Process Safety and Environmental Protection,
Год журнала:
2022,
Номер
162, С. 813 - 824
Опубликована: Апрель 28, 2022
Fluidized
bed
incinerators
are
more
efficient
and
safe
for
treating
explosive
waste
than
previous
methods
because
they
can
emit
nitrogen
oxide
(NOx)
concentrations
below
the
standard
value
(90
ppm).
However,
a
limitation
is
that
have
only
focused
on
optimizing
operating
conditions
to
minimize
NOx
emission
till
now.
In
this
situation,
it
crucial
balance
process
costs.
Therefore,
study
designed
an
incineration
performed
multi-objective
optimization.
An
artificial
neural
network
surrogate
modeling
method
vital
reduce
optimization
time.
models
with
95%
99%
accuracies
were
obtained,
reducing
calculation
time
by
90%.
Furthermore,
index
combining
costs
was
proposed
obtain
optimal
balanced
condition
of
process.
By
index,
new
obtained
could
20%
while
maintaining
within
limit.
The
data,
such
as
from
sensitivity
analysis,
would
provide
valuable
guideline
abovementioned
associated
standards.
Applied Catalysis B Environment and Energy,
Год журнала:
2023,
Номер
343, С. 123454 - 123454
Опубликована: Ноя. 9, 2023
Conventional
methods
for
developing
heterogeneous
catalysts
are
inefficient
in
time
and
cost,
often
relying
on
trial-and-error.
The
integration
of
machine-learning
(ML)
catalysis
research
using
data
can
reduce
computational
costs
provide
valuable
insights.
However,
the
lack
interpretability
black-box
models
hinders
their
acceptance
among
researchers.
We
propose
an
interpretable
ML
framework
that
enables
a
comprehensive
understanding
complex
relationships
between
variables.
Our
incorporates
tools
such
as
Shapley
additive
explanations
partial
dependence
values
effective
preprocessing
result
analysis.
This
increases
prediction
accuracy
model
with
improved
R2
value
0.96,
while
simultaneously
expanding
catalyst
component
variety.
Furthermore,
case
dry
reforming
methane,
we
tested
validity
recommendation
through
dedicated
experimental
tests.
outstanding
performance
has
potential
to
expedite
rational
design
catalysts.
Chemical Engineering Journal,
Год журнала:
2021,
Номер
431, С. 133244 - 133244
Опубликована: Окт. 30, 2021
In
wet
flue
gas
desulfurization
system,
the
resource
depletion
of
high-grade
limestone,
used
as
conventional
SOx
absorbent,
is
becoming
serious
for
capture
and
utilization.
This
paper
proposes
optimal
selection
blending
ratio
waste
seashells
an
alternative
to
limestone
using
a
deep
neural
network
(DNN)-based
surrogate
model.
Cost
optimization
proceeds
follows:
data
generation,
preprocessing,
development
DNN-based
model,
derivation
cost
ratio.
First,
process
model
developed
generate
datasets,
which
are
gypsum
purity
according
each
seashell
limestone.
addition,
mathematical
proposed
calculate
total
annualized
(TAC)
considering
pretreatment
seashell,
TAC
added
datasets
predict
well
TAC.
Second,
generated
preprocessed
intensify
prediction
performance
z-score
normalization
method.
Third,
Finally,
derived
from
2.4
billion
by
under
two
constraints:
absorbent
consumption.
As
result,
ratios
low-grade
(80.86%),
oyster
shells
(10.78%),
scallop
(0.216%),
cockle
(0.323%),
clam
(2.426%),
mussel
(5.391%),
reducing
US$788,469.
International Journal of Intelligent Systems,
Год журнала:
2021,
Номер
37(6), С. 3625 - 3653
Опубликована: Окт. 1, 2021
The
physical
properties
required
in
polypropylene
composites
(PPCs)
vary
depending
on
the
purpose
of
use.
In
manufacturing
PPCs,
it
is
crucial
to
determine
types
and
quantities
numerous
reinforcements
meet
properties.
Owing
industrial
complexity,
most
PPC
manufacturers
produce
repeatedly
until
desired
are
obtained.
Hence,
reduce
trial
error,
we
developed
prediction
models
for
PPCs
based
commercial
recipe
data.
data
included
information
about
five
manufactured
using
90
materials.
complex
environments,
because
one
usually
composed
2–12
materials,
combinations
sets
created.
It
causes
lack
same
material
combination
thus
makes
difficult
develop
a
good
performance
model.
Therefore,
novel
categorization
process
suggested
as
preprocessing
overcome
imbalance
problem.
predicting
(flexural
strength,
melting
index,
tensile
specific
gravity,
flexural
modulus)
were
random
forest,
was
improved
via
hyperparameter
optimization.
Furthermore,
effects
materials
numerically
described
through
variable
importance
analysis.
Finally,
software
implement
industry.
applied
composite
achieved
high
accuracy,
demonstrating
effectiveness
this
study.
Thus,
suggests
decision-making
solutions
save
cost
time
by
reducing
error
environment
with
complexity.
International Journal of Energy Research,
Год журнала:
2021,
Номер
46(3), С. 3409 - 3427
Опубликована: Окт. 15, 2021
In
the
pulp
and
paper
industry,
black
liquor,
which
is
a
biomass
resource,
burned
to
produce
electricity.
Black
liquor
concentrated
21
wt%
water
through
an
evaporator
before
being
in
boiler.
For
evaporator,
multiple-effect
(MEE)
mainly
used,
but
it
requires
large
amount
of
energy
cost.
Therefore,
crucial
reduce
cost
evaporation
process.
Hence,
this
study
suggested
novel
process
model
that
integrated
mechanical
vapor
recompression
(MVR)
with
MEE
increase
efficiency.
The
MVR-assisted
was
composed
preheating
processes
effectively
concentrate
liquor.
addition,
reduced
steam
consumption
by
using
MVR,
uses
relatively
inexpensive
electric
pre-evaporation
simulation
results,
steam,
electricity
consumption,
latent
heat
recovered
from
secondary
were
quantitatively
analyzed
verify
results
indicate
proposed
increases
substantial
efficiency
compared
conventional
Then,
appropriateness
evaluated
techno-economic
analysis.
total
annualized
(TAC)
determined
for
both
current
potential
future
economic
benefits.
TAC
configuration
can
be
up
77.54%.
