point of making possible the correlation of these with characteristics that can be applied in control tools.

7
  • IAGO DIÓGENES DO RÊGO
  • FFR-based ICIC powered by Q-Learning as a solution for interference in high user density areas (hotspots)

  • Líder : VICENTE ANGELO DE SOUSA JUNIOR
  • MIEMBROS DE LA BANCA :
  • ANDRÉ MENDES CAVALCANTE
  • VALDEMIR PRAXEDES DA SILVA NETO
  • VICENTE ANGELO DE SOUSA JUNIOR
  • Data: 31-ene-2020


  • Resumen Espectáculo
  • This work aims to explore interference coordination techniques (ICIC) based on fractional frequency reuse (FFR) as a solution for interference in a multicellular scenario. The system presents a dynamic variation on the density of its users. Therefore, the Q-Learning algorithm performs the dynamic configuration of the ICIC parameters. The first chapter discusses the problem of high user density and its consequences. The following chapter presents some classic solutions, leading to the introduction of Fractional Frequency Reuse (FFR) and existing FFR techniques. Prior results are presented for performance evaluation. These also serve to guide the choice of relevant parameters and to highlight the impact of high user density. The final chapters present the results regarding the use of Q-Learning in the dynamic scenario.

8
  • JOSÉ MARTINS DE CASTRO NETO
  • Coexistence Solutions for LTE and Wi-Fi Systems in Multicell Deployments

  • Líder : VICENTE ANGELO DE SOUSA JUNIOR
  • MIEMBROS DE LA BANCA :
  • ALLAN DE MEDEIROS MARTINS
  • LEONARDO AGUAYO
  • VICENTE ANGELO DE SOUSA JUNIOR
  • Data: 10-ago-2020


  • Resumen Espectáculo
  • The growth of mobile internet access from fourth generation (4G) devices, combined with increasing usage of smartphones, the upcoming fifth generation (5G) and massive usage of multimedia services, make the demand for mobile traffic reach high levels and the need for bandwidth grows. However, the electromagnetic spectrum utilized by these applications is limited, creating scarcity in the face of demand, besides the high costs and bureaucracy for granting new bands. To overcome this problem, one of the solutions is to leverage the unlicensed spectrum, for it is free of charges, presents some of its portions with bandwidth higher than the licensed spectrum, and portions with underused profile, indicating less probability of interference between technologies. In this context rises the LTE-U and LTE-LAA technologies with modifications in the medium access mechanism of LTE for leveraging the unlicensed spectrum in the Industrial, Scientific and Medical (ISM) 5 GHz band. These technologies must coexist with the most successful and consolidated technology already using this portion of the spectrum, the Wi-Fi. However, each technology implements its access mechanism, then studies must be done to point out impacts that each of these technologies has when coexist. Besides the coexistence study, the application of machine learning techniques to automatically adjust the medium access parameters, controlling the generated impact of one technology into another must be realized.   Studies in such a scenario and with single-cell environments have already been explored in literature, remaining the challenge of new solutions targeting multi-cell environments. For all it has been exposed, this work has the following goals: (1) The coexistence study of LTE-U and Wi-Fi technologies in a multi-cell scenario, with co-channel and inter-RAT (same technology) interference; (2) The application of machine learning algorithms (reinforcement learning) to adjust the parameters targeting optimizing the medium access for one, or both technologies, and consequently reach improvements in the coexistence measured in the form of data rates and decreasing packet losses. 

9
  • NILSON HENRIQUE DE OLIVEIRA CUNHA
  • Metamaterial unit cell analysis using EBG substrates for integrated circuit applications

  • Líder : JOSE PATROCINIO DA SILVA
  • MIEMBROS DE LA BANCA :
  • JOSE PATROCINIO DA SILVA
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • FRANCISCO DAS CHAGAS BARBOSA DE SENA
  • Data: 26-ago-2020


  • Resumen Espectáculo
  • With the growth in communications and the intense study of how to transmit and receive electromagnetic signals more efficiently, printed circuits have gained prominence on the world scene. The flat profile, easy construction and easy integration with other nowadays systems are attractive of these circuits that are present in almost all current devices. However, it is known that these types of elements have several disadvantages such as relative low gain, generally narrow bandwidth and low efficiency. To solve these various problems, elements that work in sets are often added, such as filters and Substrate Integrated Waveguide (SIW), but it is not always possible to apply these elements due to the physical limitations of the desired system. An alternative to work around these limitations is the application of structures with prohibited bands of electromagnetic wave propagation called EBGs (Electromagnetic Band Gap), another alternative is the application of artificial structures called metamaterials (MTMs). However depending on the format, distribution and periodicity, quasi-periodicity or non-periodicity of these techniques it is possible to obtain different results, such as increase in gain, bandwidth, impedance matching, efficiency increase, among many other parameters reported in the literature. In this context, aiming to merge the techniques mentioned above, we intend to carry out a study, from the characterization of a simple MTM-EBG cell to its application in circuits such as antennas, filters and printed lines, so as to prove the usefulness and scientific importance of these techniques for applications in current systems.

10
  • EMERSON VILAR DE OLIVEIRA
  • Performance Evaluation of LSTM Network-Based Method for Failt Classification in a Level Control Process

  • Líder : LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • MIEMBROS DE LA BANCA :
  • CLAUBER GOMES BEZERRA
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • LUIZ MARCOS GARCIA GONCALVES
  • Data: 28-ago-2020


  • Resumen Espectáculo
  • Due to the increasing demands in the operation monitoring of industrial plants, methodologies for fault detect and diagnose in the operation of these processes are gaining more and more importance, because they can contribute to more assertive and even predictive repairs in the components that generated such disturbances to the proper functioning of the system. With the growth of data-oriented approaches, Artificial Neural Networks have become considerable allies in solving these problems, and Recurrent Neural Networks, in particular, have gained strength due to their affinity in dealing with series that have temporal links between their samples, which is the case of industrial process variables monitoring. Due to this relevance, this dissertation analyzes the performance of Long Short-Term Memory (LSTM) recursive neural network for the detection and classification of faults in a pilot-scaled level control process. For the performance evaluation, a methodology based on Monte Carlo statistical tests was used, in which the influence of the LSTM network hyperparameters, such as number of layers and size of the input and regressors, was analyzed. The accuracy was the metric chosen to quantify the fault classification performance. The data set obtained from the operation of the pilot plant contained 23 situations of disturbances in this process, which resulted from disturbances applied to components such as sensor, valves and the water tank itself. The adopted methodology proved to be quite efficient to analyze both the performance and the robustness of these neural networks for the fault classification activity, in addition to indicating the best network architecture configurations.

11
  • Júlio Gustavo Soares Firmo da Costa
  • Metadata Interpretation Driven Development: promoting software development dissociated from the business domain

  • Líder : SAMUEL XAVIER DE SOUZA
  • MIEMBROS DE LA BANCA :
  • ANDERSON LUIZ DE OLIVEIRA CAVALCANTI
  • GIVANALDO ROCHA DE SOUZA
  • LUIZ MARCOS GARCIA GONCALVES
  • NELIO ALESSANDRO AZEVEDO CACHO
  • SAMUEL XAVIER DE SOUZA
  • Data: 02-sep-2020


  • Resumen Espectáculo
  • Separation of Concerns is a topic widely addressed in academia and industry. Finding ways to separate software concerns is the basis for reducing software system development costs. Although the software construction research field with the objective of obtaining a high degree of reuse is considered a relatively mature field - based on the abstraction of components - relevant changes in the software services scenario are occurring and open opportunities for new approaches related to question. The emergence of new software architectures, such as serverless computing, in the context of cloud computing, reinforces the need to think software construction possibilities based on this new scenario. It is in this context that Metadata Interpretation Driven Development (MIDD) is presented, a methodology whose purpose is to increase the degree of reuse of software artefacts that are built from their use. Its particularity, in relation to the methodologies currently employed, is that it requires a complete dissociation between the application code and the representation of the business domain, commonly put in that. That is, the application must be able to interpret the concepts of the business domain, and not implement them. As we will see, doing that, the same instances of software built with MIDD may provide software service to demands for a distinct business domain at the same time, without changing its code.

12
  • JOSÉ MARCOS LEAL BARBOSA FILHO
  • Wireless Wideband Channel Sounding: Techniques, Simulation and Data Post-processing

  • Líder : VICENTE ANGELO DE SOUSA JUNIOR
  • MIEMBROS DE LA BANCA :
  • LEONARDO HENRIQUE GONSIOROSKI FURTADO DA SILVA
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • MARCIO EDUARDO DA COSTA RODRIGUES
  • VICENTE ANGELO DE SOUSA JUNIOR
  • ÁLVARO AUGUSTO MACHADO DE MEDEIROS
  • Data: 11-sep-2020


  • Resumen Espectáculo
  • In recent years, the demand for high data rates in wireless communications has increased. The Enhanced Mobile Broadband (eMBB), one of the use cases of fifth generation of mobile communications systems (5G), requests 20 Gbps of minimum peak data throughput using carrier aggregation technique to achieve ultra-widebands. However, wideband and ultra-wideband signals are more susceptible to time and frequency spread degradation due to signal propagation over multiple paths and relative movement between transceivers, respectively. In this context, the channel sounding is of great relevance, since it allows to estimate the characteristics of the transmission medium, enabling the conception of solutions to combat the performance degradation imposed by the communication channel. From this perspective, this work develops a research on the main wideband channel sounding techniques, taking as main objective the conception in software of a channel sounding simulator and a post-processing and analysis tool for wireless wideband channel characterization. In addition, a comparative analysis of the direct cross-correlated soundings is performed, using 6 different sequences as sounding signal. Among them are the sequences PN, Frank-Zadoff-Chu and Golay.

13
  • FREDERICO AUGUSTO FERNANDES SILVEIRA
  • Smart-IoT: a DDoS protection system for the Internet of Things

  • Líder : LUIZ FELIPE DE QUEIROZ SILVEIRA
  • MIEMBROS DE LA BANCA :
  • AGOSTINHO DE MEDEIROS BRITO JUNIOR
  • EDMAR CANDEIA GURJÃO
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • RAMON DOS REIS FONTES
  • Data: 20-oct-2020


  • Resumen Espectáculo
  • The increase in the number of networked devices in the context of the Internet of Things (IoT) has driven the number of Distributed Denial of Service (DDoS) attacks in recent years. This threat takes advantage of the security limitations of these devices and their geographic locations to leverage the impact of the attacks. Developing mechanisms to detect and mitigate DDoS attacks in this new paradigm is a current challenge in network security. This work proposes a defense mechanism integrated into the IoT network controller that uses Machine Learning (AM) techniques to detect these attacks and the flexibility of Software-Defined Networks (SDN) for their mitigation. The proposed system uses random samples to perform network traffic classification and the OpenFlow (OF) protocol to apply real-time mitigation measures.  The solution was tested with four recent datasets in a controlled laboratory environment, showing to detect and mitigate DDoS attacks, with a high hit rate and low false alarm rate.

14
  • CAIO JOSÉ BORBA VILAR GUIMARÃES
  • Embedded Artificial Neural Networks Optimized for Low-cost and Low-Size-Memory Devices

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • BRAD MCDANEL
  • EDGARD DE FARIA CORREA
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • MARCELO AUGUSTO COSTA FERNANDES
  • Data: 22-oct-2020


  • Resumen Espectáculo
  • Artificial Neural Networks (ANNs) are bio-inspired systems with a high level of parallelization and almost infinite applications. However, due to the associated high computational power requirements, most application demands powerful processing characteristics and consequently, high-costs and not-so-small form-factors. This work presents an implementation of a Multilayer Perceptrons (MLPs) for 8-bit microcontrollers in two different scenarios, embedded training, and inference. Analysis of training convergence, inference time duration, and program code occupation into the internal memories and a technique to optimize this implementation to fit bigger MLP architectures. The aim of this work is to provide an overview of the feasibility of ANNs on these low-cost, low-size-memory devices, known as microcontrollers

15
  • DENIS RICARDO DA SILVA MEDEIROS
  • Proposal of Embedded Standalone and Distributed Genetic Algorithms for Low-Power, Low-Cost and Low-Size-Memory Devices

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • ALLAN DE MEDEIROS MARTINS
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • MARCELO AUGUSTO COSTA FERNANDES
  • PATRICIA DELLA MÉA PLENTZ
  • Data: 05-nov-2020


  • Resumen Espectáculo
  • This work proposes implementations of genetic algorithms targeting low-power, low-cost, and low-size-memory devices in two variants: a standalone version, to be used in one single device, and a distributed version, to be used in multiple devices simultaneously. The motivation is to adapt and optimize this important artificial intelligence (AI) technique so that it can be used in numerous applications where traditionally it was not feasible to be utilized, such as in microcontrollers. An investigation about how to optimize each separated segment of the algorithm is done and extensive details about both implementations are provided, including their source codes. Moreover, various experiments and simulations for multiple scenarios were done to validate their correct operation using Hardware-In-Loop technique, as well as to find out limitations for the algorithm parameters. The standalone implementation is compared to other work in the literature and it performs faster and using fewer resources. For the distributed version, an important analysis was done to verify that it can be faster than the standalone version but also more power-efficient when reducing the clock speed and voltage of those devices. Finally, this investigation could determine what is the threshold from where the distributed version, even with a high overhead due communication between the devices, become either faster or more power-efficient than the standalone one.

16
  • ANDREZA CRYSTINE CARVALHO DE OLIVEIRA
  • Computation of Robust Controlled Invariant Sets with Fixed Complexity Using Bilinear Optimization

  • Líder : CARLOS EDUARDO TRABUCO DOREA
  • MIEMBROS DE LA BANCA :
  • AMANDA DANIELLE OLIVEIRA DA SILVA DANTAS
  • ANDERSON LUIZ DE OLIVEIRA CAVALCANTI
  • CARLOS EDUARDO TRABUCO DOREA
  • JOSÉ MÁRIO ARAÚJO
  • Data: 06-nov-2020


  • Resumen Espectáculo
  • In this work, a methodology for the computation of robust controlled invariant polyhedra of fixed complexity, based on bilinear optimization, is proposed for discrete-time linear systems, subject to constraints on the states and control inputs and to bounded disturbances. In many real-life applications, certain variables in a system must comply with certain constraints. In general, these constraints are specified by linear inequalities that define limited and closed polyhedral sets. A set is robust controlled invariant if any state trajectory starting in this set can be maintained within it through a suitable control input, in spite of the disturbances. Thus, the calculation of a controlled invariant set is an important step in solving control problems under constraints. Conventional methods for robust invariant polyhedra computation can result in high complexity sets, defined by a large number of vertices. The methodology proposed in this work has bilinear invariance conditions and polyhedra represented by vertices, whose quantity is fixed in advance. The aim is also to maximize the volume of the robust controlled invariant polyhedra. Through numerical examples, the methodology is able to compute polyhedra with larger volumes than those obtained by recent methods which also seek for reduced complexity sets. In addition, a methodology is numerically efficient, applicable to larger systems than those treated by the methods available in the literature.

17
  • JOSÉ RAIMUNDO DANTAS NETO
  • Control Strategy Applied For Distributed Generatior Systems Through Virtual Synchronous Machine Concept.

  • Líder : RICARDO LUCIO DE ARAUJO RIBEIRO
  • MIEMBROS DE LA BANCA :
  • ALEXANDRE CUNHA OLIVEIRA
  • ALLAN DE MEDEIROS MARTINS
  • MÁRIO LÚCIO DA SILVA MARTINS
  • RICARDO LUCIO DE ARAUJO RIBEIRO
  • THIAGO DE OLIVEIRA ALVES ROCHA
  • Data: 17-dic-2020


  • Resumen Espectáculo
  • The increasing of Distributed Generation (DG) based on renewable energy sources (RES) has contributed to the reduction of several environmental, economic and social impacts caused by the burning of fossil fuels in the conventional generation system. However, the massive increase in the energy production based on FERs, also causes undesirable impacts, such as the decrease of the grid inertia, voltage oscillations and stability margins reduction. In addition, the Electrical Power System (EPS) has changed its configuration and the microgrid concept has gained interest in recent years. In this approach, more efficient and reliable control systems are required for proper EPS operation. Thereby, DG control strategies have been developed aiming, not only to provide the power from primary sources, but also provide voltage and frequency support to the power grid, increasing the stability margins and guaranteeing the energy quality. Thus, this work proposes a control system based on Virtual Synchronous Generator (VSG) approach, aiming to contribute to the stability of the electrical network, providing voltage support by regulating the power flow, using the P/ω and Q/V droop control technique. Frequency support is obtained by introducing a virtual inertia in the DG. The virtual inertia is emulated by means of transport delay techniques applied in the power loop. Unlike most works that use the VSG approach, a smooth start up scheme is developed, using the voltage vector estimation of the electrical grid, made by an SRF-PLL (Synchronous Reference Frame Phase Locked Loop). Simulation results are presented to validate the proposed method.

18
  • SILVAN FERREIRA DA SILVA JÚNIOR
  • Evaluating Human-Machine Translation with Attention Mechanisms for Industry 4.0 Environment SQL-Based Systems

  • Líder : ALLAN DE MEDEIROS MARTINS
  • MIEMBROS DE LA BANCA :
  • ALLAN DE MEDEIROS MARTINS
  • FRANCISCO DE ASSIS BRITO FILHO
  • GUSTAVO BEZERRA PAZ LEITAO
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • Data: 18-dic-2020


  • Resumen Espectáculo
  • The use of relational databases is increasingly present in the industry. Applications in medicine, IoT and Industry 4.0 are examples of this. Despite the great capacity and efficiency in data storage and retrieval, this type of database requires technical knowledge in specific query languages to access this information, which distances these types of applications from the non-specialized public. In this work, we propose an application of recent models in natural language processing that uses mechanisms to translate natural language into English into SQL applied to a database that stores sensor data, focusing on the concept of Industry 4.0. Paired examples of natural language phrases were generated with their corresponding SQL query to be used for training and validation. The model was agnostic in relation to the database schema, so that it only handles the input and output sequences regardless of the database structure. The data comes from historians of the typical process used in industrial settings. When training the deep neural network, we obtained a language model with an accuracy of approximately 99 \% in the validation set.

19
  • GABRIEL PEREIRA DE OLIVEIRA
  • Cooperative Control Applied for Power Flow Management of a DC Migrogrid Integrated with Energy Storage System

  • Líder : RICARDO LUCIO DE ARAUJO RIBEIRO
  • MIEMBROS DE LA BANCA :
  • CARLOS EDUARDO TRABUCO DOREA
  • EUZELI CIPRIANO DOS SANTOS JUNIOR
  • FLAVIO BEZERRA COSTA
  • RICARDO LUCIO DE ARAUJO RIBEIRO
  • THIAGO DE OLIVEIRA ALVES ROCHA
  • Data: 18-dic-2020


  • Resumen Espectáculo
  • The growing energy demand and the increasing social/political concern for the environment has driven the insertion of renewable energy sources (RESs) in the power system. These sourcers are inserted as distributed generation (DGs) systems in the new matrix energy model organized by microgrids (MGs). RESs are dependent on meteorological conditions which make it intermittent and stochastic. To provide the energy balance of these microgrids, it is necessary to integrate energy storage systems (ESSs). Therefore, this work proposes a cooperative control strategy to regulate the power flow of a DC MG with an interconnected ESS. The proposed strategy coordinates the voltage control of the DC bus by Interlink CC (ILC) and by ESS to provide power smoothing. To decouple the control loops from the bus, a Gaussian inference function is used based on the information of the voltage deviation on the DC bus. The proposed technique is validated through digital simulations.

20
  • JOÃO GUILHERME DOMINGOS DE OLIVEIRA
  • Analysis of Planar Structures for Applications in Sensing Systems
  • Líder : VALDEMIR PRAXEDES DA SILVA NETO
  • MIEMBROS DE LA BANCA :
  • VALDEMIR PRAXEDES DA SILVA NETO
  • ADAILDO GOMES D ASSUNCAO
  • FRED SIZENANDO ROSSITER PINHEIRO
  • ROSSANA MORENO SANTA CRUZ
  • Data: 23-dic-2020


  • Resumen Espectáculo
  • Two sensors for the characterization of the dielectric constant of different types of materials have been implemented and one of them applied in the determination of the soil water content (θ), while the other is a bio inspired antenna based on the Shiso leaf. The proposed sensors consist of in: The shape of the antenna patch is inspired by the Shiso leaf, whose scientific name is Perilla Frutescens. A complementary interdigital capacitor (CIDC) is inserted into the geometry; A microstrip antenna with a circular patch in a container printed on ABS filament, based on the complementary split-ring resonator (CSRR) structure. The operating principle is based on the displacement of the element’s resonance frequencies when the relative permittivity of the material under test (MUT) is changed. Several simulations were performed, with the data obtained, two empirical models are proposed. To validate its effectiveness, measurements were made with dielectric substrates that have the dielectric constant known in the literature. The second sensor can be applied to different types of materials, in this work it was used to determine the percentage of water contained in different types of soils. The prototypes were built and the results obtained in the measurements were compared with the results of other studies, thus validating the effectiveness of the proposed sensors.

