Energy and AI,
Journal Year:
2024,
Volume and Issue:
18, P. 100419 - 100419
Published: Aug. 29, 2024
Addressing
real-world
challenges
in
battery
diagnostics,
particularly
under
incomplete
or
inconsistent
boundary
conditions,
has
proven
difficult
with
traditional
methodologies
such
as
first-principles
and
atomistic
calculations.
Despite
advances
data
assimilation
techniques,
the
overwhelming
volume
diversity
of
data,
coupled
lack
universally
accepted
models,
underscore
limitations
these
approaches.
Recently,
deep
learning
emerged
a
highly
effective
tool
overcoming
persistent
issues
diagnostics
by
adeptly
managing
expansive
design
spaces
discerning
intricate,
multidimensional
correlations.
This
approach
resolves
previously
deemed
insurmountable,
especially
lost,
irregular,
noisy
through
specialized
network
architectures
that
adhere
to
physical
invariants.
However,
gaps
remain
between
academic
advancements
their
practical
applications,
including
explainability
computational
costs
associated
AI-driven
solutions.
Emerging
technologies
explainable
artificial
intelligence
(XAI),
AI
for
IT
operations
(AIOps),
lifelong
machine
mitigate
catastrophic
forgetting,
cloud-based
digital
twins
open
new
opportunities
intelligent
life-cycle
assessment.
In
this
perspective,
we
outline
opportunities,
emphasizing
potential
innovative
transform
demonstrated
our
recent
practice
progress
made
field.
includes
promising
achievements
both
industry
field
demonstrations
modeling
forecasting
dynamics
multiphysics
multiscale
systems.
These
systems
feature
inhomogeneous
cascades
scales,
informed
physical,
electrochemical,
observational,
empirical,
and/or
mathematical
understanding
system.
Through
efforts,
meticulous
craftsmanship,
elaborate
implementations—and
considering
wealth
spatio-temporal
heterogeneity
available
data—such
AI-based
philosophies
have
great
achieve
better
accuracy,
faster
training,
improved
generalization.
Applied Energy,
Journal Year:
2024,
Volume and Issue:
376, P. 124086 - 124086
Published: Aug. 22, 2024
Lithium-ion
batteries
have
found
their
way
into
myriad
sectors
of
industry
to
drive
electrification,
decarbonization,
and
sustainability.A
crucial
aspect
in
ensuring
safe
optimal
performance
is
monitoring
energy
levels.In
this
paper,
we
present
the
first
study
on
predicting
remaining
a
battery
cell
undergoing
discharge
over
wide
current
ranges
from
low
high
C-rates.The
complexity
challenge
arises
cell's
C-rate-dependent
availability
as
well
its
intricate
electro-thermal
dynamics
especially
at
C-rates.To
address
this,
introduce
new
definition
then
undertake
systematic
effort
harnessing
power
machine
learning
enable
prediction.Our
includes
two
parts
cascade.First,
develop
an
accurate
dynamic
model
based
integration
physics
with
capture
battery's
voltage
temperature
behaviors.Second,
model,
propose
approach
predict
under
arbitrary
C-rates
pre-specified
cut-off
limits
temperature.The
experimental
validation
shows
that
proposed
can
relative
error
less
than
3%
when
varies
between
0∼8
C
for
NCA
0∼15
LFP
cell.The
approach,
by
design,
amenable
training
computation.
Indian Chemical Engineer,
Journal Year:
2025,
Volume and Issue:
unknown, P. 1 - 18
Published: Jan. 9, 2025
In
recent
times,
artificial
intelligence
(AI)
and
machine
learning
(ML)
have
emerged
as
revolutionary
technologies
with
wide-ranging
applications
across
various
fields,
including
energy
conversion
storage
(ECS)
systems.
These
methods
utilise
large
amounts
of
data
computational
power
to
predict
material
properties,
optimise
systems,
develop
control
algorithms
for
devices.
This
literature
analysis
focuses
on
the
latest
advancements
methodologies
in
AI/ML
ECS
encompassing
design
discovery,
property
prediction,
system
optimisation.
Furthermore,
study
examines
main
challenges
integrating
ML
into
these
problems
include
issues
related
availability
quality,
model
interpretability,
transfer
learning,
experimental
integration,
ethics.
Despite
challenges,
has
potential
revolutionise
enhance
performance.
Advancements
ML-driven
sustainable
are
fostering
interdisciplinary
collaboration
research,
offering
promising
solutions
energy.
