A parameter-adaptive stochastic resonance based on whale optimization algorithm for weak signal detection for rotating machinery

Abstract

The parameters of stochastic resonance (SR) system determine its output directly. Inappropriate parameters may lead to an unsatisfactory result even an error. To search the best parameter adaptively, a novel method is proposed in this paper. In the method, the relationships between each parameter are considered. Also, the Kramers rate of noise and some information about the weak signal are considered in our method. In the novel method, what we need to determine artificially is the interval of only one parameter. The others will be searched adaptively and optimally. Therefore, the proposed method is much different with traditional parameter-adaptive SR. The effectiveness and accurateness of our method are proofed by simulation, rotor crack detection and bearing inner diagnosis detection. The results compared with some other methods demonstrate our method can work more effectively. At last, we studied an interesting phenomenon founded in the process of our study and proposed a guide that the upper limitation of the Kramers rate should be equal to the target frequency after compressing when the re-scaling technology was used in SR system. Consequently, this paper makes a great contribution in the application of SR.