Abstract: Fuzzy logic control presents a computationally efficient and robust alternative to conventional controllers for many systems. This paper presents a distributed fuzzy logic controller (FLC) structure for a flexible-link manipulator based on evaluating the importance degrees of the output variables of the system. The two velocity variables, which have higher importance degrees, are grouped together as the inputs of the Velocity FLC. The two displacement variables, which have lower importance degrees, are used as the inputs of the Displacement FLC. The outputs of those two FLCs are summed up to control the joint of the flexible link. The fuzzy rules…of the distributed importance-based FLCs are written based on the expert knowledge, and the parameters of the membership functions of the two FLCs are tuned using nonlinear programming. The distributed importance-based FLC structure is further compared with two other commonly used structures: a Linear Quadratic Regulator and a distributed PD-like FLC. The robustness of the three controllers are tested and compared under various conditions.
Show more
Abstract: Fuzzy logic control is already in use due to its computational efficiency and robustness. If a person is sufficiently familiar with a system, intuition can be used as a basis to design a fuzzy controller for it. It may be difficult to accurately describe the behavior for flexible manipulators as their variables are strongly coupled. This paper reviews existing fuzzy logic controllers for flexible manipulators and discusses some of their limitations. It presents an alternative structure using a distributed fuzzy logic controller that controls angle and link vibrations of flexible manipulators independently. While the rules and structure of the proposed…distributed controller offers significant advantages compared to those available in literature, its performance still depends on the membership functions for its input and output variables. Wrong choices of these membership functions may lead to sluggish response, excessive vibration, or instability. The paper presents a novel algorithm, Low Dimensionality Tuning Algorithm (LDTA), for tuning a fuzzy controller by changing the membership functions of its variables using nonlinear programming. LDTA uses only two variables to describe each fuzzy variable to avoid dimensionality problems associated with searching for a solution to problems with large number of variables. The paper also discusses possibilities of simplifying the tuned fuzzy controller.
Show more