Two Level Scheduling Strategy for Inter-provincial DC Power Grid Considering the Uncertainty of PV-load Prediction

In order to reduce the influence of the uncertainty of photovoltaic and load output forecasting on the economy of cross-provincial optimal dispatching, a two-level cross-provincial optimal dispatching strategy was proposed, which took into account the uncertainty of photovoltaic-load forecasting. Firstly, a vector auto-regressive (VAR) model based on the clear sky index series was used to predict the photovoltaic output. In the day-ahead time scale, the multi scene stochastic programming method was used to analyze the predicted output of two provinces day-ahead photovoltaic. In the intra-day time scale, a new trapezoid fuzzy number equivalence method was proposed to achieve the fuzzy value equivalence of the intra-day photovoltaic and load forecasting output of the two provinces. Secondly, based on the day-ahead photovoltaic output forecasting scenario set and day-ahead load forecasting curve, combined with thermal power and pumped storage energy, a day-ahead cross-provincial dispatching model was built. Based on the day-ahead dispatching results and the intra-day photovoltaic and load output fuzzy data sets, the inter-day dispatching model was built based on the DC power adjustment strategy. Finally, a two-level scheduling plan under uncertain environment was analyzed based on a cross-provincial interconnected power grid reconstructed from the actual data of a province in China. The results show that the scheduling strategy proposed in this paper can be coordinated and optimized in two provinces. It can reduce the complexity of chance constrained programming model and improve the economy of system operation. © 2021 Chin. Soc. for Elec. Eng.