Research on Hot Spot Temperature Modeling and Forecasting Method of Oil-Immersed Transformer

The hot spot temperature of transformer is the highest point of winding temperature during transformer operation, which directly reflects the heat load of winding and is a critical indicator for transformer safety and performance evaluation. Therefore, accurate prediction of this hot spot temperature is essential for estimating the transformer's service life, enhancing economic efficiency, and preventing significant thermal incidents. Based on the comprehensive analysis of the internal temperature rise process and temperature distribution characteristics of oil-immersed transformers, and combined with the measured data of the transformer after removing outliers, this paper develops a high-precision prediction model for hot spot temperature based on a BP neural network. Use particle group algorithm to optimize BP neural network and to further improve the accuracy and reliability of the model prediction.