A new type of fault diagnosis method of asynchronous motor

In view of problem of difficult parameters determination existed in fault diagnosis method of asynchronous motor based on RBF neural network, the paper proposed a fault diagnosis method of asynchronous motor based on RBF neural network optimized by differential evolution algorithm. Firstly, the method uses wavelet transformation to make de-noising process for running state signal of asynchronous motor, then it uses principle component analysis method and wavelet packet analysis method to extract feature of the de-noised signal of asynchronous motor, and at last it uses RBF neural network optimized by differential evolution algorithm to diagnose the feature. The experiment result shows that the RBF neural network optimized by the differential evolution algorithm can identify faults of asynchronous motor effectively compared with RBF neural network without optimization.