Research on locomotive motor bearing diagnosis method based on wavelet threshold denoising and correlation analysis combined with EMD decomposition and LSTM neural network

The safe operation of maglev trains is inseparable from the health status of each component, and fault diagnosis of components is of great significance. In this paper, taking the bearing of a magnetic levitation motor as an example, a fault diagnosis method based on wavelet threshold denoising and correlation analysis combined with EMD(Empirical Mode Decomposition) decomposition and LSTM(Long Short-Term Memory) neural network is proposed. This article first compares the optimized EMD decomposition with traditional Time-frequency domain analysis to show the superiority of its performance. Then, Use the IMF component with the largest correlation coefficient obtained from the optimized EMD decomposition as the input of the LSTM neural network.Set network parameters and structure to complete the detection of bearing failure status. This paper uses a certain type of SKF bearing signal as an example to verify it. The results show that the method has a high accuracy in the diagnosis of normal status, gear fault, outer race fault, inner race fault and roller fault.