Study and Prediction of Radiation Effects in Solar Power Plants using Neuro Fuzzy and Neural Network

E. N. Ganesh

Abstract

Neural and neuro-fuzzy frameworks are utilized, to figure temperature and sun powered radiation. The principle benefit of these frameworks is that they don't need any earlier information on the qualities of the information time-series to foresee their future qualities. These frameworks with various models have been prepared utilizing as information estimations of the above meteorological boundaries acquired from the National Observatory of Athens. In the wake of having reproduced a wide range of designs of neural networks and prepared involving estimations as preparing information, the best designs are chosen to assess their exhibition in connection with the presentation of a neuro-fuzzy framework. As the elective framework, ANFIS neuro fuzzy framework is thought of, on the grounds that it consolidates fuzzy rationale and neural network procedures that are utilized to acquire productivity. ANFIS is likewise prepared with similar information. The examination and the assessment of both of the frameworks are finished by their forecasts, utilizing a few blunder measurements. Neuro-fuzzy networks and all the more explicitly ANFIS it is demonstrated that the mix of etymological guidelines of fuzzy rationale with the preparation calculation utilized in neural networks, contribute in extremely subjective expectation results, which approach the ‘best’ Neural networks results.