Artificial neural network modeling and analysis of photovoltaic/thermal system based on the experimental study. (15th April 2019)

• Record Type: Journal Article
• Authors: Al-Waeli, Ali H.A.; Sopian, K.; Yousif, Jabar H.; Kazem, Hussein A.; Boland, John; Chaichan, Miqdam T.
• Is Part Of: Energy conversion and management. Volume 186(2019)
• Page Start: 368
• Page End: 379
• Publication Date: 2019-04-15
• Subjects: Artificial neural network -- Simulated multi-layer perceptron -- Hybrid PV/T system -- Nanofluid -- Nano-PCM; Direct energy conversion -- Periodicals; Energy storage -- Periodicals; Energy transfer -- Periodicals; Énergie -- Conversion directe -- Périodiques; Direct energy conversion; Periodicals; 621.3105
• DOI: 10.1016/j.enconman.2019.02.066
• Languages: English
• ISSNs: 0196-8904

Abstract

Graphical abstract: Highlights: An artificial neural network model for PV/T is presented. Discretization of parameters and equation was made. An experiment has been conducted to validate the proposed ANN models results. Systems using water, nanofluid and nano-PCM moving through the cooling pipes are tested. The comparison between different models results showed a good consistent and agreement. Abstract: A Photovoltaic/Thermal (PV/T) system combines PV and thermal collector, which is considered promising technology especially for building integrated PV/T system. The PV/T cooling systems using water, water-PCM and nanofluid/nano-PCM moves through the cooling pipes were investigated, in this study. However, this paper focuses on testing different PV/T systems (conventional PV, water-based PVT, water-nanofluid PVT, and nanofluid/nano-PCM) under the same conditions and environment using one artificial neural network (ANN) based Multi-Layer Perceptron (MLP) system. Also, investigate the differences in the efficiency of these systems on both thermal and electrical when using only one simulation system (MLP). The proposed ANN approach proved that using of nanofluid/nano-PCM was enhanced the electrical efficiency from 8.07% to 13.32% and its thermal efficiency reached 72%. Also, the voltage was improved significantly. Many measurement methods were used for validating the results of the proposed ANN model like the Mean Absolute Error (MAE), Mean Square Error (MSE), Correlation (R), and coefficient of determination (R 2 ). The proposed ANN model achieved a final MSE of 0.0229 in the training phase and 0.0282 in the cross-validation phase. The sensitivity analysis showed that the influence of solar irradiation and Amb-temp almost has a constant effect on electrical efficiency. However, the Ambient temperature had a significant impact on thermal efficiency. The results of the network were consistent with the experimental results of the current study and published works.