An Improved Multiscale Stochastic Resonance Method for Bearing Fault Diagnosis

Stochastic resonance (SR) has been widely used for bearing fault diagnosis. The multiscale noise tuning stochastic resonance (MSTSR) has been proven effective to analyze the weak bearing fault signals. However, the effect of MSTSR is not very satisfactory in practical application. The autocorrelation function (ACF) of original signals can extract the periodic components effectively. To enhance the weak signal enhancement performance, this paper proposes a method that combines the ACF and MSTSR. By applying ACF method in the system before MSTSR, a high output signal to noise ratio (SNR) can be obtained. The performance of the proposed method has been validated by comparing with the effect of MSTSR method on different types of bearing fault signals.