Akademik Veri Yönetim Sistemi
Journal of Atmospheric and Solar-Terrestrial Physics, cilt.283, 2026 (SCI-Expanded, Scopus)
Solar radiation serves as the foundation of solar energy, a critical component of renewable energy sources. Accurate prediction of solar radiation is essential for enhancing power generation, managing energy resources, and ensuring system stability. The performance of photovoltaic systems is directly dependent on incoming solar radiation, providing accurate and dependable predictions crucial for the effective operation of power generation processes. Moreover, solar radiation prediction can be applied across various fields, including irrigation scheduling in agriculture, energy management in buildings and meteorological modelling. To enhance prediction accuracy in the field through developing technology and advanced data analysis would provide more efficient utilisation of renewable energy resources. The major contribution of this study is the integration of diverse modelling approaches into the global solar radiation (GSR) prediction framework, thereby providing a distinct methodological perspective in the relevant literature. Consequently, six different Angström models were created based on these approaches, and it was observed that the fine Gaussian supported vector model (FGSVM) and quadratic supported vector model (QSVM) outperformed the other models according to the employed performance metrics. In particular, FGSVM yielded the lowest prediction errors with RPE, MPE, MAPE and SSRE values of 19.583, 15.054, 17.489 and 0.467, respectively, whereas QSVM achieved the highest coefficient of determination with an R2 value of 0.851. These findings may contribute to the development of more reliable and region-specific solar radiation prediction frameworks for renewable energy planning and related engineering applications.