Forecasting of wind parks production by dynamic fuzzy models with optimal generalisation capacity - Mines Paris Accéder directement au contenu
Communication Dans Un Congrès Année : 2003

Forecasting of wind parks production by dynamic fuzzy models with optimal generalisation capacity

Georges Kariniotakis

Résumé

On-line forecasting of the power output of wind farms is of major importance for a reliable and secure large-scale integration of wind power, especially under liberalized energy market environment. This paper presents such a prediction tool that receives on-line SCADA measurements, as well as numerical weather predictions as input, to predict the power production of wind parks 48 hours ahead. The prediction tool integrates models based on adaptive fuzzy-neural networks configured either for short-term or long-term forecasting. In each case, the model architecture is selected through non-linear optimization techniques. By this way the accuracy of the model on out of sample data (generalization) is optimized. The forecasting models are integrated in the MORE-CARE Energy Management Software (EMS) software developed in the frame of a European research project. In this EMS platform, wind forecasts and confidence intervals are used by economic dispatch and unit commitment functions. The paper presents detailed results on the performance of the developed models on a real wind farm using HIRLAM numerical weather predictions as input.
Fichier principal
Vignette du fichier
ISAP2003_paper_GK.pdf (1.21 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-00530064 , version 1 (06-02-2018)

Identifiants

  • HAL Id : hal-00530064 , version 1

Citer

Georges Kariniotakis. Forecasting of wind parks production by dynamic fuzzy models with optimal generalisation capacity. 12th Intelligent systems Application to power systems conference - ISAP 2003, Aug 2003, Lemnos, Greece. ⟨hal-00530064⟩
80 Consultations
41 Téléchargements

Partager

Gmail Facebook X LinkedIn More