Fault Diagnosis of Strong Noise in Rolling Bearings based on EEMD-RRSD

Abstract

Rolling bearings often operate under complex working conditions. When local faults occur in rolling bearings, their vibration signals contain not only periodic transient shock components related to fault information, but also harmonic components such as shaft rotation frequency and background noise. Therefore, bearings operate in high noise fault environments, and direct envelope demodulation analysis of rolling bearing vibration signals often yields poor results. A strong noise fault diagnosis method for rolling bearings based on EEMD-RRSD is proposed to address the above issues. This method first performs EEMD decomposition on the signal, and then reconstructs the signal based on spectral kurtosis and correlation coefficient criteria. Perform resonance sparse decomposition on the reconstructed signal, perform Hilbert demodulation analysis on the low resonance components based on the fast spectral kurtosis map, extract the characteristic frequency of bearing faults, and then diagnose rolling bearing faults. The analysis results of simulation signals and experimental signals show that this method can effectively extract the impact components in the vibration signals of bearing faults and diagnose bearing faults.