Applied Sciences,
Год журнала:
2024,
Номер
14(17), С. 7803 - 7803
Опубликована: Сен. 3, 2024
This
paper
introduces
a
novel
hybrid
optimization
technique
aimed
at
improving
the
prediction
accuracy
of
solar
photovoltaic
(PV)
outputs
using
an
Improved
Hippopotamus
Optimization
Algorithm
(IHO).
The
IHO
enhances
traditional
(HO)
algorithm
by
addressing
its
limitations
in
search
efficiency,
convergence
speed,
and
global
exploration.
used
Latin
hypercube
sampling
(LHS)
for
population
initialization,
significantly
enhancing
diversity
potential
process.
integration
Jaya
further
refines
solution
quality
accelerates
convergence.
Additionally,
combination
unordered
dimensional
sampling,
random
crossover,
sequential
mutation
is
employed
to
enhance
effectiveness
proposed
demonstrated
through
weights
neuron
thresholds
extreme
learning
machine
(ELM),
model
known
rapid
capabilities
but
often
affected
randomness
initial
parameters.
IHO-optimized
ELM
(IHO-ELM)
tested
against
benchmark
algorithms,
including
BP,
ELM,
HO-ELM,
LCN,
LSTM,
showing
significant
improvements
stability.
Moreover,
IHO-ELM
validated
different
region
assess
generalization
ability
PV
output
prediction.
results
confirm
that
approach
not
only
improves
also
demonstrates
robust
capabilities,
making
it
promising
tool
predictive
modeling
energy
systems.
Knowledge-Based Systems,
Год журнала:
2024,
Номер
295, С. 111737 - 111737
Опубликована: Апрель 12, 2024
This
study
proposes
a
novel
artificial
protozoa
optimizer
(APO)
that
is
inspired
by
in
nature.
The
APO
mimics
the
survival
mechanisms
of
simulating
their
foraging,
dormancy,
and
reproductive
behaviors.
was
mathematically
modeled
implemented
to
perform
optimization
processes
metaheuristic
algorithms.
performance
verified
via
experimental
simulations
compared
with
32
state-of-the-art
Wilcoxon
signed-rank
test
performed
for
pairwise
comparisons
proposed
algorithms,
Friedman
used
multiple
comparisons.
First,
tested
using
12
functions
2022
IEEE
Congress
on
Evolutionary
Computation
benchmark.
Considering
practicality,
solve
five
popular
engineering
design
problems
continuous
space
constraints.
Moreover,
applied
multilevel
image
segmentation
task
discrete
experiments
confirmed
could
provide
highly
competitive
results
problems.
source
codes
Artificial
Protozoa
Optimizer
are
publicly
available
at
https://seyedalimirjalili.com/projects
https://ww2.mathworks.cn/matlabcentral/fileexchange/162656-artificial-protozoa-optimizer.
PLoS ONE,
Год журнала:
2024,
Номер
19(8), С. e0308474 - e0308474
Опубликована: Авг. 19, 2024
This
research
article
presents
the
Multi-Objective
Hippopotamus
Optimizer
(MOHO),
a
unique
approach
that
excels
in
tackling
complex
structural
optimization
problems.
The
(HO)
is
novel
meta-heuristic
methodology
draws
inspiration
from
natural
behaviour
of
hippos.
HO
built
upon
trinary-phase
model
incorporates
mathematical
representations
crucial
aspects
Hippo's
behaviour,
including
their
movements
aquatic
environments,
defense
mechanisms
against
predators,
and
avoidance
strategies.
conceptual
framework
forms
basis
for
developing
multi-objective
(MO)
variant
MOHO,
which
was
applied
to
optimize
five
well-known
truss
structures.
Balancing
safety
precautions
size
constraints
concerning
stresses
on
individual
sections
constituent
parts,
these
problems
also
involved
competing
objectives,
such
as
reducing
weight
structure
maximum
nodal
displacement.
findings
six
popular
methods
were
used
compare
results.
Four
industry-standard
performance
measures
this
comparison
qualitative
examination
finest
Pareto-front
plots
generated
by
each
algorithm.
average
values
obtained
Friedman
rank
test
analysis
unequivocally
showed
MOHO
outperformed
other
resolving
significant
quickly.
In
addition
finding
preserving
more
Pareto-optimal
sets,
recommended
algorithm
produced
excellent
convergence
variance
objective
decision
fields.
demonstrated
its
potential
navigating
objectives
through
diversity
analysis.
Additionally,
swarm
effectively
visualize
MOHO's
solution
distribution
across
iterations,
highlighting
superior
behaviour.
Consequently,
exhibits
promise
valuable
method
issues.
Cluster Computing,
Год журнала:
2024,
Номер
27(8), С. 10589 - 10631
Опубликована: Май 8, 2024
Abstract
This
study
introduces
the
Multi-objective
Generalized
Normal
Distribution
Optimization
(MOGNDO)
algorithm,
an
advancement
of
(GNDO)
now
adapted
for
multi-objective
optimization
tasks.
The
GNDO
previously
known
its
effectiveness
in
single-objective
optimization,
has
been
enhanced
with
two
key
features
optimization.
first
is
addition
archival
mechanism
to
store
non-dominated
Pareto
optimal
solutions,
ensuring
a
detailed
record
best
outcomes.
second
enhancement
new
leader
selection
mechanism,
designed
strategically
identify
and
select
solutions
from
archive
guide
process.
positions
MOGNDO
as
cutting-edge
solution
setting
benchmark
evaluating
performance
against
leading
algorithms
field.
algorithm's
rigorously
tested
across
35
varied
case
studies,
encompassing
both
mathematical
engineering
challenges,
benchmarked
prominent
like
MOPSO,
MOGWO,
MOHHO,
MSSA,
MOALO,
MOMVO,
MOAOS.
Utilizing
metrics
such
Generational
Distance
(GD),
Inverted
(IGD),
Maximum
Spread
(MS),
underscores
MOGNDO's
ability
produce
fronts
high
quality,
marked
by
exceptional
precision
diversity.
results
affirm
superior
versatility,
not
only
theoretical
tests
but
also
addressing
complex
real-world
problems,
showcasing
convergence
coverage
capabilities.
source
codes
algorithm
are
publicly
available
at
https://nimakhodadadi.com/algorithms-%2B-codes
.
