发电侧企业群体间竞价行为的随机演化博弈  

Stochastic Evolutionary Game of Bidding Behavior for Generation Side Enterprise Groups

作　　者：杨辉[1] 莫峻[1] YANG Hui;MO Jun(Guangxi Key Laboratory of Power System Optimization and Energy Technology(Guangxi University),Nanning 530004,Guangxi Zhuang Autonomous Region,China)

机构地区：[1]广西电力系统最优化与节能技术重点实验室(广西大学),广西壮族自治区南宁市530004

出　　处：《电网技术》2021年第9期3389-3397,共9页Power System Technology

基　　金：国家自然科学基金项目(51967003)。

摘　　要：以演化博弈论为理论基础,针对确定性演化博弈模型的不足,引入高斯白噪声随机干扰项,构建了信息不对称下两个异质性发电企业群体的随机演化博弈模型。运用It?随机微分方程理论对模型进行了求解,讨论了两群体随机博弈系统趋于渐进稳定状态的条件,并结合大小型两类发电企业群体参与竞价策略博弈的具体算例对模型进行了动态演化仿真模拟。仿真结果表明,所建立的随机演化模型能更真实准确地反映主体行为策略在不确定因素干扰下的演化过程,比确定性演化博弈模型具有更强的适用性。基于仿真结果,利用熵权法对4个调控参数进行综合评价,可为政府对发电企业报价行为进行有效监管和调控提供参考。Based on the evolutionary game theory,aiming at the deficiency of deterministic evolutionary game model,this paper introduces the random interference term of Gaussian white noise and constructs a stochastic evolutionary game model for the two heterogeneous power generation enterprise groups under asymmetric information.The It?stochastic differential equation theory is used to solve the model,and the conditions for the stochastic game system with the two groups to become asymptotically stable are discussed,and the dynamic evolution simulation of the model is carried out with the specific examples of the power generation enterprises(both large and small)participating in the bidding strategy games.The simulation results show that the stochastic evolution model established in this paper can more truly and accurately reflect the evolution process of the agent’s behavior strategy under the interference of uncertain factors and that it has stronger applicability than the deterministic evolutionary game model.Based on the simulation results,the entropy weight method is used to comprehensively evaluate the four regulation parameters,which provides a reference for the government to effectively supervise and control the bidding behavior of the power generation enterprises.

关 键 词：发电侧 竞价策略 高斯白噪声 随机演化博弈 

分 类 号：TM73[电气工程—电力系统及自动化]