International Journal of Energy Research,
Год журнала:
2025,
Номер
2025(1)
Опубликована: Янв. 1, 2025
The
thermogravimetric
analysis
(TGA)
data
of
wood
present
a
very
similar
layout,
which
makes
it
difficult
to
obtain
information
on
components
or
types
biomass
with
traditional
techniques,
are
generally
based
the
decrease
in
weight
sample
at
specific
temperature
values.
In
this
work,
artificial
neural
networks
applied
as
an
innovative
technique
identify
species
differences
more
effectively.
This
work
evaluates
use
TGA
markers
(similar
genetic
markers)
biomass.
That
is
multiple
values
percentage
residual
respect
initial
one
temperatures
data.
These
achieve
classification
by
automatically
adjusting
weights
training
patterns
input
marker
By
replicating
curve
for
same
and
having
unique
characteristics,
becomes
valuable
tool
characterize
composition.
application
intelligence
techniques
possible
provide
detailed
about
improve
accuracy
classification.
results
demonstrate
that
network
successfully
classified
95%
samples
from
eight
different
accurately
determined
composition
98%
precision.
Biofuels Bioproducts and Biorefining,
Год журнала:
2024,
Номер
unknown
Опубликована: Дек. 19, 2024
Abstract
The
behavior
of
biomass
pyrolysis
can
be
predicted
by
analyzing
its
characteristics.
This
study
aimed
to
model
the
release
volatiles
across
various
temperatures,
properties,
and
heating
rates.
Palm
kernel
shells
were
pyrolyzed
at
433–773
K
with
a
rate
5
K·min
−1
using
volatile‐state
kinetic
modeling.
process
began
calculating
type
number
(
N
CT
),
which
was
used
determine
volatile
enhancement
V
E
yield
Y
VY
product
i
mass
fraction
y
).
parameters,
including
activation
energy
for
formation
ai
derived
through
fitting
process.
results
indicate
70.77%
within
devolatilization
zone,
corresponding
degradation
cellulose
hemicellulose.
increased
higher
lower
,
ranged
from
155–185
kJ·mol⁻¹
biocrude
oil
(BCO)
149–186
gas.
parameters
demonstrated
errors
below
0.2%
in
comparison
experimental
data,
confirming
model's
accuracy
reliability.
