Abstract
Solar energy is one of the cleanest renewable energy sources and has no
environmental impact. Solar radiation data is important to solar engineers, designers
and architects which is also fundamental for efficiently determining irrigation water
needs and potential yield of crops, among others. Solar energy is mainly used to meet
the growing electricity demand and decline the amount of CO2
emission thus
preserving fossil fuels and natural resources. The temperature and sunshine duration
are measured by most of the meteorological services all over the world but global solar
radiation measurements are limited due to the restricted number of solar radiation
measuring stations and some of the data are missing. In order to estimate the solar
radiation in the other areas where the meteorological stations are not established, the
theoretical solar radiation estimation models proposed by various researchers have
proved handy. One of the main important assignments is to recognize the site with high
solar energy potential for renewable power generation. This assists in accomplishing
the target of the Indian solar mission project by the year 2022. The present work aims
at the prediction of solar radiation using artificial neural network models which are
applied to four different locations across India. As validation, measured and estimated
solar radiation data were analyzed in terms of the square of the correlation coefficient
and RMSE. The outcomes of this study will play a vital role in the estimation of global
solar radiation with less percentage of error.