A novel stochastic resonance assisted deconvolution method and its application on damage identification of wind turbine bearings

Wind turbine bearings play an important role in power generation operations, and their damages will cause security risks. To identify the damages of wind turbine bearings effectively, a novel stochastic resonance assisted deconvolution method is proposed. In this method, two key parts of deconvolution are designed for extracting damage component. On the one hand, traditional second-order stochastic resonance (SSR) is improved based on parameter exploration and the defined indicator. Besides, an adjusted SSR technology is developed and fused with minimum Mahalanobis distance criterion to get the objective function, which is utilized for guiding the deconvolution orientation. On the other hand, a filter structure named iterative approximation structure is creatively constructed by the eigenvectors of covariance matrix for updating the filter coefficients of the deconvolution operation, which is utilized to recover damage component from the measured signal. The analysis results of the experimental signals and the engineering case demonstrate that this proposed method can effectively identify the damages, and the characteristic extraction ability, as well as the identification precision of this proposed method are better than those contrastive methods.