Adaptive stochastic resonance under Poisson white noise background and its application for bolt looseness detection

Abstract

Bolts are utilised extensively in machinery and often bear large loads. Reliable connection of bolts is related to the effective functioning of machinery. Therefore, it is of utmost importance to detect bolt looseness in time. However, the identification of bolt looseness is typically challenging due to the strong background noise. Compared with Gaussian white noise, only few researches were conducted on Poisson white noise. To detect the looseness of the bolt in the presence of strong Poisson white noise, we propose a novel method based on sub-harmonic resonance, time-domain averaging and adaptive stochastic resonance. The disadvantages of damaging characteristic frequencies that exist in a majority of approaches are overcome. In addition, the looseness is assessed by the quality factor derived from physical science. To verify the efficacy of the method, we propose numerical simulations and experimental validations. The results demonstrate that the proposed method has effectively detected bolt looseness under strong Poisson white noise. The detection of bolt looseness might benefit greatly by adopting the suggested approach.