Using the Concept of Fuzzy Logic and Controller System in Interpolation and Functions Approximation

Abstract

The main aim of this contribution is to present an alternative approach for interpolation and function approximation from a given data. Instead of the traditional interpolation methods we consider and propose a numerical procedure for interpolation using the concept of fuzzy logic and membership functions. The method can be used for interpolating data resulting from physical experiments, engineering, medicine, applied sciences etc. Fuzzification will be applied to the given data according to Mamdani technique and membership functions will be chosen to satisfy the interpolation mathematical condition. The defuzzification process will be implemented to get the crisp values of the interpolation. The procedure will be implemented on the mathematical code MATLAB and its Simulink fuzzy logic features. Finally the applicability and efficiency of the procedure is illustrated by numerical examples.