Inference in fuzzy models of physical processes

Abstract

General idea of the paper is comparison of different reasoning methods, which may be used in some types of fuzzy models. Different triangular norms and defuzzification methods were used. It is shown that many reasoning methods give similar results. However, many of them are not very reasonable. Some simple theorems about functions approximated by models are presented. Special attention is applied to modeling of physical processes. Examples of models used in reality are presented. Some of them are build as modifications of Takagi-Sugeno model introduced earlier by author. © Springer-Verlag 2001.