To address the complex multi-agent game problem of electric vehicles under demand response mechanisms, a scheduling method is proposed for virtual power plants and electric vehicle operators based on a master-slave evolutionary hybrid game, considering low-carbon objectives. First, an evolutionary game model for electric vehicle charging and discharging scheduling is constructed based on the logit protocol. Second, a master-slave evolutionary hybrid game model is established between the virtual power plants and electric vehicles, where the virtual power plants act as leaders to set electricity prices, and electric vehicle aggregators act as followers to determine charging and discharging strategies based on the prices. Finally, the Nash equilibrium and evolutionary equilibrium are jointly solved to obtain the optimal strategies for each agent. Simulation results verify that the proposed method can effectively reduce grid carbon emissions and mitigate load peak-valley differences while considering the interests of electric vehicle users, thereby achieving win-win outcomes among multiple parties.