Applied Sciences,
Год журнала:
2024,
Номер
14(19), С. 9123 - 9123
Опубликована: Окт. 9, 2024
Various
hydrogeological
problems
like
groundwater
inflow,
water
table
drawdown,
and
pressure
redistribution
may
be
encountered
in
the
construction
of
hydraulic
projects.
How
to
accurately
predict
occurrence
inflow
assess
drainage
effect
during
are
still
challenging
for
engineering
designers.
Taking
Jinzhai
pumped
storage
power
station
(JPSPS)
China
as
an
example,
this
paper
aims
use
different
methods
calculate
rates
underground
powerhouse
evaluate
caused
by
tunnel
construction.
The
consist
analytical
formulas,
site
rating
(SGR)
method,
Signorini
type
variational
inequality
formulation.
results
show
that
considering
stable
overestimate
caverns
drained
conditions,
whereas
SGR
method
with
available
hydro-geological
parameters
obtains
a
qualitative
hazard
assessment
preliminary
phase.
numerical
solutions
provide
more
precise
reliable
values
complex
geological
structures
seepage
control
measures.
Moreover,
effects,
including
seepage-free
surface,
pore
redistribution,
gradient,
have
been
evaluated
using
various
synthetic
cases.
Specifically,
faults
intersecting
on
significantly
change
flow
regime
around
caverns.
This
comparative
study
can
not
only
exactly
identify
capabilities
cavern
but
also
comprehensively
Agronomy,
Год журнала:
2024,
Номер
14(12), С. 3001 - 3001
Опубликована: Дек. 17, 2024
Crop
diseases
pose
a
significant
threat
to
global
food
security,
with
both
economic
and
environmental
consequences.
Early
accurate
detection
is
essential
for
timely
intervention
sustainable
farming.
This
paper
presents
review
of
machine
learning
(ML)
deep
(DL)
techniques
crop
disease
diagnosis,
focusing
on
Support
Vector
Machines
(SVMs),
Random
Forest
(RF),
k-Nearest
Neighbors
(KNNs),
models
like
VGG16,
ResNet50,
DenseNet121.
The
method
includes
an
in-depth
analysis
algorithm
performance
using
key
metrics
such
as
accuracy,
precision,
recall,
F1
score
across
various
datasets.
We
also
highlight
the
data
imbalances
in
commonly
used
datasets,
particularly
PlantVillage,
discuss
challenges
posed
by
these
imbalances.
research
highlights
critical
insights
regarding
ML
DL
detection.
A
primary
challenge
identified
imbalance
PlantVillage
dataset,
high
number
healthy
images
strong
bias
toward
certain
categories
fungi,
leaving
other
mites
molds
underrepresented.
complicates
model
generalization,
indicating
need
preprocessing
steps
enhance
performance.
study
shows
that
combining
Vision
Transformers
(ViTs)
Green
Chromatic
Coordinates
hybridizing
SVM
achieves
classification
emphasizing
value
advanced
feature
extraction
improving
efficacy.
In
terms
comparative
performance,
architectures
convolutional
neural
network
demonstrated
robust
accuracy
(95–99%)
diverse
underscoring
their
effectiveness
managing
complex
image
data.
Additionally,
traditional
exhibited
varied
strengths;
instance,
performed
better
balanced
while
RF
excelled
imbalanced
Preprocessing
methods
K-means
clustering,
Fuzzy
C-Means,
PCA,
along
ensemble
approaches,
further
improved
accuracy.
Lastly,
underscores
high-quality,
well-labeled
stakeholder
involvement,
comprehensive
evaluation
precision
are
crucial
optimizing
models,
making
them
more
effective
real-world
applications
agriculture.
Applied Sciences,
Год журнала:
2024,
Номер
14(19), С. 8814 - 8814
Опубликована: Сен. 30, 2024
In
coal-dependent
urban
economies,
the
dichotomy
between
resource
exploitation
and
ecological
conservation
presents
a
pronounced
challenge.
Traditional
remote
sensing
assessments
often
overlook
interplay
mining
activities
environmental
dynamics.
To
address
this
gap,
researchers
developed
an
innovative
Resource-Based
City
Ecological
Index
(RCEI),
anchored
in
Pressure–State–Response
(PSR)
framework
synthesized
from
six
discrete
indicators.
Utilizing
geodetic
data,
RCEI
facilitated
comprehensive
spatiotemporal
analysis
of
Jincheng
City’s
quality
1990
to
2022.
The
findings
corroborated
RCEI’s
efficacy
providing
nuanced
portrayal
state
within
regions.
(1)
predominantly
sustained
mudhopper-tier
status,
exhibiting
overarching
trend
amelioration
throughout
study
period.
(2)
Disparities
landscape
were
at
county
level,
with
Moran’s
exceeding
0.9,
signifying
clustered
pattern.
High–high
(H–H)
zones
prevalent
areas
elevated
altitude
dense
vegetation,
whereas
low–low
(L–L)
sectors.
(3)
Further,
buffer
zone
two
coal
mines,
differing
their
chronology,
geographical
positioning,
operational
elucidated
impact
exerted
over
32-year
trajectory.
These
insights
furnish
robust
scientific
technical
foundation
for
resource-centric
cities
fortify
safeguarding
advance
sustainable
development
stratagems.
Applied Sciences,
Год журнала:
2024,
Номер
14(19), С. 9123 - 9123
Опубликована: Окт. 9, 2024
Various
hydrogeological
problems
like
groundwater
inflow,
water
table
drawdown,
and
pressure
redistribution
may
be
encountered
in
the
construction
of
hydraulic
projects.
How
to
accurately
predict
occurrence
inflow
assess
drainage
effect
during
are
still
challenging
for
engineering
designers.
Taking
Jinzhai
pumped
storage
power
station
(JPSPS)
China
as
an
example,
this
paper
aims
use
different
methods
calculate
rates
underground
powerhouse
evaluate
caused
by
tunnel
construction.
The
consist
analytical
formulas,
site
rating
(SGR)
method,
Signorini
type
variational
inequality
formulation.
results
show
that
considering
stable
overestimate
caverns
drained
conditions,
whereas
SGR
method
with
available
hydro-geological
parameters
obtains
a
qualitative
hazard
assessment
preliminary
phase.
numerical
solutions
provide
more
precise
reliable
values
complex
geological
structures
seepage
control
measures.
Moreover,
effects,
including
seepage-free
surface,
pore
redistribution,
gradient,
have
been
evaluated
using
various
synthetic
cases.
Specifically,
faults
intersecting
on
significantly
change
flow
regime
around
caverns.
This
comparative
study
can
not
only
exactly
identify
capabilities
cavern
but
also
comprehensively
