Eventos Anais de eventos
COBEM 2021
26th International Congress of Mechanical Engineering
Performance and robustness comparison of fuzzy type-2 and optimal controllers in robotic joint positioning
Submission Author:
Fabrizio Leonardi , SP , Brazil
Co-Authors:
Renato Aguiar, Ivan Carlos Franco, Fabrizio Leonardi
Presenter: Renato Aguiar
doi://10.26678/ABCM.COBEM2021.COB2021-1074
Abstract
Fuzzy systems were introduced in 1965 and have been applied in many areas, such as control systems among others. Generally, fuzzy systems are based on a type-1 fuzzy set where membership functions are precise and the real value concerning specific membership function corresponds to only one membership grade. In 1975 fuzzy type-2 systems were introduced and this new theory has been also applied in many applications. Type-2 fuzzy sets possess an additional imprecision with relation to type-1 fuzzy sets, where a real value corresponds to an interval of membership degrees. It means there is uncertainty with relation to membership degree where a real value corresponds to many membership grades concerning the linguistic variable. There is an expectation that type-2 fuzzy sets can generate fuzzy controllers more efficiently in control systems applications that are linear or even nonlinear. In this work, it was designed a fuzzy controller for a position of motorized rotary joint utilizing type-2 fuzzy sets. The objective of this research is to investigate whether type-2 fuzzy controllers can present performance and robustness comparable to optimal controllers for this application which is of great interest in the robotics area. Initially, the comparison was performed numerically by simulation. Performance was evaluated employing reference inputs and disturbances of a wide-range frequency spectrum. Its robustness was investigated from the parametric and non-parametric point of view to represent parametric uncertainties but also the unmodeled dynamics, normally present in the modeling stage. The system was also evaluated experimentally in real-time for lab-scale equipment. The results observed suggest that type-2 fuzzy controllers can be used in applications like this with similar performance when compared to optimal controllers and having a more natural form of design.
Keywords
Type-2 fuzzy system, Optimal Control, Fuzzy logic
