Estimation of hourly global solar irradiation on tilted absorbers from horizontal one using Artificial Neural Network for case study of Mashhad

Today, for providing clean energy, solar capturing facilities such as photovoltaic panels (PV) or solar thermal collectors (SCTs) have been increasingly installed worldwide. On the other side, lack of solar radiation data is one of the barriers for developing these technologies locally. Short-time step calculation of solar global irradiation (SGI) on inclined planes is required regarding to predict precise performance of solar systems, leading to enhance security operation's conditions and economic cost saving. Moreover, SGI values on tilted absorbers have a nonlinear relationship with several variables such as Horizontal Solar Global Irradiation, Extraterrestrial Horizontal Global Irradiation, and number of days, collector angle, solar altitude angle and the latitude of the location. Thus computation of SGI is neither readily to obtain nor easy to forecast. This paper is proposed on estimating accurate values of SGI on tilted planes via Artificial Neural Network (ANN). Indeed, ANNs are effective tools to model nonlinear systems and are widely used simulation software incorporated in MATLAB. Mashhad the second megacity of Iran is taken into account for the case study. The ANN is developed and optimized using every 30 min of SGI data (6.00 AM until 5.00 PM) in 2013 on zero, 45° and 60° inclined planes respectively. These data have been gauged by pyranometers which are installed in Air & Solar Institute of Ferdowsi University of Mashhad. Meanwhile, the accuracies including R2 (Correlation Coefficient), MAE (Mean Absolute Error) and RMSE (Root Mean Square Error) are obtained 0.9242, 0.0284, 0.055 and 0.9302, 0.0269, 0.0549 for 60 and 45 tilted collectors respectively. Eventually it is concluded that ANN can be a reliable network and well capable for forecasting solar energy on slope solar absorbers in Mashhad.