Multivariate Empirical Wavelet Transform and Its Application to Rolling Bearings

Empirical wavelet transform (EWT) is a recently proposed algorithm for single-channel signal separation, which has been widely applied in the field of fault diagnosis. With the development of sensors, multi-sensors are increasingly applied in equipment condition monitoring, and multi-channel signals can more comprehensively characterize the fault signals of bearings. Aiming at the problem that EWT cannot deal with multi-channel signals, the multivariate empirical wavelet transform (MEWT) method is proposed in this paper. Through analyzing the simulated fault signals for bearing, the effectiveness of the proposed MEWT method is verified, and the fault types of the bearings can be accurately diagnosed.