To address the influence of large-scale wind power generation on power system stability，a new preventive control method for the transient stability of power systems was proposed.Firstly，the expression of wind power output was established based on the Weibull distribution.Secondly，in a system containing wind power，energy storage，and flexible load，taking the lowest adjustment cost as the objective function，a transient stability estimator based on a stacked ensemble deep belief network was used as the transient stability constraint to construct a preventive control model.Finally，to solve the problems of insufficient population diversity and easy trapping into local optima in the traditional grey wolf optimization algorithm，a hybrid improved grey wolf algorithm combining artificial fish swarm foraging behavior and a quadratic interpolation method was constructed.The simulation results on the IEEE 39-bus and IEEE 68-bus systems show that the proposed model and method can reduce the preventive control cost of wind power grid-connected systems and provide a guarantee for transient stability operation.