A Novel Method for Rolling Bearing Fault Diagnosis Based on VMD and SGW

Abstract

The bearing vibration signal with strong non-stationary properties is generally composed of multiple components making it complicated to extract the characteristic fault features of vibration signals of rolling bearings under the background of strong noise, how to solve this problem effectively is the focus of our research. Therefore, a new scheme based on Variational Mode Decomposition (VMD) and second-generation wavelet (SGW) is proposed in this paper. Firstly, VMD can decompose accurately and adaptively a complex multi-component signal into a set of IMF component with narrow band properties.Secondly, on the basis of kurtosis and cross-correlation analysis, the optimum signal components obtained by the VMD are selected to filter and to reconstruct the analysis signal. Then, (SGW) approach is used to eliminate the strong noise background and enhance the periodic impact in the optimum IMF components. Lastly, the accurate characteristic defect frequency can be obtained by using envelope spectrum of the reconstructing signal. The success of the proposed approach is verified by analysis the vibration signals of bearings with an outer race, an inner race and a rolling element faults, respectively. The results indicate that the scheme is feasible and useful for extracting the bearings fault features.