Research on Robust operation optimization of virtual power plant considering PV output uncertainty and load aggregation characteristics

To mitigate the supply-demand imbalance risk due to new energy output uncertainties and to devise an economical and secure energy supply strategy for virtual power plants (VPPs), this study introduces a robust operation optimization model. This model accounts for the uncertainties of photovoltaic (PV) output and load aggregation characteristics, integrating various methodologies including the regional climate model (Providing Regional Climate for Impact Studies, PRECIS), the BP neural network, the chance-constrained programming (CCP) algorithm, the Equivalent Thermal Parameters (ETP) model, and a two-level, two-stage optimization approach. The findings indicate that the model successfully predicts and calculates PV output and load changes, offering a solid data foundation for energy supply planning. It enables the generation of an optimal, economically viable, and robust day-ahead energy supply strategy, considering the maximization of benefits for various stakeholders within the VPP. Furthermore, by leveraging the rapid adjustment capabilities of energy storage equipment and air conditioning loads, the model corrects calculation errors in source-load forecasts, enhancing the security of the energy supply.