Optimal bi-level scheduling of a PV - energy storage - charging station with consideration for EV aggregators

To address the common issues of low overall operational economic efficiency, high volatility of photovoltaic (PV) output power, and insufficient local consumption capacity in PV-energy storage-charging stations during actual operation, a bi-level optimal scheduling method for station operation involving an electric vehicle (EV) aggregator is proposed. First, the EVs are clustered using the Particle Swarm Optimization-Fuzzy C-Means (PSO-FCM) algorithm, and the cluster aggregation modeling as well as constraint transformation are completed via the zonotope geometric method. Second, a bi-level optimization model is established: the upper level sets the pricing strategy with the objective of minimizing the total operating cost of the station, while the lower level adjusts the EV charging and discharging strategies to minimize the user cost. Verified through MATLAB/CPLEX simulations, the model converges after four iterations. The proposed bi-level scheduling incorporating vehicle-to-grid (V2G) reduces the station’s purchased power during evening peak hours by 25%, achieves a PV consumption rate of 90.8%, significantly lowers the station’s operating cost, and reduces its dependence on the power grid.