Machine learning-based wind speed forecasting: a comparative study.

Area of Science:

• Renewable Energy Systems
• Computational Intelligence
• Environmental Engineering

Background:

• Fossil fuel reliance necessitates transition to sustainable energy sources.
• Wind energy is a key component of global clean energy strategies.
• Efficient wind energy generation requires precise wind speed prediction for grid integration.

Purpose of the Study:

• To evaluate the performance of various machine learning (ML) techniques for accurate wind speed prediction.
• To compare the efficacy of Support Vector Machine (SVM), Random Forest, Artificial Neural Networks (ANN), and XGBoost algorithms.
• To identify the most accurate ML model for optimizing wind power generation.

Main Methods:

• Utilized a dataset for wind speed forecasting.
• Implemented and evaluated four ML algorithms: SVM, Random Forest, ANN, and XGBoost.
• Assessed model performance using Root Mean Square Error (RMSE) and Mean Absolute Error (MAE).

Main Results:

• The Support Vector Machine (SVM) model exhibited superior performance.
• SVM achieved the lowest error metrics: RMSE of 0.83609 and MAE of 0.69623.
• All evaluated ML models showed potential for wind speed forecasting.

Conclusions:

• Machine learning techniques are effective for wind speed forecasting.
• SVM is a highly accurate method for predicting wind speeds.
• Accurate forecasting supports sustainable urban development and efficient power generation.