电力拍卖市场的智能代理仿真模型

Electricity Auction Market Simulation with Intelligence Agent Model

用自适应自治的智能代理模型表示市场中的发电厂商构造了电力拍卖市场的仿真模型.该模型中发电厂商Agent采用Roth Erev强化学习算法,具有对市场结果自适应能力,在重复的拍卖市场可以发现唯一的Nash均衡.同时,该模型还可以揭示市场中发电厂商在不同的市场规则下的行为特性.基于IEEE RTS系统的数据研究表明,该模型提供了一种合适的电力拍卖市场分析工具.

A agent-based simulation was developed in this study for the electricity market, in which the generators agent learns the bidding with Roth-Erev reinforcement learning algorithm. In the repeated auction market the agent with Roth-Erev reinforcement learning algorithm can find the exclusive equilibrium. In this paper, the electricity auction emulator's structure and the learning algorithm were described and the test result with IEEE-RTS was reported, It showed that the emulator is a potential tool for studying the electricity market.