Ağaç Tabanlı ve Klasik Regresyon Modelleri Kullanılarak Temel PVT Parametrelerinin Makine Öğrenmesine Dayalı Tahmini

Öz

Küresel nüfusun sürekli artmasıyla birlikte enerji talebi hızla artmaktadır ve bu talep hala sınırlı rezervlere sahip fosil yakıtlarla karşılanmaktadır. Ancak fosil yakıtların çevresel etkileri sürdürülebilirlik açısından büyük zorluklar yaratmaktadır. Bu nedenle, enerji tüketimini azaltmak ve yenilenebilir enerji kaynaklarını teşvik etmek büyük önem taşımaktadır. Bunlar arasında güneş enerjisi, tükenmez ve çevre dostu bir kaynak olarak öne çıkmıştır. Aynı anda ısı ve elektrik üretebilen fotovoltaik-termal (PVT) paneller, iklimlendirme, ısıtma, soğutma ve kurutma gibi uygulamalar için önemli bir araştırma konusu olmuştur. PVT sistemlerinin etkili bir şekilde analiz edilebilmesi için doğru modelleme çok önemlidir. Bu çalışmada, PVT kolektörlerinin iki temel çıktı parametresi olan yüzey sıcaklığı (Ts) ve sıvı çıkış sıcaklığı (To) modellenmiş ve modellerin başarımı regresyon performans ölçütleri üzerinden değerlendirilmiştir. COMSOL Multiphysics’te sonlu elemanlar analizi ile oluşturulan bir veri seti, Python’da Karar Ağacı (KA), Rastgele Orman (RO), Hafif Gradyan Güçlendirme Makinesi (HGGM) ve Aşırı Gradyan Güçlendirme (AGG) gibi makine öğrenmesi modellerini geliştirmek için kullanılmıştır. Ek olarak, doğrusal regresyon modeli bir temel olarak kullanılmıştır. Tüm modeller yüksek tahmin doğruluğu elde etmiş olup, AGG en iyi performansı göstermiştir (Ts için R2=0.9998, RMSE=0.16, MAE=0.12; To için R2=0.9997, RMSE=0.19, MAE=0.12).

Anahtar Kelimeler

• Fotovoltaik-termal (PVT) kolektörler
• rastgele orman
• hafif gradyan güçlendirme makinesi
• aşırı gradyan güçlendirme
• sonlu elemanlar.

Machine Learning-Based Prediction of Key PVT Parameters Using Tree-Based and Classical Regression Models

Öz

With the continuous growth of the global population, energy demand is rapidly increasing, and this demand is still mainly supplied by fossil fuels with limited reserves. However, the environmental impacts of fossil fuels pose major challenges to sustainability. Therefore, reducing energy consumption and promoting renewable energy sources are of great importance. Among these, solar energy has gained prominence as an inexhaustible and environmentally friendly source. Photovoltaic-thermal (PVT) panels, capable of simultaneously generating heat and electricity, have attracted significant research interest for applications such as space conditioning, heating, cooling, and drying. To enable effective analysis of PVT systems, accurate modeling is essential. In this study, two main output parameters of PVT collectors, surface temperature (Ts) and fluid outlet temperature (To), were modeled, and the performance of the models was evaluated using regression performance metrics. A dataset generated through finite element analysis in COMSOL Multiphysics was used to develop machine learning models including Decision Tree (DT), Random Forest (RF), Light Gradient Boosting Machine (LightGBM), and Extreme Gradient Boosting (XGBoost) in Python. Additionally, a linear regression model served as a baseline. All models achieved high predictive accuracy, with XGBoost providing the best performance (R2=0.9998, RMSE=0.16, MAE=0.12 for Ts; R2=0.9997, RMSE=0.19, MAE=0.12 for To).

Anahtar Kelimeler

• Photovoltaic-thermal (PVT) collectors
• random forest
• light gradient boosting machine
• extreme gradient boosting
• finite elements.

Kaynakça

1. S. Saidur, N. A. Rahim, M. R. Islam, and K. H. Solangi, “Environmental impact of wind energy,” Renew. Sustain. Energy Rev., vol. 16, no. 1, pp. 103–112, 2012.
2. S. A. Kalogirou, “Solar thermal collectors and applications,” Prog. Energy Combust. Sci., vol. 30, no. 3, pp. 231–295, 2004.
3. A. Tiwari and R. K. Mishra, “Performance evaluation of photovoltaic/thermal solar air collector,” Energy Convers. Manag., vol. 52, no. 1, pp. 98–105, 2011.
4. T. T. Chow, “A review on photovoltaic/thermal hybrid solar technology,” Appl. Energy, vol. 87, no. 2, pp. 365–379, 2010.
5. J. K. Tonui and Y. Tripanagnostopoulos, “Improved PV/T solar collectors with heat extraction by forced or natural air circulation,” Renew. Energy, vol. 32, no. 4, pp. 623–637, 2007.
6. Y. Chaibi, M. Malvoni, T. El Rhafiki, T. Kousksou, and Y. Zeraouli, “Artificial neural-network based model to forecast the electrical and thermal efficiencies of PVT air collector systems,” Cleaner Eng. Technol., vol. 4, p. 100132, 2021.
7. T. Khatib, A. Mohamed, and K. Sopian, “A review of solar energy modeling techniques,” Renew. Sustain. Energy Rev., vol. 45, pp. 530–543, 2019.
8. A. H. Abedin, M. A. Alim, A. Islam, and M. M. Rahman, “Machine learning approaches for solar radiation prediction: A review,” Renew. Sustain. Energy Rev., vol. 121, p. 109711, 2020.

