摘要
利用多代理对大用户直购电中不同类型交易者的谈判行为进行了模拟,采取基于历史报价数据的Q学习算法增加了代理的自主学习能力,使代理能根据对手动作及时调整己方报价。此外,为保证市场竞争的公平性,提出了基于"谈判+拍卖"的两阶段谈判机制,给予因对谈判形势估计不足致使谈判破裂但又拥有成本优势的发电商再一次出价的机会,使得合同电价反映出不同发电成本间的真实差异,以此激励发电商以降低成本的方式来换取谈判中的主动权。
The negotiation actions of different traders in the negotiation process of direct power purchase with large consumers are simulated using the multi-agent technology. With the Q-learning algorithm based on historical data, an agent can strengthen its own learning capacity and timely adjust its bid price against its opponent＇ s action. Meanwhile, in order to ensure the fairness of market competition, a two-stage negotiation mechanism of ＇negotiations＋auction＇ is proposed. It gives one more opportunity to the generator agent who has a lower reserve price but fails to achieve an agreement, due to underestimation of the situation in the negotiations. It also makes the real diversity of different generating costs reflected by contract power price, and can inspire the generators to get the negotiating initiative by lowering their costs.
出处
《电力系统自动化》
EI
CSCD
北大核心
2010年第6期37-41,共5页
Automation of Electric Power Systems
基金
教育部新世纪优秀人才支持计划资助项目(NCET-08-0207)
教育部科学技术研究重点资助项目(109128)
国家社会科学基金资助项目(04CJL012)~~
关键词
一对多谈判
Q学习算法
电力市场
大用户直购电
双边合同
one-to-many negotiation
Q-learning
electricity market
direct power purchase with large consumers
bilateralcontract
作者简介
张森林（1978-），男，博士研究生，高级工程师，主要研究方向：电力市场竞价交易及双边交易理论、电力市场分析。E-mail．zhangseelin_csg@126．com
屈少青（1985-），男，硕士研究生，主要研究方向：电力市场及人工智能在电力系统中的应用。
陈皓勇（1975-），男，通信作者，博士，教授，主要研究方向：电力市场、电力系统优化规划与运行。E—mail：eehychen@scut．edu．cn
