International Journal of Pavement Engineering,
Год журнала:
2024,
Номер
25(1)
Опубликована: Окт. 23, 2024
An
essential
aspect
of
pavement
construction
sustainability
is
its
low-energy
consumption
and
emissions.
The
study
materials
workability
tests
holds
significant
importance
in
terms
achieving
well-mixed
conditions
with
consumption.
complex
components
the
material
uncertain
kinematic
behaviours
aggregates
during
mixing
make
this
process
challenging.
And,
few
studies
signal
response
have
been
found
field
civil
engineering.
For
purpose,
an
accurate
evaluation
monitoring
approach
for
are
needed.
In
paper,
a
wireless
real-time
sensing
method
used
to
monitor
dynamic
behaviour
mixing.
A
3D
digital
twin
model,
combining
data-sensing
techniques
numerical
simulation,
has
proposed
rapid
identification
material.
This
model
validated
via
data-fusion
algorithm.
application
makes
contribution
data-intensive
analysing
jobs
decision-making
tasks
ACS Omega,
Год журнала:
2023,
Номер
8(45), С. 43254 - 43270
Опубликована: Окт. 30, 2023
The
release
of
bromine-free
hydrocarbons
and
gases
is
a
major
challenge
faced
in
the
thermal
recycling
e-waste
due
to
corrosive
effects
produced
HBr.
Metal
oxides
such
as
Fe2O3
(hematite)
are
excellent
debrominating
agents,
they
copyrolyzed
along
with
tetrabromophenol
(TBP),
lesser
used
brominated
flame
retardant
that
constituent
printed
circuit
boards
electronic
equipment.
pyrolytic
(N2)
oxidative
(O2)
decomposition
TBP
has
been
previously
investigated
thermogravimetric
analysis
(TGA)
at
four
different
heating
rates
5,
10,
15,
20
°C/min,
mass
loss
data
between
room
temperature
800
°C
were
reported.
objective
our
paper
study
effectiveness
machine
learning
(ML)
techniques
reproduce
these
TGA
so
use
instrument
can
be
eliminated
enhance
potential
online
monitoring
copyrolysis
treatment.
This
will
reduce
experimental
human
errors
well
improve
process
time
significantly.
both
nonlinear
multidimensional,
hence,
regression
random
forest
(RF)
gradient
boosting
(GBR)
showed
highest
prediction
accuracies
0.999
lowest
among
all
ML
models
employed
this
work.
large
sets
allowed
us
explore
three
scenarios
model
training
validation,
where
number
samples
varied
from
10,000
40,000
for
+
hematite
under
N2
(pyrolysis)
O2
(combustion)
environments.
novelty
have
not
compounds,
while
significance
enhanced
treatment
extension
other
characterization
spectroscopy
chromatography.
Lastly,
could
greatly
benefit
applications
since
it
total
operational
costs
overall
efficiency,
thereby
encouraging
more
plants
adopt
techniques,
resulting
reducing
increasing
environmental
footprint
e-waste.
Construction and Building Materials,
Год журнала:
2024,
Номер
440, С. 137397 - 137397
Опубликована: Июль 13, 2024
Prediction
models
using
machine
learning
assume
an
important
role
in
supporting
decisions
asphalt
road
construction,
such
as
the
scheduling
of
tasks
and
control
compaction
operations.
The
development
prediction
for
physical
properties
can
be
informed
by
insights
from
adoption
specific
techniques.
However,
available
evidence
has
not
yet
been
synthesized.
To
address
this
deficit,
we
systematically
selected
analyzed
30
eligible
studies
published
peer-reviewed
journals
between
2011
2023
data
collection
preprocessing
well
training
evaluation
models.
results
establish
a
comprehensive
picture
techniques
predicting
construction.
Specifically,
review
revealed
following
findings:
(1)
large
range
input
variables
sensors
used;
(2)
pre-specified
few
that
made
feature
selection
unnecessary;
(3)
emphasis
on
Artificial
Neural
Networks
although
empirical
their
higher
performance
is
ambiguous;
(4)
low
rates
unitless
metrics,
which
are
necessary
integration
different
studies;
(5)
need
greater
completeness
clarity
reporting
test
used.
SAE technical papers on CD-ROM/SAE technical paper series,
Год журнала:
2025,
Номер
1
Опубликована: Фев. 21, 2025
<div
class="section
abstract"><div
class="htmlview
paragraph">This
paper
presents
advanced
intelligent
monitoring
methods
aimed
at
enhancing
the
quality
and
durability
of
asphalt
pavement
construction.
