Data Driven Transformer Thermal Model for Condition Monitoring

Condition monitoring of power transformers, which are key components electrical systems, is essential to identify incipient faults and avoid catastrophic failures. In this paper machine learning algorithms, i.e., nonlinear autoregressive neural networks support vector machines, proposed model the transformer thermal behavior for purpose monitoring. The models developed based on historical measurements from nine transformers comprised two 180-MVA units, four 240-MVA units three 1000-MVA units. data consist load profile, tap position, winding indicator temperature (WTI) measurement, ambient temperature, wind speed solar radiation. results validated against field measurements, it clearly demonstrated that alternative algorithms surpass IEEE Annex G model. An fault identification algorithm then successfully used an issue using taken in field. This could be alert operator plan intervention accordingly.