Research on smart home environment system with multisensor and three level data fusion
Abstract
A
three
level
data
fusion
architecture
is
proposed
to
facilitate
the
real-time
detection
and
intelligent
adjustment
of
living
environment,
which
based
on
fuzzy
theory,
back
propagation
(BP)
neural
network,
Dempster-Shafer
(D-S)
evidence
weight
optimization
method.
The
machine
learning
algorithm
used
judge
environment
with
historical
data,
so
that
current
can
make
users
feel
comfortable,
obtain
preferred
environmental
parameter
values
residents.
According
habits
residents,
transformation
parameters
from
static
dynamic
be
achieved
efficiently,
feedback
realized.
test
results
show
system
has
a
accuracy
99.96%
fast
response
speed
good
reliability
performance,
as
well
adaptive
capabilities.
Therefore,
not
only
provides
convenience
for
users,
but
also
reduces
safety
hazards
in
home
environment.
Published: April 9, 2025
Language: Английский