Abstract: In order to improve the signal enhancement ability of two-dimensional stochastic resonance, a two-dimensional asymmetric stochastic resonance system based on external signal and Gaussian function (NTABPS-EG) is proposed. Firstly, the system performance of the three asymmetric potentials is studied by taking the output signal-to-noise ratio as the measurement index, and the asymmetric potential of the well depth and well width is selected for follow-up research according to the analysis results, and the variation trend of the output signal-to-noise ratio with the noise intensity under different parameters is explored. Secondly, the fourth-order Runge-Kutta algorithm is used for numerical simulation, and the genetic algorithm is used to solve the optimal parameters, so that the system can obtain the best output response. By comparing with the two-dimensional exponential potential bistable potential system (TEPBPS), it is found that the NTABPS-EG system can drive the particle transition more effectively, and the recovery effect of weak signals is better. Finally, the NTABPS-EG system is applied to the bearing fault detection, and the results show that the system can effectively diagnose the inner and outer ring faults of the bearing, and the diagnostic ability is better than that of TEPBPS, which proves the practicability and advancement of the system.