V2V Energy Trading in Multiplayer Bayesian Games Considering Mixture of Traffic Flow and Power Flow

This article proposes a multiagent Bayesian energy trading scheme that integrates the urban traffic network to address the issues of range anxiety and instability in the power grid by uncoordinated charging in electric vehicles (EVs). Uncoordinated charging in EVs. Specifically, we establish the relationship between traffic flow and the plugging‐in time (PiT) and state of charge (SOC) of EVs to describe the flexible travel demands of EVs. Simultaneously, the constraints of protection system devices during the power trading process are considered, internalizing relevant losses as transaction costs. We formulate utility functions for the participating agents and transform the multiagent energy trading competition problem into a Bayesian game problem for solution. To achieve coordinated optimization of the integrated system, we first predict the number of transactions involved in the trading process and input it into the environment to ensure successful matching of transactions. Furthermore, we utilize a multiagent reinforcement learning method based on the fusion of long short‐term memory (LSTM) communication mechanism to solve the established model. A comparative analysis of existing research is conducted to demonstrate the practicality and effectiveness of the proposed model.