Journal of Soft Computing Paradigm,
Год журнала:
2023,
Номер
5(4), С. 417 - 432
Опубликована: Дек. 1, 2023
The
evolution
of
concrete
strength
prediction
methodologies
has
transitioned
from
empirical
formulas
based
on
experimental
data
to
contemporary
soft
computing
approaches.
Initially,
the
mix
design
was
reliant
simple
relationships
between
proportions
and
compressive
strength;
later,
early
techniques
evolved
include
statistical
models
incorporating
material
properties,
curing
conditions,
environmental
variables.
advent
computational
tools
artificial
intelligence
marked
a
paradigm
shift,
with
accurate
crucial
for
influencing
structural
integrity,
safety,
cost-effectiveness
in
construction.
article
explores
analytical
before
reviewing
application
approaches
such
as
fuzzy
logic,
genetic
algorithms,
neural
networks.
integration
these
hybrid
is
discussed
this
research
study
by
highlighting
their
effectiveness
handling
complex
within
parameters.
A
comparative
analysis
various
methods
applied
non-structural
elements
carried
out
demonstrate
diverse
applications
advantages
optimizing
designs,
enhancing
performance,
contributing
cost
time
efficiency
construction
processes.
Infrastructures,
Год журнала:
2024,
Номер
9(1), С. 10 - 10
Опубликована: Янв. 5, 2024
This
study
addresses
the
vital
issue
of
variability
associated
with
modeling
decisions
in
dam
seismic
analysis.
Traditionally,
structural
and
simulations
employ
a
progressive
approach,
where
more
complex
models
are
gradually
incorporated.
For
example,
if
previous
levels
indicate
insufficient
safety
margins,
advanced
analysis
is
then
undertaken.
Recognizing
constraints
evaluating
influence
various
methods
essential
for
improving
comprehension
effectiveness
assessments.
To
this
end,
an
extensive
parametric
carried
out
to
evaluate
response
Koyna
Pine
Flat
dams
using
solution
approaches
model
complexities.
Numerical
conducted
2D
framework
across
three
software
programs,
encompassing
different
system
configurations.
Additional
complexity
introduced
by
simulating
reservoir
dynamics
Westergaard-added
mass
or
acoustic
elements.
Linear
nonlinear
analyses
performed,
incorporating
pertinent
material
properties,
employing
concrete
damage
plasticity
latter.
Modal
parameters
crest
displacement
time
histories
used
highlight
among
selected
procedures
Finally,
recommendations
made
regarding
adequacy
robustness
each
method,
specifying
scenarios
which
they
most
effectively
applied.
International Journal for Computational Civil and Structural Engineering,
Год журнала:
2025,
Номер
21(1), С. 146 - 156
Опубликована: Март 31, 2025
Deep
learning
(DL),
a
major
part
of
artificial
intelligence
(AI)
is
considered
as
transformational
technology
in
different
areas
science,
such
structural
engineering.
This
critical
review
uncovers
the
potential
contribution
deep
solving
complex
issues
facing
engineering,
optimizing
design,
predicting
and
monitoring
material
behaviour,
real-time
health.
Through
developed
neural
network
architectures
generative
adversarial
networks
(GANs),
recurrent
(RNNs),
convolutional
(CNNs),
engineers
can
identify
solutions
based
on
traditional
deterministic
data
extraction.
However,
like
computational
requirements,
model
interpretability
scarcity
are
widely
adopted.
highlights
recent
advancements,
practical
applications,
limitations
proposing
pathways
for
future
research
to
enhance
its
efficacy
integration
real-world
scenarios
Frontiers in Materials,
Год журнала:
2024,
Номер
11
Опубликована: Окт. 22, 2024
The
thermal
power
industry,
as
a
major
consumer
of
hard
coal,
significantly
contributes
to
harmful
emissions,
affecting
both
air
quality
and
soil
health
during
the
operation
transportation
ash
slag
waste.
This
study
presents
modeling
aerated
concrete
using
local
raw
materials
ash-and-slag
waste
in
seismic
areas
through
machine
learning
techniques.
A
comprehensive
literature
review
comparative
analysis
normative
documentation
underscore
relevance
feasibility
employing
non-autoclaved
blocks
such
regions.
Machine
methods
are
particularly
effective
for
disjointed
datasets,
with
neural
networks
demonstrating
superior
performance
complex
relationships
predicting
strength
density.
results
reveal
that
networks,
especially
those
Bayesian
Regularisation,
consistently
outperformed
decision
trees,
achieving
higher
regression
values
(R
=
0.9587
R
density
0.91997)
lower
error
metrics
(MSE,
RMSE,
RIE,
MAE).
indicates
their
advanced
capability
capture
intricate
non-linear
patterns.
concludes
artificial
robust
tool
properties,
crucial
producing
curing
wall
suitable
earthquake-resistant
construction.
Future
research
should
focus
on
optimizing
balance
between
by
enhancing
properties
utilizing
reliable
models.
Sustainability,
Год журнала:
2024,
Номер
16(24), С. 11139 - 11139
Опубликована: Дек. 19, 2024
Machine
Learning
Techniques
(MLTs)
and
accurate
geographic
mapping
are
crucial
for
managing
natural
hazards,
especially
when
monitoring
the
movement
of
sand
dunes.
This
study
presents
integration
MLTs
with
information
systems
(GIS)
“R”
software
to
monitor
dune
in
Najran
City,
Saudi
Arabia
(KSA).
Utilizing
Linear
Support
Vector
(SVM),
Random
Forest
(RF),
Artificial
Neural
Networks
(ANN)
nine
dune-related
variables,
this
introduces
a
new
Drifting
Sand
Index
(DSI)
effectively
identifying
accumulations.
The
DSI
incorporates
multispectral
sensors
data
demonstrates
robust
capability
dynamics.
Field
surveys
spatial
analysis
were
used
identify
about
100
locations,
which
then
divided
into
training
(70%)
validation
(30%)
sets
at
random.
These
models
produced
thorough
encroachment
risk
map
that
areas
five
hazard
zones:
very
low,
medium,
high,
high
risk.
results
show
an
average
0.8
m/year
towards
southeast.
Performance
evaluation
utilizing
Area
Under
Curve-Receiver
Operating
Characteristic
(AUC-ROC)
approach
revealed
AUC
values
96.2%
SVM,
94.2%
RF,
93%
ANN,
indicating
RF
(AUC
=
96.2%)
as
most
effective
MLTs.
provides
valuable
insights
sustainable
development
environmental
protection,
enabling
decision-makers
prioritize
regions
mitigation
techniques
against
encroachment.