Adaptive Stochastic Resonance-Based Processing of Weak Magnetic Slippage Signals of Bearings

Slip is one of the most common forms of failure in aviation bearings, and it can pose a great threat to the stable operation of aviation bearings. Bearing cage speed monitoring methods based on weak magnetic detection can achieve nondestructive measurements. However, the method suffers from solid signal background noise due to the high sensitivity of the sensor. Therefore, in this paper, an adaptive stochastic resonance algorithm was proposed in response to the characteristics of the weak magnetic detection signal and the problem of solid noise. In addition, by adaptively adjusting the coefficients of the stochastic resonance system—by an improved moth flame optimization algorithm—the drawback in which the stochastic resonance method required artificially set parameters for extracting the feature frequencies of the weak magnetic signals was solved. In this process, we used parameters, such as general refined composite multi-scale sample entropy, as the adaptation function of the optimization algorithm. In the end, simulation and experimental outcomes verified the efficacy of the approach put forward.