21
  • JOSE GARIBALDI DUARTE JUNIOR
  • Study of a Reconfiguration Methodology Applied to the Development of Adjustable Microwave Planar Filters
  • Líder : VALDEMIR PRAXEDES DA SILVA NETO
  • MIEMBROS DE LA BANCA :
  • ADAILDO GOMES D ASSUNCAO
  • FRANCISCO DE ASSIS BRITO FILHO
  • MURILO ARAUJO ROMERO
  • VALDEMIR PRAXEDES DA SILVA NETO
  • Data: 28-dic-2020


  • Resumen Espectáculo
  • Filters play an important role in many radio frequency (RF) and microwave applications. They are used to separate or combine different frequency bands. The electromagnetic spectrum is limited and needs to be shared, so the filters have the function of limiting and / or selecting the microwave / RF signals within the assigned spectral limits. The reconfigurable, or tunable, microwave filters have attracted a lot of attention from researchers and designers due to their emerging characteristics of improving spectral capacity and performance in current and new communication systems. This work presents a study about reconfigurable microwave planar filters. A reconfiguration methodology based on the use of semiconductor devices of the varactor diode type is proposed in order to allow the adjustment of impedances along the geometry of the filter resonator, and thus, configure the frequency response of the filtering device. The designs of three reconfigurable devices are presented, being applicable to the communication standards 5G NR sub-6 GHz, with the model 01 being a band-reject filter, and the models 02 and 03, band-pass filters. Analyzes of the equivalent circuits of the filters are developed in order to establish an understanding of their respective performances depending on the geometry and the adjustment device. The initial analyzes and simulations were performed using the equivalent circuit method and the finite element method using the Ansys HFSS software, and the equivalent model of the varactor diode was also used according to its datasheet. Preliminary results are presented, which include the response obtained through simulation and measurement of the models without the use of varactors and results with the application of the reconfiguration methodology based on the varactors. Finally, the next fundamental steps for the continuation of the research are pointed out.

Tesis
1
  • RAMON AUGUSTO SOUSA LINS
  • Deep reinforcement learning one new perspective on the k-server problem

  • Líder : ADRIAO DUARTE DORIA NETO
  • MIEMBROS DE LA BANCA :
  • ADRIAO DUARTE DORIA NETO
  • FRANCISCO CHAGAS DE LIMA JUNIOR
  • GUILHERME DE ALENCAR BARRETO
  • JORGE DANTAS DE MELO
  • MARCELO AUGUSTO COSTA FERNANDES
  • SAMUEL XAVIER DE SOUZA
  • Data: 28-ene-2020


  • Resumen Espectáculo
  • The k-server problem in a weighted graph (or metric space) is defined by the need to efficiently move k servers to fulfill a sequence of requests that arise online at each graph node. This is perhaps the most influential online computation problem whose solution remains open, serving as an abstraction for a variety of applications, as buying and selling of currencies, reassign processes in a parallel processing for load balancing, online transportation service, probe management of oil production rigs, among others. Its conceptual simplicity contrasts with its computational complexity that grows exponentially with the increasing number of nodes and servers. Prior to this work, the Q-learning algorithm was used to solve small instances of the k-server problem. The solution was restricted to small dimensions of the problem because its storage structure grows exponentially with the increase in the number of nodes and servers. This problem, known as the curse of dimensionality, makes the algorithm inefficient or even impossible to execute for certain instances of the problem. To handle with larger dimensions, Q-learning together with the greedy algorithm were applied to a small number of nodes separated into different clusters (hierarchical approach). The local policy obtained from each cluster, together with
    greedy policy, were used to form a global policy satisfactorily addressing large instances of the problem. The results were compared to important algorithms in the literature, as the Work function, Harmonic and greedy. The solutions proposed so far emphasize the increase in the number of nodes, but if we analyze the growth of the storage structure defined by Cn;k ' O(nk) It can be seen that the increase in the number of servers can be quickly limited by the problem of the curse of dimensionality. To circumvent this barrier, the k-server problem was modeled as a deep reinforcement learning task whose state-action value function was defined by a multilayer perceptron neural network capable of extracting environmental information from images that encode the dynamics of the problem. The applicability of the proposed algorithm was illustrated in a case study in which different problem configurations were considered. The behavior of the agents was analyzed during the training phase and their performance was evaluated from performance tests that quantified the quality of the displacement policies of the servers generated. The results provide a promising insight into its use as an alternative solution to the k-servers problem.

2
  • WILLIANS RIBEIRO MENDES
  • Remote sensing data integration applied to variable rate irrigation systems by using fuzzy decision support systems

  • Líder : FABIO MENEGHETTI UGULINO DE ARAUJO
  • MIEMBROS DE LA BANCA :
  • PAULO ESTEVÃO CRUVINEL
  • ANDRE LAURINDO MAITELLI
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • JOSENALDE BARBOSA DE OLIVEIRA
  • RAFAEL VIEIRA DE SOUSA
  • Data: 07-feb-2020


  • Resumen Espectáculo
  • Freshwater is a scarce resource, and nowadays it is under unprecedented focus due to concerns about the sustainability of natural ecosystems and because of the high agricultural demand, which needs to feed a growing global population. These concerns make freshwater a necessary natural resource for everyone. All these issues lead us to want to manage this asset more efficiently, thus generating an increase in the demand for more intelligent, more automated and more accurate systems. These needs are the driving force behind the development and implementation of new technologies for more rational use, allowing farmers to improve when and where irrigation is applied, seeking to increase food production. In this manner, this work presents a new proposal for an intelligent variable rate irrigation system application in order to achieve that goal. It will be a site-specific irrigation management tool; in other words, a system that should assist in decisions about applying water quantity to delimited zones. We expect a higher yield from crop productivity and greater efficiency in water use. To be successful, the system has an artificial intelligence approach to automatically create optimal control maps for a center pivot irrigation system. At the heart of this system there will be the fuzzy logic, which will be able to set the reference values for the rotating speed controllers and individual opening of each central pivot sprinkler valve. The proposed system will be based on the decision making (whether applying more or less water) and will use remote sensing data, so we expect the smart irrigation system to describe the spatial variability of the crop efficiently. The results point out that the edaphoclimatic variables, when well combined with fuzzy logic, can solve uncertainties and non-linearities of an irrigation system and define a control model for high precision irrigation. However, it will not always be possible to reduce water consumption, but this technology has many uses to increase farm profitability.

3
  • ROBERTO DOUGLAS DA COSTA
  • Classification of system-based intelligent learning styles: a case study in technology-mediated education

  • Líder : RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • MIEMBROS DE LA BANCA :
  • ALINE DE PINHO DIAS
  • AQUILES MEDEIROS FILGUEIRA BURLAMAQUI
  • ELOIZA DA SILVA GOMES OLIVEIRA
  • GUSTAVO FONTOURA DE SOUZA
  • JOÃO PAULO QUEIROZ DOS SANTOS
  • RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • Data: 14-feb-2020


  • Resumen Espectáculo
  • With the growth of distance education (EAD) in Brazil in its levels of education, be it medium, technical or superior and the increase in the offer of online courses made available through the Virtual Learning Environments (AVA), there is a concern with the development Of this student during the course and their learning. In order to associate the resources and activities offered in AVA with the Learning Styles classes, this work proposes to create an algorithm that can identify the behavior patterns of the EAD students through the use of tools available in the AVA and associates them with a classification of Learning Styles in order to assist the teacher in the preparation of his / her lesson plans, suggesting resources and evaluation activities arranged in the AVA that will favor the learning of its students according to the Learning Style characteristic of each one of them.

4
  • JOSÉ CLÁUDIO VIEIRA E SILVA JUNIOR
  • Hardware Strategies Applied to the Latency Reduction on Tactile Internet

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • LUCAS M. OLIVEIRA
  • DENNIS BARRIOS ARANIBAR
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • JOSE ALBERTO NICOLAU DE OLIVEIRA
  • MARCELO AUGUSTO COSTA FERNANDES
  • Data: 28-feb-2020


  • Resumen Espectáculo
  • This work proposes to present hardware strategies applied to reduce latency in the tactile internet. The motivation is to study the challenges contained in the development of the hardware associated with the tactile devices, especially issues related to the round trip latency limit of the system components. As is known, for a tactile internet environment to work desirably, it is necessary to respect a minimum limit of round trip latency. Since some tactile applications allow some human senses to interact with the machines remotely, this means that, almost always, the minimum limit of round trip latency has a time delay in the range of milliseconds. Thus, it is clear that there is a demand for tactile devices that are quite fast. In this context, three hardware proposals are presented that have the main objective to reduce the total latency produced by this type of device. The first strategy proposed for the development of hardware is to use reconfigurable computing (on FPGA) to minimize the execution time of the algorithms associated with the device. The second hardware proposal also makes use of reconfigurable computing (on FPGA). However, the hardware is designed using another type of numerical representation. Finally, the third proposal presents a tactile glove model implemented using a variety of micro processed system. Results associated with the three proposals are presented and show the viability of the strategies, presenting better performance concerning the works that were compared.

5
  • PEDRO THIAGO VALERIO DE SOUZA
  • Spectral Sensing and Signal Detection in MIMO Systems with Impulsive Noise and Multiple Path Fading
  • Líder : LUIZ FELIPE DE QUEIROZ SILVEIRA
  • MIEMBROS DE LA BANCA :
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • VICENTE ANGELO DE SOUSA JUNIOR
  • Waslon Terllizzie Araujo Lopes.
  • ALUISIO IGOR REGO FONTES
  • GUILHERME DE ALENCAR BARRETO
  • Data: 09-mar-2020


  • Resumen Espectáculo
  • In mobile systems, information signals must be transmitted with high rates of data transmission and with high reliability. One of the possible solutions to achieve these desirable requirements is the use of systems with multiple transmitting and/or receiving antennas, featuring a MIMO system (Multiple Input-Multiple Output). However, signal processing techniques in MIMO systems depend on the Gaussianity of the transmission channel, decreasing their efficiency in non-Gaussian communication scenarios. To goal a robustness of this transmission technique, it will become imperative or to study this transmission technique in scenarios in which the channel is not exhibited by Gaussianity. In this work, a new signal detection technique in MIMO systems is proposed in scenarios characterized by non-Gaussian noise. A proposed reception technique is called Maximal Correntropy Detector (MCD) and it has been experimentally proven to treat a generalization of the MLD detector (Maximum Likelihood Detector) using the use of the correntropy complex. The MCD detector is analyzed on Gaussian and non-Gaussian channels and its performance is superior to classic detectors, without a significant increase in computational complexity. Together, this work also presents a new method of spectral sensing suitable for detecting signals that show non-linear phase variations over time. The proposed method is based on the use of theory of cyclostationary signals in waiting time, or which transformation employs signals to be a sensor in order to mitigate or effect non-linear variation of the phase. A proposed architecture is evaluated without a BPSK signal sensor and is compared to a temporal cyclostationary sensing technique. The results of the simulation tested prove the efficiency of the architecture proposal, presenting detection rates of primary users with an order of 8 dB.

6
  • FRANCISCO ARY ALVES DE SOUZA
  • Application of DGS in H Format for Suppression and Attenuation of Higher Order Modes in Microfita Antennas

  • Líder : ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • MIEMBROS DE LA BANCA :
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • JOSE PATROCINIO DA SILVA
  • MARCIO EDUARDO DA COSTA RODRIGUES
  • ALEXANDRE JEAN RENE SERRES
  • ALFREDO GOMES NETO
  • Data: 10-jun-2020


  • Resumen Espectáculo
  • This thesis aims to characterize the DGS H filter to attenuate higher-order modes in microstrip antennas. The DGS H is designed independently of the antenna. The initial dimensions of H will be calculated using the desired attenuation range. After design and analysis of the filter response, the DGS H will be integrated into the antenna. The integration allows to eliminate filters between the signal source and the antenna. The reduction of insertion losses is the main advantage among others. The H filter is investigated through the frequency response. The dimensions of H (weights and bar) were parameterized from the initial dimension proposed here, which makes it possible to control the rejection band from the modes to be attenuated. This thesis presents the characterization of the DGS filter in H format for applications in microstrip antennas as well as a method to calculate the initial dimension of H. Two antennas were designed to prove the effectiveness of the proposal. These antennas are designed for the S and C bands (2.45 and 5.80 GHz). The results show the attenuation of the higher-order modes for the Microstrip Antennas studied in this thesis. The simulated and measured results are presented and discussed. The excited modes are investigated through the modal fields and frequency response. Prototype measurements were performed using a vector network analyzer (VNA) and an anechoic camera.

7
  • TÚLIO DE PAIVA MARQUES CARVALHO
  • Sabiá: Integrated Architecture for Authentication and Data Authorization Oriented to User Consent for Health Learning Ecosystems in Brazil
  • Líder : RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • MIEMBROS DE LA BANCA :
  • RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • KARILANY DANTAS COUTINHO
  • CYNTHIA ROSAS MAGALLANES
  • CRISTINE MARTINS GOMES DE GUSMÃO
  • JAILTON CARLOS DE PAIVA
  • Data: 17-jul-2020


  • Resumen Espectáculo
  • Health information systems in Brazil have been designed and developed in a heterogeneous manner based on local regional characteristics, resulting in a lack of health information integrity. In this context, the Brazilian Ministry of Health pointed out the need for interoperability solutions of health information systems, noting the importance of integration with national databases and alignment with Brazilian data protection laws as well its application in education to aid with continuing education for health professionals. Therefore, this work presents Sabiá, a platform for authentication, authorization and data delivery based on user consent for health information systems in Brazil and currently applied in the context of health educational ecosystems. Sabiá's architecture is designed to achieve the following requirements: R1) Provide a Federated Identity; R2) Be a Federated Resource Manager; R3) Collect user data from different information systems and; R4) Deliver user data to systems based on user consent. Sabiá consists of three main components: 1) Sabiá Authorization Server, responsible for implementing Open Authentication; 2) Sabiá Collector, responsible for collecting data from different information systems and; 3) Sabiá Resource Server, responsible for delivering data previously authorized by the user to the systems. After analyzing historical data, R4 functionality was selected to be submitted to performance testing because it is the process that most affects overall system performance. The tests aimed at analyzing Sabiá's behavior in the heaviest scenario based on historical data. The results showed no flaws and indicated system stability and consistency, in which the user perceives a system reaction instantaneous, whose response time averages remained below 100 ms.

8
  • MÔNICA MARIA LEAL
  • Wavelet-Based Directional Protection Supported by Wavelet-Based Fault Classifier

  • Líder : FLAVIO BEZERRA COSTA
  • MIEMBROS DE LA BANCA :
  • FLAVIO BEZERRA COSTA
  • RICARDO LUCIO DE ARAUJO RIBEIRO
  • FELIPE VIGOLVINO LOPES
  • JOSÉ CARLOS DE MELO VIEIRA JUNIOR
  • KAI STRUNZ
  • Data: 10-ago-2020


  • Resumen Espectáculo
  • Fast and accurate protections are required to ensure the safety of the power system. In
    this research, a new directional protection based on wavelet transfom is described by using
    only the first wavelet decomposition level. The scaling coefficients are used in replacement of
    the Fourier transform in order to recreate the conventional directional protection. The torque
    equations were mathematically redefined in the wavelet domain by using sampled voltages and
    currents. Regarding the protection activation, the wavelet coefficients are used for fast detection of fault-induced transients (wavelet activators) in order to reduce the relay operating time,
    whereas the scaling coefficients are used for backup activation in accordance with the overcurrent protection (wavelet overcurrent activators). In addition, a real-time fault classification
    method using the RT-BSWT applied to modal components of the Clarke transform was developed to support the directional protection. The method uses a simple threshold-based logic
    flow instead of a sophisticated artificial intelligence-based algorithm, and requires three-phase
    current measurements from only one terminal. Furthermore, by using the wavelet coefficients
    energy of RT-BSWT, the method is fast and accurate due to the extraction of both low- and
    high-frequency components of faults. Evaluations have proved the feasibility to rebuild the
    conventional directional protection by using the RT-SWT, which has provided better performance and faster operating time than the conventional protection based on the discrete Fourier
    transform (DFT). As new functionalities to the directional protection, the wavelet-based negative sequence unit can be used to identify the fault directionality to three-phase faults, even with
    severe voltage sags without memory strategies, which is not possible with the conventional protection. The fault classifier results have shown it is accurate and fast to identify all of the ten
    fault types successfully supporting the directional protection needs


9
  • DAVI HENRIQUE DOS SANTOS
  • Methodology for comparing autonomous sailboat control systems

  • Líder : LUIZ MARCOS GARCIA GONCALVES
  • MIEMBROS DE LA BANCA :
  • LUIZ MARCOS GARCIA GONCALVES
  • PABLO JAVIER ALSINA
  • ALEXANDRE DE MORAIS AMORY
  • ESTEBAN WALTER GONZALEZ CLUA
  • JOAO MORENO VILAS BOAS DE SOUZA SILVA
  • Data: 04-sep-2020


  • Resumen Espectáculo
  • In this work we propose a methodology for estimating the performance and comparing control systems for autonomous sailboats. One of the biggest problem in mobile robotics research is the absence of standard tools and methodology for distribution/verification of implementations and the performance of the employed techiniques. Our approach consists of an framework for performance evaluation divided in three components: scope, metrics and scenarios. The scope define the parts of the system that are comparable and also presents a compreesive approach to define the basis navigation system for autonomous robot. The scenario represent situations that happen during sailboat missions such as keep the sailboat close to a desired waypoint (virutal anchoring), docking and tracking. Quantitative metrics are extracted by experiments in the scenario and are used to evaluate characteristics expected from control systems such as stability, energy consumption, robustness and efficiency, resulting in an initial classification of the control system. An initial implementation of the system is made using the Ardupilot framework, where some control techinique and metrics were implemented for first analisys. The level of analysis can be gradually improved by adding more control techinique and metrics.

10
  • LEONARDO ALVES DIAS
  • Parallel Implementation proposal of Clustering Algorithms in Hardware

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • MARCELO AUGUSTO COSTA FERNANDES
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • DANIEL SABINO AMORIM DE ARAUJO
  • NANDOR VERBA
  • CARLOS ALBERTO VALDERRAMA SAKUYAMA
  • Data: 05-oct-2020


  • Resumen Espectáculo
  • This work presents a study on data clustering algorithms implemented in dedicated hardware for applications in general, aiming to increase the processing speed. Clustering algorithms have been widely adopted to find patterns between data in different areas. However, these algorithms usually imply high processing complexity and, in addition, the amount of data currently stored is massive. Therefore, the need for high-throughput data processing has become even more critical, especially for real-time applications. One solution that has been adopted to increase processing speed is the use of parallel techniques implemented on dedicated hardware, which has proved to be more efficient compared to sequential systems. Therefore, this work proposes the fully parallel implementation of data clustering algorithms in hardware to optimize the processing time of systems in several areas, enabling applications for systems with a massive amount of data. A new proposal for implementations of the clustering algorithms K-means and Self-Organizing Maps are presented, together with an analysis of the results related to throughput and the hardware resource for different parameters. The implementations presented here point to a new direction associated with the implementation of clustering algorithms and can be used in other algorithms.

11
  • SÂMARA DE CAVALCANTE PAIVA
  • Wavelet-Based Systems for Frequency Monitoring and Islanding Detection of AC Microgrids.

  • Líder : RICARDO LUCIO DE ARAUJO RIBEIRO
  • MIEMBROS DE LA BANCA :
  • FLAVIO BEZERRA COSTA
  • JOSEP MARIA GUERRERO ZAPATA
  • KLEBER MELO E SILVA
  • LEANDRO MICHELS
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • RICARDO LUCIO DE ARAUJO RIBEIRO
  • Data: 13-oct-2020


  • Resumen Espectáculo
  • The modern distribution system widely uses renewable energy sources (RESs) as a distributed generation that introduces problems like unintentional islanding, protection concerns, and reverse power flow. The RES intermittent characteristics can cause frequency deviations that could result in systems instabilities. In general, hierarchical structures implement these microgrids' power flow control management, composed of primary, secondary, and tertiary control layers. The primary control layer consists of inner current and voltage control loops, typically implemented via droop control loops, for the purpose of maintaining a stabilization of frequency and voltage amplitudes, as well as power-sharing. The principal drawbacks of droop control approaches are steady-state frequency and voltage deviations. For that reason, the secondary control layer is introduced to mitigate these deviations. The accurate knowledge of microgrid frequency deviations permits the implementation of proper frequency restoration. Besides, those microgrids could operate in the connected or islanding modes. Islanding or fault occurrences must be detected and treated to ensure system stability. This way, the continuous wavelet transform is employed for developing a hybrid islanding detection and a frequency monitor to fulfill the microgrid's requirements related to a secure islanding detection and a frequency restoration. In this Thesis, both systems' development is carried out, with the effectiveness of proposed methods evaluated and validated through experimental results.

12
  • MAXIMILIANO ARAÚJO DA SILVA LOPES
  • t-SNE parallel: A parallel technique for data dimensionality reduction applied in Smart Cities

  • Líder : ADRIAO DUARTE DORIA NETO
  • MIEMBROS DE LA BANCA :
  • ADRIAO DUARTE DORIA NETO
  • ALLAN DE MEDEIROS MARTINS
  • ALUISIO IGOR REGO FONTES
  • CICILIA RAQUEL MAIA LEITE
  • DANIEL SABINO AMORIM DE ARAUJO
  • Data: 16-oct-2020


  • Resumen Espectáculo
  • In recent years, the urban population has been growing rapidly around the world. To adapt to this population increase, mayors need to change the way they manage these large cities. Thus, the concept of Smart Cities gains strength and comes to change the way of life of the world population. The investment in this concept of smart cities is aimed at improving people’s management and quality of life. The biggest challenges in these systems are linked to the processing, visualization and analysis of the generated data, since when they work connected these systems generate a large mass of data, called Big Data, which need to be treated differently from conventional systems. For the visualization of the data, a device that can be used are the techniques for reducing the dimensionality, which bring the data from one n dimension to two or three dimensions, thus being perceptible to human eyes. One of the main problems involving Dimensionality Reduction techniques is related to the processing time, which makes them practically unfeasible to be applied to large masses of data. In this thesis, a way to decrease the processing time
    of these algorithms is suggested, by parallelizing the t-SNE algorithm. An analysis was performed on each part of the algorithm, verifying which sections could be parallelized and which sections would continue with their conventional processing. In this way, the parallelized algorithm showed better results than its conventional version, presenting itself as a more efficient and effective technique in Reducing the Dimensionality of data in order to optimize their visualization and analysis.

