电力需求响应机制下基于多主体双层博弈的规模化电动汽车充放电优化调度

Large-scale Electric Vehicles Charging and Discharging Optimization Scheduling Based on Multi-agent Two-level Game Under Electricity Demand Response Mechanism

针对电力需求响应机制下电动汽车调度场景涉及的多元决策主体间的复杂博弈互动关系,该文提出多电动汽车聚合商分别整合规模化电动汽车入网参与电力市场竞价,并依据竞价结果指导电动汽车实时充放电优化调度的多主体双层博弈模型。首先,基于logit协议构建电动汽车充放电调度的多策略集演化博弈模型;其次,构建多电动汽车聚合商在电力市场中竞标购/售电价格的非合作博弈模型;再次,用复制者动态描述配电网运营商向各聚合商分配需求响应时段响应电量的策略演化;最后,联合求解双层博弈模型的演化均衡和纳什均衡,得到三主体的最优稳定策略。算例分析表明,提出的模型在实现负荷削峰填谷的同时,可有效平衡电动汽车用户、电动汽车聚合商和配电网运营商三者之间的经济利益。

In view of the complex game interaction between multiple decision-making subjects involved in the electric vehicle dispatching scenario under the power demand response mechanism, this paper proposes a multi-agent two-level game model in which multiple aggregators integrate large-scale electric vehicles into the grid to participate in the power market bidding, and guide electric vehicle charge and discharge optimization scheduling based on the bidding results. A multi-strategy evolutionary game model for electric vehicles charge and discharge scheduling with the logit protocol is constructed. A non-cooperative game model of multiple electric vehicle aggregators bidding purchase/sale electricity price in the electricity market is constructed. The replicator dynamics is used to guide the evolution of the distribution system operator’s strategy for allocating response power to each aggregator during each demand response period. Finally, the evolutionary equilibrium and the Nash equilibrium of the two-level game model are jointly solved, thus obtaining the optimal stable strategy. Experimental results show that the proposed two-level game model can effectively balance the economic interests of electric vehicle owners, electric vehicle aggregators and the distribution system operator while achieving peak cutting and valley filling of power load.