Due to renewable energy sources (RES) variable nature and their wide integration into power systems, setting an adequate operating power reserve is important to compensate unpredictable imbalance between generation and consumption. However, this power reserve should be ideally minimized to reduce system cost with a satisfying security level. Although many forecasting methodologies have been developed for forecasting energy generation and load demand, management tools for decision making of operating reserve are still needed. This paper deals with power reserve quantification through uncertainty analysis with a photovoltaic (PV) generator. Indeed, using an artificial neural network based predictor (ANNs), PV power and load have been forecasted 24 hours ahead, and also forecasting errors have been predicted. Through forecasting uncertainty analysis, the power reserve quantification is calculated according to various risk indexes.
展开▼
