导　　师： 张尧

学科专业： 080802

授予学位： 博士

作　　者： ;

机构地区： 华南理工大学

摘　　要： 全国跨区输电以及区内跨省输电交易和大用户直接选择发电企业购电,都属于双边交易模式的范畴。在未来开放程度越来越高的电力市场中,双边交易将大量存在。双边交易模式是未来电力市场的发展趋势,对双边交易模式下电力市场的各种问题进行分析和研究,有着必要性和现实意义。目前国内外对于双边交易的研究侧重于网损分摊、阻塞费用分摊、市场机制和规则的制定等方面,对双边交易的博弈分析、基于代理的谈判策略、阻塞管理、模拟仿真方面的深入研究都不多。 本文对电力市场双边交易若干问题进行了研究和探讨,主要工作如下: 1.利用多代理对双边交易中不同类型交易者的谈判行为进行了模拟,采取基于历史报价数据的Q学习算法增加了代理的自主学习能力,使代理能根据对手动作及时调整己方报价。此外,为保证市场竞争的公平性,提出了基于“谈判+拍卖”的两阶段谈判机制,给予因对谈判形势估计不足致使谈判破裂但又拥有成本优势的发电商再一次出价的机会,使得合同电价反映出不同发电成本间的真实差异,以此激励发电商以降低成本的方式来换取谈判中的主动权。 2.以纳什均衡理论为依据,通过引入发电商成本矩阵和购电商期望支付向量,建立了双边交易中买卖双方多个市场主体参与的竞价机制,并利用全局搜索能力较强的遗传算法对模型进行求解。仿真结果表明,该模型实用可行并总能找到使交易总成本最小的纳什均衡策略组合,市场成员以此策略进行报价将能获取最大收益;同时,该策略能很好地反映出发电商策略投标行为对市场结果的影响。 3.考虑在双边交易的不确定需求情况下,发电商以电力供应链整体利润最大化为前提,采用供应链协调的回购机制,通过对购电商行为的激励和约束,促使双方所选择的策略更接近于供应链效益最优的方法来确定双边交易的初始出价问题。该机制下,发电商根据供应链整体最优原则确定有利于自己的批发价,并通过回购策略促使购电商的订购量达到系统最优,并可以保证购售双方获取各自满意的利润,电力市场达到一种Pareto最优。 4.提出了基于可中断负荷拍卖模型的双边交易阻塞管理方法,通过市场机制来确定最优的负荷削减。调度过程可以分为如下三个阶段:调度前阶段、无约束调度阶段和考虑可中断负荷拍卖的阻塞调度阶段。构建了阻塞调度模型,综合考虑了最小化所中断的负荷量和最小化向用户支付的可中断负荷服务费两个优化目标。 5.基于Swarm平台构建了双边交易的市场模型、主体模型及智能代理学习模型,对Swarm平台在双边交易仿真中的应用进行了研究,实现对仿真过程的动态控制和实时观察。对市场中的主体构造了智能推理和RE两种学习模型,并对其收敛速度、收敛价格等特性进行了分析。 6.对南方双边交易电力市场/(西电东送双边交易和大用户直购电双边交易/)若干问题进行了研究。介绍了西电东送双边交易模式及电价形成机制的现状,并对其进行评价,设计了交易模式及电价形成机制,对不同模式下的交易价格进行分析,指出了存在的主要问题。深入分析了大用户开展直购电对电网公司、发电企业、大用户和地方政府等不同利益主体的影响,并提出了当前开展直购电的工作建议。 最后对论文中所作的研究进行简要总结,并指出了有待进一步深入研究的问题。 本论文的研究工作得到国家社会科学基金项目/(No. 04CJL012/)、新世纪优秀人才支持计划项目/(No. NCET-08-0207/)和教育部科学技术研究重点项目/(No. 109128/)的资助。 Interregional power exchange, interprovincial power exchange, and direct power purchase by large consumers, are all bilateral transactions mode. They will largely exist in a more open electricity market in the future. Bilateral transaction seems to be the trend in the near future. Therefore it will be of great significance to analyze the various issues of the bilateral transactions. Now world-wide studies primarily focus on aspects as allocation of transmission losses and congestion cost, as well as on design of market mechanism and regulation. While other studies are not adequate on the following as game playing strategy, negotiation strategy based on multi-agent, congestion management and stimulation. The paper aims to explore the bilateral transactions in electricity market and focuses its study primarily on the following aspects: 1. Negotiations between different traders of bilateral transaction are simulated by utilizing multi-agent technology. The Q-learning algorithm based on previous quote data is adopted to strengthen the agents’self-learning capacity. Hence the agents could react timely to their opposite parties. Meanwhile, in order to keep the fair competition, a two-stage negotiation mechanism of“negotiations + auction”is proposed, giving one more opportunity to the generator agent with some cost advantage but used to fail due to the underestimation of the negotiations. Also, the actual diversity of different generating cost is reflected by contract power price so as to inspire the generators to gain initiatives by lowering their costs. 