Telecommunications and Radio Engineering,
Год журнала:
2024,
Номер
83(6), С. 1 - 22
Опубликована: Янв. 1, 2024
In
order
to
deeply
study
the
characteristics
and
laws
of
downlink
beam
in
satellite
in-orbit
motion,
this
paper
proposes
a
multi-attitude
disturbing
force-GNNRL
(DF-GNNRL)
simulation
technique
based
on
calculation.
The
first
studies
motion
law
calculation
parameters
under
adjustment,
establishes
mathematical
model
with
corresponding
arithmetic
power
basis,
completes
data
flow
analysis
multiple
models
process,
further
designs
coding
algorithm,
realizes
real-time
multi-terminal
storage,
builds
DFGNNRL
digital
twin,
finally
DF-GNNRL
twin
by
means
information
technology.
application
technology
can
provide
strong
basis
support
for
virtual
conditions.
IEEE Transactions on Intelligent Transportation Systems,
Год журнала:
2024,
Номер
25(6), С. 4754 - 4772
Опубликована: Янв. 2, 2024
This
paper
provides
a
systematic
overview
of
machine
learning
methods
applied
to
solve
NP-hard
Vehicle
Routing
Problems
(VRPs).
Recently,
there
has
been
great
interest
from
both
the
and
operations
research
communities
in
solving
VRPs
either
through
pure
or
by
combining
them
with
traditional
handcrafted
heuristics.
We
present
taxonomy
studies
on
paradigms,
solution
structures,
underlying
models,
algorithms.
Detailed
results
state-of-the-art
are
presented,
demonstrating
their
competitiveness
approaches.
The
survey
highlights
advantages
learning-based
models
that
aim
exploit
symmetry
VRP
solutions.
outlines
future
directions
incorporate
solutions
address
challenges
modern
transportation
systems.
IEEE Access,
Год журнала:
2024,
Номер
12, С. 14703 - 14718
Опубликована: Янв. 1, 2024
Job-shop
scheduling
problem
(JSP)
is
a
mathematical
optimization
widely
used
in
industries
like
manufacturing,
and
flexible
JSP
(FJSP)
also
common
variant.
Since
they
are
NP-hard,
it
intractable
to
find
the
optimal
solution
for
all
cases
within
reasonable
times.
Thus,
becomes
important
develop
efficient
heuristics
solve
JSP/FJSP.
A
kind
of
method
solving
problems
construction
heuristics,
which
constructs
solutions
via
heuristics.
Recently,
many
methods
leverage
deep
reinforcement
learning
(DRL)
with
graph
neural
networks
(GNN).
In
this
paper,
we
propose
new
approach,
named
residual
scheduling,
remove
irrelevant
machines
jobs
such
as
those
finished,
that
states
include
remaining
(or
relevant)
only.
Our
experiments
show
our
approach
reaches
state-of-the-art
(SOTA)
among
known
on
most
well-known
open
FJSP
benchmarks.
addition,
observe
even
though
model
trained
smaller
sizes,
still
performs
well
large
sizes
terms
makespan.
Interestingly
experiments,
zero
makespan
gap
49
60
instances
whose
job
numbers
more
than
100
15
machines.
Logistics,
Год журнала:
2025,
Номер
9(1), С. 13 - 13
Опубликована: Янв. 16, 2025
Background:
Logistics
operations
are
integral
to
manufacturing
systems,
particularly
in
the
transportation
processes
that
occur
not
only
between
facilities
and
stakeholders
but
also
warehouses
workstations
within
a
facility.
The
design
of
functional
areas
allocating
goods
appropriate
zones
warehouse
management
system
(WMS)
critical
activities
substantially
influence
efficiency
logistics
operations.
Methods:
This
study
develops
mixed-integer
programming
(MIP)
model
optimize
material
flow
product
routing
manufacturing.
identifies
efficient
pathways,
assigns
products
routes,
determines
required
material-handling
equipment.
It
is
implemented
Python
(3.11.5)
using
Pyomo
(6.7.3)
package
CBC
solver
(2.10.11),
with
sensitivity
analysis
performed
on
constraints
decision
variables
evaluate
robustness.
Results:
findings
indicate
Material
Flow
3
Material-Handling
Equipment
1
represent
optimal
configurations
for
managing
majority
system.
Conclusions:
proposed
mathematical
supports
decision-making
process
by
enabling
adjustments
proportions
system,
ensuring
operational
flexibility
response
changing
demands.
Furthermore,
offers
managerial
insights
suggests
directions
future
research.
