多代理结构下的电动汽车充电控制协调策略研究-电气工程专业论文.docx

万方数据 万方数据 多代理结构下的电动汽车充电控制协调策略研究 摘要 随着我国污染日益严重、化石能源保有量逐渐减少，能源问题逐渐 被提上日程。电动汽车的推广与开发可以缓解能源危机和污染问题，故 其在近几年取得了飞速的发展。然而，电动汽车的自由充电一旦达到一 定数量，就会影响电网稳定性，增加其运行成本。于是，电动汽车调度 的控制策略和管理成为了当今学者研究的焦点。 在基于蒙特卡洛抽样与极大似然法对电动汽车行为进行描述的模型 上，本文以电力公司利益最大化为目标，设计了一种用于协调电动汽车 充电的四层结构多代理系统，并详细讨论各成员代理的职能。在系统架 构下，提出了一种预约模式，并针对车主可能的充电需求设计三种预约 模式，并提出相应的激励机制以引导车主充电。 最后，基于多代理系统设计一种电动汽车充电分布式协调算法。引 入 TOU 分时电价法以引导电动汽车在理想时间段充电；提出一种改良优 化算法来削去 TOU 算法带来的新的负荷尖峰；采取训练学习机制不断迭 代优化算法使负荷波形完美填谷。逐步达到理想的负荷状态。仿真结果 体现了此算法的有效性，并定量讨论本算法对电力公司利益的正向影响。 关键词：电动汽车；有序充电；多代理；分布式算法 III RESEARCH ON COORDINATION STRATEGY OF CHARGING FOR ELECTRICAL VEHICLES IN MULTI- AGENT STRUCTURE ABSTRACT As the more and more seriously problem of the pollution and the decreasing of fossil energy, the energy issue is being put on the table. The research and the industrialization of the Electrical vehicle (EV) is the key to the problems. So it has made a huge step in recent years. However, the large population of EV free charging will cause negative influence on the power Grid. Therefore, the dispatching of EVs charging has become hot topic of the research of scientists. Basing on the description model of the EV’s behavior from Monte Carlo method and the maximum-likelihood method, a four-level Multi-agent structure towards the maximum of power company’s profit is designed. And in this structure, this paper proposes a reservation method, which contains three models of reservation in connection with the EV owners demand. An excitation mechanism is also come up with based on the reservation mechanism. A distributed optimization control algorithm of EVs charging is proposed based on the MAS. Firstly, the preliminarily optimized time-of-use (TOU) IV price algorithm is used to obtain ideal charging time interval; secondly, in the obtained time interval a reformative optimization algorithm is utilized to avoid the occurrence of new peak load; finally, the training-learning mechanism is led into to achieve the effect of peak load shifting. Results of this rese