Adaptive Periodic Impulse Extraction Method and Its Application in Rolling Bearing Fault Diagnosis

Abstract

Considering that weak periodic impulse components are difficult to be separated and extracted from strong noise interference, the accuracy of fault diagnosis is greatly affected. A new adaptive periodic impulse extraction (APIE) method is proposed. First, the APIE method introduces the correlation Renyi entropy (CRE) index, which is based on the cyclo-stationarity of the signal and can accurately quantify the periodicity and impulsivity of the signal. Then, the APIE takes CRE as the objective function and realizes the accurate locking of fault feature information by iteratively updating filter parameters, and then, the redundant modes and mixed modes are eliminated by correlation coefficient (CC) index. Finally, the optimal mode number is adaptively determined by CRE index again, which improves the stability of the decomposition effect of APIE. The analysis results of rolling bearing composite simulation signal and experimental signals show that APIE method can not only effectively separate different characteristic information components but also accurately extract periodic impulse components and has outstanding noise robustness.