Rolling bearing fault diagnosis method based on VMD and LSSVM

Abstract

The vibration signal of rolling bearing is complex, it is difficult to extract fault features and diagnose accurately. In this paper, a rolling bearing fault diagnosis method based on variational mode decomposition and GWO-LSSVM is proposed. The variational mode decomposition algorithm is used to decompose the bearing vibration signal, and the fuzzy entropy of each component signal is calculated. GWO is used to optimize the parameters of LSSVM. The least square support vector machine is used to identify the bearing fault.