Estimation Of Global Solar Radiation by Using ANN and ANFIS

Abstract

Solar energy has the greatest potential among the energy sources available in the world and it is clean and universal energy source. The first parameter that should be determined carefully when planning systems based on solar energy is the solar radiation value. The solar radiation values obtained from ground observations can be estimated with various software models. Especially, estimation models designed using emerging computer technology have ease of use and accuracy. In this study, the well-known artificial intelligence techniques such as artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS) are used to estimate solar radiation. The analysis of solar radiation requires complex, lengthy and time consuming procedures and artificial intelligence techniques such as ANN and ANFIS eliminate great effort and time. A system that measures atmospheric data such as light, temperature and humidity using sensors is designed and for this study, measurements were made for over a month. The atmospheric data obtained from the climatic conditions of the province of Karaman and the surface solar radiation values measured using the Pyranometer are utilized to construct these models. As a result of testing the constructed models, the ANFIS model is found to be more successful than ANN.