Electronics,
Journal Year:
2024,
Volume and Issue:
13(14), P. 2844 - 2844
Published: July 19, 2024
Although
the
NSGA-III
algorithm
is
able
to
find
global
optimal
solution
and
has
a
good
effect
on
workshop
scheduling
optimization,
limitations
in
population
diversity,
convergence
ability
local
solutions
make
it
not
applicable
certain
situations.
Thus,
an
improved
optimization
proposed
this
work.
It
aims
address
these
of
processing
optimization.
To
solve
problem
individual
elimination
traditional
algorithm,
chaotic
mapping
introduced
generate
new
offspring
individuals
add
selected
winning
as
parent
for
next
iteration.
The
was
applied
pressure
sensor
calibration
workshop.
A
comparison
with
conducted
through
simulation
analysis.
results
show
that
can
obtain
better
performance,
improve
avoid
falling
into
solutions.
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.
Electronics,
Journal Year:
2024,
Volume and Issue:
13(18), P. 3696 - 3696
Published: Sept. 18, 2024
The
flexible
job
shop
scheduling
problem
(FJSSP),
which
can
significantly
enhance
production
efficiency,
is
a
mathematical
optimization
widely
applied
in
modern
manufacturing
industries.
However,
due
to
its
NP-hard
nature,
finding
an
optimal
solution
for
all
scenarios
within
reasonable
time
frame
faces
serious
challenges.
This
paper
proposes
that
transforms
the
FJSSP
into
Markov
Decision
Process
(MDP)
and
employs
deep
reinforcement
learning
(DRL)
techniques
resolution.
First,
we
represent
state
features
of
environment
using
seven
feature
vectors
utilize
transformer
encoder
as
extraction
module
effectively
capture
relationships
between
representation
capability.
Second,
based
on
jobs
machines,
design
16
composite
dispatching
rules
from
multiple
dimensions,
including
completion
rate,
processing
time,
waiting
resource
utilization,
achieve
efficient
decisions.
Furthermore,
project
intuitive
dense
reward
function
with
objective
minimizing
total
idle
machines.
Finally,
verify
performance
feasibility
algorithm,
evaluate
proposed
policy
model
Brandimarte,
Hurink,
Dauzere
datasets.
Our
experimental
results
demonstrate
framework
consistently
outperforms
traditional
rules,
surpasses
metaheuristic
methods
larger-scale
instances,
exceeds
existing
DRL-based
across
most
