Weak fault feature extraction of rolling bearing based on autocorrelation and energy operator enhancement

Aiming at weak impact components in rolling bearing's early fault vibration signal and difficult fault feature extraction due to noise influencing, a method for rolling bearing weak fault feature extraction based on autocorrelation and Teager energy operator enhancement was proposed. Autocorrelation calculation and empirical mode decomposition (EMD) were used to suppress random noise and low-frequency noise in the whole frequency band of bearing vibration signal, respectively and highlight fault impact period. At the same time, cross-correlation coefficient-kurtosis index based on IMF energy ratio weighting was proposed to screen out the optimal IMF for signal reconstruction and enhancing fault information in the reconstructed signal. Teager energy operator (TEO) was used to act on the reconstructed signal to obtain the instantaneous energy sequence with enhanced fault impact characteristics, and the bearing fault characteristic frequency was extracted through power spectral analysis. The analysis of inner race fault's simulated signal and rolling element fault's measured signal showed that the proposed method can effectively suppress bearing vibration signal noise and significantly enhance weak characteristics of early faults. © 2021, Editorial Office of Journal of Vibration and Shock. All right reserved.