Fault diagnosis of rolling bearing based on DWT-MFE and LSSVM

In view of the effects of non-stationary and nonlinear vibration signals on the accuracy of rolling bearing fault diagnosis, a rolling bearing fault diagnosis method was proposed based on discrete wavelet transform (DWT) and multi scale fuzzy entropy (MFE) and least squares support vector machine (LSSVM). Firstly, the original signal was decomposed by discrete wavelet transform to obtain some approximate coefficients, and after reconstructing the approximate components the optimal approximate components were selected by using the principle of correlation coefficient and correlation distance. Then the feature vector was obtained by multi scale fuzzy entropy and input into least squares support vector machine for fault recognition. Finally, the algorithm was simulated by using the bearing signal of Case Western Reserve University. The results show that the recognition accuracy of the method is 97%, which further proves the feasibility of the method.