Estimation of Diffuse Fraction of Global Solar Radiation Using Artificial Neural Networks

Abstract

Measured daily mean air temperature and relative humidity values between 1998 and 2002 for Abha city in Saudi Arabia were used for predicting diffuse fraction of solar radiation in future time domain using artificial neural networks method. The estimations of diffuse solar radiation were made using four combinations of data sets, viz. (i) day of the year and daily maximum air temperature as inputs and diffuse solar radiation as output, (ii) day of the year and daily minimum air temperature as inputs and diffuse solar radiation as output, (iii) day of the year and daily mean air temperature as inputs and diffuse solar radiation as output, and (iv) time day of the year, daily mean air temperature, and relative humidity as inputs and diffuse solar radiation as output. The measured data between 1998 and 2001 was used for training the neural networks while the remaining 250 days' data from 2002 was used as testing data. The testing data was not used in training the neural networks. Obtained results show that neural networks are well capable of estimating diffuse solar radiation from temperature and relative humidity. This can be used for estimating diffuse solar radiation for locations where only temperature and humidity data are available.