An empirical technique to predict monthly mean global solar radiation for PV applications in Indian context

Abstract

AbstractGlobal radiation is a significant input in the design and simulation of solar PV as well as low temperature solar thermal systems. Moreover, meteorological conditions of any location also greatly influence the design, simulation, and output of the solar energy system. The considerable importance of global radiation leads to a dire need for global radiation models in the lack of measurement. Consequently, to enhance the predictability of solar resources over India, models for global radiation simulation through meteorological parameters (temperature, relative humidity, and wind speed) are proposed. Data of nine stations from NASA for 10 years (2008–2017) are used in the study to calibrate the models and performance of developed models is assessed through different statistical indicators (MPE, MBE, RMSE, R2, RMSD, NSE, and TS). The investigation reveals that out of the three proposed models, the model correlating total solar radiation with relative humidity, temperature, and wind speed is found to provide the highest accuracy in global radiation simulation with least statistical errors and is selected as best fit model. Moreover, validation of the proposed models is approved through monthly average daily total solar radiation predication for the remaining four testing cities of India and also through comparison with previously available models in the literature.