Bilayer game strategy of regional integrated energy system under multi-agent incomplete information

The Journal of Engineering

  ^{1    1    1} 

[1] School of Electronic, Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai, People's Republic of China;

关键词: pricing;    power markets;    game theory;    multi-agent systems;    tendering;    scheduling;    learning (artificial intelligence);    evolutionary computation;    noncooperative bidding process;    multienergy market;    bilayer competitive game model;    bilayer game strategy;    regional integrated energy system;    multiagent incomplete information;    high coupling degree;    bilayer interaction strategy;    energy suppliers;    game interaction strategy;    scheduling;    independent system operator;    multienergy load prediction;    bidding functions;    pay-as-bid settlement protocols;    total energy cost minimization;    ISO coordinates;    quotation price;    bounded rationality;    unit characteristics;    history scheduling data;    Q-learning algorithm;    evolutionary process;    local Nash equilibrium;    distribution networks;   

DOI  :  10.1049/joe.2018.8571

来源: publisher

【 摘 要 】

In view of the high coupling degree of regional integrated energy system, a bilayer interaction strategy, consisting of energy suppliers, distribution networks, and users, is proposed. Game interaction strategy includes two aspects: scheduling and bidding. The independent system operator (ISO) coordinates all adjustable resources. Depending on the quotation price and multi-energy load prediction, ISO minimises the total energy cost, which realises the complementary of the multi-energy in the cooperative game. Under the assumption of incomplete information and bounded rationality, this study designs bidding functions and pay-as-bid settlement protocols. On this basis, according to history scheduling data and units’ characteristics, agents for energy suppliers pursue maximum interests. Also, the non-cooperative bidding process in multi-energy market is simulated by using Q-learning algorithm. Finally, the evolutionary process of the bilayer competitive game model is studied by practical example, and the existence local Nash equilibrium of the strategy is also proven.

【 授权许可】

CC BY