Trading-coupled Sequential Operation Simulation of Distribution Network with Distributed Prosumers

Considering the increasingly influence of peer-to-peer(P2P) trading on distribution network operation, a trading coupled multi-agent operation simulation method is proposed in this paper. Firstly, a bidding strategy based on disappointment avoidance theory is adopted to simulate the self-interested market behavior of prosumers. And then, a sensitivity analysis model based on deep neural network is constructed to embed network constraints which assists the dynamic pricing of network charges for various energy trading. Finally, a P2P transaction mechanism based on two-stage continuous double auction is designed to realize multi-round matching clearing. Numerical cases show that the proposed method reduces the power purchase cost by 5.12% on the premise of meeting the line congestion and voltage constraints, which is helpful to realize the economical and efficient operation of distribution network.