The detection of bearing incipient fault with maximal overlap discrete wavelet packet transform and sparse code shrinkage denoising

Abstract

This paper proposes an enhanced fault detection method incorporating the maximal overlap discrete wavelet packet transform (MODWPT) method and the sparse coding shrinkage (SCS) denoising algorithm according to the desirable property of approximate shift-invariance and the proper frequency band of noise removal and demodulation for defective bearing vibration signals. The vibration signals measured from motor bearings are used to demonstrate the performance of the proposed approach compared with traditional squared envelope spectrum (SES) and fast kurtogram (FK). The results verify the effectiveness of the method in identifying the weak fault characteristics and diagnosing defects of bearings.