DYNAMIC UNBALANCE DETECTION OF CARDAN SHATF IN HIGH-SPEED TRAIN BASED ON MODIFIED VARIATIONAL MODE DECOMPOSITION

Abstract

Aiming at the adverse influence of penalty parameter and mode number on the VMD（ variational mode decomposition）,a newly MVMD（ modification variational mode decomposition） was proposed through the information entropy difference. The change of the amount of information in each mode extraction characteristic of information entropy difference was analyzed. The information entropy difference under the condition of different penalty parameters was got through making full use of the analysis result of extraction characteristic. The excellent penalty parameter was determined in the position of jumpy information entropy difference. The Fourier spectrum of MVMD was used to detect the unbalance of Cardan shaft in high-speed train. The method and model was verified through the unbalance experiment data. The experiment results show that this method could effectively seize the basic frequency,multi-frequency fault features. With comparison to VMD,the problem of Parameter selection has been solved. With comparison to the classic ensemble empirical model decomposition,the fault detection ability has been significantly improved.

Keywords

• Cardan shaft;Dynamic unbalance;MVMD;Information entropy difference