Online MPC-Based Low-Carbon Economic Scheduling of Electric Vehicles under Uncertainty

The increasing integration of electric vehicles (EVs) presents both opportunities for carbon reduction and challenges for power grid stability. Effective management of EV charging stations (EVCSs) is essential to harness this potential, yet uncertainties in EV user behavior complicate scheduling and resource allocation. This paper proposes an online model predictive control-based framework for low-carbon economic scheduling of EVCSs under uncertainty. Unlike conventional offline aggregation models, which require precise forecasts of EV arrivals and departures, the proposed approach dynamically optimizes charging decisions and flexibility reserves using realtime information on electricity market, carbon intensity, and charging station conditions. The uncertainty of EVs' departure time is mitigated via by distributionally robust chance-constrained optimization. Comparative case studies demonstrate that the proposed method achieves the lowest total cost and highest charging completion rates. The results highlight the framework's effectiveness in enhancing system flexibility, improving overall scheduling efficiency and dealing with EVs uncertainties, making it a practical solution for EVCS operations.