【作者】 方勇；

【导师】 李渝曾； 张少华；

【作者基本信息】 上海大学 ， 控制理论与控制工程， 2004， 博士

【摘要】 电力系统的特点和市场环境下的不完全信息问题引发了发电商的各种策略性行为,从而带来了市场价格的急剧波动,影响了电力市场的安全性和经济性,资源优化配置无法实现。而稳定市场价格、合理配置资源,系统应从供需两侧加强。本论文应用机制设计理论、供应函数均衡、Cournot模型、Stackelberg模型探讨了发电侧中发电商的策略性行为、激励性发电上网竞价机制,以及需求侧中可中断负荷管理的意义和实施方法、激励性可中断负荷合同的设计和应用等问题。 首先,应用供应函数均衡模型分析了统一价格机制下信息对发电商策略性行为的影响和默契勾结持留容量的形成过程。应用供应函数均衡模型时,同时考虑发电商策略性报价和容量持留这两种实施市场力的行为,分析了发电商从单独容量持留阶段向默契勾结持留容量阶段的演变过程。分析发现,前一阶段容量持留和负荷水平、市场结构有关,发电商公布持留容量信息会降低容量持留发生的负荷域值;后一阶段的默契勾结会导致持续的高电价,负荷水平较高时将出现市场价格的飞升。进一步提出了抑制市场力作用的方法,并用算例说明了分析结果。因此,对发电商策略性行为的分析表明信息会影响发电商行为和市场价格,因而发电上网竞价机制应充分考虑市场环境下的信息不对称问题。 其次,提出了一种能合理回收容量成本的激励性发电上网竞价机制。针对电力市场中管理者和发电商之间的信息不对称而引发发电商的策略性报价问题,本论文基于机制设计理论,提出了一种具有激励相容特性的电力竞价机制。该机制能引导发电商披露真实信息,从而实现经济调度,还能根据报价信息评定机组的容量成本,引导发电容量投资。特别是该机制给予边际机组相对较多的容量成本补偿,有助于风险回避的峰机组投资商的进入。此外,新机制的电价反映出负荷变动的趋势,能激励发电商积极响应负荷的变化,增强了系统安全性。算例采用蒙特卡罗方法,其结果证实了新机制的良好特性。这是本论文工作的主要创新之一。 第三,从可置信威胁角度研究了电力公司在电力市场环境下应用可中断负荷管理的方法。电力公司实施可中断负荷管理时,电力公司与发电商的不同行动次序会产生不同的效果。基于Stackelberg、Coumot模型,本文分别分析了电力公司先行动、后行动、和与发电商同时行动的市场均衡。分析结果表明,相对于其它行动方式,电力公司先行动能中断更多负荷,也即先动优势,而更能抑制发电商的市场力。但是,如果电力公司的行动是不可置信的,均衡结果和与发电商同时行动时相同。进而分析了电力公司如何使其行动可置信,然后为电力公司在电力市场环境下实施可中断负荷管理的方法提供了一种解决方案。这是本论文工作的主要创新之一。更多还原

