Sensors,
Год журнала:
2024,
Номер
24(16), С. 5310 - 5310
Опубликована: Авг. 16, 2024
The
Industrial
Internet
of
Things
has
enabled
the
integration
and
analysis
vast
volumes
data
across
various
industries,
with
maritime
sector
being
no
exception.
Advances
in
cloud
computing
deep
learning
(DL)
are
continuously
reshaping
industry,
particularly
optimizing
operations
such
as
Predictive
Maintenance
(PdM).
In
this
study,
we
propose
a
novel
DL-based
framework
focusing
on
fault
detection
task
PdM
marine
operations,
leveraging
time-series
from
sensors
installed
shipboard
machinery.
is
designed
scalable
cost-efficient
software
solution,
encompassing
all
stages
collection
pre-processing
at
edge
to
deployment
lifecycle
management
DL
models.
proposed
architecture
utilizes
Graph
Attention
Networks
(GATs)
extract
spatio-temporal
information
provides
explainable
predictions
through
feature-wise
scoring
mechanism.
Additionally,
custom
evaluation
metric
real-world
applicability
employed,
prioritizing
both
prediction
accuracy
timeliness
identification.
To
demonstrate
effectiveness
our
framework,
conduct
experiments
three
types
open-source
datasets
relevant
PdM:
electrical
data,
bearing
datasets,
water
circulation
experiments.
Ocean & Coastal Management,
Год журнала:
2024,
Номер
251, С. 107077 - 107077
Опубликована: Март 6, 2024
The
complex
traffic
situations
are
among
the
factors
influencing
maritime
safety.
They
can
be
quantitatively
estimated
through
analysis
of
data.
This
paper
explores
impact
on
safety,
focusing
inland
waterway
traffic.
It
presents
a
big
data
analytics
method,
utilizing
from
Automatic
Identification
System
(AIS)
and
historical
accident
records.
methodology
involves
AIS
preprocessing
spatial
autocorrelation
models,
including
Moran's
index,
to
extract
evaluate
dynamic
characteristics
characteristic
includes
thorough
investigation
into
spatial-temporal
distribution
ship
average
speed
trajectory
density.
then
introduces
an
effective
model
that
evaluates
relationship
between
patterns
accidents.
study,
specifically
targeting
Nanjing
section
Yangtze
River,
reveals
variations
in
density
over
time.
identifies
several
hotspots
with
significant
local
correlation
these
factors.
Moreover,
substantial
is
found
locations
accidents
areas
increased
speed.
These
results
may
provide
insights
for
safety
management
highlight
strategies
preventing
Ocean Engineering,
Год журнала:
2024,
Номер
301, С. 117510 - 117510
Опубликована: Март 21, 2024
Monitoring
health
condition
of
offshore
jacket
platforms
is
crucial
to
prevent
unexpected
structural
damages,
where
a
prevailing
challenge
involves
translating
available
feature
information
into
damage
patterns.
Although
the
artificial
neural
network
(ANN)
models
are
popular
in
addressing
this
challenge,
they
often
fail
capture
temporal
correlations
between
and
patterns,
which
reduce
their
capability
for
discovering
laws
governing
detection.
To
bridge
research
gap,
study
proposes
novel
ensemble
deep
learning
model
enhance
extraction
improve
pattern
identification.
In
approach,
one-dimensional
Convolutional
Neural
Network
(CNN)
extracts
spatiotemporal
features
from
vibration
measurements.
Simultaneously,
SENet
attention
mechanism
introduced
select
most
informatic
features.
Subsequently,
bidirectional
long
short-term
memory
(BiLSTM)
employed
learn
mapping
extracted
Furthermore,
particle
swarm
optimization
(PSO)
algorithm
used
optimize
BiLSTM
hyperparameters
its
stability
reliability.
Both
simulations
experiments
carried
out
collect
responses
structure
different
scenarios.
The
analysis
results
demonstrate
that
proposed
method
produces
remarkable
improvement
with
respect
accuracy
robustness
identifying
damages
when
compared
ANNs.
overall
detection
CNN-BiLSTM-Attention
beyond
95%,
provides
strong
applicability
practical
monitoring
platforms.
Ocean Engineering,
Год журнала:
2024,
Номер
303, С. 117046 - 117046
Опубликована: Апрель 8, 2024
The
topology
of
planing
hulls
entails
some
the
most
innovative
specifications
found
in
modern
advanced
marine
vehicles.
Planing
hull
designs
can
vary
depending
on
their
intended
use
and
hence
sound
understanding
influence
hydrodynamics
craft
stability
performance
is
key
within
context
design
for
safety
sustainability
requirements.
motions
stepless
or
stepped
surfaces,
be
it
steady
unsteady
are
strongly
coupled
with
nonlinear
fluid
flows.
Consequently,
calm
water
performance,
seakeeping
maneuvering
waves,
idealised
by
a
diverse
array
analytical
simulation-based
models.
In
this
paper,
we
holistically
review
scholarly
studies
subject,
discuss
research
challenges
opportunities
ahead.
A
conclusion
drawn
that,
although
mathematical
models,
especially
ones
that
simulate
motions,
require
further
development
to
account
complexities
operating
real-world
environment,
they
mostly
limited
monohull
without
steps.
It
also
suggested
emergence
new-generation
artificial
intelligence
algorithms
opens
up
new
prospects
hydrodynamic
modelling
as
accounting
dynamic
motion
predictions.
holistic
optimization
hulls,
realm
yet
overlooked
hydrodynamics,
identified
an
important
interesting
future
opportunity.
Pairing
AI
methods
recommended
direction
intelligent
boat
systems.
Engineering Applications of Artificial Intelligence,
Год журнала:
2024,
Номер
133, С. 108172 - 108172
Опубликована: Март 8, 2024
Ship
Time
Headway
(STH)
is
used
in
maritime
navigation
to
describe
the
time
interval
between
arrivals
of
two
consecutive
ships
same
water
area.
This
measurement
may
offer
a
straightforward
way
gauge
frequency
ship
traffic
and
likelihood
congestion
particular
STH
an
important
factor
understanding
managing
dynamics
movements
busy
waterways.
paper
introduces
hybrid
deep
learning
method
for
predicting
domain.
The
integrates
Seasonal-Trend
Decomposition
using
Loess
(STL),
Multi-head
Self-Attention
(MSA)
mechanism
into
Long
Short-Term
Memory
(LSTM)
neural
network.
dataset
was
extracted
from
Automatic
Identification
System
(AIS)
through
trajectory
spatial
motion,
seasonal,
trend
residual
components
decomposition
were
then
determined
STL
algorithms.
MSA-LSTM
adopted
comprehensively
capture
evolving
patterns
sequence.
Comparison
studies
with
existing
methods
demonstrate
accuracy
robustness
predictions
provided
by
this
method,
indicating
that
proposed
outperforms
other
models
terms
prediction
performance
capabilities.
By
STH,
offers
potential
assist
managers
navigators
assessing
flow,
thereby
enabling
them
make
informed
decisions
on
safety
efficiency.
