Annals of Computer Science and Information Systems,
Год журнала:
2023,
Номер
37, С. 157 - 164
Опубликована: Окт. 11, 2023
Applying
population-based
metaheuristics
is
a
known
method
of
solving
difficult
optimization
problems.In
this
paper
the
search
for
best
solution
conducted
by
decentralized,
self-organized
agents,
working
in
parallel
threads,
so
called
mushroom-picking
method.The
enhanced
remembering
which
part
recently
improved
last
successful
change
took
place
and
intensifying
part.A
computational
experiment
shows
that
introducing
component
most
recent
changes
may
improve
results
obtained
model
case
JSSP
problems.
IEEE Access,
Год журнала:
2023,
Номер
11, С. 14636 - 14646
Опубликована: Янв. 1, 2023
In
the
process
control
industry,
it
is
arduous
to
some
integrating
or
unstable
processes
since
they
involve
time
delays
and
have
an
inverse
response.
Conventional
controllers
such
as
PID
cannot
provide
sufficient
performance
alone
in
of
these
systems.
This
article
proposes
a
algorithm
based
on
I-PD-based
Smith
predictor
for
time-delayed
processes.
The
controller
parameters
are
tuned
by
using
Equilibrium
optimizer
(EO)
algorithm,
which
presented
literature
2020,
proposed
approach.
EO
aims
determine
optimal
minimizing
error
signal
multi-objective
function
ITAE
criterion.
Thus,
that
will
set-point
tracking
disturbance
rejection
most
properly
can
be
determined.
Simulation
studies
conducted
different
structures
evaluate
method.
method
compared
with
from
terms
tracking,
parameter
uncertainties,
signals,
rejection.
It
seen
transient
responses
response
improved
Biomimetics,
Год журнала:
2024,
Номер
9(1), С. 31 - 31
Опубликована: Янв. 4, 2024
The
slime
mould
algorithm
(SMA)
is
a
new
swarm
intelligence
inspired
by
the
oscillatory
behavior
of
moulds
during
foraging.
Numerous
researchers
have
widely
applied
SMA
and
its
variants
in
various
domains
field
proved
value
conducting
literatures.
In
this
paper,
comprehensive
review
introduced,
which
based
on
130
articles
obtained
from
Google
Scholar
between
2022
2023.
study,
firstly,
theory
described.
Secondly,
improved
are
provided
categorized
according
to
approach
used
apply
them.
Finally,
we
also
discuss
main
applications
SMA,
such
as
engineering
optimization,
energy
machine
learning,
network,
scheduling
image
segmentation.
This
presents
some
research
suggestions
for
interested
algorithm,
additional
multi-objective
discrete
SMAs
extending
neural
networks
extreme
learning
machining.
Journal of Computational Design and Engineering,
Год журнала:
2024,
Номер
11(4), С. 83 - 108
Опубликована: Июнь 19, 2024
Abstract
In
optimization,
metaheuristic
algorithms
have
received
extensive
attention
and
research
due
to
their
excellent
performance.
The
slime
mould
algorithm
(SMA)
is
a
newly
proposed
algorithm.
It
has
the
characteristics
of
fewer
parameters
strong
optimization
ability.
However,
with
increasing
difficulty
problems,
SMA
some
shortcomings
in
complex
problems.
For
example,
main
concerns
are
low
convergence
accuracy
prematurely
falling
into
local
optimal
solutions.
To
overcome
these
this
paper
developed
variant
called
CCSMA.
an
improved
based
on
horizontal
crossover
(HC)
covariance
matrix
adaptive
evolutionary
strategy
(CMAES).
First,
HC
can
enhance
exploitation
by
crossing
information
between
different
individuals
promote
communication
within
population.
Finally,
CMAES
facilitates
exploration
reach
balanced
state
dynamically
adjusting
size
search
range.
This
benefits
allowing
it
go
beyond
space
explore
other
solutions
better
quality.
verify
superiority
algorithm,
we
select
new
original
as
competitors.
CCSMA
compared
competitors
40
benchmark
functions
IEEE
CEC2017
CEC2020.
results
demonstrate
that
our
work
outperforms
terms
jumping
out
space.
addition,
applied
tackle
three
typical
engineering
These
problems
include
multiple
disk
clutch
brake
design,
pressure
vessel
speed
reducer
design.
showed
achieved
lowest
cost.
also
proves
effective
tool
for
solving
realistic
Mathematics,
Год журнала:
2022,
Номер
10(23), С. 4608 - 4608
Опубликована: Дек. 5, 2022
In
the
job-shop
scheduling
field,
timely
and
proper
updating
of
original
strategy
is
an
effective
way
to
avoid
negative
impact
disturbances
on
manufacturing.
this
paper,
a
pure
reactive
method
for
proposed
deal
with
disturbance
uncertainty
arrival
new
jobs
in
job
shop.
The
implementation
process
as
follows:
combine
data
mining,
discrete
event
simulation,
dispatching
rules
(DRs),
take
makespan
machine
utilization
criteria,
divide
manufacturing
system
production
period
into
multiple
subperiods,
build
dynamic
model
that
assigns
DRs
subscheduling
periods
real-time;
strategies
are
generated
at
beginning
each
subperiod.
experiments
showed
enables
reduction
2–17%
improvement
2–21%.
constructed
can
assign
optimal
DR
subperiod
real-time,
which
realizes
purpose
locally
enhancing
overall
effect
system.
Osteoarthritis
(OA)
of
the
knee
is
a
chronic
state
that
significantly
lowers
quality
life
for
its
patients.
Early
detection
and
lifetime
monitoring
progression
OA
are
necessary
preventive
therapy.
In
course
therapy,
Kellgren
Lawrence
(KL)
assessment
model
categorizes
rigidity
OA.
Deep
techniques
have
recently
been
used
to
increase
precision
effectiveness
severity
assessments.
The
training
process
compromised
by
low-confidence
samples,
which
less
accurate
than
normal
ones.
this
work,
deep
learning-based
osteoarthritis
recommended
accurately
identify
condition
in
phases
designed
data
collection,
feature
extraction,
prediction.
At
first,
images
generally
gathered
from
online
resources.
given
into
extraction
phase.
A
new
implemented
predict
named
Spatial
Separable
Convolution
with
Attention-based
Ensemble
Networks
(SCAENet),
includes
stacked
target-based
pool
generation,
done
using
ResNet,
Visual
Geometry
Group
(VGG16),
DenseNet.
obtained
SCAENet.
Hence,
Hybridization
Equilibrium
Slime
Mould
Bald
Eagle
Search
Optimization
(HESM-BESO).
Here,
osteoarthritis's
prediction
performed
dimensional
convolutional
neural
network
(1DCNN)
technique.
SCAENet
validated
other
conventional
methods
show
high
performance.
