I
nte
rna
t
io
na
l J
o
urna
l o
f
P
o
wer
E
lect
ro
nics
a
nd
Driv
e
S
y
s
t
em
(
I
J
P
E
DS)
Vo
l.
16
,
No
.
2
,
J
u
n
e
20
25
,
p
p
.
851
~
8
6
3
I
SS
N:
2
0
8
8
-
8
6
9
4
,
DOI
: 1
0
.
1
1
5
9
1
/ijp
ed
s
.
v
16
.i
2
.
p
p
8
5
1
-
863
851
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
p
e
d
s
.
ia
esco
r
e.
co
m
Ens
emble lea
rnin
g
bas
ed f
a
ult
dete
ction using
PMU
da
ta in
imba
la
nced data
co
ndition
K
irut
hik
a
K
rish
na
n
1
,
Sriv
a
ni I
y
eng
a
r
2
1
D
e
p
a
r
t
me
n
t
o
f
El
e
c
t
r
i
c
a
l
En
g
i
n
e
e
r
i
n
g
,
R
a
j
a
r
a
j
e
s
h
w
a
r
i
C
o
l
l
e
g
e
o
f
E
n
g
i
n
e
e
r
i
n
g
,
B
a
n
g
a
l
o
r
e
,
I
n
d
i
a
2
D
e
p
a
r
t
me
n
t
o
f
El
e
c
t
r
i
c
a
l
En
g
i
n
e
e
r
i
n
g
,
R
.
V
.
C
o
l
l
e
g
e
o
f
E
n
g
i
n
e
e
r
i
n
g
,
B
a
n
g
a
l
o
r
e
,
I
n
d
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
an
1
1
,
2
0
2
4
R
ev
is
ed
Ma
r
7
,
2
0
2
5
Acc
ep
ted
Ma
r
2
9
,
2
0
2
5
S
ig
n
ifi
c
a
n
t
a
d
v
a
n
c
e
m
e
n
ts
in
t
h
e
e
lec
tri
c
a
l
g
rid
in
c
l
u
d
e
e
n
h
a
n
c
e
d
r
e
g
u
latio
n
,
c
o
m
m
u
n
ica
ti
o
n
,
m
e
terin
g
,
a
n
d
c
u
sto
m
e
r
in
tera
c
ti
o
n
,
d
r
iv
e
n
b
y
in
fo
rm
a
ti
o
n
c
o
m
m
u
n
ica
ti
o
n
tec
h
n
o
l
o
g
ies
(IC
Ts)
a
n
d
c
y
b
e
r
-
p
h
y
sic
a
l
sy
ste
m
s
(CP
S
).
T
h
e
a
d
a
p
tatio
n
o
f
sy
n
c
h
r
o
p
h
a
so
r
d
e
v
i
c
e
s
li
k
e
p
h
a
so
r
m
e
a
su
re
m
e
n
t
u
n
it
s
(P
M
Us
)
e
n
a
b
les
re
a
l
-
ti
m
e
m
o
n
it
o
ri
n
g
a
n
d
c
o
n
tr
o
l,
a
i
d
in
g
i
n
p
o
we
r
sy
ste
m
se
c
u
rit
y
a
ss
e
ss
m
e
n
t.
P
M
Us
re
c
o
rd
v
o
lt
a
g
e
a
n
d
c
u
rre
n
t
p
h
a
so
rs
with
G
P
S
ti
m
e
sta
m
p
s,
tran
sm
it
ti
n
g
d
a
ta
to
p
h
a
so
r
d
a
ta
c
o
n
c
e
n
trato
rs
(P
DCs
)
f
o
r
d
e
c
i
sio
n
-
m
a
k
in
g
.
