Machine learning-based wind speed forecasting: a comparative study

Abstract

Wind turbines (WTs) are increasingly replacing fossil fuel-based power plants as a primary source of energy generation due to the limited supply of fossil resources in many nations. Accurate wind speed prediction is essential to developing the efficiency of wind energy generation and ensuring grid compatibility. This research evaluates several machine learning (ML) techniques, support vector machine (SVM), random forest, artificial neural networks (ANN), and XGBoost for wind speed prediction using the selected dataset. Framework performance is evaluated using root mean square error (RMSE) and mean absolute error (MAE). The results show that the SVM model achieved the highest accuracy, with an RMSE of 0.83609 and an MAE of 0.69623. The findings show the effectiveness of applying multiple ML approaches to wind speed forecasting, supporting the development of sustainable cities and enhancing the efficiency of power generation units.