【Abstract】 In electricity markets, the distinguishing features of electric power systems and the asymmetry of information between market manager and generation companies will cause some tactful bidding, which leads to acute price and will be harmful to the security and economy of electricity markets, as well as the optimal allocation for electric power resources. In order to stabilize the market price and allocate electric power resources effectively, the system should be enforced from both supply and demand sides. Based on mechanism design theory, Supply Function Equilibrium (SFE) model, Cournot model and Stackelberg model, this dissertation discusses the generators’ strategic behaviors on the supply side, the bidding mechanism with incentive compatibility, the merit and implementation of interruptible load management on the demand side, and the design of incentive interruptible load contracts and its applications.Firstly, the SFE model is employed to analyze the influence of information upon the generators’ strategic behaviors and the formation of tacit collusion in withholding capacity under the uniform price mechanism. Considering the two generators’ strategic behaviors, i.e., strategic bidding and withholding capacity, the transformation from the withholding capacity of individual generator to the tacit collusion of all generators is developed by SFE model. The analysis shows that the amount of generators’ withholding capacity is concerned with system load level and market structure in the former period of transformation, meanwhile the load threshold value in which withholding capacity emerges will decline if generators published their information on withholding capacity; the generators’ tacit collusion will lead to durative acute prices in the latter period of transformation, and price spikes will emerge in peak load. Then the suggestions to retrain market power are put forward, and the analysis results are illustrated by numeral examples. Therefore, the analysis of generators’ strategic behaviors shows that information could influence generators’ behaviors and market price, so the generation bidding mechanism should sufficiently take the asymmetry of information in electricity markets into account.Secondly, an incentive generation bidding mechanism, which can reasonably recover capacity cost, is developed. Aimed at the generators’ strategic bidding behaviors resulting from the asymmetry of information between market manager and generation companies in electricity markets, a new generation bidding mechanism with incentive compatibility is designed. Based on mechanism design theory, the new mechanism is able to lead generators to publish their true information, so an economic dispatching can be carried out, moreover, it can evaluate the capacity cost of the generating unit using bidding information and attract investments for generation capacity. Specially, the new mechanism would compensate the marginal units more for capacity cost than other units, thus it would conduce to entities of these peak unit investor with risk disgust. In addition, the new mechanism can reflect the tendency of load variation and encourage generators to respond it, and enhance the system security. The illustration using Monte Carlo demonstrates the merits of the proposedmechanism. So this is one of the original contributions in this dissertation.Thirdly, this dissertation researches the implementation of interruptible load management by utility in electricity markets from the aspect of credible menace. The different order of utility and generators moving will lead to different outcomes when utility implements interruptible load management. Based on Stackelberg and Cournot models, this dissertation respectively analyzes the market equilibriums of utility first-move, second-move and moving together with generators. The results show that the load curtailed will be more if utility moves firstly, i.e. first-mover advantage, which will be more effective to restrain market power of generators than other moves of utility. The equilibrium outcomes, however, will be identical with the ones of utility moving together with generators if utility’s moves are not credible. Then this dissertation analyzes that utility how to make its moves credible, and then put forward a scheme to implement interruptible load management by utility in electricity markets. This part is one of the original contributions of this dissertation.Fourthly, this dissertation researches the design of incentive interruptible load contracts for distribution companies with risk preference, and the proposed model will be a new way to compete in electricity markets for distribution companies. The inelasticity of demand-side and characters of electric power systems induce the electricity price dynamics, which makes distribution companies exposed to immeasurable risk, meanwhile, distribution companies get a chance to obtain more benefit as well. Taking risk preference of distribution companies into account, this dissertation designs an incentive interruptible load contract model for discrete customer types using mechanism design. The analysis and illustration both show that the proposed model can lead customers to voluntarily reveal their true information and the electric power resources to be allocated efficiently, and in the meantime distribution companies benefit from the cost saving of electric power supply whatever the risk preference of distribution companies is. Therefore, incentive interruptible load contracts not only improve customers’ demand elasticity, but also help distribution companies to avoid market risk. Especially, the proposed model will be a powerful tool to compete and manage risk efficiently for risk-taken distribution companies in electricity markets. This part is one of main contributions of this dissertation.Finally, an incentive interruptible contract model, which can take customer maximum interruptible load constraints into account, is developed from the aspect of practicability, and then its applications are researched in electricity markets. The presence of informational asymmetry between utility and customers may cause immeasurable inefficiencies in interruptible load management program, this dissertation develops an incentive compatible contract model for interruptible load management, which can lead customers to voluntarily reveal their true outage costs information. The proposed model allows customers having maximum interruptible load constraints to be taken into account and can apply to different optimization objectives for dispatching the interruptible customers, such as maximizing the utility’s benefit or minimizing the total outage cost of customers. Then, maximizing the profit of utility or system operator, this dissertation studies three applications of the proposed model in electricity markets: electric power company bidding in a spot更多还原