The two-stage negotiation mechanism based on multi-agent using Q-learning in electricity bilateral contract

The negotiation behaviours of different traders in the process of direct power purchase by large consumers are simulated by using Multi-agent technology. With the Q-learning algorithm based on previous quotation data, he agent can strengthen its own learning capacity and timely adjust its bid price against its opponent's action. Meanwhile, in order to make sure of the justice of market competition, a two-stage negotiation mechanism of ‘negotiations + auction’ is proposed, which gives one more opportunity to the generator agent who has a lower reserve price but fails to reach an agreement, due to the underestimation of the situation of the negotiations. It also makes power contract price reflect the real diversity of different generating costs, and can inspire the generators to get the negotiating initiative by lowering their costs.