Banca de DEFESA: FLÁVIO HENRIQUES GONÇALVES

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : FLÁVIO HENRIQUES GONÇALVES
DATE: 23/08/2024
TIME: 09:00
LOCAL: Sala 02 DCA
TITLE:

DEVELOPMENT OF THE HORUS PLATFORM FOR FAILURE PREDICTION - CASE STUDY: SET OF BLADE ANGLE MOTORS IN A WIND GENERATOR.

KEY WORDS:

Wind turbines; wind farms; predictive maintenance; forecast; reliability.

PAGES: 108
BIG AREA: Engenharias
AREA: Engenharia de Energia
SUBÁREA: Fontes Renováveis de Energia
SPECIALTY: Energia Eólica
SUMMARY:

Efficient management of wind farms is essential to ensure sustainability and maximize renewable energy production. One of the main challenges is preventive maintenance and the prediction of failures in critical components, such as the set of blade angle motors in wind turbines, commonly called “pitch motors”. Predictive maintenance is an emerging field that promises greater operational efficiency and cost savings. In this context, the Hórus platform, which is the product of this work, appears as a viable alternative, aiming to contribute to the health management of wind turbines. The Hórus platform was developed to predict failures in wind turbine “pitch motors”, with a flexible architecture that allows the inclusion of other equipment and predictive maintenance techniques. The choice of the name Horus, inspired by the Egyptian god of vision, reflects the essence of the platform: offering a clear and anticipated view of maintenance needs, avoiding unexpected interruptions and maximizing operational efficiency. Hórus' technological structure combines data analysis and visualization tools, such as Power BI for processing large volumes of data and Python for complex algorithms. The visual interface is developed with HTML and CSS, providing a friendly and interactive experience. Data collection by Hórus involves multiple points, such as machine failure data, alarms, operating guidelines, wind history, among others. The integration of this data allows for a comprehensive and detailed analysis, essential for accurate forecasts adapted to the particularities of each wind farm. The user-friendly interface makes it easier for managers to interpret data, while the ability to integrate with other management systems guarantees a holistic view of operations. The failure prediction methodology includes the Jack-knife Diagram, Weibull Distribution and an empirical weight and penalty method based on the experience of the author of this work in the wind sector, allowing an effective classification and prioritization of failures. The Horus platform combines advanced and low-cost analytical tools; customization of analysis and report screens; integration of multiple data sources for analysis; Continuous evolution in beta mode. The first results of the implementation of the Hórus platform, which began in 2022, show a satisfactory assertiveness rate in the predictions made. The purpose of this work is to offer a “predictive vision” that allows managers and technical staff of wind farms to identify and resolve potential problems before they affect the production of electrical energy, thus promoting operational efficiency, reducing the cost of electrical energy and the sustainability of wind farms.

COMMITTEE MEMBERS:
Interno - 1770049 - GABRIEL IVAN MEDINA TAPIA
Interno - 346989 - JOSE TAVARES DE OLIVEIRA
Presidente - 347427 - RICARDO FERREIRA PINHEIRO
Externo à Instituição - UBIRATAN HOLANDA BEZERRA - UFPA