【作者】 陈杰；

【导师】 关新平；

【作者基本信息】 上海交通大学 ， 控制科学与工程， 2015， 硕士

【摘要】 环境污染已成为制约经济社会健康发展的重要问题之一,因此推进能源结构改革和提高能源利用率是未来电网的重要发展方向之一。作为对集中式发电的补充,分布式新能源发电的加入对推进能源绿色化和提高能源利用率方面有着至关重要的作用。同时,以混合动力汽车为代表,各类智能用户端大量接入电网中,使得负荷端不再呈现出完全刚性特征,为能源的调度与优化提供了新的思路。基于此,本文主要做了以下三个方面的工作:1、根据电力市场分级的特点,提出了基于双边竞拍方法的虚拟发电厂最优日前交易策略。把混合动力汽车按照时延需求不同进行了分类,然后根据日前电力市场的竞拍结果和新能源的实时随机特征,在满足时延要求的前提下,给出了实时电力市场中混合动力汽车的能量分配策略。2、提出了一个两层动态博弈框架来解决充电站和混合动力汽车群的利益优化问题。本文采用了演化博弈和模仿者动态来刻画下层混合动力汽车选择充电站的过程,同时采用了微分对策论方法帮助上层各个充电站设立充电价格,最后分析了博弈均衡点的存在性和求法。3、研究了一个动态“代理商-混合动力汽车”系统中代理商利益最大化问题。本文通过研究混合动力汽车用电驱动模式,建立一个等效的燃油花费,以此设定一个定价约束条件,将混合动力汽车的用油和用电比例问题转换成是否与代理商进行电能交易的问题。本文采用了李雅普诺夫优化方法来解决代理商利益最大化问题,将原来的优化问题分解成了制定充电价格和分配能量两个子问题,提出了求解这两个子问题的算法。最后进行了仿真,验证了该算法的有效性。更多还原

【Abstract】 Environmental pollution problem has become one of the important problems that restrict the healthy development of economy and society,therefore to promote the reformation of energy structure and improve energy effciency is one of the important developing directions of future smart grid. As a supplement of centralized power generation, distributed renewable energy sources play an important role of creating a greener energy structure. At the same time, the appearance of smart demand loads,take hybrid electric vehicles(HEV) as an example, have changed the rigid character of traditional loads and bring up some new ideas for power system operation. Based on the above discussion, we focus on the following three challenging works.According to the character of electricity market, we propose a double auction method to help the virtual power plant to get its optimal trading strategy in the dayahead market and classify the hybrid electric vehicles into different types according to their toleration delay time. Then based on the results of the auction and the random character of renewable energy sources, we give the energy dispatch strategy for hybrid electric vehicles while satisfying the delay time constraints.We propose a two-level game approach to describe the dynamic competition relationship between hybrid electric vehicles and charging stations. We use evolutionary game and replicate dynamics to depict the charging station selection process for the vehicles in the lower level and apply differential game to help charging stations in the upper level to set their optimal charging prices. Then we analyse the existence and uniqueness of the equilibrium for the two-level game.Then we study a real-time payoff maximization problem for the aggregator in a dynamic aggregator-HEV system. The aggregator will establish an equivalent fuel cost after studying the electricity-only driving mode of the hybrid electric vehicles. Then based on the equivalent fuel cost price, the aggregator will set a pricing constraint to convert the electricity gas proportion problem into whether choose to trade energy with the aggregator. We apply the Lyapunov optimization framework to decompose the original problem into the charging price decision problem and energy allocation problem and propose an algorithm that do not need statistical information to solve the two subproblems. Then we validate the effectiveness of our algorithm through simulation results.更多还原