Predictive Modeling of Monthly Wind Energy Production in Thailand Using Machine Learning Approaches

This study presents a predictive modeling approach for estimating monthly wind production in Thailand using advanced machine learning methodologies. Logistic Regression (LR), Extreme Learning Machines (ELM), and Support Vector Regression (SVR) were employed and evaluated through 5-fold and 10-fold cross-validation techniques. The dataset spans from 2018 to 2023, obtained from the Thai Meteorological Department and Department of Alternative Energy Development and Efficiency, encompassing monthly wind production data. Performance metrics including Mean Squared Error (MSE) and Mean Absolute Error (MAE) were used to assess model accuracy. Results indicate that SVR consistently outperforms LR and ELM in terms of predictive accuracy, demonstrating lower MSE and MAE values during both training and testing phases. Specifically, during the testing phase of the 10-fold cross-validation, the model achieved an MSE of 67.3646 and an MAE of 6.5194. This study contributes to the advancement of predictive modeling in renewable energy forecasting and provides valuable insights for policymakers and stakeholders in Thailand’s energy sector.