EMD and Zoom Envelope Spectrum for Rolling Bearing Fault Diagnosis

Abstract

ntrinsic mode function component (IMF) has been effectively decomposed by using empirical mode (EMD) decomposition method, and the high-frequency resonance frequency band the rolling bearing components is extracted. Then the high frequency component of the IMF is shifted to the lower frequency by using real frequency modulation principle, and a new envelope refining spectrum is obtained through the Hilbert transform after low frequency sampling. The method not only simplifies the design of band-pass filtering process but improves refinement computational efficiency of the envelope spectrum. The method verifies the correctness of the theory and practical method through the analysis of the fault diagnosis of rolling bearing inner ring experiment.