Analysis of the effect of linguistic variables of consequence membership function

Abstract

Fuzzy controller shows different responses depending on the size of the linguistic variables. Therefore, the choice of appropriate linguistic variable is an important part of the design of fuzzy controller. However, since fuzzy controller is nonlinear, it is difficult to analyze mathematically the affection of the linguistic variable. In this paper to analyze the effect of linguistic variables, the relationship between linguistic variables and fuzzy controller stability zone according to the variable changes in the size of each linguistic variables and the number of linguistic variables, that is calculated by Describing Function approximately, is analyzed through regression analysis. Describing Function can approximate FLC in frequency domain and we can determine the stability using nyquist plot. This study can help to understand the role of linguist variables of consequent membership function and to select them for designing FLC on the point of stability.