摘要
基于信息披露的主体框架,分析披露体系中具体的供需信息对于市场成员的引导作用。建立日前市场的双层竞价优化模型,采用强化学习的方法,通过市场中有无关键供求信息披露对于发电商报价预期的影响,推演模拟不同信息披露程度下发电商报价行为,然后从策略演化过程、策略偏离度以及策略收敛速度3个角度进行信息披露程度对市场成员交易行为影响的定量分析,验证市场供需信息的发布程度对于市场成员的引导作用,最后基于推演结果进行总结,并对我国信息披露的建设提出建议与思考。
The paper focuses on the guidance of specific supply and demand information to market participants based on the main frame of disclosure system.Firstly,the bi-level bidding optimization model for day-ahead market is established,which uses the method of reinforcement learning algorithm to simulate the effects on bidding strategies under different circumstances where the supply and demand information is disclosed or not.Then the quantitative analysis on the influence of information disclosure on trading behaviors from the perspectives of strategy evolution,strategy deviation and strategy convergence speed is given,to verify the guidance of the supply and demand information disclosure to the market participants.Finally,the thinking and suggestions on the construction of the information disclosure system in China are given based on the simulation results.
作者
高怡静
肖艳炜
杨朋朋
朱炳铨
林凯颖
王蓓蓓
GAO Yijing;XIAO Yanwei;YANG Pengpeng;ZHU Bingquan;LIN Kaiying;WANG Beibei(Southeast University,Nanjing 210096,China;State Grid Zhejiang Electric Power Company,Hangzhou 310000,China)
出处
《智慧电力》
北大核心
2020年第2期109-118,共10页
Smart Power
基金
国家自然科学基金（71873029）
国家电网公司科技项目（52110118000A）~~
关键词
信息披露
强化学习
资源配置
模拟推演
information disclosure
reinforcement learning
resource allocation
simulation
