Solar Radiation Forecasting Using Machine Learning and Ensemble Feature Selection

Accurate solar radiation forecasting is essential to operate power systems safely under high shares of photovoltaic generation. This paper compares the performance of several machine learning algorithms for solar radiation forecasting using endogenous and exogenous inputs and proposes an ensemble fe...

Full description

Bibliographic Details
Main Authors: Edna S. Solano, Payman Dehghanian, Carolina M. Affonso
Format: Article
Language:English
Published: MDPI AG 2022-09-01
Series:Energies
Subjects:
Online Access:https://www.mdpi.com/1996-1073/15/19/7049
_version_ 1797479620110450688
author Edna S. Solano
Payman Dehghanian
Carolina M. Affonso
author_facet Edna S. Solano
Payman Dehghanian
Carolina M. Affonso
author_sort Edna S. Solano
collection DOAJ
description Accurate solar radiation forecasting is essential to operate power systems safely under high shares of photovoltaic generation. This paper compares the performance of several machine learning algorithms for solar radiation forecasting using endogenous and exogenous inputs and proposes an ensemble feature selection method to choose not only the most related input parameters but also their past observations values. The machine learning algorithms used are: Support Vector Regression (SVR), Extreme Gradient Boosting (XGBT), Categorical Boosting (CatBoost) and Voting-Average (VOA), which integrates SVR, XGBT and CatBoost. The proposed ensemble feature selection is based on Pearson coefficient, random forest, mutual information and relief. Prediction accuracy is evaluated based on several metrics using a real database from Salvador, Brazil. Different prediction time-horizons are considered: 1 h, 2 h and 3 h ahead. Numerical results demonstrate that the proposed ensemble feature selection approach improves forecasting accuracy and that VOA performs better than the other algorithms in all prediction time horizons.
first_indexed 2024-03-09T21:48:25Z
format Article
id doaj.art-ddf307e1a440412f98ede3f7a8409e89
institution Directory Open Access Journal
issn 1996-1073
language English
last_indexed 2024-03-09T21:48:25Z
publishDate 2022-09-01
publisher MDPI AG
record_format Article
series Energies
spelling doaj.art-ddf307e1a440412f98ede3f7a8409e892023-11-23T20:12:16ZengMDPI AGEnergies1996-10732022-09-011519704910.3390/en15197049Solar Radiation Forecasting Using Machine Learning and Ensemble Feature SelectionEdna S. Solano0Payman Dehghanian1Carolina M. Affonso2Faculty of Electrical Engineering, Federal University of Para, Belem 66075-110, PA, BrazilDepartment of Electrical and Computer Engineering, The George Washington University, Washington, DC 20052, USAFaculty of Electrical Engineering, Federal University of Para, Belem 66075-110, PA, BrazilAccurate solar radiation forecasting is essential to operate power systems safely under high shares of photovoltaic generation. This paper compares the performance of several machine learning algorithms for solar radiation forecasting using endogenous and exogenous inputs and proposes an ensemble feature selection method to choose not only the most related input parameters but also their past observations values. The machine learning algorithms used are: Support Vector Regression (SVR), Extreme Gradient Boosting (XGBT), Categorical Boosting (CatBoost) and Voting-Average (VOA), which integrates SVR, XGBT and CatBoost. The proposed ensemble feature selection is based on Pearson coefficient, random forest, mutual information and relief. Prediction accuracy is evaluated based on several metrics using a real database from Salvador, Brazil. Different prediction time-horizons are considered: 1 h, 2 h and 3 h ahead. Numerical results demonstrate that the proposed ensemble feature selection approach improves forecasting accuracy and that VOA performs better than the other algorithms in all prediction time horizons.https://www.mdpi.com/1996-1073/15/19/7049ensemble feature selectionmachine learningphotovoltaic generationsolar radiation forecasting
spellingShingle Edna S. Solano
Payman Dehghanian
Carolina M. Affonso
Solar Radiation Forecasting Using Machine Learning and Ensemble Feature Selection
Energies
ensemble feature selection
machine learning
photovoltaic generation
solar radiation forecasting
title Solar Radiation Forecasting Using Machine Learning and Ensemble Feature Selection
title_full Solar Radiation Forecasting Using Machine Learning and Ensemble Feature Selection
title_fullStr Solar Radiation Forecasting Using Machine Learning and Ensemble Feature Selection
title_full_unstemmed Solar Radiation Forecasting Using Machine Learning and Ensemble Feature Selection
title_short Solar Radiation Forecasting Using Machine Learning and Ensemble Feature Selection
title_sort solar radiation forecasting using machine learning and ensemble feature selection
topic ensemble feature selection
machine learning
photovoltaic generation
solar radiation forecasting
url https://www.mdpi.com/1996-1073/15/19/7049
work_keys_str_mv AT ednassolano solarradiationforecastingusingmachinelearningandensemblefeatureselection
AT paymandehghanian solarradiationforecastingusingmachinelearningandensemblefeatureselection
AT carolinamaffonso solarradiationforecastingusingmachinelearningandensemblefeatureselection