Forecasting the Diffuse Solar Radiation by using a New Modeling over Saudi Arabia

Abstract

Within the arranging and execution assessment of solar energy projects, particularly within the planning and estimation of solar photovoltaic sources as elective energy sources within the future, exact evaluation of Diffuse Solar Radiation (DSR) is one of the essential and basic issues. The target of this article is to use the new model to predict DSR in four different locations in Saudi Arabia. The author explored the available meteorological and radiation data. The data covers the 25 years from 1990 to 2014 and were measured at Al-Baha, Abha, Jeddah and Taif locations in Saudi Arabia. Through detailed statistical evaluation and analysis, 19 empirical models were tested to construct the best empirical model to estimate the monthly average daily DSR of Saudi Arabia. Using widely used statistical errors, namely MBE, MPE, RMSE, U95, R, t-test and GPI the proposed correlation model was compared with 19 models provided in the literature. Through this analysis, the cubic empirical equation model is selected as the good model. Compared with these existing models, this model reveals accurate results with minimal statistical errors. Based on these results, Model 22 gives the highest GPI value. The conclusion is the cubic equation model of the diffusion fraction (i.e., D/G = 0.215+2.123 (S/So) -3.547(S/So)2+5.142 (S/So)3 is the best agreement model that has been mentioned in the previous discussion. This model is a generalized equation in Saudi Arabia, which can forecast the monthly average daily diffuse radiation on the horizontal plane at any site under similar climatic conditions without the measured climatic conditions. Experts or architects can use this model to perform site selection and technical and financial evaluation of solar energy applications and photovoltaic technology.