Data-Driven Thermal Analysis for Detecting Anomalies in Distribution Transformers

The electrification of transportation and the use of renewable energy has been growing dramatically. The electrical power system plays a significant role in supporting this growth. The stability and reliability of transformers are essential to the overall performance of electrical power systems. Monitoring the health of transformers is critical for preventing unexpected faults and ensuring continuous operation. Among various techniques, thermal condition monitoring plays a key role in mitigating overheating issues and supporting effective asset management. In this study, data-driven thermal models for transformers were developed using experimental datasets. These include a parameter-fitted IEC 60076-7 thermal model, an artificial neural network (ANN), and a long short-term memory (LSTM) network. The proposed models enhance the early detection of thermal anomalies, contributing to proactive maintenance and improved system reliability.