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

Bayesian learning model based dynamic game for pricing between power consumers and generators

电力用户从发电商直接购电是打破电力垄断经营过程的前奏。采用序贯博弈理论设计二者的交易机制,引入贝叶斯学习模型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。

The direct electricity purchasing by the large power consumer from the generator is the prelude of electricity monopoly operation breaking. In this paper, the sequential game theory was used to design the trading mechanism between the two sides, and the Bayesian learning model was introduced to enhance the two sides＇ learning ability. During the trading process, the generator and the large consumer take turns to quote their price, which enables both of them to gradually revise their awareness of the reserved price of the other side and to forecast the price more precisely. Calculation example proved the Bayesian learning model based the dynamic game method effective.