Volume 12 Issue 1 ( March 2025)

Pages_541-552

Modelling and Performance Evaluation of a Photovoltaic/thermal Solar Energy System in Oman Using Experimental Data and Artificial Neural Networks

Asil M. K. Al Khaldi, Al-Ghaliya A. S. Al-Marzouqi, Hussein A. Kazem, Miqdam T. Chaichan

[ABSTRACT ]

Photovoltaic thermal (PV/T) hybrid systems harness solar energy by simultaneously generating heat and electricity from sunlight. These systems consist of metallic panels with solar cells covered by glass, designed to convert sunlight into energy while absorbing heat. The research in this field has focused on improving PV/T efficiency, particularly in managing the heat that can negatively impact PV cell performance. This study explores the effects of temperature on the power, voltage, current, and thermal and electrical efficiencies of PV/T systems, aiming to enhance PV efficiency through experimental research. Focusing on Oman’s weather conditions, the study demonstrates that integrating thermal components significantly improves solar energy system performance, increasing the maximum output power of the PV system from 10.302 W to 11.607 W. Using empirical data, an artificial neural network model was developed with the multilayer perceptron approach, achieving high predictive accuracy. The analysis revealed minimal mean squared error values of 0.0055 for the PV system and 0.1358 for the PV/T system, along with a high coefficient of determination (R value) of 1 for the PV system and 0.9995 for the PV/T system. This study highlights the effectiveness of thermal component integration in enhancing solar energy system efficiency in Oman.

Keywords: Photovoltaic/thermal; solar energy; ANN; MLP