Neural Networks for Feedback Control

glibdoadingΤεχνίτη Νοημοσύνη και Ρομποτική

20 Οκτ 2013 (πριν από 3 χρόνια και 11 μήνες)

124 εμφανίσεις

Neural Network
s for Feedback Control

of
Robots and
Dynamical Systems



Frank L. Lewis


Moncrief
-
O'Donnell Endowed Chair

University Distinguished Scholar Professor

Automation and Robotics Research Institute

The University of Texas at Arlington

7300 Jack Ne
well Blvd. S

Ft. Worth, Texas 76118
-
7115


Abstract
: Practical

industrial
, robotic,

and aerospace
systems have
unknown
disturbances, unmodeled dynamics,
actuator constraints
,

friction,
and restricted
availability of measurements.
Such systems cannot easily

be dealt with using standard
adaptive or robust feedback control techniques.
Over the past years we have developed a
family of feedback controllers that can confront these
systems

using neural networks as
the basic control block structure. The learning
abilities of neural networks considered as
Intelligent Systems allow these controllers to learn on
-
line and improve their
performance through tuning of the weights. We will present a catalog of neural network
controllers designed based on feedback lineari
zation, backstepping, singular
pe
rturbations, and dynamic inversion techniques.
These neural network controller
s are
all tuned on
-
line in real
time based on the system e
rrors. Then, we will present some

recent result
s

on
H
-
infinity

feedback control

for c
onstrained input
nonlinear
systems.

The constraints on the input to the system are encoded via a quasi
-
norm
that

allows non
-
quadratic supply rates along with dissipativity theory to formulate the robust
output
feedback
control
problem
using Hamilton
-
Jacob
i
-
Isaac
(
HJI
)

equations.

An iterative
solution technique based on
a

game theoretic interpretation is presented.


To provide a
computationally tractable controller design method, t
he solution is approximated at each
iteration with a neural network.
The re
sult is a closed
-
loop control based on a neural net
that has been tuned
a priori

off
-
line.