Measurement Science and Technology,
Год журнала:
2025,
Номер
36(3), С. 035213 - 035213
Опубликована: Фев. 24, 2025
Abstract
The
dynamic
crack
propagation
characteristics
of
brittle
materials
are
a
crucial
focus
in
impact
dynamics.
randomness
growth
under
loading
poses
significant
challenges
to
accurately
monitoring
behavior
during
experiments.
In
this
study,
two
novel
refinement
measurement
techniques
based
on
optical
frequency
domain
reflectometry
technology
were
developed
provide
more
comprehensive
strain
distribution
data
and
capture
detailed
behavior.
Additionally,
trending
algorithm
was
introduced
process
large
datasets
pinpoint
locations.
study
begins
by
presenting
the
innovative
deployment
methods
fabrication
custom
specimens.
After
conducting
tests
specimens,
resulting
compared
with
actual
propagation.
results
demonstrated
precise
location
identification
down
millimeter
level,
extension
path
mapped
using
profiles
from
fiber-optic
measurements.
These
findings
validate
effectiveness
proposed
approach,
which
shows
great
potential
for
monitoring,
locating,
quantifying
cracks
distributed
fibers.
Credit
card
usage
is
a
vital
component
of
the
global
economy,
but
unpredictable
customer
behavior
poses
significant
challenges.
Machine
learning
(ML)
has
emerged
as
powerful
tool
for
segmentation
in
credit
industry.
This
paper
systematically
examines
different
clustering
algorith
Sensors,
Год журнала:
2024,
Номер
24(7), С. 2200 - 2200
Опубликована: Март 29, 2024
The
constant
monitoring
and
control
of
various
health,
infrastructure,
natural
factors
have
led
to
the
design
development
technological
devices
in
a
wide
range
fields.
This
has
resulted
creation
different
types
sensors
that
can
be
used
monitor
environments,
such
as
fire,
water,
temperature,
movement,
among
others.
These
detect
anomalies
input
data
system,
allowing
alerts
generated
for
early
risk
detection.
advancement
artificial
intelligence
improved
sensor
systems
networks,
resulting
with
better
performance
more
precise
results
by
incorporating
features.
aim
this
work
is
conduct
bibliometric
analysis
using
PRISMA
2020
set
identify
research
trends
machine
learning
applications
fiber
optic
sensors.
methodology
facilitates
dataset
comprised
documents
obtained
from
Scopus
Web
Science
databases.
It
enables
evaluation
both
quantity
quality
publications
study
area
based
on
specific
criteria,
trends,
key
concepts,
advances
concepts
over
time.
found
deep
techniques
Bragg
gratings
been
extensively
researched
focus
structural
health
future
research.
One
main
limitations
lack
use
novel
materials,
graphite,
designing
presents
an
opportunity
studies.
Heliyon,
Год журнала:
2024,
Номер
10(5), С. e26415 - e26415
Опубликована: Фев. 18, 2024
Skin
cancer
is
a
prevalent
form
of
that
necessitates
prompt
and
precise
detection.
However,
current
diagnostic
methods
for
skin
are
either
invasive,
time-consuming,
or
unreliable.
Consequently,
there
demand
an
innovative
efficient
approach
to
diagnose
utilizes
non-invasive
automated
techniques.
In
this
study,
unique
method
has
been
proposed
diagnosing
by
employing
Xception
neural
network
optimized
using
Boosted
Dipper
Throated
Optimization
(BDTO)
algorithm.
The
deep
learning
model
capable
extracting
high-level
features
from
dermoscopy
images,
while
the
BDTO
algorithm
bio-inspired
optimization
technique
can
determine
optimal
parameters
weights
network.
To
enhance
quality
diversity
ISIC
dataset
utilized,
widely
accepted
benchmark
system
diagnosis,
various
image
preprocessing
data
augmentation
techniques
were
implemented.
By
comparing
with
several
contemporary
approaches,
it
demonstrated
outperforms
others
in
detecting
cancer.
achieves
average
precision
94.936%,
accuracy
94.206%,
recall
97.092%
surpassing
performance
alternative
methods.
Additionally,
5-fold
ROC
curve
error
have
presented
validation
showcase
superiority
robustness
method.
