Water,
Год журнала:
2023,
Номер
15(14), С. 2623 - 2623
Опубликована: Июль 19, 2023
Permeability
characteristics
in
coarse-grained
soil
is
pivotal
for
enhancing
the
understanding
of
its
seepage
behavior
and
effectively
managing
it,
directly
impacting
design,
construction,
operational
safety
embankment
dams.
Furthermore,
these
insights
bridge
diverse
disciplines,
including
hydrogeology,
civil
engineering,
environmental
science,
broadening
their
application
relevance.
In
this
novel
research,
we
leverage
a
Convolutional
Neural
Network
(CNN)
model
to
achieve
accurate
segmentation
CT
images,
surpassing
traditional
methods
precision
opening
new
avenues
granulometric
analysis.
The
three-dimensional
(3D)
models
reconstructed
from
segmented
images
attest
effectiveness
our
CNN
model,
highlighting
potential
automation
soil-particle
Our
study
uncovers
validates
empirical
formulae
ideal
particle
size
discount
factor
soils.
robust
linear
correlation
underlying
deepens
predicts
hydraulic
under
varying
gradients.
This
advancement
holds
immense
value
soil-related
engineering
applications.
findings
underscore
significant
influence
granular
composition,
particularly
concentration
fine
particles,
on
tortuosity
water-flow
paths
factor.
practical
implications
extend
multiple
fields,
water
conservancy
geotechnical
engineering.
Altogether,
research
represents
step
hydrodynamics
where
model’s
unveils
key
into
granulometry
conductivity,
laying
strong
foundation
future
Water,
Год журнала:
2025,
Номер
17(6), С. 793 - 793
Опубликована: Март 10, 2025
Wastewater
treatment
plants
consist
of
many
biological
reactors
and
a
settler,
representing
an
example
large-scale,
nonlinear
systems.
The
wastewater
plant
in
this
study
operates
using
activated
sludge
system,
which
relies
on
processes
to
treat
effectively.
It
is
for
reason
that
iterative
process
modeling
was
used
through
the
implementation
Extended
Kalman
Filter
(EKF)
predict
height
layer
secondary
clarifiers,
where
accumulation
occurs
during
sedimentation
process.
This
technique
consists
maximum
likelihood
estimation
works
more
consistently
various
noise
scenarios.
As
result
evaluation
model
estimated
by
(EKF),
suitability
tends
be
concluded
on.
In
sense,
prediction
sewage
systems
represents
complicated
heteroscedastic
process,
can
understood
as
phenomenon
influenced
variety
factors.
Therefore,
does
not
identify
problems
estimates
thorough
examination
residuals.
state-space
increases
adaptability
adjustability
achieve
structural
optimization
plant.
approach
viable
effective
solution
efficient
management
polluting
levels
minimizing
possible
environmental
impact
out-of-control
situations
plants.
Abstract
While
the
anaerobic-anoxic-oxic
(AAO)
process
is
most
widely
applied
biological
wastewater
treatment
in
municipal
plants
(WWTPs),
it
struggles
to
meet
increasing
demands
on
toxicity
control
of
treated
effluent.
To
tackle
this
challenge,
study
develops
machine
learning
(ML)-based
models
for
optimizing
AAO
towards
improving
its
reduction
efficacy
The
water
quality
parameters,
and
information
(based
nematode
bioassay)
effluent
collected
from
122
WWTPs
China
are
used
train
models.
validated
accurately
predict
effluent’s
parameters
(average
R
2
=
0.81)
ratio
(R
0.86).
further
improve
reduction,
we
developed
a
multiple
objective
optimization
framework
optimize
via
unit
recombination.
In
short-range
combination,
four-unit
combined
processes
(up
79.8%
anaerobic-aerobic-anaerobic-aerobic)
significantly
higher
than
others.
After
optimization,
helps
average
48.6%
70.7%,
with
maximum
87.5%.
methodologies
findings
derived
work
expected
provide
foundation
expansion,
technical
transformation
WWTPs.
Sustainability,
Год журнала:
2024,
Номер
16(17), С. 7670 - 7670
Опубликована: Сен. 4, 2024
Precise
control
of
furnace
temperature
(FT)
is
crucial
for
the
stable,
efficient
operation
and
pollution
municipal
solid
waste
incineration
(MSWI)
process.
To
address
inherent
nonlinearity
uncertainty
process,
a
FT
strategy
proposed.
Firstly,
by
analyzing
process
characteristics
MSWI
in
terms
control,
secondary
air
flow
selected
as
manipulated
variable
to
FT.
Secondly,
an
prediction
model
based
on
Interval
Type-2
Fuzzy
Broad
Learning
System
(IT2FBLS)
developed,
incorporating
online
parameter
learning
structural
algorithms
enhance
accuracy.
Next,
particle
swarm
rolling
optimization
(PSRO)
used
solve
optimal
law
sequence
ensure
efficiency.
Finally,
stability
proposed
method
validated
using
Lyapunov
theory,
confirming
controller’s
reliability
practical
applications.
Experiments
actual
operational
data
confirm
method’s
effectiveness.
