With the extensive utilization of wind energy resources, more and more attention has been paid to the efficiency and safety of wind penetration. Using statistical methods to analyze the characteristics of wind speed, we can know the change rule of wind energy, so as to make reasonable distribution and scheduling of electricity, and to maximize the reduction of losses and risks. At present, there are a large number of literatures on the use of energy storage system to suppress wind farm output, but there are few researches on the introduction of conditional VaR (Value-at-Risk) as an evaluation index into wind energy resource assessment. Aiming at sloving the instability of wind speed and the serious problem of wind curtailment and power limitation, this paper proposes to configure energy storage devices with a certain capacity at the grid-connected wind farms. In the proposed method, the conditional risk value of the system is taken as the optimization objective, and energy storage devices are adopted to control the charging and discharging of wind farms, so as to achieve the purpose of stabilizing wind power fluctuations and reducing wind curtailment. We analyze the calming effect of energy storage system on wind power fluctuation through real-world examples. The quantitative and visual results show that the proposed method is of great significance in wind power peaking and valley filling.