Machine Learning-Driven Wind Resource Assessment In Cox’s Bazar, Bangladesh

Sowrav Ghosh; Md. Ashikul Islam Soron

doi:10.38032/scse.2026.4.230

Authors

Sowrav Ghosh Department of Mechanical Engineering, Chittagong University of Engineering & Technology, Chattogram-4349, Bangladesh
Md. Ashikul Islam Soron Department of Mechanical Engineering, Chittagong University of Engineering & Technology, Chattogram-4349, Bangladesh

DOI:

https://doi.org/10.38032/scse.2026.4.230

Keywords:

Machine Learning, Wind Energy, Predictive Modeling, Coastal Wind Resources, Renewable Energy

Abstract

The monsoon-induced randomness of tropical coastal winds is poorly handled by standard parametric models, making them less suitable for wind resource characterization in Bangladesh. To fill this gap, this work introduces a hybrid machine learning (ML) framework to evaluate the wind potential over Cox’s Bazar, the proposed location for the country's first coastal 60 MW Wind Power Plant (WPP). A 10-year (2002-2011) synthetic dataset, with average wind speed at a 10 m height and estimated Weibull parameters, was used. Using K-means clustering and Principal Component Analysis (PCA), four monsoon-related wind regimes were identified. Meanwhile, Gradient Boosting and Random Forest supervised learning techniques were implemented for short-term forecasting. The best-performing ML approaches achieved an R² of 0.175, suggesting limited predictive accuracy. However, when Weibull fitting was applied, the R² improved to 0.72. Power-law extrapolation to an 80 m hub height yielded an average wind speed of 3.81 m/s and a power density of 38.69 W/m², corresponding to IEC Class 1 ("Poor"). These results highlight the limitations of ML for short-term wind prediction under monsoon variability and offer insights for the strategic deployment of low-wind towers and coastal renewable energy planning.

Downloads

PDF

Downloads

Download data is not yet available.

References

[1] Victoria Agbakwuru, Peter Ofuje Obidi, Ojonimi Segun Salihu, and Ogwu Chinelo MaryJane, “The role of renewable energy in achieving sustainable development goals,” International Journal of Engineering Research Updates, vol. 7, no. 2, pp. 013–027, Nov. 2024, doi: 10.53430/IJERU.2024.7.2.0046. DOI: https://doi.org/10.53430/ijeru.2024.7.2.0046

[2] F. H. Mahmood, A. K. Resen, and A. B. Khamees, “Wind characteristic analysis based on Weibull distribution of Al-Salman site, Iraq,” Energy Reports, vol. 6, pp. 79–87, Feb. 2020, doi: 10.1016/J.EGYR.2019.10.021. DOI: https://doi.org/10.1016/j.egyr.2019.10.021

[3] L. Liu, J. Wang, J. Li, and L. Wei, “Estimation of wind speed distribution with time window and new kernel function,” Journal of Renewable and Sustainable Energy, vol. 14, no. 5, Sep. 2022, doi: 10.1063/5.0100649. DOI: https://doi.org/10.1063/5.0100649

[4] Y. Xu et al., “Artificial intelligence: A powerful paradigm for scientific research,” The Innovation, vol. 2, no. 4, p. 100179, Nov. 2021, doi: 10.1016/J.XINN.2021.100179. DOI: https://doi.org/10.1016/j.xinn.2021.100179

[5] “Bangladesh’s first large-scale centralized wind farm commences operation.” Accessed: Oct. 22, 2025. [Online]. Available: https://en.powerchina.cn/2024-03/28/c_828681.htm

[6] I. Pobočíková, M. Michalková, Z. Sedliačková, and D. Jurášová, “Modelling the Wind Speed Using Exponentiated Weibull Distribution: Case Study of Poprad-Tatry, Slovakia,” Applied Sciences 2023, Vol. 13, Page 4031, vol. 13, no. 6, p. 4031, Mar. 2023, doi: 10.3390/APP13064031. DOI: https://doi.org/10.3390/app13064031

[7] S. F. Khahro, K. Tabbassum, A. M. Soomro, L. Dong, and X. Liao, “Evaluation of wind power production prospective and Weibull parameter estimation methods for Babaurband, Sindh Pakistan,” Energy Convers Manag, vol. 78, pp. 956–967, 2014, doi: 10.1016/J.ENCONMAN.2013.06.062. DOI: https://doi.org/10.1016/j.enconman.2013.06.062

