Equivalence between Gaussian-Like and Mamdani Fuzzy Methods in the Multi-objective Design of Electronic Circuits

Abstract

The optimization of multi-objective designs is a classical problem in manufacturing electronic circuits due to the continuos miniaturization and their huge application in all type of practical issues. Classical techniques for optimizing multi-objective designs are limited when numerous nonlinear behaviors are implied, generating a rise in costs both in computational time and resources. For this reason alternative techniques applying fuzzy logic have been developed for optimizing multi-objective designs of electronic circuits, in particular the Gaussian-like method is an approach widely used for several authors. This paper explains the equivalence of a Mamdani approximation in relation with a Gaussian-like system for performing the control actions of the optimization task, the equivalence is based on a simple functional relation among the parameters of both methods. Two examples in the optimization of active circuits are presented.