Control
of
dynamic
systems
consists
two
schemes.
The
first
is
to
create
system
elements,
and
the
second
design
a
suitable
controller
for
this
meet
desired
specifications.
A
commonly
used
actuator
in
control
DC
motor.
It
provides
direct
rotational
motion
or,
when
combined
with
drums
cables,
translational
motion.
motor,
which
employed
drive
many
electromechanical
systems,
serves
as
system.
Feedback
servo
known
their
precise
angled
positions
rapid
response,
will
be
focus
discussion,
specifically
modeling
position
local
To
begin,
motor/load
feedback
model
meticulously
arranged
MATLAB/Simulink
program.
Subsequently,
Simulink
program
loaded
onto
Raspberry
Pi
card,
circuit's
executed.
Achieving
motor
positioning
requires
resetting
error
reference
command
value
measured.
Additionally,
there
desire
reset
unblocking
state
fault
resulting
from
permanent
blockage.
Other
performance
data
indicates
extremely
fast
response
times
without
overshooting
fault.
fractional-order
Proportional-Integral
has
been
designed
facilitate
these
components
coefficients
have
chosen
manually.
Scientific Reports,
Год журнала:
2024,
Номер
14(1)
Опубликована: Окт. 29, 2024
Ensuring
the
stability
and
reliability
of
modern
power
systems
is
increasingly
challenging
due
to
growing
integration
renewable
energy
sources
dynamic
nature
load
demands.
Traditional
proportional-integral-derivative
(PID)
controllers,
while
widely
used,
often
fall
short
in
effectively
managing
these
complexities.
This
paper
introduces
a
novel
approach
frequency
control
(LFC)
by
proposing
filtered
PID
(PID-F)
controller
optimized
through
hybrid
simulated
annealing
based
quadratic
interpolation
optimizer
(hSA-QIO).
The
hSA-QIO
uniquely
combines
local
search
capabilities
(SA)
with
global
optimization
strengths
(QIO),
providing
robust
efficient
solution
for
LFC
challenges.
key
contributions
this
study
include
development
application
hSA-QIO,
which
significantly
enhances
performance
PID-F
controller.
proposed
was
evaluated
on
unimodal,
multimodal,
low-dimensional
benchmark
functions,
demonstrate
its
robustness
effectiveness
across
diverse
scenarios.
results
showed
significant
improvements
quality
compared
original
QIO,
lower
objective
function
values
faster
convergence.
Applied
two-area
interconnected
system
photovoltaic-thermal
generation,
hSA-QIO-tuned
achieved
substantial
reduction
integral
time-weighted
absolute
error
23.4%,
from
1.1396
0.87412.
Additionally,
reduced
settling
time
deviations
Area
1
9.9%,
1.0574
s
0.96191
s,
decreased
overshoot
8.8%.
In
2,
improved
0.89209
4.8%.
also
demonstrated
superior
tie-line
regulation,
achieving
immediate
response
minimal
overshoot.
e-Prime - Advances in Electrical Engineering Electronics and Energy,
Год журнала:
2024,
Номер
8, С. 100601 - 100601
Опубликована: Май 20, 2024
This
paper
investigates
the
impact
of
fractional
derivatives
on
activation
functions
an
artificial
neural
network
(ANN).
Based
results
and
analysis,
a
three-layer
backpropagation
model
with
integer
employed
in
function
gradient
descent
method
learning
algorithm
has
been
proposed.
Specifically,
three
perceptrons
have
proposed
based
applied
to
algorithms.
They
are
derivative
(FDAF)
perceptron,
(IDAF)-fractional
(FDLA)
perceptron.
The
Riemann-Liouville
(RL)
derivative,
Grunwald-Letnikov
(GL)
Caputo-Fabrizio
(CF)
Caputo
(C)
Atangana-Baleanu
(AB)
employed.
these
derivatives'
fractional-order
(FO)
is
investigated
wide
range
from
0.1−0.9
testing
mean
square
error
(MSE)
time
required
train
FO-based
IO-based
compared
help
performance
metrics
such
as
MSE
models.
training
simulation
illustrate
that
derivative-based
outperform
other
derivatives.
