Research on Bayesian Game Strategy of Multi-agent Demand Response in Industrial Parks Based on Incomplete Information

With the promotion and development of China’s electricity market reform process, how to capture the maximum profits of all responding entities at the level of industrial parks to guide users to actively participate in demand response has become a new problem to be solved. Based on the situation that the market information disclosure is limited and users and aggregators freely conclude contracts in the response process, this paper proposes an optimal game strategy from the perspective of aggregators and a revenue distribution method for users. Considering the interests of users in the industrial park, a multi-agent and two-tier game structure is established with incomplete information Bayesian game as the upper layer and cooperative game as the lower layer; Secondly, in order to maximize their own interests, an algorithm combined with improved counterfactual regret minimization idea and heuristic algorithm is used to solve the game equilibrium to achieve a stable cooperative relationship in the park. Finally, a numerical example is used to verify that multiple comparison models are set up for the declaration strategy and income distribution in different scenarios. The results show that the strategy obtained under the proposed method can improve user income, and the reasonable selection of the benefit distribution mechanism can play a catalytic role in improving the enthusiasm of users to respond to their needs.