Rolling Element Bearing Fault Diagnosis Using MODWPT and MCKD

Abstract

Due to the gear meshing vibration and noise, the fault signal of the rolling element bearing is relatively weak, which makes the fault signal separation and feature extraction difficult. A rolling bearing fault diagnosis method based on maximum overlapping discrete wavelet packet transform (MODWPT) and maximum correlation kurtosis deconvolution (MCKD) was proposed. Bearing fault signal was decomposed by using MODWPT method into some components with different frequency band. It realized the separation of gear meshing signal and bearing fault signal. Based on then the kurtosis criterion, component with large kurtosis was selected to filter with maximum correlation kurtosis deconvolution (MCKD) method, which revealed the bearing fault feature. Finally, the filtered signal is analyzed with envelope demodulation, and the bearing weak fault feature was extracted from the original fault signal in order to realize the bearing defect detection and diagnosis. Simulation and experiment verify the effectiveness of the proposed method.