13
  • TIAGO ALVES DE ALMEIDA
  • Output Feedback Regulation and Constant Reference Tracking with Disturbance Rejection for Constrained Linear Systems via Controlled-Invariant Sets

  • Líder : CARLOS EDUARDO TRABUCO DOREA
  • MIEMBROS DE LA BANCA :
  • ANDRÉ FELIPE OLIVEIRA DE AZEVEDO DANTAS
  • CARLOS EDUARDO TRABUCO DOREA
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • JOÃO MANOEL GOMES DA SILVA JUNIOR
  • KURIOS IURI PINHEIRO DE MELO QUEIROZ
  • ROBERTO KAWAKAMI HARROP GALVAO
  • Data: 16-oct-2020


  • Resumen Espectáculo
  • This work is concerned with design of output feedback controllers for constrained linear discrete-time systems via set-invariance techniques. In this regard, Output-Feedback Controlled-Invariant (OFCI) polyhedra are used to ensure that state and input constraints are satisfied all time even in the presence of additive disturbances and measurement noise. Necessary and sufficient conditions for a polyhedral set to be OFCI are presented, which can be checked by the solution of a set of Linear Programming (LP) problems. Then, a dynamic output-feedback compensator (possibly nonlinear) is proposed, through the construction of an OFCI set, from a pair composed by a conditioned-invariant and a controlled-invariant polyhedron. Based on the available measurements and on the state of the compensator, which constitutes an estimate of the system state, a suitable control sequence can be computed to enforce the constraints. The uncertainty on the state is progressively reduced using information about the contraction of the conditioned-invariant set. An LP problem is formulated to compute a control action that enforces state and control constraints and minimizes, one step ahead, a guaranteed distance from the admissible states to the origin. The problem of tracking a constant reference signal in the presence of constant disturbances is also considered for which the conception of the tracking controller is motivated from the stabilizing controller. With the current approach, as illustrated through numerical examples, by embedding the estimator in the compensator structure and using the OFCI concept, it is possible to obtain solutions with larger sets
    of admissible initial states and admissible initial estimation errors, compared to other approaches available in the literature.

14
  • THIAGO MEDEIROS BARROS
  • A Data-Oriented Process for Generation of School Dropout Prediction Model


  • Líder : LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • MIEMBROS DE LA BANCA :
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • BETANIA LEITE RAMALHO
  • RAYMUNDO CARLOS MACHADO FERREIRA FILHO
  • PLACIDO ANTONIO DE SOUZA NETO
  • Data: 22-oct-2020


  • Resumen Espectáculo
  • School dropout is an extremely complex problem, as it involves not only a variety of perspectives, but also a variety of different types of dropout behavior.  Historically, the most cited school dropout models had their origin in education, however the emerging area of Data Science applied in Education is capable of developing new predictive models, with generally better results when compared to the most used traditional statistical methods.   The main objective of this thesis is the proposition of a Process for the generation of a Predictive School Dropout Model based on Data Sciences.  To this end, a sequence of steps is defined in order to model an information flow from the definition of the problem to the generation of useful information for managers and teachers. The steps consist of:  Understanding the Problem, Understanding the Data, Feature Engineering, Feature Selection, Data Balancing, Models, Evaluation and Interpretation.  The proposal’s contribution is found in the indication of which techniques and algorithms should be used in each phase of knowledge discovery, and show that the phenomenon of school dropout must be addressed as a problem of imbalanced classes, which must use tools and appropriate metrics, in order to generate a robust and easy to interpret prediction model. The proposed process was validated on educational and socioeconomic data of students at  Federal Institute of Rio Grande do Norte (IFRN).

15
  • TALES VINÍCIUS RODRIGUES DE OLIVEIRA CÂMARA
  • Automatic Modulation Classification in Impulsive Environments Based on Cyclostationary Features

  • Líder : LUIZ FELIPE DE QUEIROZ SILVEIRA
  • MIEMBROS DE LA BANCA :
  • WAMBERTO JOSÉ LIRA DE QUEIROZ
  • JOILSON BATISTA DE ALMEIDA REGO
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • MARCELO AUGUSTO COSTA FERNANDES
  • PEDRO THIAGO VALERIO DE SOUZA
  • Data: 23-oct-2020


  • Resumen Espectáculo
  • The rapid growth of applications supported by wireless communications systems drives the search for new communications systems that allow to efficiently explore the frequency spectrum, such as systems based on cognitive radio. The cognitive radio can be defined as an intelligent communications system, capable of adapting autonomously to the communication channel, through the reconfiguration of its operating parameters. An important property of cognitive radios is the ability to automatically recognize the type of modulation employed in an RF signal, thus enabling interoperability between systems, improving spectral efficiency, or even enabling electronic surveillance (in military application contexts) ). This attribute is known as automatic modulation classification (AMC). Among the AMC techniques that characterize the state-of-the-art, are those that are based on the detection of patterns obtained from the analysis of second order cycle stationary characteristics. Although very widespread, these techniques are unable to recognize some types of digital modulations, such as high-order M-QAM and M-PSK modulations. On the other hand, the higher order cycle stationary analysis techniques, used to extract singular descriptors of these modulations, have a very high computational cost and are only suitable for communication environments with AWGN noise. Although the AWGN noise model is widely used to characterize wireless communication channels, there are several practical scenarios that are better modeled by non-Gaussian distributions, such as HF communication, whose environment presents a strong contamination by impulsive noise. Recently, two new cyclostationary analysis functions, the lower order fractional cyclic autocorrelation function (FLOCAF), and the cyclic current correlation function (CCF), were defined and evaluated for the purpose of spectral sensing in impulsive environments, being the spectral sensing a less complex problem in relation to the automatic classification of modulations. In fact, knowing that there was no satisfactory solution in the literature for the automatic classification of high-order modulations in channels with impulsive noise, this problem was addressed in this work. In this work, automatic modulation classification architectures are developed based on the FLOCAF and CCF cyclostationary functions, combined with decision tree classification and logistic regression techniques. The architectures were developed for the recognition of BPSK, QPSK, 8-QAM, 16-QAM, and 32-QAM digital modulations, and evaluated in different contexts of alpha-stable additive noise contamination. The results showed that all architectures were able to operate in impulsive environments, however, architectures based on CCF were the most efficient.

16
  • ADELSON MENEZES LIMA
  • Application and Analysis of Metamaterial Cells Arrays in the Microfita Antennas Design

  • Líder : JOSE PATROCINIO DA SILVA
  • MIEMBROS DE LA BANCA :
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • FRANCISCO DAS CHAGAS BARBOSA DE SENA
  • HUMBERTO DIONISIO DE ANDRADE
  • JOSE PATROCINIO DA SILVA
  • VICENTE ANGELO DE SOUSA JUNIOR
  • Data: 23-oct-2020


  • Resumen Espectáculo
  • Obtaining and integrating solutions for technological applications in the communications area has led researchers to investigate various types of materials, with emphasis on this case for the case of metamaterials in the construction of microwave circuits. On the other hand, due to the ease of construction, microstrip antennas have used metamaterials that are obtained artificially by allowing changes in their electromagnetic properties. In this context, the present work proposes a metamaterial cell for the development of the configuration of periodic arrangements, immersed in the dielectric substrate, in the construction of the microstrip antennas. For the design of the cell and microstrip antennas, the operating frequency of 5.8 GHz was used. Simulations with the HFSS® software that uses Finite Elements as a numerical method, were performed to analyze the electromagnetic characteristics of the cell and the parameters of the antennas. Numerical results performed with the cell understudy, has a strong influence of magnetic permeability on metamaterial properties, at the desired frequency. The first case study showed greater bandwidth and better impedance matching. In the second, despite obtaining a greater bandwidth, a decrease in the resonance frequency was observed, and a change in the impedance matching when the cells were rotated. To validate the results, prototypes were built for the first case under analysis and the measured results were compared to those obtained numerically, showing a good agreement. In the second study, only simulated results were obtained, considering the angle of rotation of the cells, ranging from 0 to 90 degrees with intervals of 15 degrees. For all the studied structures, the parameters analyzed were the reflection coefficient (S11), bandwidth, Smith chart, radiation diagram (2D and 3D), gain, electric field distribution, and surface current density.

17
  • ARTHUR SALGADO DE MEDEIROS
  • Direct Control of the Rotor Voltage of a Doubly Fed Induction Generator by means of a Restricted Optimization Process in Real Time

  • Líder : MANOEL FIRMINO DE MEDEIROS JUNIOR
  • MIEMBROS DE LA BANCA :
  • JOSE TAVARES DE OLIVEIRA
  • MANOEL FIRMINO DE MEDEIROS JUNIOR
  • MARCUS VINICIUS ALVES NUNES
  • RICARDO FERREIRA PINHEIRO
  • THALES AUGUSTO DE OLIVEIRA RAMOS
  • UBIRATAN HOLANDA BEZERRA
  • Data: 30-oct-2020


  • Resumen Espectáculo
  • The constant growth in the potential for generating energy from wind sources has been highlighted on the world stage. Global behavioral changes in search of sustainable development and consumption, replacing fossil fuels aiming at improving the quality of life and preserving the planet, are important factors for the continuation of this growth, due to the increasing need for renewable energy sources. The wind energy conversion systems that use the Doubly Fed Induction Generator - DFIG, are the most popular today. Advantages such as reduced mechanical load, simple picth control, control of active and reactive power, lower power converters that mean less switching losses, reduced converter costs, less harmonic injection into the network, put it ahead of turbine topologies with synchronous generators or induction generators with squirrel cage rotors. This thesis aims to present and demonstrate the feasibility of using a new approach on voltage control, active and reactive power control and power factor control through direct control of the voltage applied to the DFIG’s rotor, by the Rotor Side Converter - RSC. This voltage is determined analytically, solving a restricted optimization process in real time. A system of equations derived from the equivalent circuit is adopted to represent the desired operating point of the DFIG, and the solution of this system defines the values of the rotor supply voltage coordinates, in steady state. A new control strategy is proposed to reach the rotor voltage reference values, without violating the limits existing in other variables. This strategy is designed so that the speed of evolution of the voltage applied to the rotor, is governed by the dynamic evolution of the mechanical speed of the rotor. An optimization process was formulated to minimize the time of convergence of the rotor speed, restricting transitory variations in the net power generated, in order to accelerate the machine, without exceeding the current limits. Following recent trends in solutions with a reduced number of sensors, only measurements obtained from the stator sensors (voltages and currents) are used. In this way, the angular velocity and rotor currents are estimated in real time. An algorithm for estimating inductance is also included, preventing deviations from the nominal value could lead to false reference voltages or changes in reactive control. In addition, a method for defining the pitch angle and the reference speed is proposed, using the Newton-Raphson numerical solution method.

18
  • PETRÚCIO RICARDO TAVARES DE MEDEIROS
  • Visual Stimulus Detection Using Multiple Foveas

  • Líder : RAFAEL BESERRA GOMES
  • MIEMBROS DE LA BANCA :
  • BRUNO MOTTA DE CARVALHO
  • ESTEBAN WALTER GONZALEZ CLUA
  • LUIZ MARCOS GARCIA GONCALVES
  • PABLO JAVIER ALSINA
  • RAFAEL BESERRA GOMES
  • ROBERTO MARCONDES CESAR JUNIOR
  • Data: 30-oct-2020


  • Resumen Espectáculo
  • The multifoving technique allows the addition of several focuses in the image, which can be explored as points of visual attention in contexts of detection, identification and / or object recognition. However, the use of the multifoving technique requires knowledge of the position of visual stimuli. In this work we propose a new approach to detect visual stimuli using the structure of multiple foveas. For this, we use mathematical strategies adapted to the context of computational vision, which take into account the distribution of the foveas to estimate the location of visual stimuli in the image. The mathematical strategies adopted were the descent of the gradient (potential field), maximum likelihood, multilateration, trilateration and barycentric coordinates. The results show that the algorithms converge for the position of the visual stimulus, with the exception of the local potential intersection algorithm due to the sensitivity to local minimums. In addition, the algorithms that use potential fields require more processing time and computational resources compared to other strategies. However, it is possible to affirm that three fóveas are sufficient to estimate the position of a visual stimulus in the image making use of the trilateration algorithms and baricentric coordinates. We conclude that the multi-panning associated with mathematical strategies can be applied in visual detection and presents convergence with at least three foveae.

19
  • REUBER REGIS DE MELO
  • RFID in IoUT Using a Dielectric Attenuation Structure for Application as Moisture Sensor


  • Líder : JOSE PATROCINIO DA SILVA
  • MIEMBROS DE LA BANCA :
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • IDALMIR DE SOUZA QUEIROZ JÚNIOR
  • JOSE PATROCINIO DA SILVA
  • VICENTE ANGELO DE SOUSA JUNIOR
  • ÁDLLER DE OLIVEIRA GUIMARÃES
  • Data: 30-oct-2020


  • Resumen Espectáculo
  • Advances in Internet of Things (IoT) applications have provided the development of new paradigms in this area, such as the Internet of Underground Things (IoUT). Research shows the potential of IoUT systems, mainly in precision agriculture, in which data it’s collected by different devices interconnected in a communication network. These data can facilitate decision-making about the management of agricultural activities. An important component in the architecture of IoUT is Underground Things (UT) responsible for sensing soil parameters, such as temperature and moisture. In view of this, the Radio Frequency Identification (RFID) technology has been shown promising for the development of UTs, mostly due to the ability of RFID tags to operate as sensors. However, the adaptation of wireless communication in IoUT systems for application in underground conditions represents a challenger yet. In this context, the proposal of the work is to develop a UT based on RFID technology for sensing soil moisture. The UT proposed is formed by a rectangular patch microstrip antenna designed to operate at Ultra High Frequency (UHF), connected to a passive UHF RFID tag chip. The RFID tag in buried in the soil and functions as a moisture sensor. To reduce the effects of soil moisture on the RFID tag, the microstrip antenna designed in the systems was encapsulated in a Dielectric Soil Attenuator (DSA). The DSA is formed by multilayers made of natural carnauba wax and applied over the antenna. Simulated and experimental results were analyzed to verify the effectiveness of the DSA on the antenna parameters at different levels of moisture and soil types. Measurements of the Received Signal Strength Indicator (RSSI) were also performed using a UHF RFID reader, and thus, finding a relationship with the moisture present in the soil. The obtained value shows that the DSA reduces variations in the antenna parameters such as gain and reflection coefficient (S11) caused by soil moisture. Statistical analyzes of the RSSI measurements showed that the proposed UT can work as a soil moisture sensor.

20
  • DANIELE MONTENEGRO DA SILVA BARROS
  • Architecture of Digital Health Based on Machine Learning for Diagnosing Eye Diseases: a study applied to glaucoma

  • Líder : RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • MIEMBROS DE LA BANCA :
  • RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • CÉSAR ALEXANDRE DOMINGUES TEIXEIRA
  • ALEXANDRE CHATER TALEB
  • FRANCISCO MILTON MENDES NETO
  • JAILTON CARLOS DE PAIVA
  • Data: 06-nov-2020


  • Resumen Espectáculo
  • The development of new technologies has been essential for the diagnosis and treatment of eye diseases. In this context, using machine learning (AM) and deep learning (DL) techniques for the classification and detection of ocular analogies has resulted in recent research. In this perspective, the present's objective is to develop a technical classification service for DL, web technologies, and cloud services, making a case study on glaucoma. Glaucoma is an asymptomatic disease that can cause irreversible blindness if diagnosed late. The type of image used in the analysis was retinography since its acquisition is inexpensive, and a non-dependent operator performs the examination. That is, it does not require the presence of the doctor. As a result, an intelligence architecture was designed that implemented and trained several DL architectures. The learning transfer technique (transfer learning) was used to reduce training time and optimize the method. Another essential strategy is the data augmentation process (data argumentation) to minimize network overfitting, thus preventing the network from generalizing in the test set. A cloud service was developed, a platform that uses pre-trained models to identify new images' according to the input's mass of data.

21
  • TIAGO DOS SANTOS BEZERRA
  • Application of EBG (Electromagnetic BandGap) on Frequency Selective Surfaces to Suppress Higher Order Modes

  • Líder : ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • MIEMBROS DE LA BANCA :
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • ERICO CADINELI BRAZ
  • JOSE PATROCINIO DA SILVA
  • LAERCIO MARTINS DE MENDONCA
  • MAURICIO WEBER BENJÓ DA SILVA
  • Data: 06-nov-2020


  • Resumen Espectáculo
  • We propose an application of electromagnetic bandgap (EBG) for suppression or reduction of higher order resonant modes in frequency selective surfaces (FSS), not yet studied in the literature.

                The study consists of an application of a rectangular periodic array of circular air holes in FSS dielectric substrate to create rejection bands and to suppress or reduce specific resonant frequency modes. Another type of unconventional air hole used was the square shaped holes, aiming to reduce the computational effort during the simulations.

                Optimizations with genetic algorithms were implemented to optimize the physical parameters of the holes, aiming at the suppression in the desired frequency range.

                In order to prove that the prohibited bands occurred in the desired frequency bands, a Brillouin dispersion diagram was constructed verifying that the structures acted as EBG

    Seven prototypes were manufactured and the results measured compared to the simulated results obtained in ANSYS HFSS. A good agreement between the results is observed. Simulations and measurements show suppression levels of up to 6 dB. The simulation time of the FSS with EBG of square holes showed a reduction of 70% in relation to the simulation time of the FSS with EBG of cylindrical holes.

22
  • EDNARDO PEREIRA DA ROCHA
  • ALGORITHM FOR A LINEAR STATE ESTIMATOR BASED ON CURRENT SUMMATION METHOD APPLIED TO THREE-PHASE POWER DISTRIBUTION NETWORKS

  • Líder : MANOEL FIRMINO DE MEDEIROS JUNIOR
  • MIEMBROS DE LA BANCA :
  • MADSON CORTES DE ALMEIDA
  • MANOEL FIRMINO DE MEDEIROS JUNIOR
  • MARCOS ANTONIO DIAS DE ALMEIDA
  • MAX CHIANCA PIMENTEL FILHO
  • UBIRATAN HOLANDA BEZERRA
  • Data: 20-nov-2020


  • Resumen Espectáculo
  • The way in which the energy companies have been operating their substations and supervising their feeders in the distribution networks has undergone a major modernization process in the past two decades. As the number of consumers and the physical extension of the infrastructure increased, there was a greater demand for equipment and monitoring techniques that would guarantee the reliability of these systems.This work presents the development of a new algorithm based on currents summation method,for implementation of a linear formulation of the state estimation problem in three-phase distribution networks, applied in the normal operation regime and under fault condition.In the normal operating regime, the modules of branch and node currents are estimated based on the weighted least squares method and a load adjustment algorithm, which makes it possible to obtain an estimative of technical losses, active and reactive power flows and the system voltage profile.For validation, data from two 13.8 kV real distribution systems were used to test the method. The meters located in the analyzed feeders provide the values of the measured quantities and an approximation of the average power factor of the loads located downstream from these measurement points, while pseudomeasures are used to make the system observable. In the proposed method, an iterative reestimation algorithm based on system scanning was implemented, where the parameters estimated in the first estimate are used to update the loads, branch currents and new variances.In the second stage of this thesis, an algorithm for fault location was developed using a system sweeping method associated with the proposed state estimator. In this case, in addition to considering the short-circuit current measured at the substation, the currents of each load were estimated during the fault to verify their influence on the fault location process.In this condition, a real distribution system was modeled in the ATP softwarein order to emulate measurements of voltages and currents at the substation, and voltage magnitudes registered by other meters during the fault. Factors such as the influence of the fault resistance, the type of fault, the system loading and the fault location were tested. The results obtained show that the developed method is able to provide the system status with adequate precision for distribution systems with a reduced number of measurement points along the feeder, besides presenting a low computational processing time and simplified modeling. The obtained results show that the developed method can provide the state of the system with adequate precision for distribution systems with a reduced amount of measurement points along the feeder, besides presenting a low computational processing time and simplified modelling. There was also a proximity between the proposed estimator and the re-estimation method, proving that the load adjustment process already provides pseudo-values close to the estimated values. Under a fault condition, the algorithm showed promise in its function, locating faults with a small error margin in most simulations.

23
  • ISRAEL EDUARDO DE BARROS FILHO
  • An anti-collision algorithm for large-scale RFID in noisy environments applied to the Industrial Internet of Things

  • Líder : IVANOVITCH MEDEIROS DANTAS DA SILVA
  • MIEMBROS DE LA BANCA :
  • CARLOS MANUEL DIAS VIEGAS
  • DANIEL GOUVEIA COSTA
  • IVAN MULLER
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • Data: 30-nov-2020


  • Resumen Espectáculo
  • The Industrial Internet of Things (IIoT) is often presented as a concept that is significantly changing the technological landscape of industries, through automation procedures and identification of relevant objects. For this, however, reliability and performance issues must be considered when providing anticipated communication services. By employing Radio Frequency Identification (RFID) in the IIoT context, several previous studies have worked to improve the efficiency of RFID communications systems, generally defining mathematical models for planning and evaluating quality. However, such models are designed based on error-free communications, which is in fact unrealistic when considering the error-prone nature of wireless communications in industrial plants. Therefore, this thesis proposes a new anti-collision algorithm for RFID together with a formal model based on Generalized Stochastic Petri Nets (GSPN) to evaluate RFID communications, modeling different possibilities of errors between readers and RFID tags. Since this proposal uses the EPCglobal UHF Class 1 Gen2 parameters as a reference, which are already adopted by the Dynamic Frame Slotted Aloha anti-collision protocol for passive RFID systems, this model can be explored to assess the performance of different RFID access protocols when assuming noisy channels, supporting better comparisons between different algorithms and protocols. The results showed that the proposed algorithm is able to present a better performability in relation to the other evaluated protocols, mainly in the presence of noisy channels if a large number of tags to be read. Simulation scenarios are defined to provide reliability and performance results when evaluating RFID tag readings, which are valuable when designing and maintaining IIoT applications.

