基于集合器分层优化的入网电动汽车控制策略研究

The large-scale integration of electric vehicles (EVs) has placed new challenges to modern power systems, but charging and discharging energy of EVs, if properly controlled, can be used as mobile energy storage to provide grid supports such as peak shaving, valley filling, voltage stabilization, power loss reduction and compensation for the power fluctuation caused by intermittent renewable energy sources. A control framework of vehicle-to-grid (V2G) operation is firstly proposed for integrating EVs into the power grid. As a large number of EVs are plugged in, the charging/discharging power of each individual EV are gathered by an EV aggregator as a central control unit on the top of a EV fleet. The optimal size and location of EV aggregations are found and the bi-directional power flow through EV charger is regulated for cost and emission minimization by using Benders decomposition and stochastic optimization tools. The V2G regulated power is also optimized with the varying load and distributed power generators in such a way that the power quality and power system reliability can be enhanced under the condition of large penetration of renewable energy sources including wind power and solar panels. The limitations of the communication and electric infrastructure for V2G operation are also modeled in the optimization algorithm. As the EV charging infrastructure alone can provide reactive power regulation services even if EVs are not parked at the charging infrastructure, both active and reactive power regulation are performed. At last, the proposed V2G control algorithm will be implemented on an experimental platform based on a group of bi-directional power converters. The simulation and experimental results will be analyzed for further improvement of V2G modeling and optimal control. The outcome of this research project will provide a suitable solution for the integrated optimization of EV aggregator and V2G operation in power systems.

经优化调控后的入网运行电动汽车（EV），可作为移动式储能设备对电网运行形成有效支撑，本项目针对入网电动汽车集群的控制策略进行深入研究。首先,将构建基于集合器的电动汽车集群分层优化控制模型；其次，在分析电动汽车充换电与电网交互特性的基础上，构建集合器配置与运行综合优化模型，并提出基于Benders分解，融合随机问题处理及Pareto排序的优化问题求解方法，设计电动汽车集群最优配置和集合器的充放电功率控制方法；其三，考虑用户需求、充放电设施和通信能力，设计电动汽车集群能量优化调度和功率快速控制算法，并利用充电设施本身具有的无功补偿能力，实现多时间尺度的充放电功率实时控制和无功电压调节；最后，构建充放电设备控制与优化算法相结合的软硬件综合实验平台，验证并进一步修正优化控制模型，总结适用于V2G建模和控制的一般规律和方法，为大规模电动汽车并网控制和集合器设计提供一种较好的解决方案。

随着电动汽车的推广普及和充换电设施建设规模不断扩大，考虑电动汽车“荷-储”双重特性和充电设施功率调节能力，对入网电动汽车集群进行充电负荷采取有效的调度和控制，可以实现电网公司、用户和充换电站运营商的多赢。项目主要围绕以电动汽车集群控制器为核心的分层优化模型、集群配置与运行综合优化及其求解算法开发、为电网提供调压调频辅助服务等内容展开研究。取得的进展如下：1、针对电动汽车集群充电负荷分布和可调节能力评估，提出了基于电动汽车集群分层优化框架下的V2G数学建模和充电集群变量可调节范围计算方法，为后续充电控制策略的设计提供基础；2、构建了电动汽车集群配置与运行综合优化模型，并提出基于Benders分解的解耦算法高效求解综合优化问题；3、提出了考虑充电设施运行特性的有功-无功混合功率控制方法，结合日前能量市场和调频服务市场提出考虑整体市场收益的电动汽车集群充电优化策略，并通过充电设施有功-无功功率控制的仿真测试，验证了电动汽车集群提供调压调频等辅助服务的可行性。研究成果可为电动汽车充电设施配置和集群控制策略的实际应用提供技术支持，为电动汽车集群参与电网经济调度和调压调频等辅助服务的分层优化的关键科学技术问题提供一种良好的解决方案，为推动电动汽车与电网灵活互动，参与电力市场调节，电动汽车充电网络与智能配电网的协调发展奠定了部分理论与技术基础。