Internet Technology Letters,
Год журнала:
2024,
Номер
unknown
Опубликована: Ноя. 12, 2024
ABSTRACT
In
the
era
of
power
Internet
Things
(PIoT),
accuracy
capacitive
voltage
transformers
(CVTs)
is
crucial
for
maintaining
reliability
measurement
and
protection
systems
in
smart
grids,
thereby
contributing
to
overall
grid
stability
efficiency.
Accurate
timely
detection
anomalies
CVT
errors
essential
preventing
equipment
failures,
reducing
maintenance
costs,
improving
system
reliability.
However,
existing
anomaly
diagnosis
methods
often
rely
on
statistical
analysis
rule‐based
approaches,
which
have
limitations
capturing
complex
patterns
adapting
evolving
types.
This
paper
proposes
a
novel
deep
learning‐based
method
error
PIoT,
called
LSTM‐CVT.
The
proposed
leverages
long
short‐term
memory
(LSTM)
neural
network
architecture
with
three
key
strategies:
bidirectional
temporal
dependency
capture,
hierarchical
feature
learning,
joint
estimation.
experimental
results
demonstrate
superior
performance
LSTM‐CVT
compared
state‐of‐the‐art
baseline
methods.