24
  • LEONARDO ANGELO VIRGINIO DE SOUTO
  • Extracting Stairs and Doors as Natural Landmarks for Mobile Robot Localization from Clouds of 3D Edge-Points

  • Líder : LUIZ MARCOS GARCIA GONCALVES
  • MIEMBROS DE LA BANCA :
  • LUIZ MARCOS GARCIA GONCALVES
  • RAFAEL BESERRA GOMES
  • BRUNO MARQUES FERREIRA DA SILVA
  • ANDERSON ABNER DE SANTANA SOUZA
  • TIAGO PEREIRA DO NASCIMENTO
  • Data: 30-nov-2020


  • Resumen Espectáculo
  • Natural landmarks are the main features in the next step of the research in localization of mobile robot platforms. The identification and recognition of these landmarks are crucial to better localize a robot. To help solving this problem, this work proposes an approach for the identification and recognition of natural marks included in the environment using images from RGB-D sensors. In the identification step, a structural analysis of the natural landmarks that are present in the environment is performed. The extraction of edge points of these landmarks is done using the 3D point cloud obtained from the RGB-D sensor. These edge points are smoothed through the $Sl_0$ algorithm, which minimizes the standard deviation of the normals at each point. Then, the second step of the proposed algorithm begins, which is the proper recognition of the natural landmarks. This recognition step is done as a real-time algorithm that extracts the points referring to the filtered edges and determines to which structure they belong to in the current scenario: stairs or doors. Finally,  the geometrical characteristics that are intrinsic to the doors and stairs are identified. The approach proposed here has been validated with real robot experiments. The performed tests verify the efficacy of our proposed approach. 

25
  • DANIEL NOBRE PINHEIRO
  • Convex fuzzy k-medoids clustering

  • Líder : DANIEL ALOISE
  • MIEMBROS DE LA BANCA :
  • CAROLINE THENNECY DE MEDEIROS ROCHA
  • DANIEL ALOISE
  • ERALDO LUIS REZENDE FERNANDES
  • MARIÁ CRISTINA VASCONCELOS NASCIMENTO ROSSET
  • SAMUEL XAVIER DE SOUZA
  • Data: 16-dic-2020


  • Resumen Espectáculo
  • The k-medoids model is one of the most popular clustering methods. In this work, we propose the Convex Fuzzy k-Medoids Problem (CFKM), which not only allows one object to be assigned to multiple clusters, but also allows a cluster to be represented by multiple medoids. The proposed model is convex and thus is robust to initialization. To evaluate the importance of CFKM, we compare it with another two fuzzy k-medoids models: the Fuzzy k-Medoids Problem (FKM) and the Fuzzy clustering with Multi-Medoids Problem (FMMdd), both solved by heuristics due to their computational complexity. Experiments with both synthetic and real-world data, along with an user survey, show that CFKM is not only more robust to the choice of parameters of fuzzy models, but also is the only able to reveal important aspects of inherently fuzzy data.

2019
Disertaciones
1
  • MARIA GRACIELLY FERNANDES COUTINHO
  • Deep Neural Network Hardware based on Stacked Sparse Autoencoder

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • ADRIAO DUARTE DORIA NETO
  • CARLOS ALBERTO VALDERRAMA SAKUYAMA
  • DIOMADSON RODRIGUES BELFORT
  • MARCELO AUGUSTO COSTA FERNANDES
  • Data: 17-ene-2019


  • Resumen Espectáculo
  • The deep learning techniques have been gaining prominence in world research in the past years. However, the deep learning algorithms have high computational cost, making it hard to apply in several commercial applications. On the other hand, new alternatives have been studying to accelerate complex algorithms, among these, those based on reconfigurable hardware has been showing very significant results. Therefore, the objective of this work is the hardware implementation of a neural network for the use of algorithms with deep learning. The hardware was developed on Field Programmable Gate Array (FPGA) and supports Deep Neural Network (DNN) trained with the Stacked Sparse Autoencoder (SSAE) technique. In order to allow DNNs with many inputs and layers on the FPGA, the systolic array technique was used in all developed hardware. The details of the architecture designed on the FPGA were evidenced, as well as the occupation data on hardware and the processing time to two different implementations. The results show that both implementations achieve high throughputs allowing the use of Deep Learning techniques in massive data problems.

2
  • YURI PEDRO DOS SANTOS
  • System of Detection and Classification of Vocal Pathologies Based on the Correntropy Spectral Density

  • Líder : LUIZ FELIPE DE QUEIROZ SILVEIRA
  • MIEMBROS DE LA BANCA :
  • ALLAN DE MEDEIROS MARTINS
  • ALUISIO IGOR REGO FONTES
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • SUZETE ELIDA NOBREGA CORREIA
  • Data: 18-ene-2019


  • Resumen Espectáculo
  • Vocal pathologies negatively affect the social and professional life of the sick, some of which may even lead to death if they are not treated quickly. Among the main procedures for the diagnosis of vocal disorders we can cite: laryngoscopy, perceptual-auditory evaluation, acoustic analysis of the voice, aerodynamic evaluation and self-evaluation of the voice by the patient. However, these procedures are usually invasive or inaccurate. Therefore, digital signal processing techniques have been used in the design of non-invasive systems to aid in the diagnosis of vocal tract pathologies. In this work a system of detection and classification of vocal pathologies is presented, using a classification technique based on descriptors obtained through the Correntropy Spectral Density (CSD) function, defined as the Fourier transform of the autocorrelation function. The descriptors obtained have information of second-order and higher-order statistical moments of the speech signal, through which vocal pathologies can be efficiently detected and classified. The classification is made by a neural network multilayer perceptron (MLP), performing a binary classification between normal and pathological voices and then between pathologies (edema and nodule). The classifier was evaluated by computer simulation, and the results indicate a high hit rate of detection and classification among pathologies.

3
  • RUTE SOUZA DE ABREU
  • A methodology for detection of causality relations among discrete time series on systems

  • Líder : LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • MIEMBROS DE LA BANCA :
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • RODRIGO SIQUEIRA MARTINS
  • Data: 25-ene-2019


  • Resumen Espectáculo
  • The need for detecting causality relations among process, events or variables is present in many areas of knowledge e.g., distributed computing, the stock market, industry, medicine, etc. This occurs because the knowledge of these relations can often be helpful in solving a variety of problems. For example, maintaining the consistency of replicated databases when writing distributed algorithms or optimizing the purchase and sale of stocks in the stock market. In this context, this dissertation proposes a new methodology for detecting causality relations in systems by using information criteria and Bayesian networks to ensemble the most probable structure of connections among discrete time series. Modeling the system as a directed graph, in which the nodes are the discrete time series and the edges represent the relations, the main idea of this work is to detect causality relations among the nodes. This detection is made using the method of transfer entropy, which is a method to quantify the information transferred between two variables, and the K2 algorithm: a heuristic method whose objective is to find the most probable belief-network structure, given a data set. Because K2 depends on the premise of having a previous structure that defines the hierarchy among the network nodes, it is proposed in the methodology the creation of the previous ordering on the nodes considering direct and indirect relations, and the modeling of these relations according to the lag between cause and effect. In addition, knowing that the K2 algorithm considers that each case of the data set occurs simultaneously, the proposed methodology modifies the original algorithm by inserting the dynamics of these lags into it. This modification provides a mechanism for comparing direct and indirect causality relations regarding its contribution to the structure. As the result, it is obtained a graph of causality relations among the series, with the relation's lags being explicit.

4
  • YURI THOMAS PINHEIRO NUNES
  • Data-Based Approach to Parametric Configuration of Industrial Alarms

  • Líder : LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • MIEMBROS DE LA BANCA :
  • CLAUBER GOMES BEZERRA
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • Data: 25-ene-2019


  • Resumen Espectáculo
  • Industrial plants are composed of processes that add up to thousands of variables. To ensure safety and quality of operation, these processes are monitored and alarms are configured to indicate a possible malfunction. Among the most common problems associated with industrial alarms we can mention: high occurrence of false alarms, missed alarms and chattering alarms, operator overload and alarm flooding. These problems are related to the process of selection of the monitored variables, the techniques of activation and deactivation of alarms, among other characteristics of the process and the alarm system. This work focuses on defining an approach to configure efficient and significant alarms for the operator. The approach proposed here is inspired by the workflow of a data scientist who initially needs to identify the characteristics of the databases used to then apply transformations that make the data more suitable allowing the extraction of valuable information. Many times the scientist is interested in creating a model that describes the data or makes predictions possible. This is a very similar task of alarm configuration where it is necessary to select the relevant variables and to configure the settings of each alarm in order to classify the operation of the process as appropriate or not and to help identify the fault. The approach proposed here consists of four parts: description of data, selection of variables, tuning and performance evaluation. During the description step, relevant information about the data is obtained, such as the presence of events, the number of different events, the duration of events, etc. In the selection stage, the relevant variables for detection of abnormalities are defined. The tuning of alarms is similar to a training process, where a model is built to describe the behavior of the data. Finally during the evaluation, the settings found are applied to a process history to asses whether the settings behave in a way that meets security and quality constraints. As a case study, an industrial alarm configuration was obtained for the Tennessee Eastman Process which is a simulator widely used by the academic community.

5
  • MAX RODRIGUES MARQUES
  • Evaluation of Wavelet-based Protection Applied in a Doubly-fed Induction Generator

  • Líder : FLAVIO BEZERRA COSTA
  • MIEMBROS DE LA BANCA :
  • CAMILA MARA VITAL BARROS
  • FLAVIO BEZERRA COSTA
  • LUCIANO SALES BARROS
  • RODRIGO PRADO DE MEDEIROS
  • THIAGO DE OLIVEIRA ALVES ROCHA
  • Data: 28-ene-2019


  • Resumen Espectáculo
  • With the globalization and the rapidly increasing global energy demand, it has been seeking the sustainability guarantee because the growing concern to preserve the planet for future generations. Therefore, investments in clean and renewable energy sources, such as wind power generation, have gained space in research groups in order to overcome their drawbacks and improve the benefits. The doubly fed induction generator (DFIG) is the predominant one in the market. Nevertheless, considering the electrical failures inherent to this generator type and the ones in the power system (PS), the protection of its elements is an important topic that still does not have comprehensive studies. This dissertation investigates the techniques used to perform the internal and external protections applied to the DFIG topology. As well as, it is presented the theoretical fundamentals about the wind energy conversion systems (WECS), the wavelet transform, and some relay-based protections. DFIG electrical signals in case of machine terminal faults is analyzed and evaluated using the real-time stationary wavelet transform with boundary effect (RT-BSWT). Furthermore, qualitative studies on wavelet overcurrent, wavelet under-voltage, directional, and differential protections will be introduced in order to assess and to validate these new protection trends applied to WECS.As also, to develop a preliminary method of identifying internal faults in a DFIG. Analysis applying these protections to real signals, collected by an experimental test-bench with DFIG, demonstrated which the used mathematical tools had good performances for protection against electrical faults at the common connection point.

6
  • KAIO MÁRCIO DA COSTA BANDEIRA
  • Multiband Microstrip Antenna for Wireless Communication Applications in IEEE 802.11ax and IEEE 802.11ah Standards

  • Líder : VALDEMIR PRAXEDES DA SILVA NETO
  • MIEMBROS DE LA BANCA :
  • VALDEMIR PRAXEDES DA SILVA NETO
  • VICENTE ANGELO DE SOUSA JUNIOR
  • CRISTHIANNE DE FATIMA LINHARES DE VASCONCELOS
  • MARCIO EDUARDO DA COSTA RODRIGUES
  • FRANCISCO DE ASSIS BRITO FILHO
  • Data: 27-jun-2019


  • Resumen Espectáculo
  • In view of the evolution of the standards of wireless communication systems, where there is a need for performance in several frequency bands, it is indispensable to design devices capable of meeting the requirements that this progress demands. The antennas are elements that play a fundamental role in these systems, in which planar microstrip antennas have been demonstrated with high evidence due to their low profile, miniaturized devices, low manufacturing cost and are easily integrated with the devices. With the objective of optimizing the occupation of physical space in the transceivers and in the end devices of the users, this work presents the proposal of a microstrip antenna based on a planar monopole with ground plane with slits for operation in multiple bands to meet new systems of wireless communications such as IEEE 802.11ax and IEEE 802.11ah. The process of design formation and design of the structure will be presented, considering the most recent research in the specialized literature. For the purposes of analysis, simulations were performed by the Finite Element method implemented through the commercial software Ansoft HFSS. Some initial prototypes were fabricated for purposes of validation of the preliminary results of this work. The experimental characterization of these prototypes was performed through a vector network analyzer. In view of the comparisons between simulated theoretical results and the preliminary experiments, it can be affirmed that a good agreement between them is observed corroborating with the studies previously developed.

7
  • MÁRIO GUILHERME FLORES FIGUEREDO
  • Performance of audio signals positioning algorithms in environments with impulsive noise

  • Líder : VICENTE ANGELO DE SOUSA JUNIOR
  • MIEMBROS DE LA BANCA :
  • FRANCISCO CARLOS GURGEL DA SILVA SEGUNDO
  • VALDEMIR PRAXEDES DA SILVA NETO
  • VICENTE ANGELO DE SOUSA JUNIOR
  • Data: 28-jun-2019


  • Resumen Espectáculo
  • Recently, with increasing processing capacity as well as the miniaturization of electronic devices, a number of commercial systems based on the location of audio signals have emerged, such as shot spotter and voice monitoring systems, and residential, among them, entertainment system and home automation (e.g. Google Home and Amazon dot). In the literature, the classical methods of Direction-of-Arrival (DoA) are commonly evaluated in acoustic environments submitted to White Additive Gaussian Noise (AWGN). However, acoustic environments are also subject to impulsive noise, under which the classical methods of DOA estimation have degraded performance. In this way, besides describing and evaluating the performance of classical parametric and non-parametric DOA methods, and demonstrating their respective performances in acoustic environments submitted to AWGN, this work also proposes to present their performance when subjected to the impulsive noise modeled by the Gaussian mixing method. As a last contribution, the methods are tested using real signals from a proprietary acquisition system.

8
  • SAMUEL BELARMINO DE PAIVA
  • Analysis and Synthesis of Bioinspired Frequency Selective Surfaces Using Neural Networks for Applications in Wireless Communication Systems

     

  • Líder : ADAILDO GOMES D ASSUNCAO
  • MIEMBROS DE LA BANCA :
  • ADAILDO GOMES D ASSUNCAO
  • ADAILDO GOMES D ASSUNCAO JUNIOR
  • GLAUCO FONTGALLAND
  • HERTZ WILTON DE CASTRO LINS
  • JOSE ALFREDO FERREIRA COSTA
  • VALDEMIR PRAXEDES DA SILVA NETO
  • Data: 01-jul-2019


  • Resumen Espectáculo
  • This work describes the analysis of Bioinspired Frequency Selective Surfaces (BFSS) for applications in wireless systems, operating in the C band, Ku band and UWB (ultra-wideband). Simple and coupled BFSS structures are considered. The BFSSs have array elements with the four-leaf clover-shaped and maple leaf-shaped, and presented dual-band response, with the operating frequencies in the C and Ku bands. For the development of BFSSs with the ultra-wideband, a cascade structure was developed, in which a FSS with elements with patches with the shape of square loops was coupled to the BFSSs. In addition, the four-leaf clover BFSS synthesis was developed using an artificial neural network with a cascade feedforward architecture and Bayesian regularization training algorithm to obtain the specifications of resonance frequency and respective desired bandwidths. The numerical values obtained by simulations for the developed prototypes were obtained by the ANSYS HFSS software. Prototypes were manufactured and experimentally characterized. The measured results were compared with the simulated ones and a good agreement was observed.

9
  • LUIZ ANDRÉ PONTAROLO
  • Adjustment of PID Controller by an Autotuning Method Based on Robustness Estimate

  • Líder : CARLOS EDUARDO TRABUCO DOREA
  • MIEMBROS DE LA BANCA :
  • ANDRE LAURINDO MAITELLI
  • CARLOS EDUARDO TRABUCO DOREA
  • OSCAR GABRIEL FILHO
  • Data: 05-jul-2019


  • Resumen Espectáculo
  • In this work a process of adjustment of PID  controllers using the auto tuning method oriented to the robustness of the system and implemented in a Programmable Logic Controller (PLC) with the use of relay experiments is presented, with possibility of application in PI-D controllers. The method was developed through the adaptation of methods directed to existing PI controllers, with the objective of using few iterative structures, allowing the implementation in PLC, as well as improving the performance in the transient response of the system. Robustness is indicated by the maximum sensitivity of the system. Relay experiments provide frequency response points that allow calculation of parameters that modify the original terms of the controller in order to remove the points from the interior of the maximum sensitivity circle. Results of two PLC applications of different didactic systems allowed comparison with an existing PI controller tuning method and the demonstration of the effectiveness of the method.

10
  • JOÃO RICARDO TAVARES GADELHA
  • Model Predictive Control Applied to a Plunger Lift Artificial Elevation System

  • Líder : CARLOS EDUARDO TRABUCO DOREA
  • MIEMBROS DE LA BANCA :
  • CARLOS EDUARDO TRABUCO DOREA
  • ANDRE LAURINDO MAITELLI
  • ANDERSON LUIZ DE OLIVEIRA CAVALCANTI
  • ANDRÉ FELIPE OLIVEIRA DE AZEVEDO DANTAS
  • Data: 12-jul-2019


  • Resumen Espectáculo
  • During the productive life of most gas wells there is an accumulation of liquid at the bottom of the well which causes a pressure contrary to the flow of the reservoir, reducing the production of the well. In this way, one of the possible artificial lift methods to solve this problem is represented by the Plunger Lift. It consists of a low-cost installation and maintenance technique that uses a piston to, among other uses, increase the efficiency of liquid removal from the system. However, this method requires a well-adjusted controller to appropriately define the opening and closing periods of a motor valve installed in the wellhead for production control. Predictive control is a technique that can be applied to perform such task, performing predictions of the future behavior of the plant in order to obtain its optimum performance. In this work, the implementation and application of a linear predictive controller to a simulated well operated by Plunger Lift was performed. The controller was applied under different conditions of draining ability, with and without, of the gas present in the annular space. Results of MPC without drainage were compared with conventional versions of Plunger Lift control in the oil industry through graphical analysis and observation of performance parameters. This controller obtained better results regarding the control of the average speed of plunger ascent and accumulated production. In the comparison of MPC with and without drainage, the version capable of manipulating the gases from the annular space was able to cause the system to operate at speeds of piston ascent in safer ranges.

11
  • RAFAEL MAGALHÃES NÓBREGA DE ARAÚJO
  • Development and Evaluation of a Self-sufficient Photovoltaic-Thermoelectric System

  • Líder : ANDRES ORTIZ SALAZAR
  • MIEMBROS DE LA BANCA :
  • ALBERTO SOTO LOCK
  • ANDRES ORTIZ SALAZAR
  • ANDRE LAURINDO MAITELLI
  • JOÃO TEIXEIRA DE CARVALHO NETO
  • Data: 29-jul-2019


  • Resumen Espectáculo
  • The generation of electric energy by a photovoltaic cell mainly depends on the operating temperature and solar irradiance that falls on it. The operating temperature can be controlled by means of thermoelectric cells working as coolers. This work proposes a self-sufficient system in which photovoltaic cells are cooled by thermoeletric cells. It also proposes an analysis on which situations a favorable energy yield can be reached. Simulations have been performed taking into account diverse sets of solar irradiance, wind speed, and electric power dedicated to the cooling. Results suggest that the proposed system reaches positive net yield in situations that environmental variables lead to higher photovoltaic panel operating temperature combined with controlled low power dedicated to the thermoeletric cells.

12
  • NAYANA LETÍCIA DE MORAIS VIANA
  • Design of a low consumption, low noise and low offset instrumentation amplifier for portable applications

  • Líder : DIOMADSON RODRIGUES BELFORT
  • MIEMBROS DE LA BANCA :
  • DIOMADSON RODRIGUES BELFORT
  • SEBASTIAN YURI CAVALCANTI CATUNDA
  • ANTONIO WALLACE ANTUNES SOARES
  • FRANCISCO DE ASSIS BRITO FILHO
  • Data: 29-jul-2019


  • Resumen Espectáculo
  • In this work is presented a low power instrumentation amplifier using 0.6 μm CMOS technology. It is initially shown the theoretical background about the chosen application: treatment of biopotentials. Next, we present the architectures of instrumentation amplifiers present in the literature focused on biopotentials, as well as low power and low noise configurations. A study is also carried out on the types of pseudo-resistors present in the literature. Once this is done, all the methodology used to perform this work is presented, such as the choice of the architecture used, how the system design is performed and what types of simulations are used to evaluate the performance of the system. Finally, it can be concluded that using integrated circuits with CMOS technology can enable numerous low-power portable applications that use the acquisition of biopotentials, thus justifying such design.

13
  • ANDREW VINÍCIUS SILVA MOREIRA
  • Photovoltaic Panel performance against partial shading using intelligent bypass diode and clustered adaptive P & O

  • Líder : LUCIANO SALES BARROS
  • MIEMBROS DE LA BANCA :
  • LUCIANO SALES BARROS
  • ANDRE LAURINDO MAITELLI
  • ANDERSON LUIZ DE OLIVEIRA CAVALCANTI
  • JOÃO TEIXEIRA DE CARVALHO NETO
  • Data: 31-jul-2019


  • Resumen Espectáculo
  • The exponential growth in the use of photovoltaic solar (PV) energy makes that the question of the performance of panels connected to the grid is extremely relevant, because its efficiency is relatively low if compared to other energy sources. Among the main studies, the following stand out: Partial Shading (PS) and Maximum Power Point Tracking (MPPT). The PS refers to the distinct levels of irradiance along the photovoltaic panel and the MPPT refers to the maximum power point tracking according to ambient conditions. In this work, a current photovoltaic panel with 72 cells, containing a modern smart bypass diode configuration was modeled and simulated. In order to verify the performance of the photovoltaic panel were simulated partials shadings and the maximum power point tracking using the simplest MPPT technique: Perturb & Observe (P&O), however modified with adaptive step and clustered.

