Research on optimal scheduling of a photovoltaic-storage-charging integrated power station based on intraday two-stage model predictive control

With the rapid development of electric vehicles, photovoltaic-storage-charging stations that supply power to electric vehicles are becoming increasingly important. To optimize the energy scheduling of integrated photovoltaic-storage-charging stations, improve energy utilization, reduce energy losses, and minimize costs, an optimization scheduling model based on a two-stage model predictive control (MPC) is proposed. The first-stage MPC aims to minimize the deviation between day-ahead scheduling values and intraday actual values, implementing intraday tracking of day-ahead values through rolling optimization. In order to further improve prediction accuracy, robust optimization-based model predictive control is used to correct deviations in photovoltaic output and electric vehicle charging power, achieving more precise tracking. The second-stage MPC, using the first-stage optimization results as input, targets the minimization of daily electricity purchase and sale costs and the fluctuation of power exchange. It employs rolling optimization and the YALMIP CPLEX solver to achieve minimized costs and power fluctuations. Finally, case study results demonstrate that the two-stage MPC method enhances system stability, reduces grid fluctuations, lowers system costs, and provides higher economic value.