<p>Wind energy has gained significant attention as a clean, renewable resource due to fossil fuels’ environmental impact. Accurate wind speed forecasting is essential to address variability and intermittency challenges. Current forecasting difficulties arise from wind speed’s high susceptibility to meteorological conditions. This study proposes a GA-based ensemble framework that combines forecasting models using genetic algorithms. We systematically compared 14 models: linear models (AR, ARMA), advanced neural networks (MLP, RBF), hybrid models, and ensembles. Models were evaluated using minute-by-minute data from five major Brazilian cities: Brasília, Florianópolis, Petrolina, Natal, and São Luís. Key findings include: I) Superior Performance: The proposed framework achieved MSE values from 0.0802 to 0.9020 and MAE values from 0.1970 to 0.6140 across all datasets; II) Robust Prediction: <InlineEquation ID="IEq1"> <EquationSource Format="TEX">$R^2$</EquationSource> <EquationSource Format="MATHML"><math> <msup> <mi>R</mi> <mn>2</mn> </msup> </math></EquationSource> </InlineEquation> values ranged from 0.7139 to 0.8723, demonstrating strong predictive capability; III) Statistical Validation: Friedman test (<InlineEquation ID="IEq2"> <EquationSource Format="TEX">$p < 0.001$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mi>p</mi> <mo><</mo> <mn>0.001</mn> </mrow> </math></EquationSource> </InlineEquation>) confirmed significant differences with perfect rank stability across all locations; IV) High Scalability: Runtimes ranged from 58,077.3 to 77,815.7 <InlineEquation ID="IEq3"> <EquationSource Format="TEX">$\mu$</EquationSource> <EquationSource Format="MATHML"><math> <mi>μ</mi> </math></EquationSource> </InlineEquation>s, determined by the base model combination; and V) Computational Efficiency: One-step-ahead forecasting requires only 0.0003 <InlineEquation ID="IEq4"> <EquationSource Format="TEX">$\mu$</EquationSource> <EquationSource Format="MATHML"><math> <mi>μ</mi> </math></EquationSource> </InlineEquation>s for weighting and combination.</p>

A genetic algorithm-based ensemble framework for wind speed forecasting

期刊论文
出版模式：

开放获取
发表日期：

2026年1月31日

查看全文

Tathiana Mikamura Barchi,
João Lucas Ferreira dos Santos,
Thiago Antonini Alves,
Paulo S. G. de Mattos Neto,
Sergio Luiz Stevan Jr.,
Fernanda Cristina Corrêa,
Marie Chantelle C. Medina,
Hugo Valadares Siqueira,
João Fausto L. de Oliveira

摘要

Wind energy has gained significant attention as a clean, renewable resource due to fossil fuels’ environmental impact. Accurate wind speed forecasting is essential to address variability and intermittency challenges. Current forecasting difficulties arise from wind speed’s high susceptibility to meteorological conditions. This study proposes a GA-based ensemble framework that combines forecasting models using genetic algorithms. We systematically compared 14 models: linear models (AR, ARMA), advanced neural networks (MLP, RBF), hybrid models, and ensembles. Models were evaluated using minute-by-minute data from five major Brazilian cities: Brasília, Florianópolis, Petrolina, Natal, and São Luís. Key findings include: I) Superior Performance: The proposed framework achieved MSE values from 0.0802 to 0.9020 and MAE values from 0.1970 to 0.6140 across all datasets; II) Robust Prediction: $R^2$ $R^{2}$ values ranged from 0.7139 to 0.8723, demonstrating strong predictive capability; III) Statistical Validation: Friedman test ( $p < 0.001$ $p < 0.001$ ) confirmed significant differences with perfect rank stability across all locations; IV) High Scalability: Runtimes ranged from 58,077.3 to 77,815.7 $\mu$ $μ$ s, determined by the base model combination; and V) Computational Efficiency: One-step-ahead forecasting requires only 0.0003 $\mu$ $μ$ s for weighting and combination.