In order to solve the problem caused by the randomness of the wind power system, the load forecasting error and the probability distribution function of wind speed are introduced to establish the stochastic model of the load and wind power. Based on the in-depth analysis of the load and the random model of wind power, the cost of the wind power access to the system to maintain the stable operation of the cost of the risk is considered, and the risk of wind power loss and risk indicators are introduced. A multi-objective optimization scheduling model for power system with wind power generation risk is constructed in order to achieve the overall optimization of the system total pollution emissions and the lowest operating cost. The model is solved by using a novel bat algorithm, and the shortcomings of the algorithm are improved. The algorithm can effectively solve the problems such as slow convergence and easy to fall into local extremum. Taking large-scale wind power system as an example, the environmental and economic dispatch simulation is carried out to verify the effectiveness of the proposed model and the algorithm.