Abstract
This
review
exploits
the
crucial
role
of
computational
methods
in
discovering
and
optimizing
materials
for
redox
flow
batteries
(RFBs).
Integration
high‐throughput
screening
(HTCS)
machine
learning
(ML)
accelerates
discovery,
guided
by
algorithms
categorizing
RFBs.
A
collaborative
exploration,
spanning
macroscopic
to
mesoscopic
scales,
combines
quantum
with
reinforcement
learning,
transfer
time
series
analysis,
Bayesian
optimization,
active
various
generative
models.
The
integration
ML
techniques
experimental
methods,
anchored
experimentally
validated
Density
Functional
Theory
(DFT)
calculations
molecular
dynamics
(MD)
simulations,
proves
indispensable
cost‐effective
Data
collection
feature
engineering
are
explored,
emphasizing
optimization
goals
precise
data
within
framework.
Feature
analysis
importance
is
highlighted,
utilizing
such
as
filter,
embedded,
wrapper
deep
efficient
energy
exploration.
Computational
perspectives
on
features
operating
conditions
encompass
membrane
characteristics,
fluid
dynamics,
temperature
dependence
pressure
sensitivity.
Time‐dependent
ML‐generated
insights
understanding
cycling
performance
intricacies,
providing
a
comprehensive
RFB
materials.
Chemical Engineering Journal,
Год журнала:
2024,
Номер
490, С. 151625 - 151625
Опубликована: Апрель 24, 2024
In
the
rapidly
evolving
landscape
of
electrochemical
energy
storage
(EES),
advent
artificial
intelligence
(AI)
has
emerged
as
a
keystone
for
innovation
in
material
design,
propelling
forward
design
and
discovery
batteries,
fuel
cells,
supercapacitors,
many
other
functional
materials.
This
review
paper
elucidates
burgeoning
role
AI
materials
from
foundational
machine
learning
(ML)
techniques
to
its
current
pivotal
advancing
frontiers
science
storage,
including
enhancing
performance,
durability,
safety
battery
technologies,
cell
efficiency
longevity,
fine-tuning
supercapacitors
achieve
superior
capabilities.
Collectively,
we
present
comprehensive
overview
recent
advancements
that
have
significantly
accelerated
development
next-generation
EES,
offering
insights
into
future
research
trajectories
potential
unlock
new
horizons
science.
Advanced Materials,
Год журнала:
2025,
Номер
unknown
Опубликована: Март 6, 2025
The
development
of
supercapacitors
is
impeded
by
the
unclear
relationships
between
nanoporous
electrode
structures
and
electrochemical
performance,
primarily
due
to
challenges
in
decoupling
complex
interdependencies
various
structural
descriptors.
While
machine
learning
(ML)
techniques
offer
a
promising
solution,
their
application
hindered
lack
large,
unified
databases.
Herein,
constant-potential
molecular
simulation
used
construct
supercapacitor
database
with
hundreds
metal-organic
framework
(MOF)
electrodes.
Leveraging
this
database,
well-trained
decision-tree-based
ML
models
achieve
fast,
accurate,
interpretable
predictions
capacitance
charging
rate,
experimentally
validated
representative
case.
SHAP
analyses
reveal
that
specific
surface
area
(SSA)
governs
gravimetric
while
pore
size
effects
are
minimal,
attributed
strong
dependence
electrode-ion
coordination
on
SSA
rather
than
size.
porosity,
respectively,
dominate
volumetric
1D-pore
3D-pore
MOFs,
pinnacling
indispensable
dimensionality.
Meanwhile,
porosity
found
be
most
decisive
factor
rate
for
both
MOFs.
Especially
an
exponential
increase
observed
ionic
conductance
in-pore
ion
diffusion
coefficient,
ascribed
loosened
packing.
These
findings
provide
profound
insights
design
high-performance
Sustainability,
Год журнала:
2024,
Номер
16(4), С. 1516 - 1516
Опубликована: Фев. 10, 2024
Supercapacitors
(SCs)
are
gaining
attention
for
Internet
of
Things
(IoT)
devices
because
their
impressive
characteristics,
including
high
power
and
energy
density,
extended
lifespan,
significant
cycling
stability,
quick
charge–discharge
cycles.
Hence,
it
is
essential
to
make
precise
predictions
about
the
capacitance
lifespan
supercapacitors
choose
appropriate
materials
develop
plans
replacement.
Carbon-based
supercapacitor
electrodes
crucial
advancement
contemporary
technology,
serving
as
a
key
component
among
numerous
types
electrode
materials.
Moreover,
accurately
forecasting
storage
may
greatly
improve
efficient
handling
system
malfunctions.
Researchers
worldwide
have
increasingly
shown
interest
in
using
machine
learning
(ML)
approaches
predicting
performance
The
driven
by
its
noteworthy
benefits,
such
improved
accuracy
predictions,
time
efficiency,
cost-effectiveness.
This
paper
reviews
different
charge
processes,
categorizes
SCs,
investigates
frequently
employed
carbon
components.
supercapacitors,
which
applications,
affected
number
capacity,
cycle
longevity.
Additionally,
we
provide
an
in-depth
review
several
recently
developed
ML-driven
models
used
substance
properties
optimizing
effectiveness.
purpose
these
proposed
ML
algorithms
validate
anticipated
accuracies,
aid
selection
models,
highlight
future
research
topics
field
scientific
computing.
Overall,
this
highlights
possibility
techniques
advancements
energy-storing
device
development.
Scientific Reports,
Год журнала:
2025,
Номер
15(1)
Опубликована: Март 28, 2025
This
paper
presents
a
machine
learning
(ML)
model
designed
to
track
the
maximum
power
point
of
standalone
Photovoltaic
(PV)
systems.
Due
nonlinear
nature
generation
in
PV
systems,
influenced
by
fluctuating
weather
conditions,
managing
this
data
effectively
remains
challenge.
As
result,
use
ML
techniques
optimize
systems
at
their
MPP
is
highly
beneficial.
To
achieve
this,
research
explores
various
algorithms,
such
as
Linear
Regression
(LR),
Ridge
(RR),
Lasso
(Lasso
R),
Bayesian
(BR),
Decision
Tree
(DTR),
Gradient
Boosting
(GBR),
and
Artificial
Neural
Networks
(ANN),
predict
The
utilizes
from
unit's
technical
specifications,
allowing
algorithms
forecast
power,
current,
voltage
based
on
given
irradiance
temperature
inputs.
Predicted
also
used
determine
boost
converter's
duty
cycle.
simulation
was
conducted
100
kW
solar
panel
with
an
open-circuit
64.2
V
short-circuit
current
5.96
A.
Model
performance
evaluated
using
metrics
Root
Mean
Square
Error
(RMSE),
Coefficient
Determination
(R2),
Absolute
(MAE).
Additionally,
study
assessed
correlation
feature
importance
evaluate
compatibility
factors
impacting
predictive
accuracy
models.
Results
showed
that
DTR
algorithm
outperformed
others
like
LR,
RR,
R,
BR,
GBR,
ANN
predicting
(Im),
(Vm),
(Pm)
system.
achieved
RMSE,
MAE,
R2
values
0.006,
0.004,
0.99999
for
Im,
0.015,
0.0036,
Vm,
2.36,
0.871,
Pm.
Factors
size
training
dataset,
operating
conditions
system,
type,
preprocessing
were
found
significantly
influence
prediction
accuracy.