【摘要】 电动汽车作为一种分散资源参与电力系统优化调节具有较大的随机性特点，提出了一种电动汽车参与虚拟电厂优化调控的数字孪生架构，在数字孪生模型中实现状态推演。虚拟电厂充分考虑了电动汽车用户的响应程度，对电动汽车用户进行分级并制定不同的合约。为实现状态推演，在充分考虑电动汽车行为特征的基础上，进行了电动汽车虚拟电厂的充电需求与调节潜力评估；结合配电网调度侧与虚拟电厂侧两个层面，提出对电动汽车的优化调控策略，并对电动汽车与虚拟电厂整体进行了状态推演，以实时推演的状态来修正日前模型的计算结果，充分挖掘并利用需求侧灵活可调的电动汽车资源。通过算例分析，验证了所提出的状态推演策略对分散电动汽车资源潜力评估与调控对配电网优化运行的有效性。更多还原

【Abstract】 As a decentralized resource, Electric vehicles have significant randomness characteristics when involved in power system optimization and regulation. This paper proposes a digital twin architecture for electric vehicles to participate in the optimization and regulation of virtual power plants, and implements state inference in the digital twin model. The virtual power plant fully considers the responsiveness of electric vehicle users, and the users are graded and different contracts are formulated. In order for state inference, the evaluation of the charging demand and regulation potential of the electric vehicle virtual power plant is conducted,taking into account the behavioral characteristics of electric vehicles. Combining the dispatching side of the distribution network and the virtual power plant side, an optimized regulation strategy for electric vehicles is proposed, and the overall state of electric vehicles and the virtual power plant is deduced in real-time to correct the results of the day ahead scheduling model, fully mining and utilizing the flexible and adjustable electric vehicle resources on the demand side. The effectiveness of the proposed state inference strategy in evaluating and regulating the potential of dispersed electric vehicle resources and optimizing the operation of distribution networks is verified through case analysis.更多还原