Sensors,
Год журнала:
2024,
Номер
25(1), С. 69 - 69
Опубликована: Дек. 26, 2024
To
solve
the
coverage
problem
caused
by
random
deployment
of
wireless
sensor
network
nodes
in
forest
fire-monitoring
system,
a
modified
marine
predator
algorithm
(MMPA)
is
proposed.
Four
modifications
have
been
made
based
on
standard
(MPA).
Firstly,
tent
mapping
integrated
into
initialization
step
to
improve
searching
ability
early
stage.
Secondly,
hybrid
search
strategy
used
enhance
and
jump
out
local
optimum.
Thirdly,
golden
sine
guiding
mechanism
applied
accelerate
convergence
algorithm.
Finally,
stage-adjustment
proposed
make
transition
stages
more
smoothly.
Six
specific
test
functions
chosen
from
CEC2017
function
benchmark
are
evaluate
performance
MMPA.
It
shows
that
this
has
good
optimization
capability
stability
compared
MPA,
grey
wolf
optimizer,
cosine
algorithm,
sea
horse
optimizer.
The
results
tests
show
MMPA
better
uniformity
node
distribution
MPA.
average
rates
highest
commonly
metaheuristic-based
algorithms,
which
91.8%
scenario
1,
95.98%
2,
93.88%
3,
respectively.
This
demonstrates
superiority
network.
Materials Testing,
Год журнала:
2024,
Номер
66(8), С. 1230 - 1240
Опубликована: Июль 5, 2024
Abstract
This
paper
introduces
a
novel
approach,
the
Modified
Electric
Eel
Foraging
Optimization
(EELFO)
algorithm,
which
integrates
artificial
neural
networks
(ANNs)
with
metaheuristic
algorithms
for
solving
multidisciplinary
design
problems
efficiently.
Inspired
by
foraging
behavior
of
electric
eels,
algorithm
incorporates
four
key
phases:
interactions,
resting,
hunting,
and
migrating.
Mathematical
formulations
each
phase
are
provided,
enabling
to
explore
exploit
solution
spaces
effectively.
The
algorithm’s
performance
is
evaluated
on
various
real-world
optimization
problems,
including
weight
engineering
components,
economic
pressure
handling
vessels,
cost
welded
beams.
Comparative
analyses
demonstrate
superiority
MEELFO
in
achieving
optimal
solutions
minimal
deviations
computational
effort
compared
existing
methods.
Scientific Reports,
Год журнала:
2024,
Номер
14(1)
Опубликована: Авг. 31, 2024
Abstract
Suspensions
containing
microencapsulated
phase
change
materials
(MPCMs)
play
a
crucial
role
in
thermal
energy
storage
(TES)
systems
and
have
applications
building
materials,
textiles,
cooling
systems.
This
study
focuses
on
accurately
predicting
the
dynamic
viscosity,
critical
thermophysical
property,
of
suspensions
MPCMs
MXene
particles
using
Gaussian
process
regression
(GPR).
Twelve
hyperparameters
(HPs)
GPR
are
analyzed
separately
classified
into
three
groups
based
their
importance.
Three
metaheuristic
algorithms,
namely
genetic
algorithm
(GA),
particle
swarm
optimization
(PSO),
marine
predators
(MPA),
employed
to
optimize
HPs.
Optimizing
four
most
significant
(covariance
function,
basis
standardization,
sigma)
within
first
group
any
algorithms
resulted
excellent
outcomes.
All
achieved
reasonable
R-value
(0.9983),
demonstrating
effectiveness
this
context.
The
second
explored
impact
including
additional,
moderate-significant
HPs,
such
as
fit
method,
predict
method
optimizer.
While
resulting
models
showed
some
improvement
over
group,
PSO-based
model
exhibited
noteworthy
enhancement,
achieving
higher
(0.99834).
Finally,
third
was
examine
potential
interactions
between
all
twelve
comprehensive
approach,
employing
GA,
yielded
an
optimized
with
highest
level
target
compliance,
reflected
by
impressive
0.999224.
developed
cost-effective
efficient
solution
reduce
laboratory
costs
for
various
systems,
from
TES
management.
Alexandria Engineering Journal,
Год журнала:
2024,
Номер
95, С. 38 - 49
Опубликована: Март 29, 2024
The
Marine
Predators
Algorithm
(MPA)
is
a
prominent
Nature-Inspired
Optimization
(NIOA)
that
has
garnered
significant
research
interest
due
to
its
effectiveness.
It
draws
inspiration
from
the
foraging
behaviors
of
marine
predators,
predominantly
using
Lévy
or
Brownian
approach
for
strategy.
Despite
acclaim,
structural
bias
within
MPA
not
been
thoroughly
investigated,
marking
gap
in
current
research.
This
absence
targeted
forms
core
rationale
behind
initiating
this
study.
Structural
recently
identified
NIOAs,
causing
population
revisit
specific
regions
search
space
without
gaining
new
information.
As
result,
it
may
lead
increased
computational
costs
and
slow
down
rate
convergence.
Therefore,
identifying
essential
better
understand
mechanism
MPA.
To
ascertain
presence
any
bias,
two
introduced
models
are
employed:
BIAS
toolbox
Generalized
Signature
Test.
These
examinations
reveal
notable
MPA,
towards
center
space.
Also,
possible
future
directions
discussed.
Our
findings
provide
valuable
insights
into
dynamics
algorithm,
fostering
development
new,
unbiased,
efficient
algorithms.
