Frontiers in Sustainable Development,
Journal Year:
2025,
Volume and Issue:
5(3), P. 227 - 237
Published: March 22, 2025
Currently,
automobile
production
in
workshops
faces
demands
for
multi-variety,
small-batch,
and
rapid
delivery.
As
a
key
auxiliary
link,
optimizing
the
performance
of
workshop
material
scheduling
system
can
enhance
efficiency
economic
benefits.
With
expansion
enterprise
scale
complexity
requirements,
multi-AGV
handling
systems
have
become
an
effective
solution
to
optimize
processes
save
costs
due
their
parallel
collaboration
advantages.
However,
NP-hard
nature
this
problem,
traditional
exact
algorithms
often
perform
poorly
when
dealing
with
complex
large-scale
problems.
Therefore,
paper
explores
applications
intelligent
such
as
genetic
algorithms,
artificial
neural
networks,
particle
swarm
optimization,
proposes
novel
efficient
solutions
methods
mixed-model
assembly
workshops.
In
addition,
address
problem
large
state
space
schemes,
also
discusses
potential
emerging
technologies
reinforcement
learning.
Through
these
studies,
it
aims
processes,
reduce
costs,
promote
development
manufacturing
industry.
IEEE Access,
Journal Year:
2024,
Volume and Issue:
12, P. 50935 - 50948
Published: Jan. 1, 2024
In
addressing
the
Flexible
Job
Shop
Scheduling
Problem
(FJSP),
deep
reinforcement
learning
eliminates
need
for
mathematical
modeling
of
problem,
requiring
only
interaction
with
real
environment
to
learn
effective
strategies.
Using
disjunctive
graphs
as
state
representation
has
proven
be
a
particularly
method.
Additionally,
attention
mechanisms
enable
rapid
focus
on
relevant
features.
However,
due
unique
structure
mechanisms,
current
methods
fail
provide
strategies
after
changes
in
scale.
To
resolve
this
issue,
we
propose
an
end-to-end
framework
FJSP.
Initially,
introduce
lightweight
model,
Graph
Gated
Channel
Transformation
(GGCT),
identify
characteristics
workpieces
being
scheduled
at
decision-making
moment,
while
suppressing
redundant
Subsequently,
address
inability
scale,
modify
expression
graph
features,
channeling
global
features
into
different
channels
capture
information
moment
effectively.
Comparative
analysis
generated
and
classical
datasets
shows
our
model
reduces
average
makespan
significantly,
from
8.243%
7.037%
10.08%
8.69%,
respectively.
Systems,
Journal Year:
2023,
Volume and Issue:
11(2), P. 103 - 103
Published: Feb. 13, 2023
Emergencies
such
as
machine
breakdowns
and
rush
orders
greatly
affect
the
production
activities
of
manufacturing
enterprises.
How
to
deal
with
rescheduling
problem
after
emergencies
have
high
practical
value.
Meanwhile,
under
background
intelligent
manufacturing,
automatic
guided
vehicles
are
gradually
emerging
in
To
disturbances
flexible
job
shop
scheduling
vehicle
transportation,
a
mixed-integer
linear
programming
model
is
established.
According
traits
this
model,
an
improved
NSGA-II
designed,
aiming
at
minimizing
makespan,
energy
consumption
workload
deviation.
improve
solution
qualities,
local
search
operator
based
on
critical
path
designed.
In
addition,
crowding
distance
calculation
method
used
reduce
computation
complexity
algorithm.
Finally,
validity
improvement
strategies
tested,
robustness
superiority
proposed
algorithm
verified
by
comparing
it
NSGA,
SPEA2.
Mathematics,
Journal Year:
2025,
Volume and Issue:
13(4), P. 545 - 545
Published: Feb. 7, 2025
This
paper
tackles
the
green
permutation
flow
shop
scheduling
problem
(GPFSP)
with
goal
of
minimizing
both
maximum
completion
time
and
energy
consumption.
It
introduces
a
novel
hybrid
approach
that
combines
end-to-end
deep
reinforcement
learning
an
improved
genetic
algorithm.
Firstly,
PFSP
is
modeled
using
(DRL)
approach,
named
PFSP_NET,
which
designed
based
on
characteristics
PFSP,
actor–critic
algorithm
employed
to
train
model.
Once
trained,
this
model
can
quickly
directly
produce
relatively
high-quality
solutions.
Secondly,
further
enhance
quality
solutions,
outputs
from
PFSP_NET
are
used
as
initial
population
for
(IGA).
Building
upon
traditional
algorithm,
IGA
utilizes
three
crossover
operators,
four
mutation
incorporates
hamming
distance,
effectively
preventing
prematurely
converging
local
optimal
Then,
optimize
consumption,
energy-saving
strategy
proposed
reasonably
adjusts
job
order
by
shifting
jobs
backward
without
increasing
time.
Finally,
extensive
experimental
validation
conducted
120
test
instances
Taillard
standard
dataset.
By
comparing
method
algorithms
such
(SGA),
elite
(EGA),
(HGA),
discrete
self-organizing
migrating
(DSOMA),
water
wave
optimization
(DWWO),
monkey
search
(HMSA),
results
demonstrate
effectiveness
method.
Optimal
solutions
achieved
in
28
instances,
latest
updated
Ta005
Ta068
values
1235
5101,
respectively.
Additionally,
experiments
30
including
20-10,
50-10,
100-10,
indicate
reduce
