Research on temperature trend forecasting of rolling electric machine based on SVM

A temperature trend forecasting algorithm based on support vector machine (SVM) was proposed to study the temperature faults of rolling electric machine. With the analysis of the monitoring system of rolling electric machine, the multi-steps forecasting model was built and the SVM algorithm was verified by a numerical example and a realistic case. The results show this algorithm has accurate forecasting ability and can help to diagnose faults in advance.