Application of Artificial Neural Network to predict the PV-Efficiency of Monocrystalline, Polycrystalline, and Amorphous photovoltaic systems

The aim of this work is to investigate in details the impacts of the temperature on the performance of three technologies of photovoltaic (PV) solar modules. Using the experimental data recorded during a year as inputs, the artificial neural network is employed to develop models to predict the efficiency of PV modules based on the effective parameters, including ambient temperature and irradiance, and having developed and validated the models, a comprehensive parametric study is conducted. The parametric study is performed to find the impacts of temperature on the power, and efficiency, as the main characteristics of a solar module. A monocrystalline, polycrystalline and amorphous solar modules with the same capacity are considered and compared together. The results show that the PV efficiency have a downward trend when the temperature increase. Moreover, in general, the monocrystalline type is found more sensitive to the temperature followed by the polycrystalline type, while the amorphous represents the type least influenced by the temperature.