Uniform Price-based Peer-to-peer Energy Trading with Data-driven Market Clearing Price Estimation

Abstract

The actual market clearing price (MCP) of the uniform price-based peer-to-peer (UPP2P) energy trading may differ from its optimal value owing to imperfect competition. Meanwhile, in the UPP2P, MCP is essential to determining optimal social welfare and thus allocating trading capacity among participants. Therefore, this study proposes UPP2P energy trading with novel data-driven MCP estimation using the finite horizon Markov decision process (MDP). Using historical data, the non-technical aspects of the UPP2P operation, such as market power, incomplete information, and bidding strategy, can be integrated into the process of predicting optimal MCP. Furthermore, a modified Kirschen network cost allocation (KNCA) method, which can estimate the effect of UPP2P participants on the network reliability and is incorporated to elucidate the optimization of MCP. The proposed UPP2P with data-driven MCP estimation is applied on a noncooperative matrix game-based UPP2P algorithm, and simulation results are given to show the effectiveness of the method.