Early fault feature extraction of rolling bearing based on optimized VMD and improved threshold denoising

Aiming at the problem of early fault signals of rolling bearing being weak to cause fault feature extraction being difficult under complex working conditions and strong background noise interference, a method of rolling bearing fault feature extraction based on optimized variational mode decomposition (VMD) and improved threshold denoising was proposed. Firstly, VMD was optimized using the whale optimization algorithm (WOA) to realize the adaptive decomposition of vibration signal, and the optimal modal components selection criteria for L-kurtosis and correlation coefficient were established. Then, the improved threshold denoising was performed on the selected optimal components. Finally, Hilbert envelope spectral analysis was performed on the de-noised signals to realize fault feature frequency extraction. The proposed method was verified to adopt simulated signals and the engineering data set of University of Western Reserve in US. At the same time, the proposed method was compared with Teager energy operator denoising method and the optimization method based on envelope entropy criterion. The results showed that the effect of the proposed method is better. © 2021, Editorial Office of Journal of Vibration and Shock. All right reserved.