电力用户从发电商购电定价的双边贝叶斯动态博弈学习模型.pdfVIP

牟或L 牵b Vo1.37 No.3 第 37 第3 期 East 胁ina EI即tric Power 2(阴Jf 3 月 Mar. 2侃 电力用户从发电商购电定价的双边贝叶斯 动态博弈学习模型 谭忠窟，柏辙，李莉，严菲 (华北电力大学电力经济研究所，北京 102206) 摘 要:电力用户从发电商直接购咆是打破电力蠢断经营过程的前奏。采用序贯博弈珊论设计二二者的交易机 制，引人贝叶斯学习模却以提高发电商和用户的学习能力。通过发电商与电力用户的多轮报价，彼此逐轮更 新对对方保留价格的认识，以更精确的预期修正报价，改进自身博弈效用，提高双方博弈过程的效率。算例表 明发电商和大用户采用贝叶斯学习模型进行博弈具有有效性。 关键询:电力市场;序贯博弈;贝叶斯动态博弈学习模型 基金项目:国家自然科学基金项目，;教育部新世纪优秀人才支持计划(NC盯峭-02-(8) ; 北京哲学社会科学规划项目(06日aJGI02)联合资助 作者简介:谭忠寓(1964-) ，男，教授，博士生导师，主要从事电力能源经济、电力企业风险管现等研究。 中团分类号:TM73 川崎7.61 文献标志码:A 文章编号:1001-9529(2ω9 )03 -0384-05 Bayesian learning model based dynamic game for pricing between powerωnsumers and generaωm TAN Zhong价l ， BAI Hω， ， U li , YAN Fei (Institute of Electricity Economics , Nonh China Electric Power Univ. , Beiji吨 102206 ， China) Abstract :11始 direct electricity purchasing by the large powerωns山ner from the generator is the prelude of electricity monopoly operation breaking. In this paper , the sequential 串串me theory was used to design the trading mechanism 00- tween the two sides , and the Bayesian leaming model was introduced to enhance the two sid盼， learning ability. 队Jr­ ing the trading proce串串， the generator and the large consumer take tums to quote their price , which enables both of them to gradually revise their aw缸eness of the reserved price of the other side and to forecast the price more preci盼蝴 ly. Calculation example proved the Bayesian learning model based the dynamic game method effective. Key words: quote price; sequential 伊