2. Based on Nash equilibrium theory and through the introduction of generators’cost matrix and vector of expectation to pay, the bidding strategies have been established for multiple market participators in bilateral transaction in electricity market. The model is solved by genetic algorithm with strong global searching ability. Simulation results show that the model is effective and practical, and can always get the Nash equilibrium strategy combination with the smallest total cost, which can maximize the benefit of each market participator. Meanwhile, the strategy also reflects the effect on market caused by generators’strategic bidding. 3. A coordination mechanism of electricity supply chain, based on the cooperative game, is proposed to make both the buyer and the seller to undertake the risk of quantity deviation between the actual demand and contract quantity together. With this mechanism, the generators made their pricing strategy according to the principle of maximizing the profits of the supply chain, and optimizing the ordr quantity by counter purchase policy. In addition, it can ensure both sides to gain satisfactory profits respectively. Pareto optimality is finally achieved in electricity markets. 4. A congestion management framework in bilateral electricity market is proposed based on interruptible load auction model. The optimal load shedding quantity was identified by market mechanism. Under the proposed framework, the procedure of dispatch can be identified as 3 stages: pre-dispatch, unconstrained dispatch and congestion dispatch considering interruptible loads. The customers could submit their interruptible load bidding curves to ISO at the third stage. The congestion dispatch model is set up, considering syntheticlly the minimization of the interrupted loads and the payment to customers who provide interruptible load services. 5. Based on the swarm platform, some models are set up as the market model of bilateral trade, the buyer and seller model, and the learning model of intelligent agent. The paper investigates the practice of swarm platform into simulation of bilateral trade so as to realize the dynamic control and the real-time observation within the simulation. It also attempts to establish the learning models as intelligence inference and RE, and analyze the convergence rate and convergence price and so on. 6. Several issues have been attentively examined as the present situation of bilateral trading modes and price formation mechanism for power transmission from the western region of South China power grid to its eastern region in southern electricity market. Besides, the paper introduces and evaluates the bilateral trading modes and price formation mechanism under the current situation of buyer’s monopoly mode to design the bilateral negotiation mode. Trading platform on different bidding mode and market-driven mode are analyzed, and main problems are pointed out.. The paper also attempts to thoroughly analyze the impacts of different interest groups such as the grid companies, generation companies, large customers and local economy by direct power purchase with larger customers. Some suggestions are then presented. Finally, several conclusions are reached briefly and future research are expected. This paper is supported by the National Social Science Foundation of China /(No. 04CJL012/), the Program for New Century Excellent Talents in University /(No. NCET-08-0207/) and the Key Project of Chinese Ministry of Education/(No. 109128/).

关 键 词： 电力市场 双边交易 谈判策略 纳什均衡 回购 阻塞管理 仿真

分 类 号： [TM711]

领　　域： [电气工程]