Banca de QUALIFICAÇÃO: GUILHERME NASCIMENTO DA SILVA

Uma banca de QUALIFICAÇÃO de MESTRADO foi cadastrada pelo programa.
STUDENT : GUILHERME NASCIMENTO DA SILVA
DATE: 10/07/2026
TIME: 08:30
LOCAL: Sala DCA001 - Laboratório de Automação
TITLE:

Classification of Internal Faults in Power Transformers Based on Fuzzy Logic and the Redundant Discrete Wavelet Transform

KEY WORDS:

Power Transformers, Internal Fault Classification, Logical System-Fuzzy, Redundant Discrete Wavelet Transform, Fuzzy-Wavelet.

PAGES: 67
BIG AREA: Engenharias
AREA: Engenharia Elétrica
SUBÁREA: Eletrônica Industrial, Sistemas e Controles Eletrônicos
SPECIALTY: Controle de Processos Eletrônicos, Retroalimentação
SUMMARY:

The reliability of electrical power systems is strongly affected by network disturbances, which can cause momentary or permanent voltage dips, increased transmission currents, and consequently, interruptions in the supply of electricity to consumers. A large part of these disruptions is due to short circuits in the network, exposing users to operational risks and causing significant losses for utilities. Among the equipment most vulnerable to faults is the power transformer, responsible for interconnecting the different segments of the electrical system and maintaining the continuity of the energy supply. In recent years, techniques based on Artificial Intelligence (AI) have been widely applied to strategies for protecting electrical systems. In this context, fault classification in power transformers presents itself as a promising area, both benefiting from the resources offered by the differential protection of large transformers and from advances in modern short-circuit analysis methodologies, allowing the identification of relevant information about the disturbance, such as the phases affected. Among AI techniques, fuzzy logic stands out for its robustness in handling the uncertainties inherent in the classification process, favoring an adequate definition of the system’s output classes. In signal proces sing, the redundant discrete wavelet transform is efficient at extracting transient features associated with the analyzed events. Based on the integration of these approaches, this work proposes the development of a fuzzy-wavelet classifier for classifying faults in power transformers, aiming to identify the phases involved in the disturbance. The methodology was validated through simulations using a wide range of short-circuit scenarios applied to the transformer’s protection zone, in a power system modeled in MATLAB/Simulink, consisting of a 300 MVA three-phase transformer with a transformation ratio of 230:69 kV. The preliminary results, considering different conditions and magnitudes of faults in the electrical network, showed that the classifier was able to quickly and efficiently iden tify all analyzed types of faults , providing detailed information about the affected phases and contributing to increased safety of the transformer’s differential protection.

COMMITTEE MEMBERS:
Presidente - 1451883 - FABIO MENEGHETTI UGULINO DE ARAUJO
Interno - 350693 - ANDRE LAURINDO MAITELLI
Interno - 1807863 - FLAVIO BEZERRA COSTA
Interno - 1141792 - RODRIGO PRADO DE MEDEIROS