基于双深度Q网络算法的多用户端对端能源共享机制研究OA北大核心CSTPCD

As a new way of energy balance and interaction in the user end energy market,peer-to-peer(P2P)power trading can effectively promote the energy sharing within the user group and improve the economic benefits of the users participating in the energy market.However,the traditional method of solving P2P power trading can not re-spond to the change of the source load among users in real time.Therefore,this paper establishes a multi-user P2P energy community trading model based on multi-type users,and introduces the deep reinforcement learning(RL)algorithm based on double deep Q network(DDQN)to solve it.The proposed method reads the environmental in-formation in the multi-user P2P energy community through the prediction network and the target network in the DDQN algorithm.The trained neural network can solve the multi-user P2P trading problem in the current communi-ty through the real-time photovoltaic,load and electricity price data.Finally,the simulation results prove that the proposed method not only promotes the sharing of P2P energy trading among users in the community,but also en-sures the economy of the multi-user P2P energy community.