A Method for Hot-spot Temperature Prediction and Thermal Capacity Estimation for Traction Transformers in High-speed Railway based on Genetic Programming

With the rapid development of China’s high-speed railway (HSR), large-scale traction transformers have been put into use. In order to realize the batch prediction of hot-spot temperature (HST) of traction transformer group and perform estimation of thermal capacity well, this article was devoted to HST prediction modeling for traction transformers based on genetic programming (GP). First, the HST, load factor, and ambient temperature data used in this article were measured from the traction transformer A and were further divided into training set and prediction set. Training set was used to driven modeling by GP. An explicit expression prediction model, which could directly predict the dynamic HST, was established. Then, it was confirmed that the model has high accuracy according to the prediction set. Furthermore, the transformers B and C that are belong to the same railway line like A were tracked and predicted in real time. It is verified that the model has high generalization performance. Simultaneously, the practical application of the model was discussed and analyzed. The research result shows it is expected that the proposed model could realize the batch accurate prediction of HST for traction transformer group. It can provide a better and more effective reference for thermal capacity estimation, train scheduling plan, and maintenance replacement plan of traction transformers.