Quantification of uncertainty of wind energy predictions

Author(s): Sameer Al-Dahidia*, Piero Baraldib, Enrico Ziob,c,d,e, Montelatici Lorenzof
a Department of Mechanical and Maintenance Engineering, School of Applied Technical Sciences, German Jordanian University, Amman 11180, Jordan
b Energy Department, Politecnico di Milano, Via La Masa 34, Milan 20156, Italy
c Aramis Srl, Via pergolesi 5, Milano 20121, Italy
d MINES ParisTech, PSL Research University, CRC, Sophia Antipolis 06560, France
e Department of Nuclear Engineering, College of Engineering, Kyung Hee University, Seoul 130-701, Korea
f Research Development and Innovation, Edison Spa, Foro Buonaparte 31, Milan 20121, Italy
International Journal of Smart Grid and Clean Energy, vol. 9, no. 2, March 2020: pp. 458-465
ISSN: 2315-4462 (Print)
ISSN: 2373-3594 (Online)
Digital Object Identifier: 10.12720/sgce.9.2.458-465

Abstract: Accurate prediction of wind energy production is of paramount importance for an affordable and reliable power supply to consumers. Prediction models are used as decision-aid tools for electric grid operators to dynamically balance the energy productions provided by a pool of diverse sources in the energy mix. The objective of the present work is the quantification of the uncertainty affecting the wind energy production predictions provided by an ensemble of Artificial Neural Networks (ANNs) models. The capability of the proposed Bootstrap (BS) technique to quantify the uncertainty affecting the predictions is verified considering a wind plant located in Italy. The obtained results show that the BS technique provides a more satisfactory quantification of the uncertainty of wind energy predictions than that of a technique based on the computation of the quantiles of the single model outcomes

Keywords: Wind energy, prediction, uncertainty quantification, prediction intervals, ensemble, bootstrap
Full Paper.pdf

Copyright © 2020 International Journal of Smart Grid and Clean Energy, All Rights Reserved