应用动态学习改进对竞争对手微增响应猜测的发电公司投标策略  

AN INPROVEMENT OF GENERATION FIRMS' BIDDING STRATEGIES BASED ON CONJECTURAL VARIATION REGULATION VIA DYNAMIC LEARNING

作　　者：宋依群[1] 焦连伟[2] 倪以信[3] 文福拴[3] 侯志俭[1] 吴复立[3] 

机构地区：[1]上海交通大学电气工程系,上海200240 [2]清华大学电机与应用电子技术系,北京100084 [3]香港大学电机及电子工程系

出　　处：《中国电机工程学报》2003年第12期23-27,共5页Proceedings of the CSEE

基　　金：国家重点基础研究专项经费项目(G1998020305);香港特别行政区政府资助基金(RGC)~~

摘　　要：该文应用微增响应猜测(Conjectural Variation,CV)方法构造一种在不完全信息情况下的寡头竞争电力现货市场中发电公司通过动态学习改进对竞争对手的微增响应猜测的新投标策略,指出当各发电公司采用该文提出的投标策略进行不断重复的投标竞争后,市场所达到的均衡点为Nash均衡点。算例计算结果说明各发电公司具有进行动态学习的动力,但是当所有公司进行学习时将会降低市场结清电价从而提高电力市场的社会总效益。算例计算结果还说明了电力需求弹性对学习过程的收敛性具有积极作用。A new bidding approach based on conjectural variation (CV) improvement via dynamic learning is developed for generation firms to construct strategic bidding in oligopolistic spot electricity market with incomplete information. It is shown that the equilibrium reached in the market is a Nash equilibrium when the proposed bidding approach is utilized, individual firms have motivations to conduct learning; learning of all firms will decrease the clearing price of spot markets and improve the social welfare in oligopolistic electricity markets, as well as the elasticity of demand has positive impacts on the convergence of learning process.

关 键 词：发电公司 投标策略 动态学习 电力市场 电力需求弹性 电价 微增响应猜测 

分 类 号：F407.61[经济管理—产业经济]