Artificial neural network model for prediction solar radiation data: application for sizing stand-alone photovoltaic power system

Abstract

The prediction of daily global solar radiation data is very important for many solar applications, possible application can be found in meteorology, renewable energy and solar conversion energy. In this paper, we investigate using radial basis function (RBF) networks in order to find a model for daily global solar radiation data from sunshine duration and air temperature. This methodology is considered suitable for prediction time series. Using the database of daily sunshine duration, air temperature and global solar radiation data corresponding to typical reference year (TRY). A RBF model has been trained based on 300 known data from TRY, in this way, the network was trained to accept and even handle a number of unusual cases. Known data were subsequently used to investigate the accuracy of prediction. Subsequently, the unknown validation data set produced very accurate estimation, with the mean relative error (MRE) not exceed 1.5% between the actual and predicted data, also the correlation coefficient obtained for the validation data set is 98.9%, these results indicates that the proposed model can successfully be used for prediction and modeling of daily global solar radiation data from sunshine duration and air temperature. An application for sizing of stand-alone PV system has been presented in this paper in order to show the importance of this modeling.