Temperature based model for predicting global solar radiation using genetic algorithm [GA]

Global solar radiation (GSR) data is the most important parameter for solar energy applications. But radiation data is not available in all the locations of India, due to higher cost and difficulty in measurements. Therefore it is necessary to develop methods to predict GSR using available meteorological parameters like sunshine data, relative humidity and temperature out of which temperature is the commonly available data. In this work, Genetic algorithm (GA) and statistical regression technique (SRT) is used to estimate the monthly mean daily global solar radiation on horizontal surface in India. For this purpose, angstrom type empirical model based on maximum and minimum temperature is developed. The empirical coefficients are determined using GA and SRT. The model results are validated against the measured GSR data of India Meteorological Department (IMD), Pune. From the results, it is found that the obtained empirical coefficients based on GA have more accuracy than the regressive model.