IEEE Transactions on Circuits & Systems II Express Briefs,
Год журнала:
2023,
Номер
71(1), С. 206 - 210
Опубликована: Авг. 4, 2023
This
paper
studies
the
security
issue
of
remote
state
estimation
cyber-physical
systems
under
replay
attacks,
where
measurement
residual
is
computed
in
smart
sensors
and
then
transmitted
to
a
estimator.
Firstly,
compromised
error
covariance
derived
analyze
effect
attacks
on
system
performance.
Next,
estimate
based
coding
detection
scheme
proposed,
which
utilizes
information
encode
decode
residual.
Compared
with
existing
works,
developed
only
uses
information,
not
introduces
extra
information.
Furthermore,
design
principle
matrix
deduced
for
sake
being
able
assist
residual-based
detector
successfully
expose
such
attacks.
Finally,
simulation
results
are
provided
verify
performance
effectiveness
scheme.
IEEE Transactions on Cybernetics,
Год журнала:
2023,
Номер
53(10), С. 6714 - 6724
Опубликована: Март 27, 2023
Security
is
a
crucial
issue
for
cyber-physical
systems,
and
has
become
hot
topic
up
to
date.
From
the
perspective
of
malicious
attackers,
this
article
aims
devise
an
efficient
scheme
on
false
data-injection
(FDI)
attacks
such
that
performance
remote
state
estimation
degraded
as
much
possible.
First,
event-based
stealthy
FDI
attack
mechanism
introduced
selectively
inject
data
while
evading
residual-based
anomaly
detector.
Compared
with
some
existing
methods,
main
advantage
it
decides
when
launch
dynamically
according
real-time
residuals.
Second,
error
covariance
compromised
system
used
evaluate
degradation
under
attacks,
larger
covariance,
more
degradation.
Moreover,
stealthiness
constraints,
optimal
strategy
presented
maximize
trace
covariance.
Finally,
simulation
experiments
are
carried
out
illustrate
superiority
proposed
method
compared
ones.
IEEE Transactions on Fuzzy Systems,
Год журнала:
2024,
Номер
32(4), С. 2352 - 2365
Опубликована: Янв. 5, 2024
In
this
article,
the
problem
of
observer-based
fuzzy
proportional–integral–derivative
(PID)
tracking
control
is
studied
for
networked
nonlinear
systems
subject
to
protocol
constraints
and
norm-bounded
noises.
The
plant
under
consideration
represented
by
an
affine
model
with
immeasurable
premise
variables.
utilization
try-once-discard
proposed
information
exchange
between
sensors
controller,
in
order
mitigate
data
transmission
burden.
An
PID
controller
put
forward
achieve
desired
task
handle
variables,
sufficient
given
both
available
measurements
reference
trajectory.
By
analyzing
dynamics
controlled
error
system
through
construction
a
piecewise
Lyapunov-like
functional,
exponential
ultimate
boundedness
ensured.
parameters
are
designed
using
convex
optimization
technique
matrix
theory,
such
that
exponentially
ultimately
bounded.
Finally,
validity
merits
developed
design
method
verified
simulation
example.
IEEE Transactions on Industrial Informatics,
Год журнала:
2024,
Номер
20(4), С. 6552 - 6562
Опубликована: Янв. 15, 2024
This
article
is
concerned
with
the
secure
distributed
economic
dispatch
(DED)
problem
of
microgrids.
A
quantized
optimization
algorithm
using
Paillier
encryption–decryption
scheme
developed.
designed
to
optimally
coordinate
power
outputs
a
collection
generators
(DGs)
in
order
meet
total
load
demand
at
lowest
generation
cost
under
DG
capacity
limits
while
ensuring
communication
efficiency
and
security.
First,
facilitate
data
encryption
reduce
release,
novel
dynamic
quantization
integrated
into
DED
algorithm,
through
which
effects
errors
can
be
eliminated.
Next,
utilizing
matrix
norm
analysis
mathematical
induction,
sufficient
condition
provided
demonstrate
that
developed
converges
precisely
optimal
solution
finite
levels
(and
even
three-level
sign
transmissions).
Moreover,
an
based
on
outputs,
ensures
confidential
by
leveraging
homomorphic
property
cryptosystem.
Finally,
effectiveness
superiority
implemented
are
confirmed
simulated
example.
IEEE/CAA Journal of Automatica Sinica,
Год журнала:
2023,
Номер
10(5), С. 1181 - 1191
Опубликована: Май 1, 2023
This
paper
investigates
the
security
issue
of
multi-sensor
remote
estimation
systems.
An
optimal
stealthy
false
data
injection
(FDI)
attack
scheme
based
on
historical
and
current
residuals,
which
only
tampers
with
measurement
residuals
partial
sensors
due
to
limited
resources,
is
proposed
maximally
degrade
system
performance.
The
stealthiness
condition
given,
then
error
covariance
in
compromised
state
derived
quantify
performance
under
attack.
strategy
obtained
by
solving
several
convex
optimization
problems
maximize
trace
subject
condition.
Moreover,
constraint
selection
principle
attacked
sensor
provided
determine
so
as
hold
most
impact
Finally,
simulation
results
are
presented
verify
theoretical
analysis.
IEEE Transactions on Industrial Informatics,
Год журнала:
2024,
Номер
20(4), С. 6413 - 6423
Опубликована: Янв. 9, 2024
This
article
is
concerned
with
domain
generalization
(DG),
a
practical
yet
challenging
scenario
in
transfer
learning
where
the
target
data
are
not
available
advance.
The
key
insight
of
DG
focused
on
robust
model
that
can
generalize
to
unseen
by
leveraging
knowledge
from
source
domain.
To
this
end,
we
propose
novel
algorithm
known
as
support-sample-assisted
Adversarial
Attacks
(SSAA)
for
DG.
In
SSAA
algorithm,
an
attack–defense
strategy
deployed
enhance
model's
generalizability
and
transferability.
includes
nontargeted
attack
stage,
during
which
samples
generated
form
pseudotarget
domains
near-realistic
covariate
shifts.
Subsequently,
defense
biclassifier
structure
used
distinguish
support
samples.
These
new
decision
boundary
encompassing
all
samples,
prompting
extension
existing
meet
these
Experimental
results
cross-domain
fault
diagnosis
tasks
suggest
outperforms
current
state-of-the-art
methods,
indicating
promising
avenue
further
development.
International Journal of Systems Science,
Год журнала:
2024,
Номер
55(6), С. 1191 - 1205
Опубликована: Янв. 23, 2024
This
paper
discusses
the
fuzzy
adaptive
consensus
secure
control
problem
for
a
class
of
stochastic
nonlinear
multi-agent
systems
(MASs)
with
unknown
false
data
injection
attacks.
Fuzzy
logic
(FLSs)
are
used
to
model
functions.
A
distributed
filter
is
proposed
obtain
information
leader
since
leaders
only
available
partial
agents.
To
deal
virtual
time-varying
gains
caused
by
attacks,
an
compensation
method
developed
via
re-codesigning
controllers.
Based
on
backstepping
methodology
and
bounded
estimation
algorithms,
controller
proposed.
It
proved
that
all
signals
in
closed-loop
probability,
output
followers
can
track
under
Finally,
simulation
results
comparative
show
effectiveness
method.
