With the large-scale accessing of various forms of distributed energy, its inherent random characteristics and multi-player conflicts of interest bring new challenges to power system dispatching. By deeply integrating a robust optimization and zero-sum game mechanism, taking nature and power dispatcher as game players, a robust dispatching collaborative planning method for a power system considering the game between dispatcher and nature is proposed. There is a problem of the uncertainty of the coupling between the traffic and energy properties of electric vehicles. To address this electric vehicles are first clustered based on their traffic attributes. Then, based on power output and operating cost characteristics of the vehicles, wind power and thermal power units, a min-max scheduling game model with multiple energy forms is established and analysed by a two-stage relaxation algorithm. The simulation results confirm that the proposed dispatching model and related strategies can realize the coordinated synergistic economic operation of wind, electric vehicles and other energy without relying on accurate wind power prediction. This provides a new approach for solving the power dispatching decision problem of uncertain distributed energy access. This work is supported by the General Program of National Natural Science Foundation of China (No. 61873292).