Coordination And Optimal Dispatch Of Multi-Stakeholders In Active Distribution Network Based On Game Theory

With the development of distributed generation,the research of active distribution network has been carried out in many countries of the world.However,with the diversification trend of the investors,the new energy connected to the distribution network is built and managed by the distributed generation operators,which has changed the pattern of the distributed generation in the past.Users who participate in demand response become the new interest entities in the distribution network.When they are independently optimized according to their respective interests,they are mutually influenced and restricted,and the benefit of each stakeholder is influenced by other stakeholders’ decision variables.Therefore,the efficient and fair operation of distribution network with multiple stakeholders in the new environment is an urgent problem to be solved.In order to coordinate the conflicts between different stakeholders in the distribution network,meet the requirement of various stakeholders with the initiative to generate and use electricity,this paper proposes an optimal scheduling strategy to coordinate the active distribution network multi stakeholder game equilibrium based on Q learning algorithm to manage and control distribution network,distributed generation operators and users who participate in demand response and achieve a fair and efficient operation of distribution network.Among them,the DG operators take the biggest economic benefit as the goal of to arrange the controllable distributed power output.The distribution company adjusts the TOU price and the tie line power with the minimum operation cost,and the users take the minimum electricity cost as the goal to adjust the power consumption.Correlated equilibrium Q-learning algorithm is adopted to solve the game model and make the game result more fair.The new stakeholders in the power grid will be willing to participate in the operation of the power grid due to the improvement of the fairness.In this paper,the correlated equilibrium Q-learning algorithm is used to simulate on the IEEE33 nodes based on the multi stakeholder game model.The coordinated optimal scheduling model based on game theory and the traditional single subject optimal scheduling model are compared and analyzed.The simulation results show that the proposed game coordination optimization mode compromise solution is obtained to coordinate the interests of all parties and is effective to promote new energy utilization and reduce network loss.At the same time,comparing the correlated equilibrium obtained by Q-learning and Nash equilibrium obtained by the best response algorithm and the results show that the revenue of the weak player in the original game will increase and the fairness of the game and the enthusiasm of players participating in the game is improved.