Hybrid AI-Based Framework for Renewable Energy Forecasting: One-Stage Decomposition and Sample Entropy Reconstruction with Least-Squares Regression DOI Creative Commons

Nahed Zemouri,

Hatem Mezaache,

Zakaria Zemali

и другие.

Energies, Год журнала: 2025, Номер 18(11), С. 2942 - 2942

Опубликована: Июнь 3, 2025

Accurate renewable energy forecasting is crucial for grid stability and efficient management. This study introduces a hybrid model that combines signal decomposition artificial intelligence to enhance the prediction of solar radiation wind speed. The framework uses one-stage strategy, applying variational mode an improved empirical method with adaptive noise. process effectively extracts meaningful components while reducing background noise, improving data quality, minimizing uncertainty. complexity these assessed using entropy-based selection retain only most relevant features. refined are then fed into advanced predictive models, including bidirectional neural network capturing long-term dependencies, extreme learning machine, support vector regression model. These models address nonlinear patterns in historical data. To optimize accuracy, outputs from all combined least-squares technique assigns optimal weights each prediction. was tested on datasets three geographically diverse locations, encompassing varying weather conditions. Results show notable improvement achieving root mean square error as low 2.18 coefficient determination near 0.999. Compared traditional methods, errors were reduced by up 30%, demonstrating model’s effectiveness supporting sustainable reliable systems.

Язык: Английский

Hybrid AI-Based Framework for Renewable Energy Forecasting: One-Stage Decomposition and Sample Entropy Reconstruction with Least-Squares Regression DOI Creative Commons

Nahed Zemouri,

Hatem Mezaache,

Zakaria Zemali

и другие.

Energies, Год журнала: 2025, Номер 18(11), С. 2942 - 2942

Опубликована: Июнь 3, 2025

Accurate renewable energy forecasting is crucial for grid stability and efficient management. This study introduces a hybrid model that combines signal decomposition artificial intelligence to enhance the prediction of solar radiation wind speed. The framework uses one-stage strategy, applying variational mode an improved empirical method with adaptive noise. process effectively extracts meaningful components while reducing background noise, improving data quality, minimizing uncertainty. complexity these assessed using entropy-based selection retain only most relevant features. refined are then fed into advanced predictive models, including bidirectional neural network capturing long-term dependencies, extreme learning machine, support vector regression model. These models address nonlinear patterns in historical data. To optimize accuracy, outputs from all combined least-squares technique assigns optimal weights each prediction. was tested on datasets three geographically diverse locations, encompassing varying weather conditions. Results show notable improvement achieving root mean square error as low 2.18 coefficient determination near 0.999. Compared traditional methods, errors were reduced by up 30%, demonstrating model’s effectiveness supporting sustainable reliable systems.

Язык: Английский

Процитировано

0