Advancing Fault Detection in Building Automation Systems through Deep Learning
Buildings,
Journal Year:
2024,
Volume and Issue:
14(1), P. 271 - 271
Published: Jan. 19, 2024
This
study
proposes
a
deep
learning
model
utilizing
the
BACnet
(Building
Automation
and
Control
Network)
protocol
for
real-time
detection
of
mechanical
faults
security
vulnerabilities
in
building
automation
systems.
Integrating
various
machine
algorithms
outlier
techniques,
this
is
capable
monitoring
anomaly
patterns
real-time.
The
primary
aim
paper
to
enhance
reliability
efficiency
buildings
industrial
facilities,
offering
solutions
applicable
across
diverse
industries
such
as
manufacturing,
energy
management,
smart
grids.
Our
findings
reveal
that
developed
algorithm
detects
with
an
accuracy
96%,
indicating
its
potential
significantly
improve
safety
However,
full
validation
algorithm’s
performance
conditions
environments
remains
challenge,
future
research
will
explore
methodologies
address
these
issues
further
performance.
expected
play
vital
role
numerous
fields,
including
productivity
improvement,
data
security,
prevention
human
casualties.
Language: Английский
Modeling and Parameter Estimation of Electric Thermal Storage utilizing Residual Components for Residential Consumer
Published: April 14, 2024
Electric
Thermal
Storage
(ETS)
systems
are
conventionally
programmed
for
participation
in
the
typical
Demand
Response
programs.
Particularly,
context
of
Dynamic
Energy
Markets
(DEMs),
ETS
enables
residential
customers
to
actively
participate
lowering
their
energy
costs.
It
is
imperative
build
a
model
ETS-based
heating
thermal
zone
achieve
precise
indoor
temperature
predictions.
This
can
also
assist
estimating
demands,
providing
advantages
during
integration
into
DEMs.
Accordingly,
this
work
introduces
grey
box
modeling
technique
predicting
temperatures.
leverages
residual
components
capture
differences
between
trained
and
experimentally
recorded
data,
thereby
highlighting
prediction
discrepancies.
Subsequently,
Least
Square
(LS)-based
parameter
estimation
brick
temperatures
utilizing
Quantile
Regression
(QR)
Huber
Loss
(HL)
functions
proposed.
Comparative
results
over
24-hour
period
with
experimental
data
proposed
method
ensemble
learning
techniques
such
as
Extra
Trees
Regressor
(ETR)
Random
Forest
(RFR)
presented.
The
performs
more
accurate
forecast
than
conventional
without
residuals
techniques.
reduction
Mean
Absolute
Error
(MAE)
under
0.4°C
compared
real
demonstrates
better
performance
Language: Английский
Thermal Control Strategy for the Sustainable Use of Large Classrooms Responding to User Demands in a School Building
Buildings,
Journal Year:
2024,
Volume and Issue:
14(12), P. 3809 - 3809
Published: Nov. 28, 2024
In
order
to
respond
the
needs
of
education,
importance
various
learning
activities
other
than
subject
courses
is
gradually
increasing
in
schools.
Therefore,
classrooms
schools
are
organized
a
variable
form
depending
on
educational
situations
and
demands,
it
necessary
improve
their
energy
efficiency
operation
without
compromising
indoor
thermal
quality.
This
study
examines
control
models
that
can
perform
cooling
heating
supply
when
using
one
large
classroom
composed
two
architectural
modules.
Through
an
adaptive
process,
proposed
model
determines
efficient
air
according
room
conditions
derived
from
occupant
schedules.
The
optimizes
condition
mitigate
users’
comfort.
Then,
results
this
process
trained
by
iterative
neural
network,
newly
improved
tested
achieve
both
use
comfort
improvement.
As
result,
confirmed
shows
about
2.78%
improvement
72.73%
consistency
as
compared
thermostat
control.
help
efficiently
operate
school
buildings
usability
classrooms.
Language: Английский
Network based Control Methodology for Improving the Indoor Thermal Environment of Sales Shops in Traditional Markets during the Change of Seasons
KIEAE Journal,
Journal Year:
2024,
Volume and Issue:
24(6), P. 61 - 67
Published: Dec. 31, 2024
Purpose:
Various
methods
for
revitalizing
traditional
markets
have
been
studied
in
several
fields
such
as
economy,
sociology,
and
engineering.
There
is
a
need
an
advanced
control
model
that
optimizes
indoor
thermal
conditions
to
improve
the
usability
mitigates
increase
energy
use.
The
aim
of
this
research
develop
effective
method
without
compromising
quality
comfort.
Method:
By
use
designed
working
hour
plan,
proposed
with
adaptive
process
controls
amount
heating
cooling
air
supply.
results
after
are
input
into
artificial
neural
network
learning
algorithm.
Then
performance
investigated
comparison
conventional
thermostat
model.
Results:
effectively
maintains
consistency
comfort
levels
by
about
61%,
reduces
3%,
respectively.
can
help
economy
independent
shops,
which
play
important
role
revitalize
urban
areas.
Language: Английский
An Adaptive Control Model for Thermal Environmental Factors to Supplement the Sustainability of a Small-Sized Factory
Sustainability,
Journal Year:
2023,
Volume and Issue:
15(24), P. 16619 - 16619
Published: Dec. 6, 2023
Effective
indoor
thermal
controls
can
have
quantifiable
advantages
of
improving
energy
efficiency
and
environmental
quality,
which
also
lead
to
additional
benefits
such
as
better
workability,
productivity,
economy
in
buildings.
However,
the
case
factory
buildings
whose
main
usage
is
produce
process
goods,
securing
comfort
for
their
workers
has
been
regarded
a
secondary
problem.
This
study
aims
explore
method
cooling
heating
air
supply
improve
by
use
data-driven
adaptive
model.
The
genetic
algorithm
using
idea
occupancy
rate
helps
model
effectively
analyze
environment
determine
optimized
conditions
comfort.
As
result,
proposed
successfully
shows
performance,
confirms
that
there
2.81%
saving
consumption
16–32%
reduction
dissatisfaction.
In
particular,
significance
this
dissatisfaction
be
reduced
simultaneously
despite
precise
air-supply
are
performed
response
building,
weather,
rate.
Language: Английский