A Flexible Energy Scheduling Strategy for Fast Charging Stations Considering V2G of EVs and PV Uncertainty

Abstract

To meet the growing electric vehicle (EV) charging demand, fast charging stations based on novel DC architecture are being widely deployed. However, the increasing number and expanding scale of EV fast charging stations introduce significant uncertainties and fluctuations, which can cause substantial disturbances to the power grid. It is necessary to design energy scheduling strategies that optimize system operation, achieving both load fluctuation mitigation and optimal economic efficiency. This paper proposes an innovative flexible energy scheduling strategy tailored for EV fast charging stations, utilizing a power forecasting model to reduce the uncertainty of scheduling and incorporating vehicle-to-grid (V2G) capabilities of EVs to reduce overall costs. An Informer forecasting model is developed to realize multi-timescale unified forecasting encompassing photovoltaic (PV) generation and EV loads with lower computing costs. On the other hand, the proposed method incorporates a certain proportion of EVs and flexible loads as virtual energy storage in system scheduling, thereby further optimizing operating costs. Finally, verification results demonstrate that the proposed method achieves high forecasting accuracy and significant cost reduction, thereby realizing optimized scheduling for EV fast-charging stations.