Uncertainty estimation of wind power forecasts: Comparison of Probabilistic Modelling Approaches
Résumé
Short-term wind power forecasting tools providing “single-valued” (spot) predictions are nowadays widely used. However, end-users may require to have additional information on the uncertainty associated to the future wind power production for performing more efficiently functions such as reserves estimation, unit commitment, trading in electricity markets, a.o. Several models for on-line uncertainty estimation have been proposed in the literature and new products from numerical weather prediction systems (ensemble predictions) have recently become available, which has increased the modelling possibilities. In order to provide efficient on-line uncertainty estimation, choices have to be made on which model and modelling architecture should be preferred. Towards this goal we proposes to classify different approaches and modelling architectures for probabilistic wind power forecasting. Then, a comparison is carried out on representatives models using real data from several wind farms.
Domaines
Energétique
Origine : Fichiers produits par l'(les) auteur(s)
Loading...