Fault Feature Extraction of Flexible Thin-walled Bearings Based on VMD and MOMEDA

Abstract

Aiming at fault feature extraction of thin-walled flexible bearing, a fault diagnosis method based on Variational Mode Decomposition (VMD) and Multipoint Optimal Minimum Entropy Deconvolution Adjusted (MOMEDA) is proposed. Firstly, the number of VMD layers is determined according to the correlation coefficient, and the kurtosis of each intrinsic mode functions (IMF) is calculated. The IMF component with the largest kurtosis is selected as the best component according to the kurtosis maximum criterion. Then the multipoint kurtosis of the best component is calculated to determine the optimal fault period and the MOMEDA is carried out. Finally, do the envelope spectrum analysis. By analyzing the fault signals of the outer and inner rings of flexible thin-walled bearings, it shows that this method can extract the fault feature frequency of flexible thin-walled bearings accurately and effectively. At the same time, this method is compared with the method based on VMD and MED (Minimum Entropy Deconvolution), and the results show that the method proposed in this paper not only has a strong noise reduction effect, but also can extract the fault feature frequency more accurately.