As
the
product
mix
in
production
changes
dramatically,
cell
reconfiguration
is
requisite
to
smoothen
process.
Meanwhile,
multiple
modes
are
simultaneously
adopted
site
promote
productivity
and
assure
flexibility,
thus
coordination
scheduling
among
them
becomes
a
challenging
problem.
To
achieve
cellular
manufacturing
system
which
no-idle
flow-line
flexible
job-shop
hybridized,
mixed
integer
linear
programming
model
formulated
an
enhanced
adaptive
multi-objective
evolutionary
algorithm
developed.
In
proposed
algorithm,
decision
tree-based
rule
selector
developed
select
most
appropriate
combination
given
scenario
hence
generate
high-quality
initial
population.
Three
crossover
operators
six
objective-oriented
local
search
designed
utilized
increase
exploration
exploitation
capability.
An
balance
mechanism
of
trained
by
Q-learning
maximize
efficiency.
addition,
adjustment
population
size
ensure
diversity
speed
up
convergence.
The
comparative
study
demonstrates
that
three
mechanisms
effective
with
significantly
outperforms
other
comparison
algorithms
solving
studied
IEEE Transactions on Intelligent Transportation Systems,
Год журнала:
2024,
Номер
25(11), С. 18078 - 18092
Опубликована: Июль 1, 2024
A
voyage
optimization
algorithm
is
an
essential
component
in
a
ship's
routing
concerning
safety,
energy
efficiency,
arrival
punctuality,
etc.
In
this
study,
predictive
integrated
with
Isochrone-based
for
energy-efficient
sailing.
Different
waypoints
generation
and
grid
partition
strategies
search
spaces
are
proposed
to
achieve
smooth
convergence
toward
the
destination,
costs
ahead
of
current
sailing
time
stages
estimated
cost
function
avoid
local
suboptimization.
Based
on
these
measures,
paper
introduces
(IPO)
method
that
can
enhanced
robust
performance
real-time
multi-objective
optimization.
The
unrealistic
routes
abrupt
turns
occur
traditional
Isochrone
graph
methods
avoided.
IPO
suggest
diverse
environments,
while
complying
punctuality
requirements
planning.
Meanwhile,
it
requires
few
computational
resources
enable
online
adjustment
during
execution,
adapting
dynamic
environments.
Its
efficiency
effectiveness
demonstrated
by
six
case
study
voyages
from
chemical
tanker
full-scale
measurements,
further
compared
other
widely
used
methods.
results
show
provide
subtle
5%
fuel
reduction
average
all
voyages,
around
40
seconds
runtime.
Frontiers in Industrial Engineering,
Год журнала:
2025,
Номер
3
Опубликована: Янв. 27, 2025
The
advent
of
Industry
4.0
and
the
emerging
5.0
have
fundamentally
transformed
manufacturing
systems,
introducing
unprecedented
levels
complexity
in
production
scheduling.
This
is
further
amplified
by
integration
cyber-physical
Internet
Things,
Artificial
Intelligence,
human-centric
approaches,
necessitating
more
sophisticated
optimization
methods.
paper
aims
to
provide
a
comprehensive
perspective
on
application
metaheuristic
algorithms
shop
scheduling
problems
within
context
5.0.
Through
systematic
review
recent
literature
(2015–2024),
we
analyze
categorize
various
including
Evolutionary
Algorithms
(EAs),
swarm
intelligence,
hybrid
methods,
that
been
applied
address
complex
challenges
smart
environments.
We
specifically
examine
how
these
handle
multiple
competing
objectives
such
as
makespan
minimization,
energy
efficiency,
costs,
human-machine
collaboration,
which
are
crucial
modern
industrial
settings.
Our
survey
reveals
several
key
findings:
1)
metaheuristics
demonstrate
superior
performance
handling
multi-objective
compared
standalone
algorithms;
2)
bio-inspired
show
promising
results
addressing
environments;
3)
tri-objective
higher-order
warrant
in-depth
exploration;
4)
there
an
trend
towards
incorporating
human
factors
sustainability
optimization,
aligned
with
principles.
Additionally,
identify
research
gaps
propose
future
directions,
particularly
areas
real-time
adaptation,
sustainability-aware
algorithms.
provides
insights
for
researchers
practitioners
field
scheduling,
offering
structured
understanding
current
methodologies
evolution
from
