Modeling of Photovoltaic/Thermal System by Artificial Neural Network Based on The Experimental Study

Abstract

In this study, Artificial Neural Network model (ANN) has been used to model the temperature dependent current, voltage and output power characteristics of uncooled and cooled photovoltaic panels. In the previous laboratory experiment, the current and voltage values produced by the photovoltaic panels in the temperature range of 20 ˚C- 65 ˚C for one hour were measured. Models have been created using the Artificial Neural Network technique with experimental data containing 60 samples for each of these three PV/T, including uncooled and two different cooled models. The combinations and features of the Artificial Neural Network model that provide the lowest model error have been achieved. The performance of the Neural Network model performed well in both the uncooled photovoltaic, cooled with flat fins/PCM and cooled with perforated fins/PCM, with RMSE model errors of 1.15e-02, 6.76e-03 and 6.10e-03, respectively. Therefore, it was suggested as a potent tool for modeling current, voltage, and generated power at all temperatures reached during the hour-long experiment.

Keywords

• Artificial neural network
• Photovoltaic/Thermal System
• PV-PCM

Deneysel Çalışmaya Dayalı Fotovoltaik/Termal Sistemin Yapay Sinir Ağı ile Modellenmesi

Öz

Bu çalışmada soğutmasız ve soğutmalı fotovoltaik panellerin sıcaklığa bağlı akım, gerilim ve çıkış gücü karakteristiklerini modellemek için Yapay Sinir Ağı modeli (YSA) kullanılmıştır. Bir önceki laboratuvar deneyinde fotovoltaik panellerin 20 ˚C- 65 ˚C sıcaklık aralığında bir saat boyunca ürettikleri akım ve gerilim değerleri ölçülmüştür. Soğutmasız ve iki farklı soğutmalı model olmak üzere bu üç PV/T'nin her biri için 60 örnek içeren deneysel verilerle Yapay Sinir Ağı tekniği kullanılarak modeller oluşturulmuştur. Yapay Sinir Ağı modelinin en düşük model hatasını sağlayan kombinasyonları ve özellikleri belirlenmiştir. Sinir Ağı modelinin performansı sırasıyla 1.15e-02, 6.76e-03 ve 6.10e-03 RMSE model hatalarıyla hem soğutmasız fotovoltaik, düz kanatçıklar/FDM ile soğutulan hem de delikli kanatçıklar/FDM ile soğutulan fotovoltaikte iyi performans gösterdi. Bu nedenle, bir saatlik deney sırasında ulaşılan tüm sıcaklıklarda akım, gerilim ve üretilen gücü modellemek için güçlü bir araç olarak önerildi.

Anahtar Kelimeler

• Yapay sinir ağı
• Fotovoltaik/Termal Sistem
• PV-FDM

References

1. Alzaabi Aa, Badawiyeh Nk, Hantoush Ho, Hamid Ak. Electrical/thermal performance of hybrid PV/T system in Sharjah, UAE. Int J Smart Grid Clean Energy 2014.
2. A.N. Celik, Artificial neural network modelling and experimental verification of the operating current of mono-crystalline photovoltaic modules, Sol. Energy 85 (2011) 2507–2517.
3. C. Renno, F. Petito, A. Gatto, Artificial neural network models for predicting the solar radiation as input of a concentrating photovoltaic system, Energy Convers. Manage. 106 (2015) 999–1012.
4. Ceylan, İlhan, et al. "The prediction of photovoltaic module temperature with artificial neural networks." Case Studies in Thermal Engineering 3 (2014): 11-20.
5. dos Santos Carstens DD, da Cunha SK. Challenges and opportunities for the growth of solar photovoltaic energy in Brazil. Energy Policy 2019;125:396–404.
6. Hasan, Ahmad, Hamza Alnoman, and Yasir Rashid. "Impact of integrated photovoltaic-phase change material system on building energy efficiency in hot climate." Energy and Buildings 130 (2016): 495-505.
7. Hiyama, Takashi, and Ken Kitabayashi. "Neural network based estimation of maximum power generation from PV module using environmental information." IEEE Transactions on Energy Conversion 12.3 (1997): 241-247.
8. Huang, Chao, et al. "Improvement in artificial neural network-based estimation of grid connected photovoltaic power output." Renewable Energy 97 (2016): 838-848.

9. Huang, M. J., et al. "Natural convection in an internally finned phase change material heat sink for the thermal management of photovoltaics." Solar Energy Materials and Solar Cells 95.7 (2011): 1598-1603.
10. K. Hornik, M. Stinchcombe, H. White, Multilayer feedforward networks are universal approximators, Neural Network. 2 (1989) 359–366, https://doi.org/10.1016/0893-6080(89)90020-8.
11. Kazem HA, Chaichan MT. Effect of humidity on photovoltaic performance based on experimental study. Int J Appl Eng Res (IJAER) 2015;10(23):43572–7.
12. M.M. Bayat, E. Buyruk, A. Can, “Use of PCM wıth aluminum fins to improve solar panel performance” 26th International Conference on Heating, Cooling and Air-conditioning (2023).
13. M. Mohanraj, S. Jayaraj, C. Muraleedharan, Applications of artificial neural networks for refrigeration, air-conditioning and heat pump systems, a review, Renew. Sustain. Energy Rev. 16 (2012) 1340–1358.
14. M. Sardarabadi, M. Passandideh-Fard, S. Zeinali Heris, Experimental investigation of the effects of silica/water nanofluid on PV/T (photovoltaic thermal units), Energy 66 (2014) 264–272.
15. Ma T, Yang H, Zhang Y, Lu L, Wang X. Using phase change materials in photovoltaic ystems for thermal regulation and electrical efficiency improvement: A review and outlook. Renew Sustain Energy Rev 2015; 43:1273–84DIN 1946-4, ‘’VAC Systems in Buildings and Rooms Used in The Health Care Sector’’, 2008
16. Mellit, Adel, and Soteris A. Kalogirou. "ANFIS-based modelling for photovoltaic power supply system: A case study." Renewable energy 36.1 (2011): 250-258.
17. Park, Jungwoo, Taeyeon Kim, and Seung-Bok Leigh. "Application of a phase-change material to improve the electrical performance of vertical-building-added photovoltaics considering the annual weather conditions." Solar Energy 105 (2014): 561-574.
18. Sardarabadi, Mohammad, et al. "Experimental study of using both ZnO/water nanofluid and phase change material (PCM) in photovoltaic thermal systems." Solar Energy Materials and Solar Cells 161 (2017): 62-69.
19. Torun, Y., & Doğan, H. (2021). Modeling of Schottky diode characteristic by machine learning techniques based on experimental data with wide temperature range. Superlattices and Microstructures, 160, 107062.
20. Voyant, Cyril, et al. "Machine learning methods for solar radiation forecasting: A review." Renewable energy 105 (2017): 569-582.