Multi-Entity Interactive Operation Strategy of Vehicle-Station-Net Driven by Intelligent Network Data Fusion

Abstract

The electric vehicle charging behavior involves multiple subjects such as traffic and charging stations, and contains a large number of uncertainty factors, such as uncertainty of traffic road conditions and uncertainty of queuing time at charging stations, resulting in a strong uncertainty of electric vehicle charging behavior. In order to better deal with these stochastic variables, firstly, a reasonable charging tariff is formulated by the method of cooperative game of multiple subjects to satisfy the charging station operator side, and the grid side optimal revenue decision. Secondly, for the EV charging guidance method considering charging tariff and the optimal scheduling strategy of charging stations on EV charging behavior, this paper adopts Reinforcement Learning (RL) method. Finally, experimental simulations are conducted to validate the proposed algorithm with a certain number of typical charging stations in each administrative region of Tianjin, and the realized results verify that the proposed algorithm can significantly reduce the peak load value of EV centralized charging, reduce the impact of large-scale EV charging on the grid, and effectively improve the utilization rate of the grid and charging facilities.