Feature Extraction of Mixed faults of Rotating Machinery Based on ICA -R and Stochastic Resonance

Abstract

Aiming at the problem of extracting the fault features of the rotating machinery under the situation of the mixed faults and low signal-to-noise ratio (SNR), a method of fault feature extraction based onindependent component analysis with reference (ICA-R) and stochastic resonance (SR) algorithm is proposed. Firstly, the improved fault signal pre-processing is carried out by using the improved ICA-R, and the expected fault signal is extracted. Then, combining the time-domain analysis method with the artificial bee colony algorithm, the scale adaptive SR algorithm is employed to further extract fault features. The experimental results show that the proposed method is effective in diagnosing the mixed faults of rotating machinery.