STUDY ON FAULT DIAGNOSIS METHOD BASED ON MORLET WAVELET-SVD AND VPMCD

How to extract characteristic parameters from vibration signals with noise is a key problem of bearing fault diagnosis. A novel method based on Morlet wavelet-Singular Value Decomposition（ SVD） and Variable Predictive Model based Class Discriminate（ VPMCD） was proposed in this paper aiming to solve this problem. Firstly,Morlet wavelet transform was used to pre-process the signals in the time domain to obtain a time-frequency coefficient matrix,then SVD was applied to the matrix to remove noise and extract the weak fault information in the corresponding dimensions according to the singular value curvature spectrum; Secondly,the signal components near the optimal meature were selected,and the Shannon energy entropy were used as the characteristic parameters to construct the feature vectors,which were then used to establish the fault identification model based on VPMCD. Finally,5-fold cross validation method and Jackknife test method were adopted to verify the proposed method,and the results have demonstrated its effectiveness.