Laser simulator logic: A novel inference system for highly overlapping of linguistic variable in membership functions

Abstract

A new approach for Fuzzification and Defuzzification processes of a high degree of overlapping between the linguistic variables through proportional and relative-dynamic-scale membership functions is presented in this paper. Most of related works that utilized type-1 and 2 fuzzy logic algorithms are confused and questionable from two points of view: (1) the overlapping between linguistic variables values starts almost from the midst of each other; so, what will be happened if it starts before the midst i.e., increasing the degree of overlapping, - (2) the linguistic variables are commonly represented by symmetrical membership functions form. It has been found that the current type 1 and 2 fuzzy logic are no longer suitable to deal with highly inference between linguistic variables. Thus, a novel algorithm, called laser simulator logic, is introduced in this paper to deal with the issue of highly inference of linguistic variables membership functions, which depends mainly on a proportional and relative-dynamic scale of membership function that will change the scale of membership continuously based on the crisp input/output. Results show that the laser simulator performs better than both type-1 and 2 fuzzy logic when there is a high degree of overlapping between the linguistic variables.