A novel application of an analog ensemble for short-term wind power forecasting. (April 2015)

• Record Type: Journal Article
• Authors: Alessandrini, S.; Delle Monache, L.; Sperati, S.; Nissen, J.N.
• Is Part Of: Renewable energy. Volume 76(2015)
• Page Start: 768
• Page End: 781
• Publication Date: 2015-04
• Subjects: Analog ensemble -- Short-term wind power forecasting -- Probabilistic predictions -- Uncertainty quantification -- Ensemble verification; Renewable energy sources -- Periodicals; Power resources -- Periodicals; Énergies renouvelables -- Périodiques; Ressources énergétiques -- Périodiques; 333.794
• DOI: 10.1016/j.renene.2014.11.061
• Languages: English
• ISSNs: 0960-1481

Abstract

Abstract: The efficient integration of wind in the energy market is limited by its natural variability and predictability. This limitation can be tackled by using the probabilistic predictions that provide accurate deterministic forecasts along with a quantification of their uncertainty. We propose as a novelty the application of an analog ensemble (AnEn) method to generate probabilistic wind power forecasts (WPF). The AnEn prediction of a given variable is constituted by a set of measurements of the past, concurrent to the past forecasts most similar to the current one. The AnEn performance for WPF is compared with three state-of-the-science methods for probabilistic predictions over a wind farm and a 505-day long period: a wind power prediction based on the ensemble wind forecasts from the European Centre for Medium-Range Weather Forecasts (ECMWF) Ensemble Prediction System (ECMWF-EPS), the Limited-area Ensemble Prediction System (LEPS) developed within the COnsortium for Small-scale MOdelling (COSMO-LEPS) and a quantile regression (QR) technique. The AnEn performs as well as ECMWF-EPS, COSMO-LEPS and QR for common events while it exhibits more skill for rare events. A comparison with the performances obtained with a deterministic forecasting method based on a Neural Network is also carried out showing the benefits of using AnEn. Highlights: We obtain probabilistic wind power forecasts with the analog ensemble model. The analog ensemble is tested with the data of a wind farm in Italy. The performances of the analog ensemble are comparable to those of other methods for common events. The use of analog ensemble is beneficial particularly for forecasting rare events. Compared to other ensemble meteorological models, the cpu time of the analog ensemble is 1/3 lower.