Soil and Sediment Contamination An International Journal,
Год журнала:
2024,
Номер
unknown, С. 1 - 19
Опубликована: Янв. 21, 2024
The
problem
of
soil
heavy
metal
pollution
in
decommissioned
sites
has
become
an
environmental
threat
and
challenge
faced
by
countries
around
the
world.
Establishing
a
high-precision
3D
model
contaminants
is
essential
for
risk
assessment
accurate
monitoring
contaminated
sites.
In
this
study,
geological
SSA-XGBoost
are
proposed
to
predict
concentration
site.
These
models
can
effectively
improve
prediction
accuracy
metals,
RMSE
XGBoost
optimized
SSA
algorithm
reduced
24.3%-34.3%.
Compared
with
other
machine
learning
models,
optimal
performance
improving
metals.
It
suitable
areas
significant
spatial
heterogeneity
Using
model,
distribution
characteristics
metals
determined.
pollutants
ranked
as
As>Pb>Mo,
overall
degree
decreases
gradually
from
top
bottom.
mainly
distributed
production
workshop
area
southwest
site,
miscellaneous
fill
layer
main
that
needs
be
remediated.
Journal of Renewable and Sustainable Energy,
Год журнала:
2024,
Номер
16(5)
Опубликована: Сен. 1, 2024
In
the
era
of
burgeoning
renewable
integration,
shift
toward
low-carbon
energy
hubs
is
a
pivotal
developmental
trajectory.
Amidst
this
paradigm,
operational
challenges
posed
by
inherent
uncertainty
variable
sources,
such
as
wind
and
solar
power,
alongside
stochastic
load
fluctuations,
must
be
reckoned
with.
Herein,
we
present
an
innovative,
economically
viable
strategy
that
embraces
fuzzy
opportunity
constraints,
thereby
accommodating
dual-sided
impact
on
hubs.
First,
advanced
optimization
framework
developed
for
hub
holistically
couples
electricity,
cooling,
gas,
heat
sectors.
Leveraging
conversion
technologies,
it
amplifies
complementary
interaction
among
diverse
sources
implements
integrated
demand
response
model
to
mitigate
variability.
Subsequently,
ladder-type
carbon
trading
green
certificate
mechanisms
are
incorporated,
designed
pare
down
both
emissions
expenditures.
Addressing
unpredictability
grid-connected
resources,
introduces
chance
constraints.
These
transform
rigid
deterministic
system
limitations
into
more
flexible
constraints
encapsulating
variables
employing
trapezoidal
parameters
elucidate
their
nature.
The
robustness
practical
utility
proposed
substantiated
through
meticulous
case
analyses.
Scientific Reports,
Год журнала:
2025,
Номер
15(1)
Опубликована: Март 15, 2025
The
rapid
increase
in
carbon
emissions
from
the
logistics
transportation
industry
has
underscored
urgent
need
for
low-carbon
solutions.
Electric
vehicles
(ELVs)
are
increasingly
being
considered
as
replacements
traditional
fuel-powered
to
reduce
urban
logistics.
However,
ELVs
typically
limited
by
their
battery
capacity
and
load
constraints.
Additionally,
effective
scheduling
of
charging
management
duration
critical
factors
that
must
be
addressed.
This
paper
addresses
low
energy
consumption
(LECS)
problem,
which
aims
minimize
total
heterogeneous
with
varying
capacities,
considering
availability
multiple
stations
(CSs).
Given
complexity
LECS
this
study
proposes
a
attention
model
based
on
encoder-decoder
architecture
(HAMEDA)
approach,
employs
graph
network
introduces
novel
decoding
procedure
enhance
solution
quality
learning
efficiency
during
encoding
phases.
Trained
via
deep
reinforcement
(DRL)
an
unsupervised
manner,
HAMEDA
is
adept
at
autonomously
deriving
optimal
routes
each
ELV
specific
cases
presented.
Comprehensive
simulations
have
verified
can
diminish
overall
utilization
no
less
than
1.64%
compared
other
heuristic
or
learning-based
algorithms.
excels
maintaining
advantageous
equilibrium
between
execution
speed
solutions,
rendering
it
exceptionally
apt
expansive
tasks
necessitate
swift
decision-making
processes.
Industrial Management & Data Systems,
Год журнала:
2025,
Номер
unknown
Опубликована: Март 26, 2025
Purpose
This
study
explores
optimizing
high-speed
railway
(HSR)
meal
services,
a
unique
logistical
challenge
requiring
precise
alignment
with
train
departure
times.
Unlike
standard
delivery
systems,
HSR
services
demand
strict
on-time
delivery,
balancing
the
conflicting
costs
of
earliness
and
tardiness
while
accounting
for
stochastic
nature
preparation
processes.
Design/methodology/approach
A
single-machine
scheduling
model
is
developed
to
minimize
expected
in
delivery.
The
problem
formulated
as
two-stage
mixed-binary
program,
incorporating
uncertainties
intermodal
coordination.
surrogate
algorithm
proposed
enhance
computational
efficiency,
particularly
large
sizes.
Extensive
numerical
experiments
based
on
real-world
scenarios
are
conducted
validate
algorithm.
Findings
significantly
improves
efficiency
maintaining
high
solution
accuracy.
It
outperforms
commercial
solvers
sample
sizes
highlights
importance
uncertainties.
Particularly,
size
increases,
this
can
even
match
optimal
(i.e.
0%
performance
gap)
63.594%
reduction
computation
time.
Originality/value
bridges
gap
integrating
synchromodal
logistics
principles
into
services.
provides
innovative
methodologies
synchronizing
operations
across
transport
modes,
addressing
both
cost
objectives
system
findings
offer
actionable
insights
time-sensitive,
industry
beyond.
