Artificial neural network model for the estimation of direct solar radiation in the Indian zone

This paper presents a self-consistent model for the estimation of direct solar radiation in the Indian zone. It takes into account the atmospheric transmittance modified in accordance with the climate zone and calculates solar radiation at normal incidence using Hottel's clear day model. The regional weather phenomena are taken into account with the help of variables such as relative humidity, mean duration of sunshine per hour and the rainfall, and a composite parameter referred to as sky clearness index (CI) is determined using artificial neural network analysis. The CI is finally applied to the modified Hottel's clear day model to predict the grey day solar irradiance. The model predictions for the Indian region are found to be in good agreement with the measurements. The variability of sky CI is represented by the contours of constant value in Indian region, which in turn would enable the present model to be used in a self-consistent manner.