The study of the influence of potential well parameters on the output of a nonlinear energy harvester system is conducive to the design of the high-performance energy harvester system. Meanwhile，the stochastic resonance phenomenon in the corresponding electromechanical coupling dynamics model of the energy harvester system can be used to enhance the characteristics of weak faults，so as to effectively identify weak faults. This paper proposes a decoupled saddle-point-degradation bistable potential function，and the electromechanical dynamic model is introduced. The bifurcation diagram under different excitation amplitudes is obtained to discuss the effect of the barrier width and the barrier height on the responses （periodic response and chaotic response）. According to the methods of the Poincaré map，the frequency spectrum analysis，and the Lyapunov exponent，the periodic response and the chaotic response are examined at a fixed excitation amplitude，which is consistent with that obtained from the bifurcation diagram. Based on the electromechanical dynamic model perturbed by the random noise，the stochastic-resonance-based method is proposed for fault diagnosis，which achieves the enhancement of the simulated and experimental bearing fault characteristics.