Research Square (Research Square),
Год журнала:
2023,
Номер
unknown
Опубликована: Дек. 6, 2023
Abstract
Accurate
measurement
of
cable
tension
is
crucial
for
real-time
monitoring
bridge
systems,
preventing
potential
risks,
and
ensuring
safety
continuous
operation.
However,
traditional
often
faces
the
challenge
accuracy
when
dealing
with
complex
elastic
boundary
conditions.
This
article
uses
9
finite
element
model
suspension
cables
conditions
as
data
force
identification,
heuristic
algorithms
to
achieve
identification
goal
minimizing
frequency
actual
frequency.
Based
on
recognition
results
process,
reasons
inaccurate
forces
under
boundaries
were
analyzed,
a
mutual
fusion
mechanism
was
proposed
improve
identification.
The
show
that
reduces
maximum
relative
error
in
by
12.6%,
significantly
improving
accuracy,
most
initial
5%,
meeting
needs
practical
engineering.
In
addition,
non
parametric
test
statistical
method
also
proves
introduction
has
significant
impact
value
tension.
Finally,
verified
through
from
three
engineering
meet
requirements.
provides
new
technical
solution
intelligent
accurate
long
beams,
broad
application
prospects.
Facta Universitatis Series Mechanical Engineering,
Год журнала:
2023,
Номер
21(3), С. 529 - 529
Опубликована: Окт. 31, 2023
The
quadratic
assignment
problem
(QAP)
is
an
NP-hard
with
a
wide
range
of
applications
in
many
real-world
applications.
This
study
introduces
discrete
rat
swarm
optimizer
(DRSO)algorithm
for
the
first
time
as
solution
to
QAP
and
demonstrates
its
effectiveness
terms
quality
computational
efficiency.
To
address
combinatorial
nature
QAP,
mapping
strategy
introduced
convert
real
values
into
values,
mathematical
operators
are
redefined
make
then
suitable
problems.
Additionally,
improvement
based
on
local
search
heuristics
such
2-opt
3-opt
proposed.
Simulations
test
instances
from
QAPLIB
library
validate
DRSO
algorithm,
statistical
analysis
using
Wilcoxon
parametric
confirms
performance.
Comparative
other
algorithms
superior
performance
quality,
convergence
speed,
deviation
best-known
making
it
promising
approach
solving
QAP.
Scientific Reports,
Год журнала:
2024,
Номер
14(1)
Опубликована: Июнь 11, 2024
In
recent
years,
many
researchers
have
made
a
continuous
effort
to
develop
new
and
efficient
meta-heuristic
algorithms
address
complex
problems.
Hence,
in
this
study,
novel
human-based
algorithm,
namely,
the
learning
cooking
algorithm
(LCA),
is
proposed
that
mimics
activity
of
humans
order
solve
challenging
The
LCA
strategy
primarily
motivated
by
observing
how
mothers
children
prepare
food.
fundamental
idea
mathematically
designed
two
phases:
(i)
learn
from
their
(ii)
chef.
performance
evaluated
on
51
different
benchmark
functions
(which
includes
first
23
CEC
2005
functions)
2019
compared
with
state-of-the-art
algorithms.
simulation
results
statistical
analysis
such
as
t-test,
Wilcoxon
rank-sum
test,
Friedman
test
reveal
may
effectively
optimization
problems
maintaining
proper
balance
between
exploitation
exploration.
Furthermore,
has
been
employed
seven
real-world
engineering
problems,
tension/compression
spring
design,
pressure
vessel
design
problem,
welded
beam
speed
reducer
gear
train
three-bar
truss
cantilever
problem.
demonstrate
LCA's
superiority
capability
over
other
solving
Processes,
Год журнала:
2024,
Номер
12(2), С. 400 - 400
Опубликована: Фев. 17, 2024
Particle
swarm
optimization
(PSO)
has
been
extensively
used
to
solve
practical
engineering
problems,
due
its
efficient
performance.
Although
PSO
is
simple
and
efficient,
it
still
the
problem
of
premature
convergence.
In
order
address
this
shortcoming,
an
adaptive
particle
with
state-based
learning
strategy
(APSO-SL)
put
forward.
APSO-SL,
population
distribution
evaluation
mechanism
(PDEM)
evaluate
state
whole
population.
contrast
using
iterations
just
state,
spatial
more
intuitive
accurate.
PDEM,
center
position
best
for
calculation
are
calculation,
greatly
reducing
algorithm’s
computational
complexity.
addition,
(ALS)
proposed
avoid
population’s
ALS,
different
strategies
adopted
according
ensure
diversity.
The
performance
APSO-SL
evaluated
on
CEC2013
CEC2017
test
suites,
one
problem.
Experimental
results
show
that
compared
other
competitive
variants.
Scientific Reports,
Год журнала:
2024,
Номер
14(1)
Опубликована: Март 15, 2024
In
order
to
improve
the
accuracy
of
concrete
dynamic
principal
identification,
a
identification
model
based
on
Improved
Dung
Beetle
Algorithm
(IDBO)
optimized
Long
Short-Term
Memory
(LSTM)
network
is
proposed.
Firstly,
apparent
stress-strain
curves
containing
damage
evolution
were
measured
by
Split
Hopkinson
Pressure
Bar
(SHPB)
test
decouple
and
separate
rheology,
this
system
was
modeled
using
LSTM
network.
Secondly,
for
problem
low
convergence
easy
fall
into
local
optimum
(DBO),
greedy
lens
imaging
reverse
learning
initialization
population
strategy,
embedded
curve
adaptive
weighting
factor
PID
control
optimal
solution
perturbation
strategy
are
introduced,
superiority
IDBO
algorithm
proved
through
comparison
optimization
with
DBO,
Harris
Hawk
Optimization
Algorithm,
Gray
Wolf
Fruit
Fly
combination
built
construct
IDBO-LSTM
homeostasis
model.
The
final
results
show
that
can
recognize
material
without
considering
damage;
in
case
damage,
prediction
basically
match
SHPB
curves,
which
proves
feasibility
excellence
proposed
method.
