The reliability optimization design method under stochastic winds was established. First, due to the fluctuating characteristic of natural winds, the fluctuating winds of a moving point shifting with high-speed trains were simulated based on Cooper theory and harmonic superposition method, and an algorithm was proposed to calculate the unsteady aerodynamic loads of high-speed trains under stochastic crosswinds. Then the multi-body system dynamics model of a high-speed train was established to calculate the operation safety of high-speed trains, and the computational formula of the probability of failure of high-speed trains under stochastic winds was further shown. On this basis, the multi-objective genetic algorithm NSGA-II was used for the automatic optimization, then the reliability optimization design model of kinetic parameters of high-speed trains was established. In the model, the kinetic parameters were taken as the optimization variables, and the probability of failure and wheelset lateral force were taken as optimization objectives. After optimization, the probability of failure was reduced from 0.4884 to 0.1406, and the wheelset lateral force was reduced from 45.13kN to 43.01kN. The operation safety of high-speed trains under stochastic winds was significantly improved by optimizing the kinetic parameters.