Solar photovoltaic power generation using machine learning considering weather conditions: A case study of Biret, Mauritania

The increasing adoption of renewable energy sources positions solar energy as a vital solution, especially in remote areas. Forecasting energy production based on climate variables is essential to align supply with demand and improve the management of photovoltaic systems. Artificial Intelligence (A...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Engineering applications of artificial intelligence Ročník 162; s. 112621
Hlavní autoři: Rchid, Abdellahi Moulaye, Attia, Moussa, Mohamed Mahmoud, Mohamed Elmamy, Aoulmi, Zoubir, Mahmoud, Abdelkader Ould
Médium: Journal Article
Jazyk:angličtina
Vydáno: Elsevier Ltd 26.12.2025
Témata:
Artificial intelligence
Gradient boosting algorithm
Hyperparameter optimization
Machine learning
Photovoltaic systems
Photovoltaic systems
Hyperparameter optimization
Gradient boosting algorithm
Artificial intelligence
Machine learning
ISSN:0952-1976
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:The increasing adoption of renewable energy sources positions solar energy as a vital solution, especially in remote areas. Forecasting energy production based on climate variables is essential to align supply with demand and improve the management of photovoltaic systems. Artificial Intelligence (AI) algorithms are necessary to predict electrical capacity and power, providing accurate estimations. This research investigates the accuracy of forecasts using data from the village of Biret, Mauritania, over one month. It evaluates several machine learning models, including Gradient Boosting, Random Forest, Adaptive Boosting (AdaBoost), Support Vector Regression (SVR), and Elastic Net. The study applies advanced machine learning techniques, such as Gradient Boosting and Random Forest, to enhance the accuracy of solar energy predictions. Grid Search for Hyperparameter Optimization (GridSearchCV) improved performance by reducing the mean square error (MSE). The optimization allowed Gradient Boosting to achieve the coefficient of determination (R2) value of 0.99, with Random Forest reaching 0.97 and AdaBoost reaching 0.88. The results indicate that the full models, especially Gradient Boosting and AdaBoost, achieve high accuracy, facilitating better matching of electricity production to actual needs.
ISSN:0952-1976
DOI:10.1016/j.engappai.2025.112621