The high proportion of renewable energy connected to the grid puts forward higher requirements for the flexibility and regulation ability of the power system. As a flexible regulation resource, energy storage will gradually replace the traditional thermal power units in the new power system and undertake more tasks of frequency modulation and peak load regulation. However, the high price of energy storage limits its large-scale application. Therefore, from the technical and economic point of view, this paper considers the energy storage capacity planning under the background that the wind storage combined system participates in the peak shaving and frequency modulation auxiliary service. Firstly, in order to deal with the uncertainty of wind power and load output, based on the historical data of wind power and load, this paper uses K-means clustering method to generate their own typical scenarios, and generates wind load typical joint scenarios and their corresponding probabilities by Cartesian product. Secondly, combined with the characteristics of peak shaving and frequency modulation demand, the various costs and benefits of the wind storage system in the process of participating in auxiliary services are quantitatively modeled, including the benefits of auxiliary peak shaving and frequency modulation, the initial investment cost of energy storage, and the life attenuation cost of energy storage. Finally, a wind farm in a certain area is taken as an example to verify the feasibility and effectiveness of the proposed method.