The
continuous
release
of
pollutants
has
posed
a
significant
threat
to
both
public
health
and
the
sustainability
environment.
Consequently,
governments
across
globe
have
introduced
more
stringent
rules
pertaining
vehicle
emissions
address
this
issue.
In
addition,
Portable
Emission
Measurement
System
(PEMS)
is
now
most
widely
used
approach
for
detecting
NOx
PN
from
heavy-duty
vehicles
on
real
roads.
However,
technology
associated
with
requirements
in
terms
human
labor
resources,
making
it
expensive
time-consuming.
Hence,
imperative
put
up
novel
surveillance
actual
on-road
emissions.
Prior
research
indicated
potential
employing
soft
sensors
as
an
alternative
physical
purpose
CO2
monitoring.
there
exists
scarcity
scholarly
investigations
monitoring
PN.
present
study
presents
sensor
based
combination
genetic
algorithm
(GA)
gated
recurrent
unit
(GRU)
real-time
nitrogen
oxides
(NOx)
particle
number
(PN)
within
on-board
diagnostics
(OBD)
system.
This
addresses
existing
gap
development
specifically
designed
study,
we
evidence
that
described
exhibits
exceptional
R2
values
outperforms
other
conventional
models.
Our
findings
illustrate
operate
by
effectively
eliminating
outliers
accurately
promptly
consistently
tracking
predictions
over
lifespan
vehicle.
Moreover,
our
model
high
degree
reusability
domain
prediction
possesses
addressing
emission
emerging
gas
pollutant
constituents
forthcoming
times.
Multiple
active
sites
have
been
confirmed
to
exist
in
the
Cu-SSZ-13,
and
three
types
of
reached
a
consensus.
To
accurately
simulate
reaction
kinetics
NH3-SCR,
constructing
global
kinetic
model
with
responsible
for
NH3
storage
is
key
element,
moreover,
precisely
modeling
ammonia
adsorption
desorption
most
basic
core.
In
this
study,
several
TPD
experiments
were
carried
out
at
various
temperatures,
presence
was
demonstrated
effects
temperature
on
explored.
Then,
characteristics
different
investigated
by
conducting
peak
fitting.
Finally,
three-sites
based
over
Cu-SSZ-13
developed
simulating
actual
release.
The
identified
experiment
data,
average
fitting
accuracy
can
95%
above
while
comparing
simulation
results
results,
confirming
that
well
describing
covering
initial
temperature.