Machine learning–based estimation of Weibull distribution parameters for wind energy assessment: A case study in Türkiye
Journal of Atmospheric and Solar-Terrestrial Physics, cilt.287, 2026 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 287
- Basım Tarihi: 2026
- Doi Numarası: 10.1016/j.jastp.2026.106973
- Dergi Adı: Journal of Atmospheric and Solar-Terrestrial Physics
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Artic & Antarctic Regions, Compendex, INSPEC, Academic Search Ultimate (EBSCO), Engineering Source (EBSCO)
- Anahtar Kelimeler: Machine learning, Support vector regression, Weibull distribution, Wind energy
- Çukurova Üniversitesi Adresli: Evet
Özet
The growing global demand for energy has underscored the critical importance of renewable resources, with wind energy emerging as a key contributor to sustainable energy systems. Nevertheless, the full exploitation of wind energy potential remains constrained by region-specific challenges such as site selection and installation costs which directly influence the efficiency of wind power plants (WPPs). Despite the significance of these challenges, a universally accepted methodology for WPP site selection has yet to be established. In this study, the wind energy potential of the Çanakkale region which is characterised by its rich wind resources, was evaluated using observational data from the General Directorate of Meteorology for the period 2017–2021. The analysis was carried out in MATLAB which employs the Weibull distribution function (WDF) to estimate both mean wind speeds and expected wind power outputs. Various approaches were applied to determine the Weibull coefficients, and their performance was validated against measured data through a comprehensive set of statistical error metrics. For the first time in this region, a machine learning (ML)–based approach, namely the Linear Support Vector Regression (LSVR) method, was integrated into the analysis to enhance the determination of WDF parameters. This approach significantly improves the accuracy of coefficient estimation and offers a perspective on regional wind energy assessment. The findings demonstrate the applicability of LSVR for Weibull parameter estimation in Çanakkale and support its use in preliminary regional wind-resource assessment. Overall, the findings deliver valuable insights for future academic research and practical applications in wind energy forecasting which contributes to more effective utilisation of renewable resources in diverse geographical contexts.