The Impact of Weather Conditions on Renewable Energy Production: A Machine Learning Approach

In recent decades, obtaining energy from wind, solar, and hydro energy has brought hope for a sustainable future. Machine learning (ML) significantly impacts the field of energy systems. ML techniques have emerged as a powerful tool for analyzing and optimizing energy systems. This approach can potentially reduce the cost of analyzing a device by focusing solely on the data generated during its operation. This paper investigates the application of regression models to predict renewable energy production based on weather conditions, aiming to enhance the reliability and efficiency of renewable energy systems. Experiments are performed using Decision Tree (DT), k-nearest Neighbors (KNN), Multiple Linear (MLR), Polynomial (PR), Random Forest (RF), and Support Vector (SVR) Regression algorithms. The performances of ML models are compared based on the evaluations obtained through the Coefficient of Determination (R2), and error measures namely Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), Mean Squared Error (MSE), and Root Mean Square Error (RMSE). RF is the most successful algorithm in predicting renewable energy production depending on weather conditions as it has the highest R2 value and the lowest error measures.