Proceedings of the Institution of Mechanical Engineers Part I Journal of Systems and Control Engineering,
Год журнала:
2024,
Номер
238(10), С. 1763 - 1784
Опубликована: Июль 27, 2024
This
paper
mainly
discussed
the
highly
efficient
iterative
identification
methods
for
bilinear
systems
with
autoregressive
moving
average
noise.
Firstly,
input-output
representation
of
is
derived
through
eliminating
unknown
state
variables
in
model.
Then
based
on
maximum-likelihood
principle,
a
gradient-based
(ML-GI)
algorithm
proposed
to
identify
parameters
colored
noises.
For
improving
computational
efficiency,
original
model
divided
into
three
sub-identification
models
smaller
dimensions
and
fewer
parameters,
hierarchical
(H-ML-GI)
by
using
principle.
A
(GI)
given
comparison.
Finally,
algorithms
are
verified
simulation
example
practical
continuous
stirred
tank
reactor
(CSTR)
example.
The
results
show
that
effective
identifying
noises
H-ML-GI
has
higher
efficiency
faster
convergence
rate
than
ML-GI
GI
algorithm.
International Journal of Adaptive Control and Signal Processing,
Год журнала:
2024,
Номер
38(10), С. 3268 - 3289
Опубликована: Июль 29, 2024
Summary
This
paper
deals
with
the
problem
of
parameter
estimation
for
feedback
nonlinear
output‐error
systems.
The
auxiliary
model‐based
recursive
least
squares
algorithm
and
stochastic
gradient
are
derived
estimation.
Based
on
process
theory,
convergence
proposed
algorithms
proved.
simulation
results
indicate
that
can
estimate
parameters
systems
effectively.
International Journal of Adaptive Control and Signal Processing,
Год журнала:
2024,
Номер
unknown
Опубликована: Сен. 9, 2024
ABSTRACT
This
article
is
aimed
to
study
the
parameter
identification
of
ExpARX
system.
To
overcome
computational
complexity
associated
with
a
large
number
feature
parameters,
separation
scheme
based
on
different
features
model
introduced.
In
terms
phenomenon
that
coupling
parameters
lead
inability
algorithms,
separable
synchronous
interactive
estimation
method
introduced
eliminate
and
perform
in
accordance
hierarchical
principle.
For
purpose
achieving
high‐accuracy
performance
reducing
complexity,
gradient
iterative
algorithm
derived
by
means
search.
order
improve
accuracy,
multi‐innovation
proposed
introducing
theory.
convergence
speed,
conjugate
Finally,
simulation
example
real‐life
piezoelectric
ceramics
are
used
verify
effectiveness
algorithm.
Proceedings of the Institution of Mechanical Engineers Part I Journal of Systems and Control Engineering,
Год журнала:
2024,
Номер
238(10), С. 1763 - 1784
Опубликована: Июль 27, 2024
This
paper
mainly
discussed
the
highly
efficient
iterative
identification
methods
for
bilinear
systems
with
autoregressive
moving
average
noise.
Firstly,
input-output
representation
of
is
derived
through
eliminating
unknown
state
variables
in
model.
Then
based
on
maximum-likelihood
principle,
a
gradient-based
(ML-GI)
algorithm
proposed
to
identify
parameters
colored
noises.
For
improving
computational
efficiency,
original
model
divided
into
three
sub-identification
models
smaller
dimensions
and
fewer
parameters,
hierarchical
(H-ML-GI)
by
using
principle.
A
(GI)
given
comparison.
Finally,
algorithms
are
verified
simulation
example
practical
continuous
stirred
tank
reactor
(CSTR)
example.
The
results
show
that
effective
identifying
noises
H-ML-GI
has
higher
efficiency
faster
convergence
rate
than
ML-GI
GI
algorithm.
