Abstract: In order to solve the problem of resource waste and cost increase caused by blind investment,the hybrid model of self-built and lease under tiered cost is applied to the cloud energy storage configuration in park. Firstly,the characteristics of cloud energy storage are analyzed,and service mode of cloud energy storage with a large number of photovoltaic users participating in the park is constructed. Secondly,a bi-level optimization model considering different time scales and two subjects is established,the upper layer solves the planning problem of cloud energy storage in long time scale,and the lower layer solves the operation problem of user group in a short time scale. Thirdly,the two-layer model is transformed into a single-layer model by Karush-Kuhn-Tucker（KKT）condition,and the resulting single-layer model is linearized by Big-M. Finally,with the analysis of three different scenarios,the results demonstrate the effectiveness of proposed cloud energy storage allocation model,it can reduce the investment costs of energy storage,improve the utilization rate of energy storage resources and save electricity cost of users.