In view of the high cost and insufficient resource utilization of energy storage invested independently by users, the introduction of shared energy storage in the park is conducive to reducing the power cost of users in the park and promoting the consumption of distributed renewable energy. In order to achieve a win-win situation between shared energy storage service provider and park users, a Stackelberg game model in which shared energy storage service provider dominates and park users follow is established. The shared energy storage service provider sets the capacity price and power price and uses conditional value-at-risk to evaluate the income risk caused by the uncertainty of photovoltaic output. According to the price of shared energy storage and the forecast of their own load and photovoltaic output, park users decide the purchased energy storage capacity and charging and discharging power strategy. Through KKT optimality condition and the dual theorem of linear programming, the above Stackelberg game problems can be transformed into mixed integer linear programming problems. Finally, based on the example of multiuser and multi-scenario, the influence of energy storage price on the game result is analyzed, and the economy of shared energy storage is compared with that of fixed energy storage configuration and no energy storage confiodel.