Banca de DEFESA: ISRAEL DA SILVA FÉLIX DE LIMA

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : ISRAEL DA SILVA FÉLIX DE LIMA
DATE: 14/07/2026
TIME: 09:00
LOCAL: Sala virtual Google Meet
TITLE:

Optimization of an Interval Type-2 Fuzzy Servo Controller for State-Feedback Control of Nonlinear Systems Using a Quantum-Inspired Genetic Algorithm

KEY WORDS:

Nonlinear Control, Interval Type-2 Fuzzy Logic, Quantum-Inspired Genetic Algorithm, State Feedback, MAGLEV.

PAGES: 81
BIG AREA: Engenharias
AREA: Engenharia Elétrica
SUBÁREA: Eletrônica Industrial, Sistemas e Controles Eletrônicos
SPECIALTY: Controle de Processos Eletrônicos, Retroalimentação
SUMMARY:

Currently, the control of nonlinear and inherently unstable dynamic systems, such as magnetic levitation (MAGLEV) systems, imposes significant challenges on engineering. The presence of complex dynamics and uncertainties makes it difficult to obtain exact mathematical models. In this scenario, controllers based on Interval Type-2 Fuzzy Logic (IT2-FLC) emerge as robust tools to handle uncertainties while preserving the analytical transparency and interpretability that are lost in “black box” approaches, such as neurofuzzy techniques. However, when this logic is structured into multiple local models via Parallel Distributed Compensation (PDC) to handle nonlinearity, a considerable structural bottleneck arises: the high complexity involved in the empirical tuning of the closed-loop poles associated with each rule and the limitations of classical metaheuristics, which frequently struggle with premature convergence and trapping in local minima. To address these gaps, this work proposes the optimization of Interval Type-2 fuzzy servo controllers for state feedback in nonlinear systems, using a Quantum-Inspired Genetic Algorithm (QGA). The proposed servo control architecture aims to ensure accurate trajectory tracking of the MAGLEV, while the QGA performs the pole allocation tuning, mitigating the search inefficiency of conventional algorithms. For validation, the exploratory performance of the QGA is compared to that of a classical Genetic Algorithm (GA), using the ITAE index as the objective function, and the IT2-FLC controller is contrasted with its Type-1 counterpart. The results obtained through computational simulations indicated promising performance, showing an advantage of the quantum-inspired approach under the adopted simulation protocol. The optimized system achieved robust control and rapid stabilization in the evaluated scenario, reinforcing the proposed architecture as a practical, interpretable, and efficient solution within the scope of computational validation.

COMMITTEE MEMBERS:
Presidente - 1451883 - FABIO MENEGHETTI UGULINO DE ARAUJO
Interno - 1242315 - PABLO JAVIER ALSINA
Externo ao Programa - 3374361 - JEAN MARIO MOREIRA DE LIMA - UFRNExterno à Instituição - MÁRIO SÉRGIO FREITAS FERREIRA CAVALCANTE - UFRN