IET Intelligent Transport Systems,
Год журнала:
2025,
Номер
19(1)
Опубликована: Янв. 1, 2025
ABSTRACT
A
novel
scheme
is
proposed
for
the
distributed
multi‐ship
collision
avoidance
(CA)
problem
with
consideration
of
autonomous,
dynamic
nature
real
circumstance.
All
ships
in
envisioned
scenarios
can
share
their
decisions
or
intentions
through
route
exchange,
allowing
them
to
make
subsequent
based
on
planning
each
iteration.
By
leveraging
CA
involves
iterations
negotiation,
and
regarded
as
a
staged
cooperative
game
under
conditions
complete
information.
The
concept
closest
spatio‐temporal
distance
(CSTD)
introduced
more
accurately
assess
risk
between
ships.
coordinated
mechanism
established
when
identified,
which
further
incorporates
considerations
including
stand‐on/give‐way
relationships,
negotiation
rounds,
re‐planning
calculation,
well
cost
factor
evaluation.
Nash
bargaining
solution
(NBS)
elaborated
achieve
Pareto‐optimal
routes
scenarios.
In
model,
while
individual
interest
ship
are
maximized,
economic
fairness
global
optimization
overall
system
also
maintained.
Simulation
results
indicate
that
NBS
shows
good
flexibility
adaptability,
all
comply
solution,
bring
out
normal
solutions
within
limited
number
iterations.
JOIV International Journal on Informatics Visualization,
Год журнала:
2024,
Номер
8(1), С. 158 - 158
Опубликована: Март 31, 2024
This
review
article
looks
at
the
developing
field
of
artificial
intelligence
and
machine
learning
in
maritime
marine
environment
management.
The
industry
is
increasingly
interested
applying
advanced
AI
ML
technologies
to
solve
sustainability,
efficiency,
regulatory
compliance
issues.
paper
examines
applications
using
a
deep
literature
case
study
analysis.
Modeling
ship
fuel
consumption,
which
impacts
operating
expenses,
top
responsibility.
demonstrates
that
approaches
such
as
Random
Forest
Tweedie
models
can
estimate
use.
Statistical
analysis
model
beats
regarding
accuracy
consistency.
For
training
testing
datasets,
has
high
R2
values
0.9997
0.9926,
indicating
solid
match.
Low
Root
Mean
Square
Error
(RMSE)
average
absolute
relative
deviation
(AARD)
suggest
accurately
reflects
use
variability.
While
still
performing
well,
lower
higher
RMSE
AARD
values,
suggesting
reduced
precision
consumption
prediction.
These
findings
provide
light
on
potential
Advanced
analytics
enables
decision-makers
analyze
patterns
better,
increase
operational
decrease
environmental
impact,
thus
improving
sustainability.
Journal of Marine Science and Engineering,
Год журнала:
2024,
Номер
12(8), С. 1351 - 1351
Опубликована: Авг. 8, 2024
The
accurate
prediction
of
vessel
trajectory
is
crucial
importance
in
order
to
improve
navigational
efficiency,
optimize
routes,
enhance
the
effectiveness
search
and
rescue
operations
at
sea,
ensure
maritime
safety.
However,
spatial
interaction
among
vessels
can
have
a
certain
impact
on
accuracy
models.
To
overcome
such
problem
predicting
trajectory,
this
research
proposes
novel
hybrid
methodology
incorporating
graph
attention
network
(GAT)
long
short-term
memory
(LSTM).
proposed
GAT-LSTM
model
comprehensively
consider
spatio-temporal
features
process,
which
expected
significantly
robustness
prediction.
Automatic
Identification
System
(AIS)
data
from
surrounding
waters
Xiamen
Port
collected
utilized
as
empirical
case
for
validation.
experimental
results
demonstrate
that
outperforms
best
baseline
terms
reduction
average
displacement
error
final
error,
are
44.52%
56.20%,
respectively.
These
improvements
will
translate
into
more
trajectories,
helping
minimize
route
deviations
collision
avoidance
systems,
so
effectively
provide
support
warning
about
potential
collisions
reducing
risk
accidents.
Ocean Engineering,
Год журнала:
2023,
Номер
286, С. 115687 - 115687
Опубликована: Авг. 31, 2023
Accurate
vessel
traffic
flow
(VTF)
prediction
can
enhance
navigation
safety
and
economic
efficiency.
To
address
the
challenge
of
inherently
complex
dynamic
growth
VTF
time
series,
a
new
hierarchical
methodology
for
is
proposed.
Firstly,
original
data
reconfigured
as
three-dimensional
tensor
by
modified
Bayesian
Gaussian
CANDECOMP/PARAFAC
(BGCP)
decomposition
model.
Secondly,
matrix
(hour
✕
day)
each
week
decomposed
into
high-
low-frequency
matrices
using
Bidimensional
Empirical
Mode
Decomposition
(BEMD)
model
to
non-stationary
signals
affecting
results.
Thirdly,
self-similarities
between
within
high-frequency
are
utilised
rearrange
different
one-dimensional
series
solve
weak
mathematical
regularity
in
matrix.
Then,
Dynamic
Time
Warping
(DTW)
employed
identify
grouped
segments
with
high
similarities
generate
more
suitable
tensors.
The
experimental
results
verify
that
proposed
outperforms
state-of-the-art
methods
real
Automatic
Identification
System
(AIS)
datasets
collected
from
two
areas.
potentially
optimise
relation
operations
manage
traffic,
benefiting
stakeholders
such
port
authorities,
ship
operators,
freight
forwarders.