Ho
we
v
e
r,
e
n
su
rin
g
t
h
e
sta
b
il
it
y
a
n
d
se
c
u
rit
y
o
f
t
h
is
m
e
th
o
d
a
g
a
in
st
c
y
b
e
rse
c
u
rit
y
t
h
re
a
ts
is
c
ru
c
ial
d
u
e
to
it
s
re
li
a
n
c
e
o
n
I
n
tern
e
t
P
r
o
to
c
o
l
(I
P
)
n
e
two
rk
s.
D
y
n
a
m
ic
se
c
u
rit
y
a
ss
e
ss
m
e
n
t
u
ti
li
z
e
s
P
M
U
d
a
ta,
re
p
o
rte
d
u
p
t
o
3
0
–
6
0
ti
m
e
s
p
e
r
se
c
o
n
d
,
t
o
e
v
a
lu
a
te
p
o
we
r
sy
ste
m
sa
fe
ty
.
To
a
d
d
re
s
s
se
c
u
rit
y
issu
e
s,
a
P
y
th
o
n
-
b
a
se
d
fa
u
l
t
d
e
tec
ti
o
n
sy
ste
m
e
m
p
lo
y
in
g
a
sta
c
k
e
n
se
m
b
le
lea
rn
in
g
a
lg
o
rit
h
m
is
d
e
v
e
lo
p
e
d
.
Th
is
a
p
p
ro
a
c
h
c
o
n
siste
n
tl
y
o
u
tp
e
rfo
rm
s
trad
it
io
n
a
l
m
e
th
o
d
s,
p
ro
d
u
c
i
n
g
sa
ti
sfa
c
to
ry
re
su
lt
s
wi
th
s
u
p
e
ri
o
r
A
UC
-
ROC
c
u
rv
e
s,
v
a
li
d
a
te
d
th
r
o
u
g
h
c
o
rre
c
tn
e
ss
c
h
e
c
k
s
a
n
d
g
ra
p
h
ica
l
a
n
a
ly
sis.
Th
e
d
a
tas
e
t
in
c
lu
d
e
s
b
o
t
h
n
a
t
u
ra
l
a
n
d
m
a
n
-
m
a
d
e
se
c
u
rit
y
t
h
re
a
ts,
fa
c
il
it
a
ti
n
g
c
o
m
p
re
h
e
n
siv
e
a
ss
e
ss
m
e
n
t
a
n
d
m
it
ig
a
ti
o
n
stra
teg
ies
.
Th
e
e
n
se
m
b
le
lea
rn
i
n
g
a
lg
o
rit
h
m
p
e
rfo
rm
e
d
b
e
tt
e
r
th
a
n
t
h
e
i
n
d
i
v
id
u
a
l
a
l
g
o
r
it
h
m
s
b
y
o
b
tai
n
in
g
9
5
%
in
t
h
e
AUC
-
ROC c
u
rv
e
.
K
ey
w
o
r
d
s
:
AUC
-
R
O
C
cu
r
v
e
E
n
s
em
b
le
lear
n
in
g
L
o
g
is
tic
r
eg
r
ess
io
n
PMU
f
au
lt d
etec
tio
n
SMOT
E
Stack
en
s
em
b
le
lear
n
in
g
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Kir
u
th
ik
a
Kr
is
h
n
an
Dep
ar
tm
en
t o
f
E
lectr
ical
E
n
g
i
n
ee
r
in
g
,
R
ajar
ajesh
war
i
C
o
lleg
e
o
f
E
n
g
in
ee
r
i
n
g
R
am
o
h
alli Cro
s
s
,
My
s
o
r
e
R
d
,
Ku
m
b
alg
o
d
u
,
B
an
g
alo
r
e
,
Kar
n
atak
a
5
6
0
0
7
4
,
I
n
d
ia
E
m
ail:
k
ir
u
th
i.k
m
2
1
@
g
m
ail.
c
o
m
1.
