Research Into Power Market Bidding Strategies Based On Agent

Following the release of n Several Opinions on Further Deepening the Reform of the Power System"(also known as "file 9")by the State Council,the domestic power market set off a new wave of reform,and the domestic power market structural reform has advanced step by step.Due to the particularity of electricity,the trial and error cost of the power market reform is very high,therefore it has been particularly important to simulate the existing market model and to explore the future of the market model.It is very important to build a power market simulation system which covers many market modes to simulate and deduce the reform.The power market simulation system is mainly simulating the main body behavior in the market.Under the market environment,main activity for main body in the market to participate in the market competition is achieving independent bidding,therefore the simulation of bidding process and choice of bidding strategies is very important.The work of this paper as part of the development of power market simulation system,This paper is focusing on the study of the bidding strategies of market main body in the power market simulation system.Main work contents and conclusions include:(1)To study the generation company agent bidding model based on cost,the bidding behavior of the generation company agent is fed back based on the cost price and without considering the opponent's bid based on the combination of RE learning algorithm,finally after several rounds of study to select the optimal strategy.Simulation results show that the algorithm can be used to select the optimal strategy from the strategy set;(2)On the basis of cost price,this paper studies the optimal bidding model of generation company agent based on the opponent's strategy.Based on consideration of the opponent's bid,use the concept of fuzzy set membership function to solve the problem of the profit of each strategy under the opponent's bidding strategy function based on the opponent's bid,and select their own offer strategy based on the membership function.Simulation results show that the method can be used to select the optimal bidding strategies under the condition of considering opponent's bid;(3)To deal with the bidding problem of the agent in the direct purchase of the big customer and study the bidding strategy in the bilateral transaction between the generation company and the big customer,first determine the transaction form and the bidding procedure in the direct purchase of the big costumer,and then the Rubinstein bargaining model in economics applied to both sides of the negotiations.Next,we set up the Bayesian learning model on the basis of Rubinstein model with consideration of the actual situation.Simulation results show that using the Bayesian learning algorithm,the generator agent and the big customer agent can greatly shorten the negotiation time and promote the negotiations.