Research Square (Research Square),
Год журнала:
2024,
Номер
unknown
Опубликована: Март 20, 2024
Abstract
The
RIME
optimization
algorithm
(RIME)
represents
an
advanced
technique.
However,
it
suffers
from
issues
such
as
slow
convergence
speed
and
susceptibility
to
falling
into
local
optima.
In
response
these
shortcomings,
we
propose
a
multi-strategy
enhanced
version
known
the
improved
(MIRIME).
Firstly,
Tent
chaotic
map
is
utilized
initialize
population,
laying
groundwork
for
global
optimization.
Secondly,
introduce
adaptive
update
strategy
based
on
leadership
dynamic
centroid,
facilitating
swarm's
exploitation
in
more
favorable
direction.
To
address
problem
of
population
scarcity
later
iterations,
lens
imaging
opposition-based
learning
control
introduced
enhance
diversity
ensure
accuracy.
proposed
centroid
boundary
not
only
limits
search
boundaries
individuals
but
also
effectively
enhances
algorithm's
focus
efficiency.
Finally,
demonstrate
performance
MIRIME,
employ
30
CEC2017
test
functions
compare
with
11
popular
algorithms
across
different
dimensions,
verifying
its
effectiveness.
Additionally,
assess
method's
practical
feasibility,
apply
MIRIME
solve
three-dimensional
path
planning
unmanned
surface
vehicles.
Experimental
results
indicate
that
outperforms
other
competing
terms
solution
quality
stability,
highlighting
superior
application
potential.
Journal Of Big Data,
Год журнала:
2024,
Номер
11(1)
Опубликована: Май 8, 2024
Abstract
The
Fennec
Fox
algorithm
(FFA)
is
a
new
meta-heuristic
that
primarily
inspired
by
the
fox's
ability
to
dig
and
escape
from
wild
predators.
Compared
with
other
classical
algorithms,
FFA
shows
strong
competitiveness.
“No
free
lunch”
theorem
an
has
different
effects
in
face
of
problems,
such
as:
when
solving
high-dimensional
or
more
complex
applications,
there
are
challenges
as
easily
falling
into
local
optimal
slow
convergence
speed.
To
solve
this
problem
FFA,
paper,
improved
Fenna
fox
DEMFFA
proposed
adding
sin
chaotic
mapping,
formula
factor
adjustment,
Cauchy
operator
mutation,
differential
evolution
mutation
strategies.
Firstly,
mapping
strategy
added
initialization
stage
make
population
distribution
uniform,
thus
speeding
up
Secondly,
order
expedite
speed
algorithm,
adjustments
made
factors
whose
position
updated
first
stage,
resulting
faster
convergence.
Finally,
prevent
getting
too
early
expand
search
space
population,
after
second
stages
original
update.
In
verify
performance
DEMFFA,
qualitative
analysis
carried
out
on
test
sets,
tested
newly
algorithms
three
sets.
And
we
also
CEC2020.
addition,
applied
10
practical
engineering
design
problems
24-bar
truss
topology
optimization
problem,
results
show
potential
problems.
Geosciences,
Год журнала:
2025,
Номер
15(1), С. 8 - 8
Опубликована: Янв. 3, 2025
Noise
profoundly
affects
the
quality
of
electromagnetic
data,
and
selecting
appropriate
hyperparameters
for
machine
learning
models
poses
a
significant
challenge.
Consequently,
current
denoising
techniques
fall
short
in
delivering
precise
processing
Wide
Field
Electromagnetic
Method
(WFEM)
data.
To
eliminate
noise,
this
paper
presents
an
data
approach
based
on
improved
dung
beetle
optimized
(IDBO)
gated
recurrent
unit
(GRU)
its
application.
Firstly,
Spatial
Pyramid
Matching
(SPM)
chaotic
mapping,
variable
spiral
strategy,
Levy
flight
mechanism,
adaptive
T-distribution
variation
perturbation
strategy
were
utilized
to
enhance
DBO
algorithm.
Subsequently,
mean
square
error
is
employed
as
fitness
IDBO
algorithm
achieve
hyperparameter
optimization
GRU
Finally,
IDBO-GRU
method
applied
WFEM
Experiments
demonstrate
that
capacity
conspicuously
superior
other
intelligent
algorithms,
surpasses
probabilistic
neural
network
(PNN)
accuracy
Moreover,
time
domain
processed
more
line
with
periodic
signal
characteristics,
overall
significantly
enhanced,
electric
field
curve
stable.
Therefore,
adept
at
sequence,
application
results
also
validate
proposed
can
offer
technical
support
inversion
interpretation.
Mathematics,
Год журнала:
2025,
Номер
13(3), С. 405 - 405
Опубликована: Янв. 26, 2025
The
Kepler
optimization
algorithm
(KOA)
is
a
metaheuristic
based
on
Kepler’s
laws
of
planetary
motion
and
has
demonstrated
outstanding
performance
in
multiple
test
sets
for
various
issues.
However,
the
KOA
hampered
by
limitations
insufficient
convergence
accuracy,
weak
global
search
ability,
slow
speed.
To
address
these
deficiencies,
this
paper
presents
multi-strategy
fusion
(MKOA).
Firstly,
initializes
population
using
Good
Point
Set,
enhancing
diversity.
Secondly,
Dynamic
Opposition-Based
Learning
applied
individuals
to
further
improve
its
exploration
effectiveness.
Furthermore,
we
introduce
Normal
Cloud
Model
perturb
best
solution,
improving
rate
accuracy.
Finally,
new
position-update
strategy
introduced
balance
local
search,
helping
escape
optima.
MKOA,
uses
CEC2017
CEC2019
suites
testing.
data
indicate
that
MKOA
more
advantages
than
other
algorithms
terms
practicality
Aiming
at
engineering
issue,
study
selected
three
classic
cases.
results
reveal
demonstrates
strong
applicability
practice.