The
study
focuses
on
two
critical
tasks:
foreign
object
detection
uniform
application
tack
coat
oil.
For
recognition,
YOLOv5
algorithm
is
employed,
which
provides
real-time
capabilities
essential
for
construction
environments
where
timely
decisions
are
crucial.
A
meticulously
annotated
dataset
comprising
4,108
images,
created
with
LabelImg
tool,
ensures
accurate
objects
such
as
leaves
cigarette
butts.
By
utilizing
pre-trained
weights
during
model
training,
research
achieved
significant
improvements
in
key
performance
metrics,
including
precision
recall
rates.</div><div
paragraph">In
addition
to
detection,
explores
color
space
analysis
through
HSV
(Hue,
Saturation,
Value)
effectively
differentiate
between
coated
uncoated
areas
following
Statistical
analyses,
mean
standard
deviation
calculations
values,
provide
insights
into
differences
that
inform
establishment
threshold
settings
effective
identification.
also
addresses
various
challenges
posed
by
environmental
factors,
steam
smoke,
can
interfere
visual
recognition
operations.
To
mitigate
these
challenges,
an
innovative
automated
mechanical
system
was
designed
stabilize
camera,
ensuring
consistent
data
acquisition
significantly
reliability
tasks.
improving
identification
accuracy
overall
quality,
this
contributes
development
more
efficient
methodologies
maintenance
procedures.
implications
work
suggest
adoption
technologies
vital
facilitating
reliable
processes,
ultimately
leading
better
long-term
surfaces.
This
aims
establish
a
foundation
future
monitoring,
promoting
continuous
improvement
practices
within
industry.</div></div>
SAE technical papers on CD-ROM/SAE technical paper series,
Год журнала:
2025,
Номер
1
Опубликована: Фев. 21, 2025
<div
class="section
abstract"><div
class="htmlview
paragraph">Temperature
segregation
significantly
affects
the
compaction
of
asphalt
mixtures
and
durability
pavement
layer.
Uneven
cooling
mixture
during
transportation
is
a
key
factor
contributing
to
temperature
segregation.
This
study
uses
finite
element
simulations
analyze
temporal
spatial
evolution
mixtures.
A
evaluation
index
(TSI<i>v</i>)
proposed
assess
significance
various
factors
affecting
Support
vector
regression
(SVR),
random
forest
(RFR),
extreme
gradient
boosting
(XGBoost)
models
are
employed
predict
changes
optimize
predictive
models.
The
results
indicate
that
proportion
areas
with
difference
less
than
10°C
consistently
highest,
followed
by
greater
25°C,
then
those
differences
in
ranges
10-16°C
16-25°C.
Higher
discharge
temperatures,
higher
convective
heat
transfer
coefficients,
lower
air
temperatures
associated
In
early
stages
transportation,
has
slightly
effect
transfer,
whereas
later
stages,
plays
most
significant
role.
Both
SVR
RFR
can
effectively
distribution
transportation.</div></div>
Journal of Civil and Hydraulic Engineering,
Год журнала:
2024,
Номер
2(2), С. 100 - 108
Опубликована: Июнь 13, 2024
Recent
observations
of
global
warming
phenomena
have
necessitated
the
evaluation
service
performance
asphalt
pavements,
which
is
substantially
influenced
by
surface
temperature
levels.
This
study
employed
twelve
distinct
machine
learning
algorithms—K-neighbors,
linear
regression,
multi-layer
perceptron,
lasso,
ridge,
support
vector
decision
tree,
AdaBoost,
random
forest,
extra
gradient
boosting,
and
XGBoost—to
predict
pavements.
Data
were
sourced
from
Road
Weather
Information
System
Iowa
State
University,
comprising
12,581
data
points
including
air
temperature,
dew
point
wind
speed,
direction,
gust,
pavement
sensor
temperature.
These
segmented
into
training
(80%)
testing
(20%)
datasets.
Analysis
model
outcomes
indicated
that
Extra
Tree
algorithm
was
superior,
exhibiting
highest
R$^2$
value
0.95,
whereas
Support
Vector
Regression
recorded
lowest,
with
an
0.70.
Furthermore,
Shapley
Additive
Explanations
utilized
to
interpret
results,
providing
insights
contributions
various
predictors
outcomes.
The
findings
affirm
algorithms
are
effective
for
predicting
temperatures,
thereby
supporting
management
systems
in
adapting
changing
environmental
conditions.
