An optimized VMD method and its applications in bearing fault diagnosis

By Hua Li, Tao Liu, Xing Wu, Qing Chen. 2020

338 citations recorded by the source corpus.

Abstract Variational mode decomposition (VMD) is a newly proposed signal processing method which is diffusely used in fault diagnosis of rotating equipment. It has two issues that need to be addressed: (1) its impact parameters need to be determined in advance; (2) its sensitive intrinsic mode function (IMF) needs to be chosen from the multiple IMFs generated by VMD. The purpose of this paper is to introduce effective solutions to the above two issues respectively, to get an optimized VMD method. The method can be characterized as follows: the envelope kurtosis maximum is first used as an indicator to optimize and determine the mode number of VMD in this paper. Then, a novel method to realize the selection of the optimal IMF(s) of VMD which containing abundant fault information based on frequency band entropy (FBE) is introduced. Finally, the envelope power spectrum analysis is performed on the selected IMF(s) to pick up the fault feature frequency to identify the bearing fault type. The feasibility and ascendancy of the presented optimized VMD method are confirmed by the simulation signal analysis and the actual case analysis.

Subfields: Control and Systems Engineering, Mechanical Engineering, Mechanics of Materials

View Paper: https://www.semanticscholar.org/paper/d3048349620395b003dc1fc0c044502c717b3fa8