Technical gazette, Vol. 26 No. 3, 2019.
Original scientific paper
https://doi.org/10.17559/TV-20160702220418
Determining Optimum Tilt Angles of Photovoltaic Panels by Using Artificial Neural Networks in Turkey
Mustafa Şahin
; Afyon Kocatepe University, Technology Faculty, Electrical and Electronic Engineering, Gazlıgöl, 03200, Afyonkarahisar, Turkey
Abstract
Sun is the most important energy source of the world. To make use of this energy source effectively, the sun’s angle of incidence to earth must be known. The angle, however, between rotation axis and orbital plane of the world is not constant and it changes continuously. Depending on this change, the incidence angle of sun beams also change. For purpose of increasing the light amount falling on solar panels, light beams must be adjusted according to their angles of incidence. The difference of this study from the studies in literature realized to determine the optimum tilt angle by means of mathematical methods is the determination of optimum tilt angles with artificial neural networks. In the study, not each province within boundaries of Turkey but the whole country as a system was introduced to the artificial neural network. Thanks to this, monthly optimum tilt angle of the system to be installed in any place within boundaries of Turkey was determined by the artificial neural network. The installed solar panels were adjusted according to this optimum tilt angle. Thanks to this, it was ensured to be obtained maximum output from solar panels. Mounting of the system according to an angle as predicted by artificial neural network has caused 34% increase in energy amount obtained from fixed solar panel systems. Consequently, in this study, in prediction of optimum tilt angle of fixed solar panels, in what extent the artificial neural networks are successful were observed. It was determined that correct results have been obtained in respect of both economy and utilization.
Keywords
Artificial Neural Networks; Optimum Tilt Angle; Renewable Energy; Solar Cell
Hrčak ID:
220980
URI
Publication date:
12.6.2019.
Visits: 2.465 *