The increasing disorderly charging of electric vehicles (EVs) is challenging the stability of community power systems. However, microgrids offer a pathway for integrating renewable energy and managing EV loads. To optimize capacity allocation in such scenarios, this paper proposes a two-level optimization strategy based on a real-time pricing demand response mechanism. The upper level plans energy storage, photovoltaic, and wind power capacities to minimize annual comprehensive costs, while the lower level optimizes daily scheduling plans. To address renewable energy, conditional generative adversarial networks and an improved k-medoids algorithm are employed for scenario generation and simplification. Simulation results demonstrate that this strategy reduces microgrid operating costs by 5.3% and diminishes peak-to-off-peak load differences by 158.85 kW. Analysis of the impact of the electric vehicle participation scale in vehicle-to-grid interactions on system dispatch effectiveness identified an optimal participation level of 120 electric vehicles. This approach was validated for enhancing economic benefits and promoting sustainable energy utilization.
Huang et al. (Sun,) studied this question.