Nuclear power control system products must be firmly fixed in the cabinet to ensure that the control system can work normally under earthquake and other working conditions. The locking mechanism of a nuclear power control products were analyzed by FTA, and the quantitative probabilistic safety analysis of the key failure mode locking mechanism loosening event was carried out. Firstly, the Finite element model of locking mechanism was established by using ANSYS, conduct a deterministic simulation analysis under seismic conditions. Then, based on the reliability theory, the failure mode interference equation was established according to the deformation of the locking hook of the locking mechanism, Monte Carlo reliability algorithm program and ANSYS joint simulation interface program were developed, for the failure of locking mechanism loosening, Monte Carlo sampling statistical joint simulation calculation was carried out by using the ANSYS joint simulation interface program, the probability of loosening failure event of locking mechanism under different confidence was calculated. This paper explores a new method for quantitative probabilistic safety analysis of locking mechanism, which has reference value for quantitative probabilistic safety analysis of similar mechanisms. The locking mechanism loosening event probability in the fault tree of probabilistic safety analysis was accurately calculated by using this method, through the introduction of reliability algorithm, it focuses on the evaluation core processing of uncertainty, which is an improvement of probabilistic safety analysis method.