Distinguish coherence resonance and stochastic resonance in bearing fault evaluation

Abstract

Similar to the famous stochastic resonance (SR), coherence resonance (CR) is an ordered dynamic behavior induced by the noise in a nonlinear system. Inspired by the dynamical characteristics of CR, the importance of distinguishing CR and SR in bearing fault evaluation is presented. It solves the problem that the characteristic frequency cannot be determined when using the SR for bearing fault diagnosis. The general scale transformation is introduced to process the vibration signal further to realize adaptive CR at different frequencies. The autocorrelation function and the resonance factor are key measurements of CR. Especially, the resonance factor index can reflect the output of the nonlinear system transforming from CR to SR and then realize the detection of the characteristic frequency and then identify the bearing fault. Numerical and experimental verifications show the advantages of the CR method in the periodic potential system in bearing fault diagnosis are illustrated. By distinguishing CR and SR successfully, we can improve the accuracy of bearing fault diagnosis.