9. A. Mellit and A. M. Pavan, “A 24-h forecast of solar irradiance using artificial neural network: Application for performance prediction of a grid-connected PV plant at Trieste, Italy,” Sol. Energy, vol. 84, no. 5, pp. 807–821, 2010.
10. H. Gharaee, M. Erfaniantti, and A. M. Bahman, “Machine learning development to predict the electrical efficiency of photovoltaic-thermal (PVT) collector systems,” Energy Convers. Manag., vol. 315, p. 118808, 2024.
11. Y. Wang et al., “A hybrid CFD and machine learning study of energy performance of photovoltaic systems with a porous collector: Model development and validation,” Case Stud. Therm. Eng., vol. 69, p. 105998, 2025.
12. A. K. Tripathi et al., “Advancing solar PV panel power prediction: A comparative machine learning approach in fluctuating environmental conditions,” Case Stud. Therm. Eng., vol. 59, p. 104459, 2024.
13. S. Diwania et al., “Machine learning-based thermo-electrical performance improvement of nanofluid-cooled photovoltaic–thermal system,” Energy Environ., vol. 35, no. 9, pp. 1799–1817, 2024.
14. M. H. Ahmadi et al., “Machine learning prediction models of electrical efficiency of photovoltaic-thermal collectors,” Preprints, 2019.
15. R. Khan, A. M. Mohsen, and A. Alizadeh, “AI-integrated design of nanofluid-based PVT systems containing nano-enhanced PCM: Combining optimized GPR models with grasshopper optimization,” Case Stud. Therm. Eng., 2025.
16. Y. Li, A. Basem, A. Alizadeh, P. K. Singh, and S. Dixit, “Artificial neural networks with multi-objective thermal exchange optimization and PROMETHEE decision-making to improve PCM-based photovoltaic thermal systems,” Case Stud. Therm. Eng., 2025.
17. A. Nazeri, A. Taheri, and Z. S. Zomorodian, “Predicting photovoltaic-thermal panel output in urban contexts using machine learning methods,” Res. Sq. (preprint), 2024.
18. M. Jabari, A. Rad, M. A. Nasab, and M. Zand, “Parameter identification of PV solar cells and modules using bio dynamics grasshopper optimization algorithm,” IET Gener. Transm. Distrib., 2024.
19. S. F. Ali and D. Rakshit, “Genetic algorithm and grasshopper optimization algorithm with metaoptimization and RL-based parameter fine-tuning and their comparison for optimal thermal systems,” J. Comput. Civ. Eng., 2025.
20. H. Moayedi and A. Mosavi, “Electrical power prediction through a combination of multilayer perceptron with water cycle, ant lion and satin bowerbird searching optimizers,” Sustain., vol. 13, no. 4, p. 2336, 2021.
21. M. Jalal, I. U. Khalil, A. U. I. Haq, and A. Flah, “Advancements in PV array reconfiguration techniques,” IEEE Access, 2024.
22. K. N. Çerçi, “Energy, exergy, economic and enviro-economic analysis and artificial neural network modeling of an air-cooled PVT collector with NACA 8412 airfoils,” Appl. Therm. Eng., vol. 277, p. 126955, 2025.
23. J. R. Quinlan, “Learning decision tree classifiers,” ACM Comput. Surv., vol. 28, no. 1, pp. 71–72, 1996.
24. S. Alous, M. Kayfeci, and A. Uysal, “Experimental investigations of using MWCNTs and graphene nanoplatelets water-based nanofluids as coolants in PVT systems,” Appl. Therm. Eng., vol. 162, p. 114231, 2019.
25. L. Breiman, “Random forests,” Mach. Learn., vol. 45, no. 1, pp. 5–32, 2012.
26. G. Ke et al., “LightGBM: A highly efficient gradient boosting decision tree,” Adv. Neural Inf. Process. Syst., vol. 30, 2017.
27. T. Chen and C. Guestrin, “XGBoost: A scalable tree boosting system,” in Proc. 22nd ACM SIGKDD Int. Conf. Knowl. Discov. Data Min., pp. 785–794, 2016.