Fault diagnosis of rolling element bearings based onoptimal Morlet wavelet and hidden Markov model
Ruige Zhang1, Yonghong Tan2
(1. School of Electronic Engineering, Xidian University, Xi’an 710071, China2. College of Information, Mechanical and Electronic Engineering, Shanghai Normal University, Shanghai 201814, China)
Abstract:Abstract: This paper presents a new approach for the fault diagnosis of rolling element bearings using the optimal Morlet wavelet and the statistic characteristic of the wavelet coefficients. It has been shown that the optimization of the wavelet parameters would benefit to extract the effective features. Thus, the criteria of the minimal Shannon entropy and technology of the singular value decomposition are applied to optimize the parameters of the Morlet wavelet. The feature extraction firstly divides the Morlet wavelet coefficients into a series of segments. Next, the infinite-norm of the covariance matrix for each segment is calculated, which is also applied to construct the observation vectors of the hidden Markov models. Finally, the experimental results on bearing faults identification and isolation are illustrated, and all the identification accuracies are greater than 93%.