Study Of The Fault Diagnosis Based On Wavelet And Fuzzy Neural Network For The Motor

In the fault diagnosis of the motor, the vibration signals can fully reflect the status of the motor. In this paper, on the basis of wavelet packet fault feature extraction, a new approach for motor fault diagnosis based on wavelet packet analysis and fuzzy RBF neural network was presented.The method gains the energy of characteristic channel of bearing failure vibration signals of asynchronous motor, which adopts the technology of wavelet packet analysis. It also composes the characteristics of the vector as input of fuzzy RBF neural network, used to diagnose the induction motor bearing failures. The method overcomes the slow convergence, a long training time, local minimum problems when using BP neural network. Experimental results shows that using fuzzy RBF neural network can improve the accuracy of the motor fault diagnosis.