Abstract: Shared energy storage addresses the challenges of high cost and low utilization through the reuse of energy storage resources. Furthermore, rapidly developing demand-side resources have the potential to be applied in the shared energy storage, but the issue of their uncertainty requires urgent resolution. A virtual energy storage model for electric vehicles and thermal control loads is introduced, integrated with the physical energy storage, this model is employed to construct a comprehensive shared energy storage model that takes uncertainties into consideration, along with the corresponding optimization algorithms to determine the optimal capacity configuration of the physical energy storage. Shared energy storage operators optimize the configuration of multiple types of energy storages based on user demands and design the satisfaction compensation for virtual energy storage holders to safeguard their user experience and economic interests. Additionally, the Wasserstein distance is used to characterize the uncertainty associated with electric vehicles and temperature-controlled loads, in conjunction with the utilization of a risk-valuebased distributionally robust chance-constrained algorithm for optimization. The results of the case study demonstrate that the utilization of the generalized shared energy storage model and the distributionally robust optimization algorithm allow for a comprehensive consideration of uncertainty, leading to a substantial reduction in energy consumption costs for users and energy storage configuration costs for operators.