电力市场环境下备用容量的集中和分散优化决策

针对电力市场环境下，应该采用何种方式确定和获取所需要的备用容量这个有普遍争议的问题，提出了集中优化方法和应用保险理论的优化方法。集中优化方法是以从发电公司购买成本和用户停电损失之和最小化为目标，通过集中方式确定所需的总的运行备用容量及其价格并在相关的发电公司间分配这些容量的方法。应用保险理论的分散优化方法是用户通过购买保险，把停电可能引起的损失转移给独立调度机构(ISo)，从而引导ISO对备用容量进行有效的管理；同时，当发电公司没有投入所成交的备用容量时，IsO会对其进行高昂的惩罚，从而激励发电公司很好地履行合同职责。文中先导出了与备用容量相关的各个发电公司的成本函数和用户的效用函数，然后给出了相应的分散优化和集中优化的数学模型，以及求解算法，证明了在完全竞争的市场中分散优化方法可以得到和集中优化方法相同的最优解，即最大化社会福利。最后以有14个发电公司参与的模拟电力市场为例进行了仿真计算，说明所提出方法的基本特征。

CENTRALIZED AND DECENTRALIZED OPTIMAL DECISION-MAKING ON RESERVE CAPACITY PROCUREMENT IN ELECTRICITY MARKETS

In the electricity market environment, how to determine and procure reserve capacity is still a problem with extensive debates and has an important impact on the secure and economic operation of power systems. A systematic research is made in this paper, and both a centralized optimization approach and a decentralized one are presented. The objective of the centralized optimization approach is to minimize the sum of the procurement cost and the outage loss, and through an optimization procedure of the overall reserve capacity required, the price associated together with the allocation of the reserve capacity among related generation companies could be obtained. The developed decentralized optimization approach is based on the insurance theory. With this approach, consumers are able to transfer their financial risk of loss-of-supply to the ISO(independent system operator) , and hence incentives could then be induced for the ISO to manage the reserve capacity in an efficient manner. At the same time, a penalty system is also introduced. Under the penalty system each generation company is required to pay a high penalty if it does not commit the required reserve capacity provision, which enforces the liability of generation companies to provide operating reserve. Detailed mathematical models and solution procedures are presented for both the centralized and decentralized optimization methods. It is shown that with properly defined reserve market rules the decentralized approach could yield the same optimal solution as its centralized counterpart aiming at maximizing social welfare in the reserve market. Finally, numerical test results on a simulated electricity market with 14 generation companies participated are presented to illustrate the essential features of the proposed model and approach.