I
NT
RO
D
UCT
I
O
N
C
y
b
er
attac
k
s
o
r
ig
in
atin
g
f
r
o
m
g
lo
b
al
h
ac
k
e
r
s
h
av
e
h
ad
a
s
u
b
s
tan
tial
im
p
ac
t
o
n
th
e
f
in
a
n
cial
well
-
b
ein
g
o
f
i
n
d
iv
id
u
als.
I
n
r
ec
en
t
tim
es,
cy
b
er
attac
k
s
h
av
e
em
er
g
ed
as
a
s
ig
n
if
ican
t
to
o
l
f
o
r
d
is
r
u
p
tin
g
v
a
r
io
u
s
g
o
v
er
n
m
en
t
ag
e
n
cies.
T
h
is
th
r
ea
t
ex
ten
d
s
to
p
o
wer
s
y
s
tem
s
th
a
t
u
tili
ze
p
h
aso
r
m
ea
s
u
r
e
m
en
t
u
n
its
(
PMUs),
as
th
ese
d
ev
ices
ar
e
co
n
n
ec
ted
to
th
e
in
ter
n
et.
No
tab
ly
,
m
ac
h
i
n
e
lear
n
in
g
alg
o
r
ith
m
s
h
a
v
e
d
is
p
lay
ed
a
cr
u
cial
r
o
le
in
ad
v
an
cin
g
th
e
d
etec
tio
n
o
f
cy
b
er
attac
k
s
with
in
th
e
p
o
wer
s
y
s
tem
en
v
ir
o
n
m
en
t.
A
r
ep
o
r
t
o
n
cy
b
er
-
attac
k
s
u
g
g
ests
th
at
2
6
%
o
f
th
e
in
ci
d
en
ts
o
f
cy
b
e
r
-
attac
k
s
ar
e
b
ase
d
o
n
s
p
ea
r
p
h
is
h
in
g
,
a
way
o
f
m
ak
in
g
th
e
v
ictim
b
eliev
e
th
at
t
h
e
m
ail
is
c
o
m
in
g
f
r
o
m
th
e
o
r
ig
i
n
al
s
o
u
r
ce
,
b
u
t
a
ctu
ally
n
o
t
f
r
o
m
th
e
o
r
i
g
in
al
s
o
u
r
ce
[
1
]
.
A
r
ec
o
r
d
o
f
1
5
9
7
0
0
cy
b
er
i
n
cid
en
ts
o
cc
u
r
r
ed
in
th
e
y
ea
r
2
0
1
7
alo
n
e.
A
cy
b
er
-
p
h
y
s
ical
s
y
s
tem
is
cr
ea
ted
b
y
co
m
b
in
in
g
cy
b
er
s
y
s
tem
s
with
p
h
y
s
ical
p
o
wer
g
r
id
s
i
n
m
o
d
er
n
s
m
ar
t
g
r
id
s
.
T
im
e
-
s
y
n
c
h
r
o
n
ized
m
ea
s
u
r
em
en
t
d
ata
is
tr
an
s
m
itted
to
th
e
cy
b
e
r
s
y
s
tem
f
r
o
m
th
e
p
h
y
s
ical
g
r
id
b
y
u
s
in
g
p
h
aso
r
m
ea
s
u
r
em
e
n
t
u
n
its
(
PMUs).
B
y
r
etu
r
n
in
g
th
e
r
eq
u
ir
ed
c
o
m
m
a
n
d
s
to
th
e
PMUs,
th
e
s
y
s
tem
o
p
er
ato
r
s
(
SO)
at
th
e
cy
b
e
r
b
an
d
ev
alu
ate
b
o
t
h
th
e
o
n
an
d
o
f
f
lin
e
f
o
r
m
ats
o
f
th
e
g
en
er
ated
d
ata
an
d
g
u
ar
a
n
tee
th
e
g
r
id
'
s
r
eliab
ilit
y
an
d
s
ec
u
r
ity
.
