Optimal scheduling for regional power grid is an important means to increase the consumption of renewable energy, enhance the reliability of power systems, and reduce operating costs. Therefore, focusing on a regional power grid including electric vehicle (EV) charging stations, an auction-based two-stage optimal scheduling method is proposed. In the first stage, the regional power grid is scheduled one day ahead based on the predicted values of local renewable energy generation and consumption. In the second stage, the day-ahead scheduling results are adjusted according to the actual measured values of generation and consumption. For the second stage, an electricity auction method is designed to achieve real-time supply-demand balance of the regional power grid by using EV charging and discharging energies. It is proved that the proposed auction method has the property of incentive compatibility, which ensures that EVs submit true energy demand or supply information in the auction. Simulation results show that the proposed two-stage method can reduce the operating cost of the regional power grid and make EVs obtain the highest utilities only when true information is submitted.