Motor rolling bearing fault diagnosis based on MVMD energy entropy and GWO-SVM

For motor rolling bearing fault diagnosis, vibration signal analysis is a common method to extract sensitive fault characteristics. In this paper, a newly signal processing method, multivariate variational mode decomposition (MVMD), is proposed to extract features from motor rolling bearings. The MVMD was carried out on the motor rolling bearings state signals of different categories, and the prior parameter K value which had a great influence on the decomposition effect was analyzed. Each component obtained by decomposition was measured in the form of energy entropy (EE), and the measured feature information was classified and identified by support vector machine (SVM) classifier. Meanwhile, the grey wolf optimization (GWO) was used to optimize the parameters of the classifier network to further improve the recognition accuracy. Through the simulation results, it is found that the scheme can achieve 100 % effect on the diagnosis rate of normal working condition, outer ring fault, inner ring fault and rolling element fault under the condition of different load and speed of the motor rolling bearing.