In order to efficiently solve the problem of asynchronous motor faulta fault diagnosis system of asynchronous motor based on back-propagation (BP) neural network and wavelet packet energy analysis is proposed. The stator current signal is used as the fault signal of asynchronous motorand the collected stator current signal is analyzed by wavelet packet energy analysis to extract the corresponding fault feature vector. In order to improve the accuracy of diagnosisthe time-domain and frequency-domain features of signals are extracted and input into BP neural network for training and learning. After enough trainingthe accuracy is tested with test samples. Through the proposed methodthe faults of asynchronous motor can be eliminated and corrected in timeand the economic benefits of the factory can be improved.