Application of Hilbert-Huang Transformation to fault diagnosis of rotary machinery

摘要

The vibration signal of a rotor bearing system is usually nonlinear and non-stationary. Fourier transform is hard to analyze these signals. A new method based upon empirical mode decomposition (EMD) and Hilbert spectrum is proposed for fault diagnosis of roller bearings. We get vibration signals from 6205-type ball bearings with inner-race faults and with outer-race faults, then analyzing its local Hilbert spectrum and local Hilbert marginal spectrum. Comparing the results with theory value, we can diagnose the fault of rotary machinery fault. In this study, we find that local Hilbert spectrum and local Hilbert marginal spectrum are very useful. Hilbert Transformation is introduced to confirm the HHT method is fit to process nonlinear and non-stationary signals.