Vibration generated by incipient faults of rolling element bearing is usually with low energy and dispersed frequency distribution, then, it is easy merged in strong disturbances from other vibration sources. Performing the Hilbert Huang Transition (HHT) directly may fail to remove the noise from the observed modulated signals by separating the Intrinsic Mode Function (IMF) accurately since the vibrations are lower at the signal-to-noise ratio and more frequency modulation. In this paper, a new fault-diagnosis approach is proposed based on the ICA and the HHT. In the approach, the vibration signals are decomposed by the Empirical Mode Decomposition (EMD) to get stable IMF components at first. Then, the IMFs which contain the fault information of the rolling element bearing mostly are selected and the corresponding envelopes are extracted by band-pass filter and the Hilbert transform. Subsequently, the independent component analysis (ICA) is employed to separate the envelopes to independent components (ICs) according to the independent of vibration sources. Finally, the envelope spectrums of the ICs are calculated respectively and compared with the fault characteristics of the rolling element bearing to realize the accurate faults diagnosis of rolling element bearings. Simulations and tests verified the feasibility of the method.
TANG Xian-guang;GUO Yu;DING Yan-chun .
Application of hilbert huang transition and independent components analysis on rolling element bearing faults features extraction[J]. Journal of Vibration and Shock, 2011, 30(10): 45-49