Abstract:Stochastic resonance has the advantages of extracting bearing fault feature frequency and converting noise energy into signal energy. To solve the problem of unsatisfactory performance caused by inadequate selection of models and parameters, a model of asymmetric piecewise linear stochastic resonance system is proposed. The model is assumed to have two steady states. The analytical relationship and the output signaltonoise ratio formula of the model are derived. Meanwhile, the model is compared with the symmetric piecewise linear stochastic resonance system in terms of formula analysis and numerical simulation. The advantages of the system are proved. Then, the model is applied to bearing fault detection, and the adaptive genetic algorithm is used to optimize the parameters of the system. In the inner and outer fault detection for the symmetric system, the amplitude frequency values are 5079 and 1303, respectively. In the inner and outer fault detection for the asymmetric system, the amplitude frequency values are 3918 and 1893, respectively. Results show that the model has excellent detection effect, which proves that the model has great potential in fault detection.