Water Practice & Technology,
Journal Year:
2024,
Volume and Issue:
19(8), P. 3330 - 3349
Published: Aug. 1, 2024
ABSTRACT
The
majority
of
the
environmental
outputs
from
gas
refineries
are
oily
wastewater.
This
research
reveals
a
novel
combination
response
surface
methodology
and
artificial
neural
network
to
optimize
model
oil
content
concentration
in
Response
based
on
central
composite
design
shows
highly
significant
linear
with
P
value
<0.0001
determination
coefficient
R2
equal
0.747,
R
adjusted
was
0.706,
predicted
0.643.
In
addition
analysis
variance
flow
effective
parameters
other
optimization
results
verification
revealed
minimum
8.5
±
0.7
ppm
when
initial
991
ppm,
temperature
46.4
°C,
pressure
21
Mpa,
flowrate
27,000
m3/day
which
is
nearly
closed
suggested
ppm.
An
(ANN)
technique
employed
this
study
estimate
treatment
process.
remarkably
accurate
at
simulating
process
under
investigation.
A
low
mean
squared
error
(MSE)
relative
(RE)
1.55
×
10−7
2.5,
respectively,
were
obtained
during
training
phase,
whilst
testing
demonstrated
high
(R2)
0.99.
Analytical Sciences,
Journal Year:
2025,
Volume and Issue:
unknown
Published: Jan. 21, 2025
In
this
research,
a
green
approach
utilizing
deep
eutectic
solvent
liquid-liquid
microextraction
is
combined
with
smartphone
digital
image
colorimetry
for
the
determination
of
boron
in
nut
samples.
A
camera
was
used
to
capture
analyte
extract
located
custom-made
colorimetric
box.
Using
ImageJ
software,
images
were
split
into
RGB
channels,
channel
identified
as
optimum.
The
distance
between
cuvette
containing
and
detection
determined
be
8
cm,
while
brightness
light
source
30%.
All
obtained
at
585
nm
monochromatic
positioned
background
source.
extraction
achieved
450
µL
1:4
choline-chloride
phenol
mole
ratio
within
60
s
another
minute
centrifugation.
limits
quantification
found
0.02
0.06
µg
mL-1,
respectively.
method
linearity,
indicated
by
relative
coefficient,
greater
than
0.9955
standard
deviations
below
5.4%.
Lastly,
application
chemometrics
form
artificial
intelligence
(AI)-based
models
hybrid
machine
learning
methodologies
has
been
incorporated
SDIC
quantitative
simulation
parameters.
results
gathered
showed
that
these
are
capable
predicting
Water,
Journal Year:
2023,
Volume and Issue:
15(19), P. 3515 - 3515
Published: Oct. 9, 2023
The
need
for
reliable,
state-of-the-art
environmental
investigations
and
pioneering
approaches
to
address
pressing
ecological
dilemmas
nurture
the
sustainable
development
goals
(SDGs)
cannot
be
overstated.
With
power
revolutionize
desalination
processes,
artificial
intelligence
(AI)
models
hold
potential
global
water
scarcity
challenges
contribute
a
more
resilient
future.
realm
of
has
exhibited
mounting
inclination
toward
modeling
efficacy
hybrid
nanofiltration/reverse
osmosis
(NF–RO)
process.
In
this
research,
performance
NF–RO
based
on
permeate
conductivity
was
developed
using
deep
learning
long
short-term
memory
(LSTM)
integrated
with
an
optimized
metaheuristic
crow
search
algorithm
(CSA)
(LSTM-CSA).
Before
model
development,
uncertainty
Monte
Carlo
simulation
adopted
evaluate
attributed
prediction.
results
several
statistical
criteria
(root
mean
square
error
(RMSE)
absolute
(MAE))
demonstrated
reliability
both
LSTM
(RMSE
=
0.1971,
MAE
0.2022)
LSTM-CSA
0.1890,
0.1420),
latter
achieving
highest
accuracy.
accuracy
also
evaluated
new
2D
graphical
visualization,
including
cumulative
distribution
function
(CDF)
fan
plot
justify
other
evaluation
indicators
such
as
standard
deviation
determination
coefficients.
outcomes
proved
that
AI
could
optimize
energy
usage,
identify
energy-saving
opportunities,
suggest
operating
strategies.
Additionally,
can
aid
in
developing
advanced
brine
treatment
techniques,
facilitating
extraction
valuable
resources
from
brine,
thus
minimizing
waste
maximizing
resource
utilization.