14
  • ALESSANDRO DE SOUZA LIMA
  • Analysis of Photovoltaic Panels Facing Partial Shading Using Magic Square and Adaptive P & O.

  • Líder : ANDRE LAURINDO MAITELLI
  • MIEMBROS DE LA BANCA :
  • ANDRES ORTIZ SALAZAR
  • ANDRE LAURINDO MAITELLI
  • JOÃO TEIXEIRA DE CARVALHO NETO
  • LUCIANO SALES BARROS
  • Data: 31-jul-2019


  • Resumen Espectáculo
  • Partial shading decreases the power generated by solar panel arrangements. In this condition, some panels end up receiving solar radiation different from others, causing multiple peaks in P-V and I-V, resulting in losses. This work aims to reduce the losses caused by partial shading from reconfigurations of the panel arrangements, in conjunction with an adaptive P & O MPPT algorithm. The modeling of the solar panels is carried out in a Matlab & Simulink simulation environment. The modeling of the adaptive MPPT P&O method is also performed. The tested arrangements are of the Total Cross-Tied (TCT), Total Cross-Tied (SP-TCT), Bridge-Linked - Total Cross-Tied (BL-TCT), Bridge-Linked-Honey-Comb (BL-HC), plus a Magic Square (MS) equivalent for each of the arrangements presented. Magic Square is a way to rearrange the solar panels in different positions, in order to mitigate the losses caused by shading. The performances of all configurations are tested for different shading situations and demonstrated from V-P graphics. In addition to the MS, in order to make the system more efficient, it implements the adaptive MPPT P&O that has a faster response time than the standard P&O.

15
  • ADRIANA BENÍCIO GALVÃO
  • Collaborative Tool for Data Analysis in Public Health: Application in the Time Series Study for the Project "Syphilis No"

  • Líder : RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • MIEMBROS DE LA BANCA :
  • RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • AQUILES MEDEIROS FILGUEIRA BURLAMAQUI
  • KENIO COSTA DE LIMA
  • ROBINSON LUIS DE SOUZA ALVES
  • Data: 16-ago-2019


  • Resumen Espectáculo
  • Nowadays, there is a trend towards the use of data analysis in order to implement changes in health systems, which allow, among other benefits, cost reduction, disease prevention and better service delivery to the population. One of the initiatives of the Brazilian government, in this sense, is the project "Syphilis No", which induces actions aimed at the control of syphilis in health care networks.     The term "Big Data in Health" has been widely used in several studies that seek, in general, to find and evaluate the possible benefits of Big Data for health care. Many of these studies, however, are based on theoretical and qualitative analysis, presenting limitations in the validation of the proposed solutions and highlighting the skills in the development of competences for the use of Big Data technologies. In this context, the present work has the objective of elaborating and implementing a collaborative tool, which facilitates the analysis of big data, which will contribute, in particular, to the study of time series regarding the reports of acquired, congenital, and registered syphilis in the Notification of Injury Information System (SINAN).

        A framework was developed that allows collaboration between professionals with different skills to develop experimental data analysis programs using a preconfigured big data environment. The framework and its implemented functions make up the tool developed in this work, which allows to import time series and execute methods for the extraction of characteristics involving data from one or more Health Information Systems. Finally, through the tool developed, epidemiological indicators were obtained for the monitoring of Syphilis.
16
  • MARCELLA ANDRADE DA ROCHA
  • A text as unique as fingerprint: The use of intelligent systems for authorship recognition

  • Líder : RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • MIEMBROS DE LA BANCA :
  • AQUILES MEDEIROS FILGUEIRA BURLAMAQUI
  • ELOIZA DA SILVA GOMES OLIVEIRA
  • KARILANY DANTAS COUTINHO
  • RICARDO ALEXSANDRO DE MEDEIROS VALENTIM
  • ROBINSON LUIS DE SOUZA ALVES
  • Data: 16-ago-2019


  • Resumen Espectáculo
  • The Authorship Attribution, the science of inferring an author for a particular text based on their writing characteristics is a problem with a long history. In this work, it is being proposed the study of the problem of attribution of authorship in order to make it a tool of use in the distance teaching platform of the Ministry of Health (MS), AVASUS, and will be presented the techniques of analysis of text and authors' stylistic characteristics that allow authorship to be determined in significantly better indexes, in which the texts are greater than 140 characters. This proposal targets AVASUS, where students take the courses of the platform, share their interests and thoughts in the form of messages in the forums and do activities that require writing on certain topics in the area of health, these written productions are the focus of the application of attribution of authorship. The techniques studied as a proposal are a two-stage process, where in the first stage, stylometric information is extracted from the collected data set and in the second stage different classification algorithms are trained and lexical analysis techniques are applied to predict the authors of the texts. The effort is to maximize the accuracy of predictions with optimal amount of data and users under consideration. 

17
  • LUÃ SILVA CARDOSO
  • Virtual Impedance Concept to Decouple P/Q of Distributed Generation Systems. 

  • Líder : RICARDO LUCIO DE ARAUJO RIBEIRO
  • MIEMBROS DE LA BANCA :
  • CARLOS EDUARDO TRABUCO DOREA
  • CECÍLIO MARTINS DE SOUSA NETO
  • JOSE RENES PINHEIRO
  • RICARDO LUCIO DE ARAUJO RIBEIRO
  • THIAGO DE OLIVEIRA ALVES ROCHA
  • Data: 30-ago-2019


  • Resumen Espectáculo
  • The distributed generation systems based on renewable sources has increased due to environmental problems caused by non-renewable sources, fast return on investment and advances in interconnection systems. In Brazil, the distributed generation energy capacity is growing. Before a new electric power system model, distributed generation systems participation, methods to regulate the power flow in generation units can directly contribute to electric system stability, especially in low-voltage grids, which is the predominantly resistive and the active power can cause overvoltages. In addition, power flow strategies used in conventional generation systems cannot be used in low-voltage systems due to active and reactive power coupling. In this work is proposed the power flow control adjust in a distributed generation photovoltaic system controlled in voltage mode based on virtual impedance concept. In order to overcome the active and reactive power coupling problem, the proposed method is implemented by means a virtual resistor. For analysis purposes, the virtual resistor is implemented by two ways. In the first one, a positive virtual resistor is implemented to accentuate the resistive profile of the system, performing the P/V and Q/θ control. In the second one, a negative virtual resistor is implemented in order to make the network profile predominantly inductive, and then perform the traditional P/θ and Q/V control. The virtual impedance concept technique is evaluated through simulation and experimental results.

18
  • MARCOS SÉRGIO RODRIGUES LEAL
  • FPGA-Based Real-Time Power System Simulation for Traveling-Wave-Based Protection Validation

  • Líder : FLAVIO BEZERRA COSTA
  • MIEMBROS DE LA BANCA :
  • FLAVIO BEZERRA COSTA
  • MARCELO AUGUSTO COSTA FERNANDES
  • RICARDO LUCIO DE ARAUJO RIBEIRO
  • THIAGO DE OLIVEIRA ALVES ROCHA
  • WASHINGTON LUIZ ARAUJO NEVES
  • Data: 06-sep-2019


  • Resumen Espectáculo
  • In this work is proposed a real-time digital simulator of power systems using a low-cost custom platform based on FPGA (field-programmable gate array) proper to perform real-time validation of traveling-wave-based transmission line protections. The operational view of the simulator is introduced by means the modeling, implementation, and simulation steps of a transmission system, which is used to highlight the simulator capability to represent high-frequency transient phenomenon taking place transmission lines. Hence, at first, the mathematical models of the power system used in the case study are presented as well as the solver design, which is developed based on the electromagnetic transients program (EMTP) proposed by Dommel. Then, the simulator characteristics, such as the hardware architecture, development software, communication strategies, graphical interface, input/output, and data export, are introduced, as well as the implementation stages of the test system. Moreover, it addresses the implementation of a relay prototype using a hardware based on DSP (digital signal processor), running a traveling-wave-based protection scheme, besides its closed-loop integration with the simulation.  A GUI (graphical user interface) is developed to set the simulation parameters, including the conditions for applying an electrical fault, and to monitor the dynamic of power system used as a case study. Off-line simulations obtained from Matlab/Simulink are used to validate the real-time results.

19
  • VICTOR HUGO FREITAS DE OLIVEIRA
  • Application speedup characterization: modeling parallelization overhead and variations of problem size and number of cores

  • Líder : SAMUEL XAVIER DE SOUZA
  • MIEMBROS DE LA BANCA :
  • SAMUEL XAVIER DE SOUZA
  • CALEBE DE PAULA BIANCHINI
  • CARLOS AVELINO DE BARROS
  • EDSON BORIN
  • Data: 13-sep-2019


  • Resumen Espectáculo
  • To make efficient use of multi-core processors, it is important to understand the performance behavior of parallel applications. Modeling this behavior can enable the use of online approaches to optimize throughput or energy, or even guarantee the desired Quality of Service (QoS). Accurate models would avoid having to probe different runtime configurations, which causes extra overhead. Throughout the years, many speedup models were proposed. Most of them based on Amdahl’s law or Gustafson’s law. However, many of those make considerations such as a fixed parallel fraction, or a parallel fraction that varies linearly with problem size, and inexistent parallelization overhead. Although such models aid in the theoretical understanding of parallel computing, these considerations do not hold in real environments, which makes the modeling unsuitable for accurate characterization of parallel applications. The model proposed in this dissertation estimates the speedup taking into account the overhead caused by the parallelization and the variation of the parallel fraction according to problem size and number of cores used. Using ten applications from the PARSEC benchmark suite, the proposed model was able to estimate speedups more accurately than other models in recent literature. Tests were made in two servers, each one with a different hardware architecture.

20
  • YURI IOHANSSEN RIBEIRO DAMASCENO
  • Control Strategy for Soft Transition Between Grid-Connected and Islanding Modes Applied for Low-voltage Distributed Generation Without Energy Storage System

  • Líder : RICARDO LUCIO DE ARAUJO RIBEIRO
  • MIEMBROS DE LA BANCA :
  • RICARDO LUCIO DE ARAUJO RIBEIRO
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • THIAGO DE OLIVEIRA ALVES ROCHA
  • CECÍLIO MARTINS DE SOUSA NETO
  • RODRIGO PRADO DE MEDEIROS
  • Data: 04-oct-2019


  • Resumen Espectáculo
  • The world energy demand is expected to rise in the coming decades. Renewable Energy Source Distributed Generation (RES-DG) is being used to meet this increased demand. RES-DG systems can serve local loads on an isolated system (islanded operation) or connected to the power grid (grid-connected operation). Ideally, RES-DG systems should have the ability to transition smoothly between modes so that a continuous supply of power is achieved regardless of the grid state. This work proposes a communication-less control strategy for a low-voltage RES-DG system capable of operating in islanded as well in grid-connected mode and of achieving a smooth transfer between modes. An energy storage system is not required for operation in islanded mode since the energy balance is maintained by acting in the power generated by the RES. The proposed control strategy is divided in three parts: (a) the VSI control, which constitutes the inner control loops of the Voltage Source Inverter (VSI) (current, voltage and power control loops), (b) mode-related control loops (DC link, voltage at the PCC and DG system angular frequency control loops) and (c) the synchronization and islanding detection methods. The control strategy is validated through simulation of a three-phase RES-DG system. The proposed strategy presents a smooth transition between modes, with little to no perturbation on the PCC voltage, and a fast synchronization process from islanded to grid-connected mode. In addition, it also presents a fast DC link voltage control response time that is capable of maintaining the transitory small, even for large power imbalances during mode transition.

     

21
  • DEYVID LUCAS LEITE
  • Channel characterization using Unmanned Aerial Vehicle

  • Líder : PABLO JAVIER ALSINA
  • MIEMBROS DE LA BANCA :
  • MARCIO EDUARDO DA COSTA RODRIGUES
  • PABLO JAVIER ALSINA
  • VICENTE ANGELO DE SOUSA JUNIOR
  • ÁLVARO AUGUSTO MACHADO DE MEDEIROS
  • Data: 31-oct-2019


  • Resumen Espectáculo
  •  

    The use of Unmanned Aerial Vehicle (UAV) to provide services such as internet, goods delivery and air taxi has become a reality in recent years. The use of these aircraft requires secure communication between the control station and the UAV, a task made easier by knowing the characteristics of the communication channel. This work is intended to present a measurement setup using an unmanned aircraft, and to use the measured data for RF channel characterization. Control ground. The collected data were analyzed in order to characterize the large scale fading (path loss and shading) and the small scale fading (multipath and Doppler) in Caatinga scenarios, flight over a lake and a region mixed with the two previous environments. The following work presents the data collected in the measurement campaigns, the processing performed by the developed scripts and the results of the RF Channel characterization.


22
  • MYCHAEL JALES DUARTE

  • ANALYSIS AND DESIGN OF 2.5D AND 3D STRUCTURES OF FREQUENCY SELECTIVE SURFACES FOR APPLICATIONS IN MODERN WIRELESS COMMUNICATION SYSTEMS

  • Líder : VALDEMIR PRAXEDES DA SILVA NETO
  • MIEMBROS DE LA BANCA :
  • ADAILDO GOMES D ASSUNCAO
  • ADAILDO GOMES D ASSUNCAO JUNIOR
  • CRISTHIANNE DE FATIMA LINHARES DE VASCONCELOS
  • FRANCISCO DE ASSIS BRITO FILHO
  • VALDEMIR PRAXEDES DA SILVA NETO
  • Data: 12-dic-2019


  • Resumen Espectáculo
  • This work presents a study on Selective Frequency Surfaces (FSS) 2,5 D and 3D, for applications in modern wireless systems. Initially, 2.5D, patch-like structures are analyzed. The simple FSS are formed by elements in square and spiral turns based on the Minkowski fractal. The use of the fractal curve resulted in a miniaturization of the circuit, the appearance of a new transmission band and a decrease in bandwidth. The 2,5D FSS are formed from the simple structures that, for the formation of their elements, the coils were sectioned in eight parts, where four were printed on the upper face of the dielectric, four on the underside of the dielectric and were connected with cylindrical jumpers of 0.5 mm in diameter. The 2.5D surfaces provided miniaturization of the circuit without the appearance of new transmission bands and without decreasing the relative bandwidth. All HSS were stable with respect to the angle of incidence of the plane wave on the surface of the circuit. The simulated numerical results for the projected prototypes were obtained by Ansoft HFSS software. The prototypes were constructed and the experimental measurements of the transmission coefficients were performed, the values obtained were compared and discussed with the results of the simulations, which showed good agreement.

23
  • CARLOS YURI FERREIRA SILVA
  • Implementation of a neural velocity estimator in a three phase induction motor

  • Líder : ANDRES ORTIZ SALAZAR
  • MIEMBROS DE LA BANCA :
  • ALBERTO SOTO LOCK
  • ANDRES ORTIZ SALAZAR
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • JOSE ALVARO DE PAIVA
  • Data: 16-dic-2019


  • Resumen Espectáculo
  • The three phase induction motor is the most used type of machine for electric drives. Low cost, construction simplicity and reliability are the motives that explains this statement. Nowadays, this motor is also highlighting for it's applicability in the automotive area, with the electric vehicle ascension which, besides being a economic alternative to fuel consumption compared to ordinary vehicle, are also a solution for the emission of polluting gases to the environment. From this scenario, the objective of this work is to perform the speed control of an induction motor without the use of a mechanical sensor attached to the rotor, avoiding some inconveniences such as a periodic maintenance and increased complexity and number of equipment are always part of the engine running. For the induction motor control was applied the control strategy known as Field Oriented Control and for the estimation of the motor rotor speed was used an artificial neural network, which observer electrical and mechanical system variables. The practical results were satisfatory, since the neural network worked as both observer and speed estimation and it was possible to perform the control os the motor with the estimated speed in the range of 100 to 170 rad/s.

24
  • JOÃO ALEXANDRE DA SILVA NETO
  • Using Metamaterial Resonators to Improve Planar Filter Performance

  • Líder : ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • MIEMBROS DE LA BANCA :
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • IRADILSON FERREIRA DA COSTA
  • MARCIO EDUARDO DA COSTA RODRIGUES
  • Data: 26-dic-2019


  • Resumen Espectáculo
  • This work proposes the insertion of Complementary Split-ring Resonators (CSRR) in the planar filter ground. The intention is to improve some filter parameters found in the literature that have practical application, considering the growing demand for higher performance, smaller size, lower weight and cost of the devices that integrate the communication systems. The filters chosen were a bandpass filter that uses matrioska geometry for applications in the 2.45 GHz Industrial Scientific and Medical (ISM) band. The other filter was a low pass filter for digital TV applications. The idea is to insert the CSRR into the structure ground plane to increase the free attenuation of the matrioska band-reject filter and to reduce the size of the low-pass filter. In the work the numerical and experimental results are presented, for the purpose of validation of the proposed technique. A good agreement between the results is observed, these results show that the proposals presented were achieved.

Tesis
1
  • PATRIC LACOUTH DA SILVA
  • Research and Development of a New Synthesis Technique of Microwave Active Planar Circuits

  • Líder : ADAILDO GOMES D ASSUNCAO
  • MIEMBROS DE LA BANCA :
  • ADAILDO GOMES D ASSUNCAO
  • ADAILDO GOMES D ASSUNCAO JUNIOR
  • ALFREDO GOMES NETO
  • HERTZ WILTON DE CASTRO LINS
  • JOSE ALFREDO FERREIRA COSTA
  • LAERCIO MARTINS DE MENDONCA
  • Data: 14-mar-2019


  • Resumen Espectáculo
  • The work developed in this thesis has as main objective to contribute to the development
    of reconfigurable structures through the application of optimization techniques inspired
    by bee swarm behavior. The main structures studied are planar microwave antennas
    and frequency selective surfaces both for use in systems requiring frequency reconfiguration capability. We present new structures obtained with the classical algorithms and with the modified algorithm developed during the thesis. Simulated and measured results are obtained certifying the accuracy of the projects developed.

2
  • DÊNIS KEUTON ALVES
  • Real-time estimation of power and impedance estimation of the grid through the stationary packet wavelet transform

  • Líder : FLAVIO BEZERRA COSTA
  • MIEMBROS DE LA BANCA :
  • ALLAN DE MEDEIROS MARTINS
  • EDSON HIROKAZU WATANABE
  • FLAVIO BEZERRA COSTA
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • PAULO FERNANDO RIBEIRO
  • RICARDO LUCIO DE ARAUJO RIBEIRO
  • Data: 23-abr-2019


  • Resumen Espectáculo
  • The estimate of power quality indices has as primary goal provides to the monitoring systems (measurement, control, protection, etc.) information about proper levels of voltage and power quality. Besides power quality monitoring strategies, the knowledge of the grid impedance is fundamental importance to the understanding of the deviation of the electric power quality indices, providing relevant information to control and stability of the power system. In this context, in this thesis, it proposes the application of the stationary discrete wavelet packet transform (SDWPT) to estimate effective voltages and currents, as well as primary power quantities (active, total apparent, nonactive power, and power factor), and distortion power. Additionally, the analysis tools of signal processing developed in this work allowed the development of a method for estimating the grid impedance. The SDWPT provides a uniform frequency band and possesses the timeinvariance property, which is ideal for power estimation and grid impedance estimation in the real-time. The method of estimation of power quality indices was assessed and compared to the IEEE Standard 1459-2010 under different operational scenarios, where stationary synthetic waveforms and with both stationary and nonstationary power quality disturbances were emulating in an experimental setup. Regarding the grid impedance estimation, the current study was developing to identify the impedance of low voltage electric grids interconnected with distributed power generation systems. The primary goal is providing parameters to delimit the threshold of system instability in terms of the penetration level and identify islanding conditions. The validation of the proposed method was accomplished in a microgrid whose distributed generation is implemented through the photovoltaic plant of 8 kWp connected to the point of common coupling of a three-phase feeder composed by a transformer of 15 kVA. 


3
  • DIEGO HABIB SANTOS NOLASCO
  • Fuzzy Hierarchical Architecture with Additional Deffuzufication Leyer and Applications to the Power Quality Diagnosis 

  • Líder : FLAVIO BEZERRA COSTA
  • MIEMBROS DE LA BANCA :
  • BENJAMIN RENE CALLEJAS BEDREGAL
  • DANTON DIEGO FERREIRA
  • DENIS VINICIUS COURY
  • EDUARDO SILVA PALMEIRA
  • FLAVIO BEZERRA COSTA
  • LUIZ AFFONSO HENDERSON GUEDES DE OLIVEIRA
  • RICARDO LUCIO DE ARAUJO RIBEIRO
  • Data: 30-abr-2019


  • Resumen Espectáculo
  • Among various existing decision-making methods, hierarchical fuzzy methods have
    emerged as a suitable tool for dealing with complex applications which have many input
    variables and a high degree of subjectivity. In this context, the product of electric
    energy stands out. In general, the diagnosis of energy quality is a difficult practice due
    to the subjectivities inherent to the analysis process, nuances among different standards
    existing in the world, and uncertainties of evaluation parameters. This thesis proposes a
    new methodology for the power quality diagnosis based on the hierarchical fuzzy theory
    with a cascade-type architecture. The proposed method analyzes the quality parameters
    in steady-state electrical systems based on different existing standards in the world and
    performs a linguistic/quantitative diagnosis in which the contributions of the analyzed indices
    are weighted on the power quality of the evaluated system. Firstly, the diagnosis
    method was implemented from two hierarchical fuzzy architectures known (conventional
    and defuzzification free). Posteriorly, a new proposed architecture with additional
    defuzzification of layers was developed to aggregate the main advantages of conventional
    and defuzzification free in order to make the diagnosis method more complete and
    robust. This study proposes that the output of each subsystem obtained from primary
    decision-making process is transferred directly between the hierarchical layers, without
    loss of linguistic information, to obtain a resultant power quality diagnosis. In addition,
    a secondary decision-making process is performed together with an additional defuzzification
    method in order to obtain a complementary specific diagnosis at the out of each
    hierarchical subsystem. The diagnosis method based on the proposed fuzzy architecture
    presented satisfactory results when compared with the two existing architectures. After
    validation of the diagnosis method and hierarchical fuzzy architecture, both presented in
    this thesis, at the end of research, an adaptive wavelet-fuzzy system with generic inference
    method based on extended overlap functions is proposed as a new tool able of monitoring
    the power quality in renewable energy systems.

