Optimal Scheduling Strategy of Park-level Virtual Power Plant for Demand Response

The park-level virtual power plant (PVPP) can aggregate diversified and flexible resources with the industrial parks to fully explore demand response potential, thus improving the profits of the PVPP. In order to optimize the demand response strategy of the PVPP in the electricity carbon joint market, a two-stage optimal scheduling model of the PVPP is developed here, in which, carbon emission costs are included in the optimization objects. The developed model includes two stages, i.e., a day-ahead stage and a real-time stage, and can be solved by the GAMS software. In the day-ahead stage, photovoltaic (PV) generation power is predicted by the improved radial basis function neural network, and the demand response bidding capacities are determined aiming to maximize the net benefits of the PVPP resulted from the demand response. In the real-time stage, the day-ahead schedules are revised according to real PV generation power, thus minimizing the negative effects of the PV power generation forecasting errors on the net benefits of the PVPP resulted from the demand response. Finally, the simulation analysis based on a PVPP is carried out. The results demonstrate that the strategy developed in this paper can maximize the net benefits of the PVPP resulted from the demand response in the context of considering the carbon emission costs.