Wind Speed Forecasting using Machine Learning, Deep Learning, and Explainable AI: A Case Study of Phuket, Thailand
Article Sidebar
Main Article Content
Abstract
This study aims to forecast hourly wind speed in Phuket Province to support wind power generation planning by applying and comparing machine learning and deep learning models in conjunction with explainable artificial intelligence (XAI) techniques. The dataset consists of historical hourly meteorological data covering a 12-month period from January to December 2024. The data were divided into an eight-month training set and a four-month testing set. Experimental results indicate that the machine learning model, particularly linear regression, outperformed deep learning models, achieving a mean absolute error (MAE) of 1.5679, a root mean square error (RMSE) of 2.0739, and a coefficient of determination (R²) of 0.8113. In contrast, the deep learning models yielded R² values ranging from 0.2581 to 0.5760. XAI-based analysis reveals that short-term lagged wind speed variables and temporal features are the most influential factors affecting wind speed, reflecting the daily wind patterns characteristic of coastal areas. The findings confirm that the linear regression model is well suited for hourly wind speed forecasting in terms of both predictive accuracy and interpretability. The resulting wind speed forecasts can be effectively applied to wind energy management, storage planning, and enhancing the operational stability of wind power generation systems.
Article Details
References
Ackermann, T. (2012). Wind power in power systems. Wiley.
Ahmad, T., Zhang, D., and Huang, C. (2021). Wind speed prediction using machine learning techniques. Energy Reports, 7, 576–586.
Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5–32.
Dawan, P., Sriprapha, K., Kittisontirak, S., Boonraksa, T., Junhuathon, N., Titiroongruang, W., and Niemcharoen, S. (2020). Comparison of power output forecasting on the photovoltaic system using adaptive neuro-fuzzy inference systems and particle swarm optimization–artificial neural network model. Energies, 13(2), 1-18.
Department of Alternative Energy Development and Efficiency. (2024). Thailand renewable energy outlook 2024. Department of Alternative Energy Development and Efficiency.
Electricity Generating Authority of Thailand. (2024). Power development plan 2567–2580 (PDP 2024). Electricity Generating Authority of Thailand.
Friedman, J. H. (2001). Greedy function approximation: A gradient boosting machine. The Annals of Statistics, 29(5), 1189–1232.
Global Wind Energy Council. (2024). Global wind report 2024. GWEC.
Hochreiter, S., and Schmidhuber, J. (1997). Long short-term memory. Neural Computation, 9(8), 1735–1780.
Hong, Y., Zhou, Z., andamp; Zhao, J. (2023). Can we trust explainable AI in wind power
forecasting? Energy and AI, 12, 100241.
Hyndman, R. J., and Athanasopoulos, G. (2021). Forecasting: Principles and practice. 3rd ed. OTexts.
International Energy Agency. (2023). Net zero by 2050: A roadmap for the global energy sector. International Energy Agency.
Khortsriwong, N., Boonraksa, P., Boonraksa, T., Fangsuwannarak, T., Boonsrirat, A., Pinthurat, W., and Marungsri, B. (2023). Performance of deep learning techniques for forecasting PV power generation: A case study on a 1.5 MWp floating PV power plant. Energies, https://doi.org/10.3390/en16052119
Khortsriwong, N., Boonraksa, P., Boonraksa, T., Fangsuwannarak, T., Boonsrirat, A., Pinthurat, W., and Marungsri, B. (2023). Performance of deep learning techniques for forecasting PV power generation: A case study on a 1.5 MWp floating PV power plant. Energies, 16(5), 2119. https://doi.org/10.3390/en16052119
Li, G., and Shi, J. (2010). Application of forecasting methods in wind power integration. Renewable Energy, 35(12), 2850–2859.
Li, X., Chen, Y., and Sun, W. (2023). Short-term wind speed forecasting using optimized BiLSTM networks. Applied Energy, 334, 120653.
Lundberg, S. M., and Lee, S. I. (2017). A unified approach to interpreting model predictions. In Advances in neural information processing systems, pp. 4765–4774.
Sankari, S., Paramasivan, S. K., Arockiasamy, K., andamp; Senthivel, S. (2023). Deep learning for wind speed forecasting using Bi-LSTM with selected features. Intelligent Automation andamp; Soft Computing, 35(3), 3829–3844. https://doi.org/10.32604/iasc.2023.030480
Shapley, L. S. (1953). A value for n-person games. In H. W. Kuhn and A. W. Tucker (Eds.), Contributions to the theory of games. Princeton University Press.
United Nations. (2015). Paris agreement. United Nations Treaty Series, 3156, 79.
Wang, Q., Li, K., and Liu, S. (2022). Short-term wind speed prediction using PCA and LSTM. Energy Conversion and Management, 254, 115256.
Zhang, Y., Liu, H., and Wang, J. (2024). Interpretable short-term wind speed forecasting using LSTM and SHAP. Energy, 286, 128640.