4
  • DANILO DE SANTANA PENA
  • Localization based on Acoustic Signals subject to Impulsive Noise 
  • Líder : ALLAN DE MEDEIROS MARTINS
  • MIEMBROS DE LA BANCA :
  • ALLAN DE MEDEIROS MARTINS
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • VICENTE ANGELO DE SOUSA JUNIOR
  • WALTER DA CRUZ FREITAS JÚNIOR
  • ÁLVARO AUGUSTO MACHADO DE MEDEIROS
  • Data: 14-may-2019


  • Resumen Espectáculo
  • The acoustic channel has received much attention in recent years due to many applications and some emerging technologies. As a result, researchers have considered realistic acoustical channels for studying of the source localization methods. Therefore, this work presents an approach of time difference of arrival (TDOA) for the impulsive acoustic channels based on non-linear data transform. The TDOA methods are evaluated in different scenarios using synthetic and real data. Moreover, the non-Gaussian impulsive noise models are analyzed in the various environment with measurement using an experimental setup.
5
  • PAULO CÉSAR LINHARES DA SILVA
  • Daubechies Wavelets Application in Conjunction with the Vectorial Beam Propagation Method in the Analysis of Photonics Structures.  

  • Líder : JOSE PATROCINIO DA SILVA
  • MIEMBROS DE LA BANCA :
  • JOSE PATROCINIO DA SILVA
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • VICENTE ANGELO DE SOUSA JUNIOR
  • ANTONIO RONALDO GOMES GARCIA
  • IDALMIR DE SOUZA QUEIROZ JÚNIOR
  • Data: 24-may-2019


  • Resumen Espectáculo
  • Wavelets are mathematical tools that allow the decomposition, description or representation of a given function. Among the various types of wavelets, Daubechies has a peculiar property of having the compact support, which allows describing the behavior of functions with discontinuities or abrupt variations of values in frequency and/or time domain. It is possible to obtain its coefficients, integrals, and derivatives by means of numerical procedures. In this context, the propagation of signals in devices that propagate electromagnetic waves can be numerically analyzed with the aid of these types of wavelets. In this work, we use the Daubechies wavelets as base functions for joint application with the Beam Vector Propagation Method (VBPM). In this case, these functions are obtained by means of the change in the translation and resolution of the wavelets and by the use of the moment generating function, obtained as part of this study. From the obtained wavelet base, an algorithm was developed to calculate elementary matrices, specific to the VBPM, which is based on the finite element (FEM) method. As a convergence test of the VBPM with the new set of base functions obtained in this work, we analyzed the wave propagation in an electromagnetically coupled guide and the transfer of energy between a conventional optical fiber (COF) and a near-photonic crystal optical fiber (FQCF).

     

6
  • KENNEDY REURISON LOPES
  • Expert System for Industrial Environment Based on Self-Learning Rules

  • Líder : ANDRE LAURINDO MAITELLI
  • MIEMBROS DE LA BANCA :
  • ANDRE LAURINDO MAITELLI
  • CARLOS EDUARDO TRABUCO DOREA
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • GILBERT AZEVEDO DA SILVA
  • OSCAR GABRIEL FILHO
  • Data: 24-may-2019


  • Resumen Espectáculo
  • This work presents a methodology of how to acquire and represent knowledge through automatic logic rules for a simulated industrial plant. The initial knowledge about an industrial process can be acquired through a specialist who interprets situations present in the plant and can describe what is occurring. In the work, a way of acquiring statistical knowledge of the plant during the execution of its processes is presented, using a method of online clustering known as TEDA-Cloud being modified to improve performance. The representation of knowledge is described through the manipulation of a neural network known as CILP and its own symbology is described to represent the logical variables drawn from the process signals. The results show an efficiency in interpreting the rules and acceleration in the process of clustering and classifications of the standards that define the rules.

7
  • ANDRÉ NASCIMENTO DA SILVA
  • New Techniques for the Impedance Matching Optimization of Printed Antennas for Wireless Communication Systems

  • Líder : ADAILDO GOMES D ASSUNCAO
  • MIEMBROS DE LA BANCA :
  • ADAILDO GOMES D ASSUNCAO
  • ADAILDO GOMES D ASSUNCAO JUNIOR
  • EVANDRO CONFORTI
  • FRED SIZENANDO ROSSITER PINHEIRO
  • JOABSON NOGUEIRA DE CARVALHO
  • LAERCIO MARTINS DE MENDONCA
  • Data: 27-may-2019


  • Resumen Espectáculo
  • The work presented in this dissertation consists in the development of new techniques for the impedance matching optimization of printed antennas for wireless communication systems. The techniques proposed are based on the insertion of slots in the antenna feeding line or below it (in the ground plane), enabling to keep unchanged the patch geometry of the investigated microstrip antenna. The study was performed using a parametric analysis and enabled the development of three new impedance matching techniques. In the first, the insertion of the slot is carried out in the conducting strip of the microstrip line, being symmetrically placed. In the second, the slot is introduced into the ground plane but it is inserted below and parallelly to the conducting strip of the microstrip line. In the third, the slot is also positioned in the ground plane, but is aligned parallel to the lower edge of the conducting patch, i.e. perpendicularly to the conducting strip of the microstrip line. The effect of the input impedance matching on the performance of rectangular patch microstrip antennas printed on glass fiber substrates (FR-4) was investigated. The antennas were fed by microstrip lines and designed for operation at 1.8 GHz, 2.45 GHz, and 3.5 GHz. Simulation and parametric analyses were performed using Ansoft Designer and HFSS softwares, that implement the method of moments (MoM) and the finite element method (FEM), respectively. In addition, the wave concept iterative procedure (WCIP) was also used in the analysis, enabling a better comparison between the obtained simulated results. For validation purpose, several prototypes were fabricated and measured. The good agreement observed between simulated and measured results confirms the efficiency and accuracy of the new impedance matching techniques. The proposed techniques are easy to implement and allow considerable reductions in the reflection coefficient value, | S11 | (dB), indicating that the antenna efficiency was significantly increased. In the first case, for the slot inserted in the conducting strip of the microstrip, the value of | S11 | (dB) was reduced from -8.81 dB to -28.09 dB in the analysis for 2.45 GHz. In the second case, for the slot inserted in the ground plane, below and parallel to the conducting strip of the microstrip, the value of | S11 | (dB) was reduced from -7.62 dB to -37.17 dB in the analysis for 2.45 GHz. In the third case, for the slot inserted in the ground plane, parallel to the lower edge of the conducting patch, the value of | S11 | (dB) was reduced from -10.11 dB to -34.1 dB in the analysis for 1.8 GHz. The use of slots, as proposed in this work, enabled a good impedance matching between the antenna and the microstrip feeding line, avoiding the use of high impedance microstrip lines, abrupt transitions in the width of microstrip lines, changes in the shape of the antenna conducting patch, or increasing the antenna size. The proposed optimization techniques present an excellent potential for applications in the development of other microwave integrated circuits, such as filters, power dividers, and directional couplers, in planar technology.

8
  • JAN'' ERIK MONT'' GOMERY PINTO
  • PLC Implementation of Piecewise Affine PI Controller Applied to Industrial Systems with Constraints

  • Líder : ANDRE LAURINDO MAITELLI
  • MIEMBROS DE LA BANCA :
  • ANDERSON LUIZ DE OLIVEIRA CAVALCANTI
  • ANDRE LAURINDO MAITELLI
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • FABIO SOARES DE LIMA
  • OSCAR GABRIEL FILHO
  • Data: 31-may-2019


  • Resumen Espectáculo
  • In this work the design of a piecewise affine proportional integral (PWA-PI) controller algorithm based on invariant set and multiparametric programming for constrained systems is proposed. We implemented the algorithm in a programmable logic controller (PLC) to control an industrial constrained level plant and analyze its behavior. Structured text routines were programmed and validated while controlling two systems with PLC. The results show that the constraints on the error, integral of the error, system output and control action are respected because PWA-PI controllers are tuned from the solution of an optimization problem. The evaluated performance indexes (such as mean square error, Goodhart,  overshoot and settling time) show that PWA-PI can be adjusted for better performance than proportional integral (PI) controller tuned by Ziegler–Nichols (Z–N) rules. In the analyzed cases, a settling time of 108 s was obtained, whereas PI controller with Z–N rules presented a 179 s settling time. All of the analyzed performance indexes that we used to evaluate both controllers show PWA-PI as a better controller for constrained systems.

9
  • MARCELO DAVID SILVA DE MESQUITA
  • SYNTHESIS OF A NEW CONDUCTIVE INK FOR MICROWAVE PRINTED CIRCUITS ON GLASS AND FIBERGLASS SUBSTRATES

  • Líder : ADAILDO GOMES D ASSUNCAO
  • MIEMBROS DE LA BANCA :
  • ADAILDO GOMES D ASSUNCAO
  • ADAILDO GOMES D ASSUNCAO JUNIOR
  • JEFFERSON COSTA E SILVA
  • JOAO BOSCO LUCENA DE OLIVEIRA
  • LAERCIO MARTINS DE MENDONCA
  • MARIA THEREZA MIRANDA ROCCO GIRALDI
  • Data: 31-may-2019


  • Resumen Espectáculo
  • This thesis describes the work carried out for the synthesis of a new conductive ink, with high concentration of silver particles, and the investigation on the possibility of its use in the manufacture of integrated circuits for wireless communication systems. The use of the new conductive ink is proposed as an alternative to the use of conventional microwave integrated circuits manufacturing techniques. The great interest in the use of conductive ink is associated with the ease, efficiency, versatility and reduction of manufacturing costs of the integrated circuits, provided by the use of small paint brushes for the application on different dielectric materials. The conductive ink obtained from the synthesis of nitrocellulose, which, in ethyl acetate solution, acts as a bonding agent of the pigment (silver metallic powder) to form a conductive film, one of the contributions of this work. The synthesized conductive ink was used in the development of microstrip antennas, bioinspired frequency selective surfaces (FSS), DC blocks, and directional couplers. The last two microwave integrated circuits were developed using parallel coupled microstrip lines. The simulation and design of these circuits were performed using Ansoft Designer and Ansoft HFSS softwares. In the analyses, glass and fiberglass substrates, very thin conductive plates made out of copper and silver, and surfaces painted with silver ink were considered. The frequency behavior of the scattering parameters of these circuits were investigated, enabling the evaluation of different aspects related to the transmission, reflection, coupling, radiation, and electromagnetic wave filtering. In order to validate the simulation results and to compare the measurement results, prototypes of microstrip antennas were fabricated and measured on fiberglass substrates with a copper laminate and with a surface painted with conductive ink. Also, prototypes of microstrip antennas were manufactured on glass substrate. The measurement of the reflection coefficient and the input impedance of these antennas were performed using a network analyzer. In the case of FSSs, prototypes were simulated and manufactured with elements inspired on the leaf of the Oxalis triangularis plant on fiberglass substrates, with copper clad laminate and surfaces painted with the synthesized conductive ink. The transmission coefficient measurement of these FSSs were performed using a network analyzer and two horn antennas. In the cases of DC block and directional coupler, prototypes were simulated on fiberglass substrates, with copper clad laminate and with surface painted with conductive ink, providing the opportunity to investigate the effect on the electromagnetic coupling between the painted parallel microstrip lines. The measurement of the reflection and transmission coefficients of these circuits were carried out using a network analyzer. The good agreement observed between the simulated and measured results for the microwave integrated circuit prototypes painted with silver ink confirms the efficiency and versatility of the painting procedure with the synthesized conductive ink.

10
  • TIAGO FERNANDO BARBOSA DE SOUSA
  • Equalization and Beamforming Proposals for Modern Communication Systems

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • MARCELO AUGUSTO COSTA FERNANDES
  • JOSE PATROCINIO DA SILVA
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • DALTON SOARES ARANTES
  • TOKTAM MAHMOODI
  • Data: 07-jun-2019


  • Resumen Espectáculo
  • This work aims to present new proposals for equalization and beamforming associated with modern digital communication systems. Three proposals are presented in which the first is characterized by an innovative neural equalizer structure called here the Butterfly Neural Equalizer (BNE). The BNE can be applied to linear and non-linear problems involved in intensity-modulated and direct-detection (IM-DD). The second proposal is characterized by an adaptive beamforming scheme applied to Orthogonal Frequency-Division Multiplexing (OFDM) systems. The OFDM technique and variants are being used in various communication systems such as LTE, digital TV, and others. Finally, the third proposal presents a neural beamforming structure called Butterfly Neural Beamforming (BNB) that can be applied to several communication systems. Results associated with the three proposals are presented and show the feasibility of the techniques in several channel scenarios and digital modulation schemes.

11
  • AYLANNA RAQUEL DA COSTA OLIVEIRA
  • Analysis of relief functions to estimate branch loadings after Corrective Switchings

  • Líder : MANOEL FIRMINO DE MEDEIROS JUNIOR
  • MIEMBROS DE LA BANCA :
  • CLOVIS BOSCO MENDONCA OLIVEIRA
  • MANOEL FIRMINO DE MEDEIROS JUNIOR
  • MARCOS ANTONIO DIAS DE ALMEIDA
  • MAX CHIANCA PIMENTEL FILHO
  • MELINDA CESIANARA SILVA DA CRUZ
  • Data: 29-jun-2019


  • Resumen Espectáculo
  • There are several studies proposing Corrective Switching technique for eliminating overloads in transmission branches. Some proposed Corrective Switching algorithms are based on electrical circuits theory or Relief Function methodology. This work presents a new linear methodology that proposes to compare the results obtained from the execution of a Relief Function applied to Corrective Switching technique in contingency cases with the results of other methodologies previously developed and with the load flow tests in transmission systems. This Relief Function has been determined analytically and it is based on a Two-ports network model. When carrying out the studies throughout this Thesis, contingencies were created from line disconnections in the studied transmission systems - this alteration occasionally caused multiple branch overloads in the electric network, but they were analyzed one at a time. Another step is to verify the nonlinearities of the switching measures: to analyze if the electric quantities obtained from the application of a linear technique can present great variations with respect to the obtaining from a nonlinear model, meaning, an exact load flow, and to comprehend the influence that certain parameters present in the estimated loadings can exert on the nonlinearity of these operation maneuvers. It is important to emphasize that the new network topology can be identified quickly, since the calculations necessary to estimate the new load are linear, which requires little computational effort. This enables an on-line application to find a feasible solution in Energy Management Centers.

12
  • IGOR GADÊLHA PEREIRA
  • Dynamic Synchronization of Distributed Multimedia Systems Using Machine Learning

  • Líder : LUIZ FELIPE DE QUEIROZ SILVEIRA
  • MIEMBROS DE LA BANCA :
  • ALLAN DE MEDEIROS MARTINS
  • ANDERSON ABNER DE SANTANA SOUZA
  • COSIMO DISTANTE
  • GUIDO LEMOS DE SOUZA FILHO
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • LUIZ MARCOS GARCIA GONCALVES
  • Data: 01-jul-2019


  • Resumen Espectáculo
  • The exchange of multimedia data streams in a distributed scenario represents a challenge mainly because of two problems generally present in this context: the synchronization between several streams and the management of heterogeneous hardware present in the system. The difficulty in maintaining the time limits necessary for the reproduction of time-dependent multimedia data stream increases with disturbances introduced by the communication channel, which is directly related to the quality of service. On the Internet, streaming real-time multimedia data requires efficient synchronization and storage strategies, which can add a significant amount of latency and interfere with the end-user quality of experience (QoE). Among the applications that require low latency to ensure an acceptable QoE, the composition and reproduction of songs through the Internet represent the biggest challenge. Towards this direction, we propose an effective method for low latency synchronization of several streams in a distributed global multimedia network, such as the Internet. Throughout the experiments we made to verify our proposed systems, we identified good accuracy levels on the synchronization of audio and video streams. We specifically emphasize the accuracy levels of 72% for the synchronization of video streams and up to 93% for the synchronization of raw audio sequences. In order to achieve these results, we employed statistical analysis of superior order and neural networks, both convolutional and recurrent, to compute a time relationship between multimedia flows.

13
  • ELIEL POGGI DOS SANTOS
  • Development of a numerical technique for the analysis of frequency selective surfaces of the type absorb-transmitt.

  • Líder : ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • MIEMBROS DE LA BANCA :
  • ALFREDO GOMES NETO
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • ERICO CADINELI BRAZ
  • MAURICIO WEBER BENJÓ DA SILVA
  • VICENTE ANGELO DE SOUSA JUNIOR
  • Data: 18-jul-2019


  • Resumen Espectáculo
  • Modern wireless communication systems use various types of devices such as antennas, modulators, filters, detectors, absorbers and many other types for the most diverse applications emerging in the field of telecommunications. In particular, electromagnetic wave absorbers have become critical components in security, detection and imaging systems. The fundamentals of the absorbers are based on configurations that allow the control of the absorption of electromagnetic waves and that by increasing the performance requirements of microwave absorption, good performance in civil applications and the emergence of new types of materials absorbers of microwaves, attracted considerable attention from the researchers resulting in considerable progress.

    In this sense, the full-wave methods and numerical methods are usually used for the analysis of several parameters in the telecommunications structures. They are used in many cases in the core of commercial software as well as in hybrid methods. The use of simpler numerical techniques allows an analysis of reflection and transmission behavior in a suitable way with good approximation mainly in cascade structures.

    In this work the absorber structure is composed of a conductive frequency selective surface (FSS) and a resistive FSS in which there is a d spacing between the FSS. The conductive FSS is formed by a FR4 dielectric with an electric permittivity of 4.4 and the geometries formed by conductive copper strips. Resistive FSS have an FR4 dielectric with an electrical permittivity of 4.4 and the geometries are formed by OhmegaPly 1A50PT (35 micron - 50 ohms / sq) resistive materials.

    The thesis proposal is to develop a numerical technique for the analysis of integrated microwave absorbers to selective surfaces in frequency. A model was used for conductive FSS and one for resistive FSS both using the equivalent circuit technique based on the Marcuvitz equations and for analysis of the set or cascade of the structures, the scattering matrix technique was used. Attempts have been made to use other techniques, however, this combination has resulted in better results. Internal parameters of the method such as coefficients, resistance, inductive and capacitive reactances of the geometries were analyzed in order to allow a better approximation with respect to the simulations and measurements.

    The designed absorbers used cross-dipole geometry and square loop both designed and measured in the laboratory. The numerical analysis technique showed good agreement with the measured and simulated results of the author's own structures as well as with other measured and simulated absorbers presented in the literature. Improvements and modifications can be tuned for other applications, integrate with optimization techniques, hybrid algorithms and other projects of periodic structures.

14
  • EMANOEL RAIMUNDO QUEIROZ CHAVES JUNIOR
  • Strategies for Fault Estimation in Actuators and Sensors in Non-Linear Processes with Uncertainties.
  • Líder : ANDRE LAURINDO MAITELLI
  • MIEMBROS DE LA BANCA :
  • ANDERSON LUIZ DE OLIVEIRA CAVALCANTI
  • ANDRE LAURINDO MAITELLI
  • CARLOS EDUARDO TRABUCO DOREA
  • MARCELO ROBERTO BASTOS GUERRA VALE
  • OSCAR GABRIEL FILHO
  • Data: 26-jul-2019


  • Resumen Espectáculo
  • Over the years, control processes have become more complex, containing a large number
    of components that work in an integrated manner. Any of these components is subject
    to defects or malfunctions. All these factors are defined as faults, which are unexpected
    variations of the properties of a given component to its nominal operating condition. The
    consequences of failures may cause economic losses and risk the life of the workers in
    the enclosure. A fault tolerant control system is able to keep the control process running
    with proper performance even in the presence of failures. In its active approach, the nominal
    control strategy is reconfigured so that the effect of the fault is accommodated. This
    reconfiguration is usually based on the estimate of the fault, which is obtained by means
    of an observer. Generally the effectiveness of an observer is related to the degree of knowledge
    about the process by the designer. An unforeseen change in system parameters
    or the presence of uncertainties may adversely affect the performance of the observer.
    This work proposes two state observer design techniques which are able to perform the
    simultaneous estimation of states and faults in actuators and sensors in nonlinear systems
    in discrete time with uncertainties. The operation of this method is verified by means of
    computational simulations based on case studies involving crane, liquid level processes
    and control of a flexible articulated robotic manipulator.

