A novel adaptive stochastic resonance method based on coupled bistable systems and its application in rolling bearing fault diagnosis

Stochastic resonance (SR), as a typical noise-assisted signal processing method, has been extensively studied in weak signal detection by virtue of the advantage of using noise to enhance the feature of periodic signal. And how to better apply the SR method in engineering signal processing has always been the research hotspot. In this paper, a novel adaptive SR method based on coupled bistable systems (ACBSR) is proposed to further improve the performance of SR method for weak signal detection. Compared with the cascaded bistable SR system, the suggested coupled bistable SR (CBSR) system can enhance the target signal while suppressing the low-frequency and high-frequency interferences, whose output behavior exhibits good bandpass filtering characteristic. In addition, a weighted spectral peak signal-to-noise ratio (WPSNR) index is constructed by using spectral peak SNR, spectral correlation coefficient and zero-crossing ratio to serve as the measurement index of SR for periodic signal detection. The constructed WPSNR index is simple in form and easy to calculate, which can quantitatively evaluate the resonance effect of periodic signal, even though the target signal frequency cannot be estimated accurately. And the combination of WPSNR index and the CBSR system is beneficial to improve the performance of SR method in rolling bearing fault diagnosis. Experiments and engineering application are performed to verify the effectiveness and superiority of the proposed ACBSR method in comparison with the traditional SR method.