The Application of Self Correlation Function-cyclic Spectrum Analysis Method in Bearing Fault Diagnosis

At present, the research of rolling bearing fault diagnosis method has become a hot topic. In order to solve the problem of optimal resonant band selection, noise reduction and frequency extraction in fault diagnosis, this paper presents a method of rolling bearing fault diagnosis based on self correlation function-cyclic spectrum analysis (SCF-CSA). Firstly the method selects the appropriate resonance band signal through the band pass filter based on the spectral kurtosis (SK). And then makes the self correlation analysis of the signal with the original vibration signal to reduce noise. Finally, the rolling bearing fault diagnosis is realized by cyclic spectrum analysis. It proved by experiment that this method not only overcomes the interference of noise caused by human factors and signal superimposed reconstruction to the diagnosis result, but also can be more targeted to analyze the slice energy of the fault feature frequency and realize the fault diagnosis of the rolling bearing.