15
  • BETOVEN OLIVEIRA DE ANDRADE
  • Analysis and Design of Microstrip Patch Antennas  with substrate PBG printed 3D

  • Líder : LAERCIO MARTINS DE MENDONCA
  • MIEMBROS DE LA BANCA :
  • LAERCIO MARTINS DE MENDONCA
  • ADAILDO GOMES D ASSUNCAO
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • JOSE PATROCINIO DA SILVA
  • HUMBERTO DIONISIO DE ANDRADE
  • WELLINGTON CANDEIA DE ARAUJO
  • Data: 26-jul-2019


  • Resumen Espectáculo
  • After a few decades of the first publications on microstrip antennas, research continues to expand. This is the consequence of several of its attractive features, especially the possibility of mounting in the same plane of the transceiver circuits. However, this advantage presents some problems - the waves produced by the circuits reflect and refract inside the substrate, since it has different refractive indices, which causes surface currents in the earth plane, leading to a narrow bandwidth and also to lower radiation efficiency and low gain, since some of the energy is wasted. In order to attenuate such negative characteristics, several techniques have been investigated and employed, one of the most promising being the use of PBG (Photonic Band Gap) substrates. PBGs are periodic structures in a dielectric or conductive material. They propagate the electromagnetic waves at a frequency specific to all states of polarization and angles of incidence. In this thesis, it is proposed an adequate methodology to study new structures and techniques aimed at the use of PBG material in the form of three-dimensional periodic (3D) holes in the substrate in the manufacture of microstrip patch antennas for the range between 1 to 4 GHz. This thesis presents some contributions, as described below: (i) proposition of new antennas, including antennas with printed substrates (3D); (ii) theoretical development of coherent antenna design, avoiding resonance frequency shifts, caused by changes in the permittivity after insertion of the holes; (iii) demonstration that different hole geometries are not related to resonance frequency variation; (iv) analysis that an unconventional geometry can promote large efficiency improvements in computational design time; (v) applying an artificial neural network capable of substantially improving the time to obtain the return losses; (vi) proposition of low-cost techniques for the development of antennas with PBG substrate. In order to validate the proposed antennas, simulations and measurements are presented and discussed.

16
  • ÍCARO BEZERRA QUEIROZ DE ARAÚJO
  • Maximum correntropy criterion applied to structure selection and parameter estimation of NARX models

  • Líder : FABIO MENEGHETTI UGULINO DE ARAUJO
  • MIEMBROS DE LA BANCA :
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • ALLAN DE MEDEIROS MARTINS
  • JOILSON BATISTA DE ALMEIDA REGO
  • EVANDRO DE BARROS COSTA
  • JOSE BEZERRA DE MENEZES FILHO
  • Data: 26-jul-2019


  • Resumen Espectáculo
  • In the last decades, due to the growing complexity of dynamic systems and the growing demand for better performance, the area of systems identification has emphasized the use of non-linear models to represent dynamic systems. In this context, Non-linear autoregressive with exogenous inputs models (NARX) are heavily used due to to their simplicity, flexibility and capacity of better representation. However, such models rely heavily on structure selection and the most traditional algorithms have limitations when the data is contaminated by non-gaussian distribution noises. Noting this, in this thesis, the objective is to present a new identification method called simulated correntropy maximization with pruning which uses concepts of learning based on information theory. In this work basic concepts about systems identification and correntropy, methods based on orthogonal least squares and simulated error reduction, and the new proposed methodology. The proposed method is applied and compared to the traditional methods in some study cases. The first experiment is composed by three SISO numeric dynamic systems in the presence of bimodal noise. The second study case is a set taken from a benchmark system called Silver Box. The third is a real dynamic system. The fourth and last one is a real dynamic MIMO coupled-tanks system. The obtained results validate the performance of the proposed method when compared to other algorithms of structure detection and parameter estimation, showing that the proposed method presents a better and more robust performance in the presence of non-gaussian distribution noise.

17
  • CLÁUDIO PEREIRA DA COSTA
  • Analysis of New Topologies of CPS, SIEW and FSS-EBG Filters for Wireless Communication Systems

  • Líder : ADAILDO GOMES D ASSUNCAO
  • MIEMBROS DE LA BANCA :
  • ADAILDO GOMES D ASSUNCAO
  • ADAILDO GOMES D ASSUNCAO JUNIOR
  • CRISTHIANNE DE FATIMA LINHARES DE VASCONCELOS
  • GERVASIO PROTASIO DOS SANTOS CAVALCANTE
  • JOABSON NOGUEIRA DE CARVALHO
  • VALDEMIR PRAXEDES DA SILVA NETO
  • Data: 30-jul-2019


  • Resumen Espectáculo
  • The objective of this thesis is to present the analysis of new topologies of CPS, SIEW and FSS-EBG filters for wireless systems applications. The coplanar stripline (CPS) structure consists of two conducting strips loaded by two split ring resonators (SRRs), forming a CPS filter. The SIEW filter is based on the integrated E-plane waveguide substrate technology, being composed of several dielectric layers stacked with two resonant elements inside it, enabling selectivity with horizontal polarization. The third filter configuration consists of two periodic arrays formed by an FSS and an EBG, composing an integrated structure. The investigation of these structures was carried out by numerical analysis of the frequency response due to the variation of the physical parameters of the investigated geometries, identifying the attractive potentialities for microwave applications. The simulated results were obtained with the commercial software Ansoft HFSS and exhibited attractive characteristics as: reconfiguration, wide bandwidth, double band, simple design, light weight and easy manufacturing on a large scale. Some prototypes of the proposed devices were manufactured and the experimental results confirmed the validity of the used computational models. The experimental results were also compared with the simulated results, exhibiting a good agreement between them.

18
  • JOÃO PAULO FERREIRA GUIMARÃES
  • Complex Correntropy: Definition, Properties and Applications

  • Líder : ALLAN DE MEDEIROS MARTINS
  • MIEMBROS DE LA BANCA :
  • ALLAN DE MEDEIROS MARTINS
  • ADRIAO DUARTE DORIA NETO
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • RICARDO VON BORRIES
  • ALUISIO IGOR REGO FONTES
  • Data: 30-sep-2019


  • Resumen Espectáculo
  • Recent studies have demonstrated that correntropy is an efficient tool for analyzing higher-order statistical moments in non-Gaussian noise environments. Although corren- tropy has been used with complex data, no theoretical study was pursued to elucidate its properties, nor how to best use it for optimization. By using a probabilistic inter- pretation, this work presents a novel similarity measure between two complex random variables, which is defined as complex correntropy. It’s properties are studied as well as a new recursive solution for the Maximum Complex Correntropy Criteria (MCCC) and two algorithms are derived, one based in the ascendent gradient and a second one on a fixed-point solution. Simulations were made in order to evaluate how the robust this new measure is to impulsive noise in different problems: liner system identification, chan- nel equalization and in a compressive sensing problem. It is also shown the application of complex correntropy as a tool to analyse the similarity between angles. The results demonstrate prominent advantages of the proposed method when compared with the clas- sical algorithms from the literature.

19
  • ITALO AUGUSTO SOUZA DE ASSIS
  • Intra-node and Inter-node load balancing and other scalable approaches for high-performance seismic processing

  • Líder : SAMUEL XAVIER DE SOUZA
  • MIEMBROS DE LA BANCA :
  • JOAO MEDEIROS DE ARAUJO
  • JORGE DANTAS DE MELO
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • REYNAM DA CRUZ PESTANA
  • SAMUEL XAVIER DE SOUZA
  • Data: 14-oct-2019


  • Resumen Espectáculo
  • Seismic modeling, reverse time migration (RTM), and multi-scale waveform inversion (MFWI) are three of the most important techniques in seismic surveying. Seismic modeling simulates the wave propagation, RTM generates an image of the subsurface, and MFWI produces a wave propagation velocity model. These methods demand intensive computational cost due to a large amount of data they process and the complexity of their algorithms. Because of that, they are only implemented for parallel systems in practical. Although there are efficient parallel implementations of modeling, RTM, and MFWI in the literature, further improvement can be achieved by better exploring the parallelism in these methods and the characteristics of the current parallel systems. This research proposes coupled multi-scale waveform inversion (CMFWI), an alternative method to MFWI, which improves parallel scalability by reducing the parallel dependency between the processing of different frequency content of the data. An implementation of CMFWI using the coupled local minimizers method (CLM) is presented. L2-norm results showed that CMFWI had an inferior performance when compared to MFWI. These experiments indicate that further research is necessary to implement CMFWI as it compares data with different frequency contents. This work also introduces an auto-tuning strategy for properly choosing the optimal chunk size that reduces the runtime of a 3D RTM algorithm in shared memory systems. A coupled simulated annealing method (CSA) is employed to adjust the chunk size of work that parallel loops assign dynamically to worker threads. Experiments show that the proposed method is consistently better than two default OpenMP loop schedulers being up to 44% faster. This thesis also introduces the cyclic token-based work-stealing (CTWS) for distributed memory systems. The novel cyclic token approach reduces the number of failed steals, avoids communication overhead, and simplifies the victim selection and the termination strategy. Results obtained by applying the proposed technique to balance the workload of a 3D RTM present a factor of 14.1% speedup and reductions of the load imbalance of 78.4% when compared to the conventional static distribution. Finally, an implementation of a 2D visco-acoustic modeling is presented.

20
  • DANIEL GUERRA VALE DA FONSECA
  • Explicit Formulation for Generalized Predicitve Control: a Multiparametric Approach

  • Líder : ANDRE LAURINDO MAITELLI
  • MIEMBROS DE LA BANCA :
  • ANDERSON LUIZ DE OLIVEIRA CAVALCANTI
  • ANDRÉ FELIPE OLIVEIRA DE AZEVEDO DANTAS
  • ANDRE LAURINDO MAITELLI
  • CARLOS EDUARDO TRABUCO DOREA
  • PERICLES REZENDE BARROS
  • Data: 30-oct-2019


  • Resumen Espectáculo
  • Generalized Predictive Control (GPC) is one of the most traditional and popular Model-based Predictive Control (MPC) techniques in industry and academia and has been applied over decades in several systems to improve the control performance. Among the application areas, GPC can be found in petrochemical, energy, food, automotive and aerospace. This type of controller uses process model information to predict future system behavior. The prediction is made by minimizing a functional cost that produces optimal control actions capable of leading the controlled variables to the desired values. In addition, GPC can deal directly with both MIMO systems and process constraints. However, when considering the constraint set, the controller needs to solve a Quadratic Programming (QP) in real time, which can be prohibitive in certain cases, such as for embedded systems. This work uses multiparametric programming (mp) to generate an Explicit Piece-wise Affine (PWA) control law for GPC (mp-GPC) which holds the same control performance without the need to keep solving the optimization problem at each sample time. Hence, initially, the proposed formulation is compared with GPC based on online QP. The results show that the performance is maintained, reducing the computational time to calculate the control action. Then, a new format is proposed, which differs from the last one by the amount of parameters needed in the mp formulation. Both propositions are applied in three different situations: a MIMO system, a process with input-output delays and a underactuated system. A comparison is made by checking the computational time spent to calculate the control signal, as well as the time required for mp resolution. Finally, studies were initiated involving a Hybrid Multiparametric GPC formulation, which makes use of the resolution of a multiparametric Mixed-Integer Linear Programming (mp-MILP). A nonlinear valve is used as a case study, in which its nonlinear characteristics are treated as discrete dynamics for the optimization problem, in order to minimize its effects. Preliminary results are satisfactory, however, tests on real applications need to be done in future works.

21
  • VIRGINIA PINTO CAMPOS
  • System for Automatic Audio-description Generation from Video Content Analysis

  • Líder : LUIZ MARCOS GARCIA GONCALVES
  • MIEMBROS DE LA BANCA :
  • LUIZ MARCOS GARCIA GONCALVES
  • RAFAEL BESERRA GOMES
  • BRUNO MARQUES FERREIRA DA SILVA
  • NEY ROBINSON SALVI DOS REIS
  • TIAGO MARITAN UGULINO DE ARAUJO
  • Data: 28-nov-2019


  • Resumen Espectáculo
  • Audio description is an accessibility feature designed to make visual information accessible to blind or low vision people. To increase the range of audio description tracks in digital video applications, we propose a system for automatic audio description generation of movies. The system can use as source of information about the film the original script or the video itself. As a proof of concept, we developed a prototype that automatically generates audio description based on actions taken from the script and objects recognized in the video. The experiments contemplated the application of the solution in fiction films and surveillance videos. For fiction films, an evaluation was made with blind people. The results indicated that through the automatic audio description generated by the solution, it was possible to provide contextual information that can help the user in the general understanding of the story. For surveillance videos, a performance evaluation was made using the delay time of each component. Results indicate that a solution has the potential to be used in contexts that require real-time AD.

22
  • THALES AUGUSTO DE OLIVEIRA RAMOS
  • An Efficient System of Electric Machines for Wind Power Generation driven by an Electromagnetic Frequency Regulator

  • Líder : MANOEL FIRMINO DE MEDEIROS JUNIOR
  • MIEMBROS DE LA BANCA :
  • JOSE TAVARES DE OLIVEIRA
  • JOÃO PAULO ABREU VIEIRA
  • MANOEL FIRMINO DE MEDEIROS JUNIOR
  • PAULO VITOR SILVA
  • RICARDO FERREIRA PINHEIRO
  • Data: 29-nov-2019


  • Resumen Espectáculo
  • A new topology has recently been developed to drive generators to enable hybridization of two power sources. Particularly, if one of these sources is wind power, it also serves the purpose of avoiding power electronic devices directly connected to the grid. The system consists of a squirrel cage rotor induction machine and a rotational armature with a three-phase winding that is powered by a secondary source. This new machine frame has been named the Electromagnetic Frequency Regulator (EFR). The objective established in a first research was to convert a variable speed imposed by the wind turbine to the armature, in a constant speed to be developed by the cage rotor, driving a synchronous generator shaft. A first objective of this Thesis was to mathematically model the topology and simulate it inserted in a large wind turbine. Then, it was proposed to use an induction generator instead of a synchronous generator, aiming to find the maximum extraction point of available wind power. From steady state analyzes, two possible control systems are proposed: a scalar control, which aims to obtain a desired speed for the system rotor, and a control that aims at the strategy of obtaining the maximum possible efficiency for the REF. Permanent regime simulations are performed in order to compare the two control systems presented. Thus, it is verified that the control by maximum efficiency strategy obtains better results when compared with the correspondents obtained from the scalar control. Dynamic simulations performed for the two proposed topologies show that both are viable and support variations in wind speed and disturbances in the power grid.

23
  • HEITOR MEDEIROS FLORENCIO
  • Analytical Model of Link Stability in Industrial Wireless Sensor Networks.

  • Líder : ADRIAO DUARTE DORIA NETO
  • MIEMBROS DE LA BANCA :
  • ADRIAO DUARTE DORIA NETO
  • ANDRES ORTIZ SALAZAR
  • DANIEL LOPES MARTINS
  • DENNIS BRANDAO
  • JORGE DANTAS DE MELO
  • Data: 09-dic-2019


  • Resumen Espectáculo
  • Wireless Sensor Networks (WSN) have emerged in the industrial environment with the purpose of increasing flexibility and deploying measurement and actuation elements in hard to reach areas. However, due to the peculiarities of wireless communication in the industrial environment, the use of metrics to evaluate the performance of communication links is crucial. An essential factor in assessing network behavior is the stability of links. The stability is a property, but several studies define the stability model using different parameters of the link. This thesis presents an analytical model of link stability based on received signal strength and packet delivery rate. The results obtained through tests performed with different scenarios showed that the analytical model can identify link installability situations when it comes to high variation in received signal strength and low packet delivery. Then, an ISA 100.11a network control system was developed to validate the model proposed in this work. The results obtained with the application of the model in the level control system presented steady state error values below the predetermined maximum limit, while in the control tests that did not consider the link stability factor the error exceeded the limit.

24
  • MISSILENE DA SILVA FARIAS
  • State Feedback Type-2 Fyzzy Rule-Based Applied to Nonlinear Systems Servocontrol

  • Líder : FABIO MENEGHETTI UGULINO DE ARAUJO
  • MIEMBROS DE LA BANCA :
  • ANDRE LAURINDO MAITELLI
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • JOSE BEZERRA DE MENEZES FILHO
  • OSCAR GABRIEL FILHO
  • WALLACE MOREIRA BESSA
  • Data: 09-dic-2019


  • Resumen Espectáculo
  • Many researches use T-S fuzzy models to accurately represent nonlinear dynamic systems. However, T-S fuzzy makes the implementation of fuzzy controller more complex as system order and nonlinearities increase. Thus, the present work is aimed to overcome these limitations by using an Interval Type-2 Fuzzy Rule-Based System in which the membership functions and the number of rules can be freely chosen simplifying the implementation of the technique. To this end, it is established a direct state feedback control with reference tracking to generate the nonlinear control action using parallel distributed compensation techniques with no need to include T-S fuzzy models to describe the dynamic system. The proposed strategy is applied to a synchronous generator and also to a magnetic levitation system. From the results, it was verified that SBRF T2Is are able to stabilize the systems analyzed at different equilibrium points with higher performance and less settling times, given the uncertainties in the linearized model. In fact, the SBRF T2I proved to be a proper way to accomplish this task, because fuzzy logic control itself does not depend on an accurate model.

25
  • ISAAC BARROS TAVARES DA SILVA
  • Development of NiFe2O4 Ferrimagnetic Material for Application in Microwave Communication Systems

  • Líder : ADAILDO GOMES D ASSUNCAO
  • MIEMBROS DE LA BANCA :
  • ADAILDO GOMES D ASSUNCAO
  • IDALMIR DE SOUZA QUEIROZ JÚNIOR
  • JOABSON NOGUEIRA DE CARVALHO
  • JOAO BOSCO LUCENA DE OLIVEIRA
  • LAERCIO MARTINS DE MENDONCA
  • MARCELO DAVID SILVA DE MESQUITA
  • VALDEMIR PRAXEDES DA SILVA NETO
  • Data: 16-dic-2019


  • Resumen Espectáculo
  • Due the need to develop new wireless technologies, the researches involving communication devices has become intense, mainly in characteristics like device miniaturization and reconfigurability. With this objective, this work presents a proposal for the development of ferrimagnetic compounds aiming their application in wireless communication devices. The project was performed starting at the nickel ferrite (NiFe2O4) through Sol-Gel modified by ICR technique. A calcination process was applied to eliminate organic compounds from Sol-Gel, as the large particle size production, over 300 nm. The calcination was performed at 1100 °C, for ten hours, with 10 °C/min step. The obtained particles were characterized by SEM, XRD and EDS. Magnetic behavior (hysteresis loop) was also obtained, as complex electric permittivity parameter. All performed morphological characterization were applied in a study of Sol-Gel process efficiency, according to nitrate concentration on chitosan matrix. The nitrate concentration was modified to verify which concentration promote better process efficiency, so that 120 g/L concentration could be verified as the more suitable for a high-quality product. Posteriorly, one of the proposed devices was manufactured: ferrite powder encapsulated with low permittivity wood material, aiming FRA application. Electric characterization was performed in all four types of applied woods. Reflection coefficient was determined and discussed. Then, ferrite bulk materials were manufactured using a 15-ton uniaxial press with propanone binding agent, followed by heating process before sintering. This process was performed with same temperature parameters of the powder production, aiming to not change microscopical characteristics. Electric characterization of ferrite bulk was performed and its application as microstrip patch antenna substrate and FRA was verified. Magnetic permeability was estimated based on previous papers characterization of NiFe2O4 powder. With electric characterization, simulations of FRA, with three different types of excitation, and proposed antennas was performed by HFSS software, using FEM method and the reflection coefficient was experimentally determined to validate the simulations. The experimental results presented good agreement with all performed process of production and characterization, reaching a 0.82% error in resonance frequency based on simulated results. The nickel ferrite potential as FRA and microstrip patch antenna is evidenced along the results discussion, being possible to expand the applications possibilities in other electrical engineering fields.

26
  • FRANCISCO SALES DE LIMA FILHO
  • Smart Defender: A DoS / DDoS Attack Detection and Mitigation System Using Machine Learning

  • Líder : LUIZ FELIPE DE QUEIROZ SILVEIRA
  • MIEMBROS DE LA BANCA :
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • AGOSTINHO DE MEDEIROS BRITO JUNIOR
  • PAULO SERGIO DA MOTTA PIRES
  • GENOVEVA VARGAS-SOLAR
  • PLACIDO ANTONIO DE SOUZA NETO
  • Data: 20-dic-2019


  • Resumen Espectáculo
  • Denial of service (DoS) attacks, and their distributed variant (DDoS), represents one of the most significant risks to the availability of services operating on the Internet. This cyber threat continues to grow worldwide and is detrimental even as network protection technologies advance. Developing mechanisms that can detect and mitigate the impacts of malicious traffic generated by DDoS attacks is a current challenge in network security, with a direct impact on the lives of the world's population. This paper proposes the Smart Defender system as a solution, consisting of Smart Detection (SD) and Smart Protection (SP) tools that operate in an integrated and distributed manner. The proposed detection system makes inferences based on pre-existing signatures using Machine Learning (ML) techniques to classify network traffic. The protection system, in turn, applies rules to control useless traffic and mitigate the effects of DDoS, considering the information shared by the detection system and the security policy of the local network. In the experiments performed, modern reference data sets and semi-controlled laboratory environment tests were used. The results show that the proposed solution is capable of early detection of high hit rate and low false alarm rate DDoS attacks, as well as isolating the threat in the first minute of attack.

27
  • ANDERSON COSTA SILVA DOS SANTOS
  • System to assist installation and maintenance of control-focused industrial wireless networks

     

  • Líder : ADRIAO DUARTE DORIA NETO
  • MIEMBROS DE LA BANCA :
  • ADRIAO DUARTE DORIA NETO
  • ANDRE LAURINDO MAITELLI
  • DANIEL LOPES MARTINS
  • DANIELLE SIMONE DA SILVA CASILLO
  • JORGE DANTAS DE MELO
  • Data: 27-dic-2019


  • Resumen Espectáculo
  • Nowadays, wireless communication has aroused interest and is a tendency in industry environments. In part this interest is due to the ease of deployment and maintenance of this network, which dispenses sophisticated designs and execution of works of deploying cables, which in industrial environment often prohibitively expensive, and enabling the addition of new applications when compared to their wired counterparts. Despite its high degree of applicability, a wireless network faces some technical (reliability, energy consumption, environment interference) and cultural challenges present as the main negative points for the deployment of large-scale industrial networks in the industry. To reduce some these problems, this work presents a system with multiobjective modeling to aid installation and maintenance of devices - sensors and actuators - in industrial wireless networks, allowing the use of controllers in these networks, with emphasis on WirelessHART and ISA100.11a standards, which presents themselves as the most promising solution for industrial wireless networks.

