Abstract: Aiming at the problem that the traditional stochastic resonance system cannot adjust the structural parameters adaptively, and that stochastic resonance system using the adaptive genetic algorithm and swarm intelligence algorithm is easy to cause the defect of local optimization with the increase of the population number, a fault diagnosis method of stochastic resonance based on adaptive bat algorithm is proposed. The method is based on the model of bistable stochastic resonance system, combined with the idea of twice sampling frequency transformation, and output the structural parameters of bistable stochastic resonance by adaptive bat algorithm, so as to enhance the effect of fault signal and realize the fault diagnosis of rolling bearing. In order to verify the feasibility and superiority of this method, the simulation data analysis and the application diagnosis of bearing inner and outer ring fault are carried out. The results show that the model of the method is simple, the parameters of the algorithm are few, the convergence speed is fast, and it can accurately and efficiently realize the fault diagnosis of bearing inner and outer rings.