The solar radiation data are a very important parameter to forecast generation from any solar PV (Photo Voltaic) plant to be installed at a suitable location, though it is not readily available for many locations because of the costly measurement instruments, and their maintenance, and calibration issues. The solar irradiation data available from NASA and Meteonorm has the prediction topology of consolidating 20–25 years old irradiation data which has the question of reliability. Furthermore, at various locations in India, the irradiance value of Meteonorm is overestimated than actual. In this research study an attempt has been made to develop an algorithm for the anisotropic solar radiation model by coalescing Gopinathan empirical parameter, Klucher anisotropic model, and Collares-Pereira & Rabl (GKCR) empirical parameter which gives the best results of forecasted daily and monthly solar radiation at almost every location in India. The computed results are compared with Meteonorm and the NASA database for three distinct locations (New Delhi, Chennai, and Kolkata) at different azimuths and tilt angles. The results obtained i.e., solar tilted radiation and tilt gain were very close to that simulated using PVsyst software using Meteonorm and NASA database. The computed results were compared using various statistical tools with measured irradiance values for four distinctly located solar PV plants. The computed result shows the least error compared to actual data for two locations. Employing the GKCR empirical algorithm mentioned in the research article, hourly irradiance data for the next day, for any specific tilt and azimuth angle can be estimated using available weather forecast data (sunshine duration) and can circumvent the penalty levied by the Load dispatch Centre. Furthermore, by using monthly data anticipated by the algorithm, the question of data reliability of chargeable Meteo sources working on coalescing data of antiquated years like NASA and Meteonorm can be solved.
