Prediction of irradiation: A comparative study of ANFIS

The accurate information of solar isolation data and its components are fundamental specifications for the purpose of designing, installation and monitoring of solar power plant. The beam components is easily available but the diffuse component is rarely available for each location around the globe. During the past decades, the most of work is done on prediction of global and diffusion radiation on plane surface by using metrological and geographical data. In this study, predicting the global solar isolation on inclined surface by using Adaptive Neural Fuzzy Interference System (ANFIS) method with different member function on the bases of input parameters such as global irradiance, extraterrestrial isolation, incident angle and zenith angle. This study is based on Liu & Jorden isotropic diffuse radiation model. The statistical indices such as RMSE, RMBE and R
²
are used to predict the solar isolation on tilted surface. The results show that the Triangular-shaped member function estimate the better prediction as compare to other member functions at location Jodhpur, Bharat.