Bearing Fault Diagnosis Based on EMD and Smoothed Pseudo Wigner-Ville Spectrum Entropy

A method of fault diagnosis for rolling bearings based on empirical mode decomposition(EMD) and smoothed pseudo Wigner-Ville distribution(SPWVD) spectral entropy is proposed. In this method, the nonlinear and nonstationary characteristics of the signal in the EMD method, which has a great advantage in signal filtering and de-noising, are fully reserved. The SPWVD spectral entropy is used to quantitatively characterize the time-frequency energy distribution of the vibration signals in different states of the bearing. The intelligent model is designed based on the least square support vector machines(LS-SVM). The automatic classification of bearing state and identification of fault type of the bearing are realized. Through the mutual comparison of the SPWVD spectral entropy method and spectral kurtosis method, the effectiveness of the SPWVD spectral entropy is verified. The results show that this method can effectively extract the characteristics of the bearing fault information and improve the rate of bearing fault diagnosis.