Complex
heavy
equipment
manufacturing
involves
multiple
stakeholders
and
requires
a
platform
for
interconnection
collaborative
manufacturing.
Networked
under
cloud
environment
have
emerged
as
solution
to
enhance
efficiency
innovation
in
industries.
However,
attracting
users
join
such
platforms
addressing
strategic
tactical
decision-making
challenges
are
critical
concerns.
To
address
these
concerns,
we
develop
an
evolutionary
game
model
investigate
the
user
acquisition
strategy
quality
decisions.
We
obtain
two
ideal
equilibrium
points:
First,
when
offers
high-quality
services,
find
designers
still
actively
even
if
they
face
higher
risk
of
knowledge
loss
lower
synergy
benefits;
manufacturers
more
likely
get
benefits
or
suffer
prospective
profit-and-loss,
pay
service
fee.
Second,
provides
low-quality
that
choose
because
low
commission
fee
by
high
offer
fair
distribution.
then
further
impact
government
subsidies
subsidy
policy
can
effectively
encourage
provide
but
it
should
be
noted
charged
is
too
high,
may
hinder
from
joining.
also
discuss
managerial
implications
our
results.
IEEE Internet of Things Journal,
Journal Year:
2024,
Volume and Issue:
11(14), P. 25382 - 25393
Published: April 30, 2024
Industrial
Internet-of-things
(IIoT)
is
considered
an
emerging
infrastructure
for
enhancing
manufacturing
efficiency
by
facilitating
the
sharing
of
resources
across
multiple
factories.
With
increasing
requirements
on
customized
production
in
IIoT,
tasks
and
objectives
flexible
job
shop
scheduling
problem
different
orders
vary
greatly,
leading
to
repetitive
algorithm
adjustment
time-consuming
solver
invocation.
To
accelerate
scenarios,
this
paper
proposed
a
transfer
optimization
method
based
pointer
networks
improved
Transformer
(Trans-Ptr-Nets).
A
historical
solution
selection
strategy
accompanied
with
dataset
retrieve
solutions
similar
current
scenario
are
established.
Then,
Trans-Ptr-Nets
designed
candidate
new
that
feasible
target
scenario.
Subsequently,
introduced
as
additional
population
evolutionary
process.
Experimental
results
conducted
four
scenarios
show
can
realize
at
most
50%
reduction
running
time
while
improving
quality
least
10%,
compared
six
typical
algorithms
three
learning
networks.