LAUSR

Abstract In order to evaluate different modeling techniques for Unglazed Transpired Collectors (UTC), not only mathematical modeling method for UTC based on heat transfer expressions to estimate the various heat transfer coefficients for the UTC components and empirical relationship, but also grey modeling (GM) approach, Artificial Neural Networks (ANN) and Adaptive Network based Fuzzy Inference System (ANFIS) methods have been designed and introduced, in this study. Thermal performance experiments of UTC have been carried out on an optimized experimental setup. Firstly, obtained experimental results have been compared with the mathematical model. To constitute a common point, output temperature of the UTC has been selected as the output variable. Secondly, the GM(1,1) approach has been used to forecast the output temperature with higher accuracy with the aid of simple mathematical equations. Then, an ANN has been designed to estimate the output temperature using measured inputs variables. Next, ANFIS has been designed and used to predict the output temperature. Finally, obtained results have been compared and comparison results have been illustrated in both graphical and tabular form. GM(1,1) is the simplest method to forecast the output temperature with high accuracy, while ANFIS technique will be the best solution to predict the output temperature.