Wa. Kwong et Km. Passino, DYNAMICALLY FOCUSED FUZZY LEARNING CONTROL, IEEE transactions on systems, man and cybernetics. Part B. Cybernetics, 26(1), 1996, pp. 53-74
Citations number
52
Categorie Soggetti
Controlo Theory & Cybernetics","Computer Science Cybernetics","Robotics & Automatic Control
A ''learning system'' possesses the capability to improve its performa
nce over time by interacting with its environment. A learning control
system is designed so that its ''learning controller'' has the ability
to improve the performance of the closed-loop system by generating co
mmand inputs to the plant and utilizing feedback information from the
plant. Learning controllers are often designed to mimic the manner in
which a human in the control loop would learn how to control a system
while it operates. Some characteristics of this human learning process
may include: (i) a natural tendency for the human to focus their lear
ning by paying particular attention to the current operating condition
s of the system since these may be most relevant to determining how to
enhance performance; (ii) after learning how to control the plant for
some operating condition, if the operating conditions change, then th
e best way to control the system may have to be relearned; and (iii) a
human with a significant amount of experience at controlling the syst
em in one operating region should not forget this experience if the op
erating condition changes. To mimic these types of human learning beha
vior, we introduce three strategies that can be used to dynamically fo
cus a learning controller onto the current operating region of the sys
tem. We show how the subsequent ''dynamically focused learning'' (DFL)
can be used to enhance the performance of the ''fuzzy model reference
learning controller'' (FMRLC) [1]-[5] and furthermore we perform comp
arative analysis with a conventional adaptive control technique. A mag
netic bah suspension system is used throughout the paper to perform th
e comparative analyses, and to illustrate the concept of dynamically f
ocused fuzzy learning control.