| Keywords |
Photovoltaic system ; Module temperature ; Data mining ; Artificial neural network ; Multiple regression analysis |
| Abstract |
In recent studies, it has been confirmed that the output performance of the photovoltaic system decreased as module temperature increased. Therefore, it will have a positive effect if it reflects the effects of module temperatures when it increases the efficiency of the design of the Photovoltaic system or the predicted accuracy of system power. To reflect this effect on the system design process, in this study, photovoltaic module temperature was predicted based on artificial neural networks(ANN) according to outlet solar irradiation, ambient air temperature and wind speed as well as multiple regression analysis. In addition, it was compared that the accuracy of the Data mining model with the prediction formula proposed in the previous studies. It shows that the accuracy of Data mining model is approximately 1.94% higher than that of the forecasting model of previous researches, among others, the ANN model is about 2.53% more accurate. |