Multi-time-scale scheduling strategy of a virtual power plant based on model predictive control

To address the grid volatility issues caused by the integration of electric vehicles and distributed energy storage, a multi-timescale real-time rolling optimization scheduling strategy for virtual power plants based on model predictive control (MPC) is proposed. Considering the particularity of the electric vehicle units aggregated in the virtual power plant, an analysis of electric vehicle travel behavior is conducted to enable effective participation of electric vehicles in the charging and discharging scheduling plan of the virtual power plant. A multi-timescale day-ahead and intra-day operational framework for the virtual power plant is constructed, considering the constraints involved in system operation. An objective function incorporating intra-day balancing market and incentive-based demand response price incentives is introduced, and a multi-objective, multi-timescale optimization scheduling model based on the MPC rolling optimization algorithm is established for day-ahead and intra-day scheduling. By employing the MPC algorithm for real-time rolling optimization, the optimal day-ahead and intra-day scheduling plan for the virtual power plant is obtained. Simulation results indicate that the proposed strategy improved economic performance by 6.7%, increased carbon-credit revenues by 43.6%, reduced photovoltaic curtailment by 8291 kW, and decreased the output variability of distributed energy resources by 32%. The multi-timescale rolling optimization scheduling strategy presented in this paper addresses the volatility challenges of integrating distributed energy into the power grid. It offers theoretical value and serves as a practical reference for studying the large-scale integration of distributed energy resources on the user side.