Research on EEMD-MCKD Method of Bearing Vibration Feature Extraction

Abstract

Aiming at the difficulty of identifying the characteristics of rolling bearing vibration signal faults under strong noise interference, a signal decomposition-selection-filtering method for extracting bearing fault features is proposed. First, the integrated EEMD is used to preprocess the signal for noise reduction. The kurtosis and correlation coefficient are used as evaluation indicators to select IMF; the MCKD can heighten the fault impact ingredient in the sensitive IMF signal, and further improve the signal-to-noise ratio; finally, the fault ingredient is extracted from the envelope spectrum, and the characteristic frequency of the failure is identified. Experiments indicate that the aforesaid method can effectively improve the accuracy of the fault characteristics extraction of rolling bearings.