[8] M. I. Faraz, “Strategic analysis of wind energy potential and optimal turbine selection in Al-Jouf, Saudi Arabia,” Heliyon, vol. 10, no. 20, p. e39188, Oct. 2024, doi: 10.1016/J.HELIYON.2024.E39188. DOI: https://doi.org/10.1016/j.heliyon.2024.e39188

[9] A. Younis et al., “Wind Speed Forecast for Sudan Using the Two-Parameter Weibull Distribution: The Case of Khartoum City,” Wind 2023, Vol. 3, Pages 213-231, vol. 3, no. 2, pp. 213–231, Apr. 2023, doi: 10.3390/WIND3020013. DOI: https://doi.org/10.3390/wind3020013

[10] “Wind Speed Analysis for Coastal Regions of Pakistan using Extended Generalized Lindley Distribution.” Accessed: Oct. 22, 2025. [Online]. Available: https://www.researchgate.net/publication/359601484_Wind_Speed_Analysis_for_Coastal_Regions_of_Pakistan_using_Extended_Generalized_Lindley_Distribution

[11] “(PDF) Optimizing Wind Energy Predictions Using Supervised Machine Learning Models.” Accessed: Oct. 21, 2025. [Online]. Available: https://www.researchgate.net/publication/384466661_Optimizing_Wind_Energy_Predictions_Using_Supervised_Machine_Learning_Models

[12] C. M. Rosca and A. Stancu, “A Comprehensive Review of Machine Learning Models for Optimizing Wind Power Processes,” Applied Sciences (Switzerland), vol. 15, no. 7, Apr. 2025, doi: 10.3390/APP15073758/S1. DOI: https://doi.org/10.3390/app15073758

[13] A. A. Martins, D. C. Vaz, T. A. N. Silva, M. Cardoso, and A. Carvalho, “Clustering of Wind Speed Time Series as a Tool for Wind Farm Diagnosis,” Mathematical and Computational Applications 2024, Vol. 29, Page 35, vol. 29, no. 3, p. 35, May 2024, doi: 10.3390/MCA29030035. DOI: https://doi.org/10.3390/mca29030035

[14] J. S. Ramírez and N. D. Duque, “Evaluation of Unsupervised Machine Learning Algorithms with Climate Data,” Ingeniería y Desarrollo, vol. 40, no. 02, pp. 131–165, Dec. 2022, doi: 10.14482/INDE.40.02.622.553. DOI: https://doi.org/10.14482/inde.40.02.622.553

[15] B. Li, P. D. Williams, J. M. Mbugua, and Y. Hiraga, “Recent Advances in Long-Term Wind-Speed and -Power Forecasting: A Review,” Climate 2025, Vol. 13, Page 155, vol. 13, no. 8, p. 155, Jul. 2025, doi: 10.3390/CLI13080155. DOI: https://doi.org/10.3390/cli13080155

[16] G. Gualtieri, “A comprehensive review on wind resource extrapolation models applied in wind energy,” Renewable and Sustainable Energy Reviews, vol. 102, pp. 215–233, Mar. 2019, doi: 10.1016/J.RSER.2018.12.015. DOI: https://doi.org/10.1016/j.rser.2018.12.015

[17] S. Jahandideh, G. Ozavci, B. W. Sahle, A. Z. Kouzani, F. Magrabi, and T. Bucknall, “Evaluation of machine learning-based models for prediction of clinical deterioration: A systematic literature review,” Int J Med Inform, vol. 175, p. 105084, Jul. 2023, doi: 10.1016/J.IJMEDINF.2023.105084. DOI: https://doi.org/10.1016/j.ijmedinf.2023.105084

[18] “Classes of Wind Power Density [34] | Download Table.” Accessed: Oct. 22, 2025. [Online]. Available: https://www.researchgate.net/figure/Classes-of-Wind-Power-Density-34_tbl1_274713877

[19] M. J. Khan, M. T. Iqbal, and S. Mahboob, “A wind map of Bangladesh,” Renew Energy, vol. 29, no. 5, pp. 643–660, 2004, doi: 10.1016/J.RENENE.2003.10.002. DOI: https://doi.org/10.1016/j.renene.2003.10.002

[20] S. E. Reza, N. Reza, Z. T. Zohura, and A. Ahammad, “An assessment incorporating Weibull distribution method on wind energy prospective areas of Bangladesh,” 2nd International Conference on Electrical Engineering and Information and Communication Technology, iCEEiCT 2015, Oct. 2015, doi: 10.1109/ICEEICT.2015.7307359. DOI: https://doi.org/10.1109/ICEEICT.2015.7307359