A COMPARATIVE STUDY OF PREDICTION OF HOURLY SLOPE IRRADIATION

These days, the installation of solar photovoltaic panel is increasing rapidly around the world. Solar irradiation data is very essential for modeling of solar panels but lack of solar insolation become barrier. To improve the performance of solar panel, shortest time interval calculation of solar irradiation measurement and prediction on inclined surface are required. Solar global irradiation on inclined surface have a nonlinear relations with input variables such as extraterrestrial irradiance, global solar irradiance on flat surface, zenith angle and incident angle. In this paper, Decision Tree Model, Random Forest Model, Generalized Linear Models, Artificial Neural Network, Linear Regression Model and Adaptive Neural Fuzzy Interference System model are implemented to predict the value of solar global irradiance on inclined surface, so as to perform a comparative study. Here, rattle package in R software is used to achieve the desired accuracy at location Jodhpur ( Rajasthan, Bharat).