Also,
perceptron
better
terms
least
MSE.
Systems Science & Control Engineering,
Год журнала:
2025,
Номер
13(1)
Опубликована: Янв. 12, 2025
This
research
introduces
a
novel
dual
Fast
Grey
Wolf
Optimizer
(FGWO)
combined
with
Radial
Basis
Function
Neural
Networks
(RBFNN)
for
Fractional-Order
PID
(FOPID)
controller
applied
to
helicopter
simulator.
The
proposed
FGWO
improves
the
standard
(GWO)
by
enhancing
hunting
during
exploitation
phase
and
increases
robustness
in
convergence
minimum
value.
optimizes
FOPID
parameters
using
objective
function.
RBFNN
is
integrated
address
nonlinearities
uncertainties,
while
block
mitigates
coupling
effects.
performance
of
characterized
two
simulation
scenarios.
first
scenario
involved
nine
benchmark
functions
across
thirty
trials.
Results
demonstrated
that
offered
superior
terms
proximity
global
compared
GWO.
second
applying
controllers
helicopter.
evidenced
dual-FOPID-FGWO
(DRF-FG)
achieved
4.3363%
faster
response
1.8199%
higher
precision
than
GWO-based
(DRF-G).
DRF-FG
showed
trajectory
tracking
based
on
Ant
Lion
(DRF-A)
Whale
Optimization
Algorithm
(DRF-W).
improved
average
regulation
1.702%
0.152%
DRF-G.
International Journal of Robust and Nonlinear Control,
Год журнала:
2024,
Номер
34(9), С. 5996 - 6020
Опубликована: Март 12, 2024
Abstract
The
model
parameter
uncertainty
and
controller
gain
disturbance
of
the
factory
servo
system
are
challenges
that
affect
robustness
control
performance
system.
In
this
paper,
a
class
systems
with
non‐integer
order
is
studied.
stable
boundary
trajectory
method
fractional
used
to
determine
stability
domain
makes
stable.
An
optimal
trade‐off
design
for
time‐varying
PID
()
proposed.
time
function
introduced
as
adjustment
formula
realize
adaptive
gain.
Lyapunov
theorem
analyzes
method.
At
same
time,
an
ameliorated
gazelle
optimization
algorithm
(AGOA)
proposed
optimize
parameters
controller,
weight
relationship
changed
set
objective
obtain
combination
after
optimization.
benchmark
test
completed.
Statistical
analysis
shows
AGOA
can
enhance
global
search
ability,
prevent
acquisition
local
optimum,
have
faster
convergence
speed.
final
simulation
results
show
scheme
promising
alternative
improve
performance.
Research Square (Research Square),
Год журнала:
2024,
Номер
unknown
Опубликована: Авг. 22, 2024
Abstract
The
paper
summarizes
the
pattern
framework
of
some
optimization
algorithms,
and
proposes
four
directions
within
this
algorithmic
pattern:
search
agent
particles,
direction
perturbation,
step
size.
Based
on
pattern,
introduces
an
algorithm
called
"Detector",
name
is
just
for
convenient.In
algorithm,
as
well
size,
it
performs
simple
processing
only
considers
iterative
particles.
For
current
position
a
function
designed
to
evaluate
position.
If
value
greater
than
threshold,
particle
will
not
explore
around
that
position,
but
move
another
location.This
without
any
natural
inspirations,
in
order
prove
effectiveness
proposed
framework.
It
also
analyzes
limitations
class
algorithms.The
tested
CEC2013,
CEC2014,
CEC2015,
CEC2017,
CEC2022
test
suites,
compared
with
9
other
focus
high-cost
CEC2013
suite.
most
functions
suites.
uses
very
small
number
parameters
steps,
inspirations.
