电力市场环境下基于多智能体的风电接入成本分摊研究

发布时间：2019-06-10 15:30

【摘要】：近年来,我国电力市场化改革逐步深入,电力市场初步完成由“垂直一体化”向“厂网分开、竞价上网”的转变,发电权交易、辅助服务补偿机制等多种电力市场化实践逐步推行。各区域电力市场出台了不同的辅助服务管理实施方案,补偿原则主要依据机组的容量和成本,无法区分不同机组参与辅助服务的贡献。随着我国对能源保障和节能减排的重视,大力发展风电等可再生能源是能源战略的重要部分,风电在电网电源中比例的不断提高。风电具有间歇性、不确定性,且输出功率有别于常规电源的有功出力。对于以火电调峰为主的电网,风电的反调峰特性增加了调度机构的调峰压力。当风电接入容量超过电网的接纳能力时,要弃风以保证负荷平衡。风电并网替代火电机组发电,获得发电收益的同时增加了电网的辅助服务费用。一种可行的方法是从风电并网的收益取出一部分,作为对水电、火电机组的辅助服务的补偿。本文以调峰为例,从量化风电接入引起的机组调峰成本和确定基于合作博弈的成本分摊方法两个方面,研究分析风电并网的调峰成本补偿。首先,本文研究了常规机组的调峰成本,采用名义成本和实际成本的概念,剔除机会成本的干扰因素,得到风电并网导致的火电机组的能耗增量。算例构建了一个包含供热机组、非供热火电机组和水电机组的电力系统,模拟大规模风电接入火电调峰为主的电网,量化分析了风电接入后调峰增加的实际成本。其次,本文采用多智能体的方法分析了机组调度,以只有发电商和调度机构的电力系统为例,在满足系统功率平衡的条件下,以运行成本最小化为目标,分析了调度Agent和发电Agent的行为方式,给出了求解机组联盟的思路。不同机组联盟的调峰成本不同,成本最小的机组组合方式为最优调度方式。最后,本文分析了合作博弈方法对调峰成本分摊的适用性,夏普利值分摊法的依据是将收益按照所有的边际贡献进行分摊,更能体现分摊方案的公平性。算例分析了两个火电机组和一个水电机组的调峰成本,假设对辅助服务的补偿等于机组组合方案中最高的成本,采用夏普利值分摊法确定各机组的调峰补偿,突出了水电机组的调峰价值,为风电接入后对常规机组的调峰补偿分摊提供了思路。
[Abstract]:In recent years, the reform of electricity marketization in China has gradually deepened, and the power market has initially completed the transformation from "vertical integration" to "separation of power plants and networks, bidding for the Internet," and the trading of power generation rights. Auxiliary service compensation mechanism and other power marketization practices are gradually implemented. Different implementation schemes of auxiliary service management have been introduced in each regional power market. The compensation principle is mainly based on the capacity and cost of the unit, and it is impossible to distinguish the contribution of different units participating in the auxiliary service. With the attention paid to energy security and energy saving and emission reduction in our country, the development of renewable energy such as wind power is an important part of energy strategy, and the proportion of wind power in power supply of power grid is increasing. Wind power is intermittent and uncertain, and the output power is different from the active power output of conventional power supply. For the power grid dominated by thermal peak shaving, the reverse peak shaving characteristics of wind power increase the peak shaving pressure of dispatching mechanism. When the wind power access capacity exceeds the acceptance capacity of the power grid, the wind should be abandoned to ensure load balance. Wind power connected to the grid replaces thermal power generation, which not only obtains the power generation income, but also increases the auxiliary service cost of the power grid. One feasible method is to withdraw part of the revenue from wind power grid connection as compensation for auxiliary service of hydropower and thermal power units. Taking peak shaving as an example, this paper studies and analyzes the peak shaving cost compensation of wind power grid connection from two aspects: quantifying the peak shaving cost caused by wind power access and determining the cost allocation method based on cooperative game. Firstly, this paper studies the peak shaving cost of conventional units, adopts the concepts of nominal cost and actual cost, eliminates the interference factors of opportunity cost, and obtains the energy consumption increment of thermal power units caused by wind power grid connection. An example is given to construct a power system including heating unit, non-heating thermal power unit and hydropower unit, which simulates the large-scale wind power to connect to the power grid dominated by thermal power peak shaving, and quantitatively analyzes the actual cost of peak shaving after wind power access. Secondly, the multi-agent method is used to analyze the unit dispatching. Taking the power system with only generator and dispatching mechanism as an example, under the condition of satisfying the power balance of the system, the minimum operation cost is taken as the goal. The behavior mode of dispatching Agent and power generation Agent is analyzed, and the idea of solving unit alliance is given. The peak shaving cost of different unit coalitions is different, and the unit combination mode with the lowest cost is the optimal dispatching mode. Finally, this paper analyzes the applicability of cooperative game method to peak-shaving cost allocation. The basis of Sharpley value sharing method is to share the benefits according to all marginal contributions, which can better reflect the fairness of the allocation scheme. The peak shaving cost of two thermal power units and one hydropower unit is analyzed by an example. Assuming that the compensation for auxiliary service is equal to the highest cost in the unit combination scheme, the split value sharing method is used to determine the peak shaving compensation of each unit. The peak shaving value of hydropower units is highlighted, which provides an idea for the allocation of peak shaving compensation for conventional units after wind power access.
【学位授予单位】：华北电力大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F275.3;F426.61

【参考文献】