2018
Disertaciones
1
  • BRUNO SÁTIRO DA SILVA
  • Complementary Frequency Selective Surface (CFSS): Concept and Design

  • Líder : ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • MIEMBROS DE LA BANCA :
  • ALFREDO GOMES NETO
  • ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
  • MARCIO EDUARDO DA COSTA RODRIGUES
  • Data: 24-ene-2018


  • Resumen Espectáculo
  • Wireless communication is an integral part of modern society, having fundamental role in the process of globalization and technological development. Therefore, several researches are carried out in order to achieve constant optimization in this area. Frequency Selective Surfaces (FSS) have been approached for decades because of its telecommunications functionalities. Due to evolution of technology, new demands arise every day with specific answers for each application, thus, the FSSs need to be modeled to meet the most diverse requirements. In this context, the present work aims to show a specific model of FSS. The Complementary Frequency Selective Surfaces (CFSS) have very interesting characteristics for modern communication systems, such as multi-band frequency behavior, high selectivity in transmission bands and great angular stability, being useful in a wide range of applications, such as aircrafts, radomes and satellite communication. As part of this work, new unit-cell geometries for CFSS will be proposed and analyzed, comparing the simulated results with those of constructed prototypes and indicating possible applications.

2
  • WINNIE DE LIMA TORRES
  • Detection of Vocal Deviations Using Auto Regressive Models and the KNN Algorithm

  • Líder : ALDAYR DANTAS DE ARAUJO
  • MIEMBROS DE LA BANCA :
  • ALDAYR DANTAS DE ARAUJO
  • ALLAN DE MEDEIROS MARTINS
  • ADEMAR GONÇALVES DA COSTA JÚNIOR
  • Data: 30-ene-2018


  • Resumen Espectáculo
  • Some fields of science propose to study disorders in the vocal tract from analyzes on patterns of vibration of the voice. In general, the importance of these researches is to identify, in a more specific stage, diseases of greater or lesser severity, to be healed with vocal therapy or that require more attention, even generating the need for surgical procedures for its control. Although there are already indications in the literature that digital signal processing allows non-invasive diagnosis of laryngeal pathologies, such as vocal diseases that cause edema, nodule and paralysis, there is no definition of the most indicated method and the most characteristic features or parameters to detect the presence of vocal deviations. Thus, in this work an algorithm is proposed for the detection of vocal deviations through the analysis of speech signals. To perform this work, we used constant data in the database Disordered Voice Database, developed by the Massachusetts Eye and Ear Infirmary (MEEI), due to its use in research in the acoustic area of voice. We used 166 signals contained in this database, with signs of healthy voices and pathological voices affected by edema, nodule and paralysis in the vocal folds. From the voice signals, Auto Regressive models (AR and ARMA) were generated to represent these signals and, using the parameters of the obtained models, the KNN algorithm was used to classify the analyzed signals. In order to analyze the efficiency of the algorithm proposed in this study, the results obtained from this algorithm were compared with a detection method considering only Euclidean distance between the signals. The results show that the method proposed in this work presents a good result, generating a rating rate above 71% (greater than 31% from the use of Euclidean distance). In addition, the method used is easy to implement and can be used in simpler hardware. Therefore, this research has the potential to generate an inexpensive and affordable classifier for large-scale use by health professionals as a non-invasive pre-analysis alternative for the detection of otorhinolaryngological pathologies that affect the voice.

3
  • PEDRO YOCHINORI GUSHIKEN
  • Second Level Adaptation as a Parameter Estimation Technique and its Application to Model Reference Adaptive Control

  • Líder : ALDAYR DANTAS DE ARAUJO
  • MIEMBROS DE LA BANCA :
  • ALDAYR DANTAS DE ARAUJO
  • JOSENALDE BARBOSA DE OLIVEIRA
  • FRANCISCO DAS CHAGAS DA SILVA JÚNIOR
  • Data: 31-ene-2018


  • Resumen Espectáculo
  • Adaptive control is used when there is parametric uncertainty in the controlled system. There are two distinct strategies: direct and indirect. Direct adaptive control estimates the controller parameters in real time. In indirect adaptive control the plant parameters are estimated in real time through an identification model and these estimates are used to calculate the controller parameters. There are several identification models that can be used such as parallel, series-parallel and linear regression models. There are several parameter estimation methods that can be used to update these models that are suitable to be used in conjunction with indirect adaptive control such as the least squares method and normalized gradient method. In this context a new method for parameter estimation based on multiple identification models was recently proposed, known as second-level adaptation. In second level adaptation the estimates of the plant parameters are obtained by convex combinations of the estimates from multiple identification models. The coefficients for these convex combinations are also estimated, as a second level of adaptation. In this dissertation we demonstrate the concept of second level adaptation as a parameter estimation method for the case of a plant of degree unity, the case of a plant of degree n with single input and output available for measurement (SISO) and the case of a plant of degree n with p inputs whose state vector is available for measurement (MIMO). We propose a modified form of the adaptive law for second level adaptation. In all cases simulation studies show that the estimates reach their true values faster with second level adaptation compared to individual identification models and that the proposed modification is even faster and also smoother in this regard. Finally, we apply second level adaptation based on linear regression identification models updated through the normalized gradient method to the problem of model reference adaptive control(MRAC) of a degree n SISO plant with relative degree one. The closed-loop stability of the system is demonstrated. Simulation results show that the control signal generated with second level adaptation yields better results of model reference tracking compared to individual identification models.

4
  • ISAAC DANTAS ISIDÓRIO
  • Variable Structure Adaptive Pole Placement Control for Supression of Chaos in the Lorenz System

  • Líder : ALDAYR DANTAS DE ARAUJO
  • MIEMBROS DE LA BANCA :
  • ALDAYR DANTAS DE ARAUJO
  • CAIO DORNELES CUNHA
  • ODAILSON CAVALCANTE DE OLIVEIRA
  • Data: 31-ene-2018


  • Resumen Espectáculo
  • Two main methods highlight for the design of adaptive controllers: the model reference adaptive control (MRAC) and the adaptive pole placement control (APPC). In MRAC, a reference model is chosen to generate a trajectory, which must be followed by the output of the plant to be controlled. This type of project may involve cancellation of the plant zeros, not applicable to non-minimum phase plants. APPC, in turn, is considered the most general type of adaptive control, presenting a design methodology for the controller and adaptation law quite flexible, besides not involving  plant zero-pole cancellations. The combination of the conventional adaptive methods, that use integral adaptive laws, with variable structure systems (VSC) based in switching laws, allows to aggregate fast transient and robustness to disturbances and parametric uncertainties. In this context, based on a class of pole placement control schemes (PPC), the variable structure adaptive pole placement control (VS-APPC) arises. Here, such as in APPC, the control law is generated for the case with known parameters (transfer function coefficients), replacing them by their estimates. The estimation process identifies the adaptive method used, so that APPC uses integral laws while the VS-APPC makes use of switching laws. In this work is proposed, based on the variable structure adaptive pole placement control theory, a control system able to control the chaos phenomenon in Lorenz system, using only the measured plant's output and input variables. The parameters of system are considered unknown and is considered the disturbance presence in input system. It is shown that in closed-loop system, the output tracks the reference trajectory and the state vector converges to the equilibrium state, presenting good transient behavior and robustness to the parametrics uncertainties as also in presence of disturbance in input plant.


5
  • HUGO RAFAEL GONÇALVES CAVALCANTE
  • Acquisition of frequency in low-orbit satellite by maximum correntropy criterion.

  • Líder : LUIZ FELIPE DE QUEIROZ SILVEIRA
  • MIEMBROS DE LA BANCA :
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • SAMUEL XAVIER DE SOUZA
  • ALUISIO IGOR REGO FONTES
  • JOSÉ MARCELO LIMA DUARTE
  • Data: 16-feb-2018


  • Resumen Espectáculo
  • Low-earth orbit (LEO) satellites move at very high speeds and can circumscribe the planet several times in a single day. These high dynamics environments induce considerable Doppler shift and Doppler rate values in spatial communications signals, which is known as dynamic Doppler shift. Establishing communication with these systems is a multi-task process and, due to the dynamic Doppler shift, a critical task is the fine acquisition of the carrier frequency. Several studies have investigated frequency estimation techniques of satellite signals, considering communication environments characterized by additive white Gaussian noise (AWGN). However, the effects of impulsive noise should also be considered in a more accurate characterization of the LEO satellite communication scenario. This work introduces a new method of frequency acquisition for low-orbit satellites based on similarity calculus by maximum correntropy criterion. Due to some of its properties, correntropy is very efficient in the processing of non-Gaussian signals, especially in impulsive noise environments. It was investigated the robustness of the technique proposed in this work on a communication environment characterized by impulsive noise, and compare its performance with that obtained by a classical approach, based on the correlation coefficient. The carrier frequency acquisition method for LEO satellites proposed in this work can be used in the development of a decoder for signals from the Brazilian Environmental Data Collection System (SBCDA) and Franco-American Envi-ronmental Data Collection System (ARGOS).

6
  • RAFAEL FERREIRA ALVES DE ASSIS
  • Design of a NTC Thermistor-Based Thermal Anemometer with Linearized Feedback Model

  • Líder : SEBASTIAN YURI CAVALCANTI CATUNDA
  • MIEMBROS DE LA BANCA :
  • DIOMADSON RODRIGUES BELFORT
  • JUAN MOISES MAURICIO VILLANUEVA
  • SEBASTIAN YURI CAVALCANTI CATUNDA
  • Data: 01-jun-2018


  • Resumen Espectáculo
  • The velocity or volumetric flow measurement of laminar or turbulent flow meet various engineering applications. Thus, scientific research for the development of architectures and measurement techniques for these quantities is of great relevance and is constantly expanding. Among the existing architectures for anemometric applications in an indistinct flow regime, thermal anemometers denotes several advantages such as excellent bandwidth, spatial resolution and low cost. This is basically a thermistor connected to one of the arms of a modified Whetstone bridge. Although it has simple configuration, it has some limitations imposed by the non-linear relation between the temperature and the quantity applied to the sensor. In addition, it contributes as a limiting non-ideality of the operational amplifier that makes up the architecture. Therefore, it is proposed the design and implementation of a thermal anemometer architecture with linearization of the relation with the temperature by voltage feedback and using a thermoresistive sensor of the NTC type. We also promote comparisons between the classic anemometer and the one proposed in this work. For this, the validation of the proposed architecture is conducted through simulations, as well as experimental comparison with classical and other instrument

7
  • CARLOS EDUARDO DE BARROS SANTOS JÚNIOR
  • Proposed implementation of MD5 and SHA-1 hashes algorithms on reconfigurable hardware

  • Líder : MARCELO AUGUSTO COSTA FERNANDES
  • MIEMBROS DE LA BANCA :
  • ADRIANO MAURO CANSIAN
  • CARLOS EDUARDO DA SILVA
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • MARCELO AUGUSTO COSTA FERNANDES
  • VALENTIN OBAC RODA
  • Data: 20-jun-2018


  • Resumen Espectáculo
  • This work proposes two Application Specific System Processor (ASSP), one to the MD5 algorithm and other to the SHA-1 algorithm implemented on Field Programmable Gate Array (FPGA) Xilinx Virtex 6 xc6vlx240t-1ff1156. The throughput and the occupied area were analyzed for several implementations on parallel instances of those algorithms. The results showed that the hardware proposed for MD5 achieved a throughput of 0.8 Gbps for a single instance and slightly more than 140 Gbps for 320 instances in a single FPGA. Already for the SHA-1 algorithm, it has achieved a throughput of the 0.644 Gbps for a single instance and a little more than 28 Gbps for 48 instances on a single FPGA. Several applications such as password recovery, password validation, and high volume data integrity checking can be performed efficiently and quickly with an ASSP for MD5 and SHA-1. This work also presents a comparative analysis of the energy consumption associated with execution of the MD5 and SHA-1 algorithms for three different hardware platforms, a microprocessor (uP) of 8 bits and 32 bits and the specific application hardware designed for each algorithm. Results of consumption estimation from the processing time (measured in the laboratory) show that the use of dedicated hardware presents significant gains in energy savings.

8
  • TAIANE VIANA DE CARVALHO
  • Data Clustering Methods Applied to Image Segmentation

  • Líder : ALLAN DE MEDEIROS MARTINS
  • MIEMBROS DE LA BANCA :
  • AGOSTINHO DE MEDEIROS BRITO JUNIOR
  • ALLAN DE MEDEIROS MARTINS
  • ALUISIO IGOR REGO FONTES
  • PABLO JAVIER ALSINA
  • Data: 28-jun-2018


  • Resumen Espectáculo
  • Due to the great technological advance,  the computer is increasingly being used to extract information in images. Segmentation is one of the most main and complex process in  image processing. This paper presents a new technique of images automatic segmentation based on textures through a clustering application method. This method is composed of a set of distinct techniques and divided into two steps. The first step uses vector quantization to group the data set (pixels) in auxiliary centers and the second step group the auxiliary centers that have more common properties in a single group.  The main differential of this proposal is the use the method Quadratic Renyi’s Negentropy Separation (QRNS) which creates a hierarchical structure between  groups found. Validation of the method is performed with different images and its performance compared to other methods.

     

9
  • CAMILA NASCIMENTO DOS SANTOS
  • Image processing system for computing Transfer Function using Mason formula

  • Líder : ALLAN DE MEDEIROS MARTINS
  • MIEMBROS DE LA BANCA :
  • ALLAN DE MEDEIROS MARTINS
  • PABLO JAVIER ALSINA
  • AGOSTINHO DE MEDEIROS BRITO JUNIOR
  • ALUISIO IGOR REGO FONTES
  • Data: 28-jun-2018


  • Resumen Espectáculo
  • Several daily activities utilize complex control systems. Many of these require the users, skill and time for their representation. These systems can be modeled and represented in the form of signal flow diagrams to represent the causal relationship between their components. One method used to calculate the equivalent Transfer Function (FT) of a diagram is the Mason Formula. This requires that all paths be found, and if this does not happen, the calculated final FT is incorrect. 
    The objective of this work was to describe an image processing system, which is able, from a hand drawn diagram, to obtain the final transfer function more efficiently. Where from a hand-drawn signal flow diagram, a directed graph is automatically obtained, which represents the drawing with all the direct paths and loops present. The developed tool was shown to be able to help students and professionals correctly calculate the final FT for a given diagram, following the proposed conditions.
10
  • WERBET LUIZ ALMEIDA DA SILVA
  •  
     
    Automation of a Pilot Plant for Flow  Estimation in Multizoned Injection Wells from Temperature Profiles.
  • Líder : ANDRES ORTIZ SALAZAR
  • MIEMBROS DE LA BANCA :
  • ANDRES ORTIZ SALAZAR
  • ANDRE LAURINDO MAITELLI
  • CARLA WILZA SOUZA DE PAULA MAITELLI
  • JEFFERSON DOOLAN FERNANDES
  • Data: 29-jun-2018


  • Resumen Espectáculo
  • A pilot system was built at the Laboratório de Avaliação de Medição em Petróleo (LAMP) to evaluate estimation methods of flow distribution in wells by heat transmission models. In this work, it is intended to develop an automation system with a sensor structure, actuators and a monitoring software, to control the variables needed to the process simulation of water injection in multizones. Equipment to be used: flow, level and temperature sensors, and actuators like valves and pumps, all connected to a programmable logic controller by WEG, model: TPW03 60HT-A, that will be able to concentrate the data and communicate with a monitoring system implemented in Elipse SCADA software. This structure will acquire data to evaluate the flow estimation method from temperature measurements throughout the simulated injection tubing. The flow sensors will allow the comparison of the measured and estimated values with the mathematical model. It's intended to make adjustments in the original designed structure and to repeat the experiments in order to reduce the measurement uncertainty obtained in previous tests.
11
  • ANDRÉ LUIZ AGUIAR SOUSA
  • Multivariable Fuzzy System Applied to the Control of the Composition of Liquefied Petroleum Gas

  • Líder : FABIO MENEGHETTI UGULINO DE ARAUJO
  • MIEMBROS DE LA BANCA :
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • ANDRE LAURINDO MAITELLI
  • JOSE MEDEIROS DE ARAUJO JUNIOR
  • Data: 02-jul-2018


  • Resumen Espectáculo
  • In the present work, artificial intelligence techniques were applied in a simulated natural gas processing plant, composed of two distillation columns, a deethanizer column and a debutanizer column. In this process, the background product of the deethanizer, known as LGN, is supplied to the debranching column, where it is fractionated. The lighter components are evaporated giving rise to LPG (Liquefied Petroleum Gas), while the heavier fractions, denominated of C5+, remain in the liquid state. Ideally, LPG is composed only of butane and propane, but in practice this is not the case, since contaminants such as pen- tane and ethane are always present. In this work it is proposed to simultaneously regulate the amount of pentane and ethane in the LPG composition by means of the dynamic gene- ration of set points (SP) for the present controller’s layer of the instrumentation known as regulatory control. For this, a multivariate fuzzy system is used that will adjust the values of these SPs, from the comparison of the molar fraction of pentane and ethane present at the exit of the plant (LPG) and its desired quantities. The columns were simulated in the HYSYS® chemical process simulation software designed with real characteristics of the columns present in the Natural Gas Processing Unit, located in the municipality of Guamaré-RN.

12
  • FÁBIO RICARDO DE LIMA SOUZA
  • Identification and Control of Dynamic Systems using Fuzzy Wavelet Neural Network

  • Líder : FABIO MENEGHETTI UGULINO DE ARAUJO
  • MIEMBROS DE LA BANCA :
  • ANDRE LAURINDO MAITELLI
  • FABIO MENEGHETTI UGULINO DE ARAUJO
  • JOSE MEDEIROS DE ARAUJO JUNIOR
  • Data: 03-jul-2018


  • Resumen Espectáculo
  • Mathematical modeling is a task of vital importance for the development of science and technology, since its objective is to obtain a mathematical description of a real phenomenon. Regardless of their nature, real systems need to be studied and their dynamics must be known in order for their functioning to take place as desired. The classical way of obtaining a mathematical representation is by analyzing the physical laws governing systems. However, in this case, in addition, they are (often) difficult or even impossible to follow this path, and a plausible alternative is the identification of systems. The identification of dynamic systems aims to obtain a mathematical representation of the dynamics of a system based on input and output data. Real dynamic systems are ultimately nonlinear. In some applications, as linear approximations are sufficient, however, when linear representations do not express a dynamic process, it is necessary to use a nonlinear model. In the last decades, as neural networks have been installed as one of the main tools for an identification of nonlinear dynamic systems, as they have characteristics that make them attractive for identification, due to nonlinearity and generalization and learning. From the possession of a model that satisfactorily represents the dynamics of a system, it can be used for applications such as control, prediction, inference, among others. The present work proposes the application of a Fuzzy Wavelet Neural Network in the identification of nonlinear dynamic systems. This hybrid technique combines the multiresolution features of the Wavelet theory with the learning ability and the generalization of the neural networks together with the ability to deal with Fuzzy logic uncertainties. The obtained model is used for the tuning of a PID controller. In the course of the work, the concepts and techniques necessary to perform identification and controller tuning will be presented. The results obtained attest to the efficiency of the technique used.

13
  • DANIEL RODRIGUES DE LUNA
  • Performance Evaluation of new waveforms for underwater submarine channels

  • Líder : VICENTE ANGELO DE SOUSA JUNIOR
  • MIEMBROS DE LA BANCA :
  • JOSE PATROCINIO DA SILVA
  • VICENTE ANGELO DE SOUSA JUNIOR
  • ÁLVARO AUGUSTO MACHADO DE MEDEIROS
  • Data: 05-jul-2018


  • Resumen Espectáculo
  • This master’s degree working plan aims to perform a study on the performance of modulation and multiplexing techniques over submarine acoustic channel. Initially, what is underwater communication and its existing forms of communication is introduced. Then, the main effects that a submarine channel is subject and their respective formulations are presented, as well as the set of models well accepted by the academy, Bellhop models. The work is based on the proposal to using the OFDM and GFDM waveforms as an alternative to the classical waveform used in submarine communications, the FSK. A link simulator was developed and, through it, value-added analyzes are presented for different waveform configurations and channel manifestations implemented. Performance is accessed by means of bit rate error curves for some levels of Eb/N0. Finally, the work produced and its perspectives are presented, elucidating the conformity between the simulation and the theoretical curves, as well as the challenging work that is to transmit in high rate of transmission in the submarine channel.

14
  • HILÁRIO JOSÉ SILVEIRA CASTRO
  • A methodology for hybrid indoor location to mobile systems

  • Líder : IVANOVITCH MEDEIROS DANTAS DA SILVA
  • MIEMBROS DE LA BANCA :
  • IVANOVITCH MEDEIROS DANTAS DA SILVA
  • LUIZ FELIPE DE QUEIROZ SILVEIRA
  • MARCELO HENRIQUE RAMALHO NOBRE
  • SILVIO COSTA SAMPAIO
  • Data: 09-jul-2018


  • Resumen Espectáculo
  • In recent years, technological development has allowed the emergence of new concepts for new services, among the new proposals the indoor location systems. Since conventional solutions such as the Global Positioning System do not result in good results, several areas encourage the generation of new solutions for location indoor, in order to enhance its services. With the motivation to provide a new solution, will be demonstrated in this work a proposal of methodology of location indoor, able to find the location of the target in a hybrid way in terms of propagation of signals in indoor environments, using the trilateration method,