Abstract: With the increase of wind power penetration level, it is of a great practical significance to accurately grasp the characteristics and laws of wind power output forecast errors for improving the safety and economy of power grid operation. Based on the Pair-Copula theory, an uncertainty model considering the correlation and conditional distribution characteristics of output forecast errors of multiple wind farms is proposed in this paper, in order to effectively improve the accuracy of wind power output uncertainty modeling. Then, a stochastic unit commitment model incorporating the wind power risk cost and the system safety constraints is presented. By introducing the system reserve and network security constraints considering the boundary of wind power output interval, the system is able to safely accommodate arbitrary fluctuations of wind power within the set confidence interval, ensuring the feasibility of the intra-day scheduling scheme. In addition, the wind power interval boundaries are introduced as decision variables into the optimization process of the model, and the risk chance constraints are set to restrict the probability of load shedding risk and the wind curtailment risk to effectively improve the flexibility of scheduling decision. Finally, the effectiveness of the proposed model is verified in the simulation test of specific examples.