Banca de DEFESA: INGRID THAÍS AZEVÊDO DANTAS

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : INGRID THAÍS AZEVÊDO DANTAS
DATE: 07/03/2025
TIME: 14:00
LOCAL: Online
TITLE:

Power Transformer Fault Classification Using Industry 4.0-Compliant Machine Learning Under Data Missing Conditions

KEY WORDS:

Machining learning, Power transformers, fault classification, wavelet transform, data missing conditions, industry 4.0.

PAGES: 50
BIG AREA: Engenharias
AREA: Engenharia Elétrica
SUBÁREA: Sistemas Elétricos de Potência
SPECIALTY: Medição, Controle, Correção e Proteção de Sistemas Elétricos de Potência
SUMMARY:

Faults in power transformers can occur due to insulation degradation in their windings and must be quickly cleared by protective systems to preserve the transformer’s integrity. In addition to transformer protection, classifying faults is also an essential task since it allows a better understanding of the transformer’s health over time. This work presents an innovative approach for fault classification in power transformers, combining the Real-Time Boundary Stationary Wavelet Transform (RT-BSWT) with Machine Learning techniques. The proposed method stands out for its robustness, even in the face of real-time data missing, a critical challenge in fault detection for these systems. The approach employs machine learning algorithms for fault classification, even when data quality and availability are compromised due to transmission failures or possible interference in the connection circuit between the current transformers and the transformer protective relay. Its ability to adapt to data loss makes it highly suitable for real-world industrial applications, aligning with Industry 4.0 principles—particularly in environments where data integrity is essential for real-time analysis and decision-making. The approach's effectiveness was validated through a comprehensive evaluation of a diverse dataset, covering various fault types and conditions. The results demonstrated high accuracy rates in both complete data scenarios and situations with missing data, reinforcing the method’s viability for challenging industrial operations.

COMMITTEE MEMBERS:
Interno - 1807863 - FLAVIO BEZERRA COSTA
Interno - 1543191 - LUIZ FELIPE DE QUEIROZ SILVEIRA
Externo ao Programa - 1141792 - RODRIGO PRADO DE MEDEIROS - UFRNExterno à Instituição - MÁRIO OLESKOVICZ - USP