Robust dispatching model of active distribution network considering PV time-varying spatial correlation

Abstract

With a high proportion of photovoltaic (PV) connected to the active distribution network (ADN), the correlation and uncertainty of the PV output will significantly affect the grid dispatching operation. Therefore, this paper proposes a novel robust ADN dispatching model, which considers the dynamic spatial correlation and power uncertainty of PV. First, the dynamic spatial correlation of PV output is innovatively modeled by dynamic conditional correlation (DCC) generalized autoregressive conditional heteroskedasticity (DCC-GARCH) model. DCC can accurately represent and forecast the spatial correlation of the PV output and reflect its time-varying characteristics. Second, a time-varying ellipsoidal uncertainty set constructed using the DCC, is introduced to bound the uncertainty of the PV outputs. Subsequently, the original mixed integer linear programming (MILP) model is transformed into the mixed integer robust programming (MIRP) model to realize robust optimal ADN dispatching. Finally, a numerical example is provided to demonstrate the effectiveness of the proposed method.