Water and Environment Journal,
Год журнала:
2024,
Номер
38(4), С. 554 - 572
Опубликована: Июль 8, 2024
Abstract
Wastewater
treatment
plants
(WWTPs)
are
high‐energy
consumers
and
major
Greenhouse
Gas
(GHG)
emitters.
This
review
offers
a
comprehensive
global
overview
of
the
current
utilization
machine
learning
(ML)
to
optimize
energy
usage
reduce
emissions
in
WWTPs.
It
compiles
analyses
findings
from
over
hundred
studies
primarily
conducted
within
last
decade.
These
organized
into
five
primary
areas:
consumption
(EC),
aeration
(AE),
pumping
(PE),
sludge
(STE)
greenhouse
gas
(GHG).
Additionally,
they
further
categorized
based
on
type,
scale
application,
geographic
location,
year,
performance
metrics,
software,
etc.
ANNs
emerged
as
most
prevalent,
closely
trailed
by
FL
RF.
While
GA
PSO
predominant
metaheuristic
approaches.
Despite
increasing
complexity,
researchers
inclined
towards
employing
hybrid
models
enhance
performance.
Reported
reductions
or
GHG
spanned
various
ranges,
falling
0–10%,
10–20%
>20%
brackets.
Environmental Science & Technology,
Год журнала:
2024,
Номер
58(29), С. 12989 - 12999
Опубликована: Июль 10, 2024
The
denitrifying
sulfur
(S)
conversion-associated
enhanced
biological
phosphorus
removal
(DS-EBPR)
process
for
treating
saline
wastewater
is
characterized
by
its
unique
microbial
ecology
that
integrates
carbon
(C),
nitrogen
(N),
(P),
and
S
biotransformation.
However,
operational
instability
arises
due
to
the
numerous
parameters
intricates
bacterial
interactions.
This
study
introduces
a
two-stage
interpretable
machine
learning
approach
predict
conversion-driven
P
efficiency
optimize
DS-EBPR
process.
Stage
one
utilized
XGBoost
regression
model,
achieving
an
ACS ES&T Engineering,
Год журнала:
2023,
Номер
4(1), С. 139 - 152
Опубликована: Май 26, 2023
Industrial-scale
garage
dry
fermentation
systems
are
extremely
nonlinear,
and
traditional
machine
learning
algorithms
have
low
prediction
accuracy.
Therefore,
this
study
presents
a
novel
intelligent
system
that
employs
two
automated
(AutoML)
(AutoGluon
H2O)
for
biogas
performance
Shapley
additive
explanation
(SHAP)
interpretable
analysis,
along
with
multiobjective
particle
swarm
optimization
(MOPSO)
early
warning
guidance
of
industrial-scale
fermentation.
The
stacked
ensemble
models
generated
by
AutoGluon
the
highest
accuracy
digester
percolate
tank
performances.
Based
on
optimal
parameter
combinations
were
determined
in
order
to
maximize
production
CH4
content.
conditions
involve
maintaining
temperature
range
35–38
°C,
implementing
daily
spray
time
approximately
10
min
pressure
1000
Pa,
utilizing
feedstock
high
total
solids
Additionally,
should
be
maintained
at
liquid
level
1500
mm,
pH
8.0–8.1,
inorganic
carbon
concentration
greater
than
13.8
g/L.
software
developed
based
was
successfully
validated
warning,
MOPSO-recommended
provided.
In
conclusion,
described
could
accurately
predict
guide
operating
condition
optimization,
paving
way
next
generation
industrial
systems.
