Research Square (Research Square),
Journal Year:
2024,
Volume and Issue:
unknown
Published: Feb. 14, 2024
Abstract
The
Marine
Predators
Algorithm
(MPA)
is
recognized
as
one
of
the
optimization
method
in
population-based
algorithm
that
mimics
foraging
strategy
dominated
by
optimal
theory,
which
encounter
rate
policy
between
predator
and
prey
marine
ecosystems
for
solving
problems.
However,
MPA
presents
weak
point
towards
premature
convergence,
stuck
into
local
optima,
lack
diversity,
specifically,
real-world
niche
problems
within
different
industrial
engineering
design
domains.
To
get
rid
such
limitations,
this
paper
an
Improved
(IMPA)
to
mitigate
above
mentioned
limitations
deploying
self-adaptive
weight
dynamic
social
learning
mechanism
performs
well
challenges
tough
multimodal
benchmark-functions
CEC
2021
benchmark
suite,
compared
with
state-of-the-art
hybrid
algorithms
recently
modified
MPA.
experimental
results
show
IMPA
outperforms
better
precision
attainment
robustness
due
its
enjoying
equalized
exploration
exploitation
feature
over
other
methods.
In
order
provide
a
promising
solution
highlight
potential
useful
tool
This
study
has
implemented
four
highly
representative
problems,
including
Welded
Beam
Design,
Tension/Compression
Spring
Pressure
Vessel
Design
Three
Bar
Design.
also
proved
efficiency
successfully
solve
complex
Scientific Reports,
Journal Year:
2025,
Volume and Issue:
15(1)
Published: Jan. 16, 2025
Unmanned
aerial
vehicle
(UAV)
path
planning
is
a
constrained
multi-objective
optimization
problem.
With
the
increasing
scale
of
UAV
applications,
finding
an
efficient
and
safe
in
complex
real-world
environments
crucial.
However,
existing
particle
swarm
(PSO)
algorithms
struggle
with
these
problems
as
they
fail
to
consider
dynamics,
resulting
many
infeasible
solutions
poor
convergence
optimal
solutions.
To
address
challenges,
we
propose
spherical
vector-based
adaptive
evolutionary
(SAEPSO)
algorithm.
This
algorithm,
based
on
vectors,
directly
incorporates
dynamic
constraints
introduces
improved
tent
map
reverse
learning
enhance
diversity
distribution
initial
Additionally,
nonlinear
factors
are
integrated
balance
exploration
exploitation
capabilities.
avoid
local
optima
highly
environments,
acceleration
strategy
for
particles,
programming
incorporated
further
improve
capability.
Finally,
conducted
comparative
studies
six
benchmark
scenarios
varying
threat
levels,
results
demonstrated
that
proposed
algorithm
outperforms
others
solution
effectiveness,
final
accuracy,
stability,
scalability.
Scientific Reports,
Journal Year:
2024,
Volume and Issue:
14(1)
Published: June 6, 2024
Abstract
The
Marine
Predator
Algorithm
(MPA)
has
unique
advantages
as
an
important
branch
of
population-based
algorithms.
However,
it
emerges
more
disadvantages
gradually,
such
traps
to
local
optima,
insufficient
diversity,
and
premature
convergence,
when
dealing
with
complex
problems
in
practical
industrial
engineering
design
applications.
In
response
these
limitations,
this
paper
proposes
a
novel
Improved
(IMPA).
By
introducing
adaptive
weight
adjustment
strategy
dynamic
social
learning
mechanism,
study
significantly
improves
the
encounter
frequency
efficiency
between
predators
preys
marine
ecosystems.
performance
IMPA
was
evaluated
through
benchmark
functions,
CEC2021
suite
problems,
including
welded
beam
design,
tension/compression
spring
pressure
vessel
three-bar
design.
results
indicate
that
achieved
significant
success
optimization
process
over
other
methods,
exhibiting
excellent
both
solving
optimal
parameter
solutions
optimizing
objective
function
values.
performs
well
terms
accuracy
robustness,
which
also
proves
its
successfully
problems.
