Research on data-driven charging strategy for electric vehicles in provincial areas

This paper proposes a data-driven rolling optimization strategy based on multiple time windows for the problem of uneven spatial and temporal distribution of electric vehicle charging demand in the provincial area. Firstly, a system architecture integrating transportation road network, charging facilities, and Telematics terminals is constructed, and a charging demand prediction model considering real-time traffic status and battery characteristics is established. Secondly, we analyze the coupling relationship between charging station queuing dynamics and path selection, form an optimization model to minimize the total time consumption, and consider the utilization rate of charging stations. Finally, simulation verification is carried out based on the 111-node highway network in Qinghai Province, which is constructed from actual data. The results show that the proposed strategy reduces the average charging time during peak hours by $29.7 \%$, the standard deviation of the charging station utilization rate by $37 \%$, and the system load balance is optimized to 0.12, which verifies the validity of the modeland its engineering applicability in the dynamic traffic environment.