Solar energy prediction using linear and non-linear regularization models: A study on AMS (American Meteorological Society) 2013–14 Solar Energy Prediction Contest

S.K. Aggarwal and L.M. Saini

Energy, 2014, vol. 78, issue C, 247-256

Abstract: In 2013, American Meteorological Society Committees on AI (artificial intelligence) Applications organized a short-term solar energy prediction competition aiming at predicting total daily solar energy received at 98 solar farms based on the outputs of various weather patterns of a numerical weather prediction model. In this paper, a methodology to solve this problem has been explained and the performance of ordinary LSR (least-square regression), regularized LSR and ANN (artificial neural network) models has been compared. In order to improve the generalization capability of the models, more experiments like variable segmentation, subspace feature sampling and ensembling of models have been conducted. It is observed that model accuracy can be improved by proper selection of input data segments. Further improvements can be obtained by ensemble of forecasts of different models. It is observed that the performance of an ensemble of ANN and LSR models is the best among all the proposed models in this work. As far as the competition is concerned, Gradient Boosting Regression Tree has turned out to be the best algorithm. The proposed ensemble of ANN and LSR model is able to show a relative improvement of 7.63% and 39.99% as compared to benchmark Spline Interpolation and Gaussian Mixture Model respectively.

Keywords: Artificial neural network; Ensemble learning; Least square regression; Regularization; Solar energy forecasting; Variable segmentation (search for similar items in EconPapers)
Date: 2014
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (13)

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S036054421401161X
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:energy:v:78:y:2014:i:c:p:247-256

DOI: 10.1016/j.energy.2014.10.012

Access Statistics for this article

Energy is currently edited by Henrik Lund and Mark J. Kaiser

More articles in Energy from Elsevier
Bibliographic data for series maintained by Catherine Liu ().