The frequency security problem of power system is highlighted as wind power penetration increases yearly, the eigenvalue analysis method based on the deterministic model is difficult to accurately evaluate the small signal stability of power system. The Weibull probability distribution is used to describe the uncertainty of wind speed. In this paper, the shape and scale parameters of the Weibull distribution are calculated by using the maximum likelihood approach and combining with the measured data, and the stability of small signal with large-scale wind power access is analyzed by the probability distribution method. Meanwhile, considering that the lack of inertia of the system caused by high-penetration wind power access can easily lead to the problem of frequency stability, it is proposed to take the maximum damping ratio of regional oscillation mode as optimization objective, virtual inertia and pitch angle control parameters as variables, and frequency stability as constraints, an small signal stability optimization model of wind power access to power system is established according to the proposed probabilistic small signal stability model. The sensitivity algorithm is used to obtain the optimal solution for the virtual inertia and pitch angle control parameters, and the validity of the proposed method and model is verified by two-area four-machine examples. Simulation results show that by optimizing the virtual inertia and pitch angle parameters of wind turbines, the small signal stability of the high-penetration wind power access system can be improved under the condition of ensuring sufficient inertia of the system, which provides a certain theoretical basis for the safe and reliable operation of wind power large-scale cluster access power system.