Jixie chuandong (Jan 2018)
Fault Feature Extraction of Rolling Bearing based on Blind Separation Noise Reduction by ITD and KICA
Abstract
The fault signal and information of rolling bearing is often drowned in the noise,the traditional extraction of bearing fault feature methods are difficult to effectively extract the bearing fault feature. The SNR of blind source separation,combing the inherent time scale decomposition（ITD） and independent component analysis（ICA） method is used to reduce noise. The fault signal is decomposed by ITD. The PRC component will be decomposed by correlation coefficient criterion restructuring as the virtual noise channel. The KICA is used to realize the separation of fault signal and noise signal,then analysis on the effective fault signal envelope is carried out. The simulation and comparison of experiments of rolling bearing fault analysis indicates that this method is effective to extract the bearing fault feature.
