The charging and discharging demand of electric vehicles is closely related to the load balance of the power grid. In order to effectively alleviate the load pressure caused by the disorderly connection of large-scale electric vehicle (EV) to the power system during the peak period of electric vehicle charging and discharging, a research on electric vehicle charging and discharging scheduling based on comprehensive energy grid load balance is proposed. In vehicle-to-grid (V2G) mode, the architecture and requirements of EV charging and discharging scenarios are analyzed. The upper layer takes the minimum peak valley difference as the objective function, and the lower layer takes the minimum charging cost as the objective function. The corresponding constraints are set and a dual layer scheduling model is established for electric vehicle charging and discharging. Innovatively using the roulette wheel selection algorithm combined with the crisscross optimization algorithm (CSO) to solve the dual layer scheduling model of electric vehicle charging and discharging, and obtain the optimal electric vehicle charging and discharging scheduling scheme. The experimental results show that the proposed method can effectively reduce the peak-valley difference and scheduling time consuming, and obtain more satisfactory scheduling scheme.