Artificial Neural Network Fitting Tool Based Prediction of Solar Radiation for Identifying Solar Power Potential

Abstract

In this study details implementation of artificial neural network fitting tool is presented for prediction of solar radiation which can be used for assessment of power potential. Solar radiation data from 14 cities of Himachal Pradesh India is used in Network. The input variables are latitude, longitude, atmospheric pressure and wind speed and solar radiation as output parameter. The correlation coefficients are higher than 91%, proving high accuracy of model for solar energy potential. The procedure of using neural network fitting tool (nftool) and matlab code are developed for prediction of solar radiation to identify the solar power potential at any site worldwide.