Nev
er
th
eless
,
a
v
ar
iety
o
f
p
h
y
s
ical
o
cc
u
r
r
en
c
es,
in
clu
d
in
g
cy
b
er
attac
k
s
,
f
r
eq
u
en
cy
ev
e
n
ts
,
tr
an
s
f
o
r
m
er
ev
en
ts
,
an
d
lin
e
-
to
-
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
6
9
4
I
n
t J Po
w
E
lec
&
Dr
i Sy
s
t
,
Vo
l.
16
,
No
.
2
,
J
u
n
e
20
25
:
8
5
1
-
8
6
3
852
g
r
o
u
n
d
f
ailu
r
es,
ca
n
r
esu
lt
in
d
ev
iatio
n
s
in
th
e
m
ea
s
u
r
em
en
ts
th
at
th
e
SO
r
ec
eiv
es,
wh
ic
h
is
k
n
o
wn
as
"b
a
d
d
ata.
"
T
h
e
SO
m
ay
th
e
n
ch
o
o
s
e
th
e
in
co
r
r
ec
t r
esto
r
ativ
e
o
r
m
itig
atio
n
ap
p
r
o
ac
h
as
a
r
esu
l
t
o
f
th
ese
in
ac
cu
r
at
e
d
ata.
Fo
r
th
e
g
r
id
to
o
p
er
ate
s
af
ely
an
d
o
p
tim
ally
,
p
r
ec
is
e
b
ad
d
ata
d
etec
tio
n
an
d
id
en
tific
atio
n
o
f
th
e
p
r
o
p
er
b
ad
d
ata
k
in
d
ar
e
th
er
ef
o
r
e
ess
en
tial.
Mic
r
o
g
r
id
in
teg
r
atio
n
with
s
m
ar
t
i
n
f
r
astru
ctu
r
e,
s
u
ch
as
s
en
s
o
r
s
,
co
m
m
u
n
icatio
n
,
an
d
m
o
n
ito
r
i
n
g
d
ev
ices,
h
as led
to
th
e
ev
o
l
u
tio
n
o
f
th
e
co
n
ce
p
t o
f
th
e
s
m
ar
t g
r
id
(
SG)
,
wh
ich
o
f
f
er
s
ad
d
itio
n
al
lev
els
o
f
r
o
b
u
s
tn
ess
,
d
ep
en
d
ab
ilit
y
,
an
d
ef
f
icien
t
o
p
er
atio
n
.
T
h
e
p
h
y
s
ica
l
lay
er
o
f
an
MG
is
m
ad
e
u
p
o
f
in
te
r
co
n
n
ec
ted
co
m
p
o
n
en
ts
,
wh
ile
th
e
c
y
b
er
lay
er
is
m
ad
e
u
p
o
f
i
n
ter
co
n
n
ec
te
d
s
m
ar
t
co
m
m
u
n
icatio
n
s
y
s
tem
s
,
m
ete
r
in
g
d
ev
ices,
an
d
m
o
n
ito
r
i
n
g
e
q
u
ip
m
en
t.
Sin
ce
th
e
p
h
y
s
ical
l
ay
er
is
co
n
s
tr
u
cted
u
p
o
n
th
e
c
y
b
er
lay
e
r
,
th
e
en
tire
SG
is
a
cy
b
er
-
p
h
y
s
ical
s
y
s
tem
(
C
PS
)
.
Ph
aso
r
m
ea
s
u
r
em
en
t
u
n
its
(
PMU)
,
wh
ich
o
f
f
er
q
u
ic
k
er
r
ep
o
r
tin
g
r
ates
an
d
m
o
r
e
d
ep
e
n
d
ab
le
an
d
m
o
r
e
s
ec
u
r
e
s
y
s
tem
m
o
n
ito
r
in
g
th
an
tr
ad
itio
n
al
s
u
p
er
v
is
o
r
y
c
o
n
tr
o
l
a
n
d
d
ata
a
cq
u
is
itio
n
(
SC
ADA)
s
y
s
tem
s
,
ar
e
o
n
e
o
f
th
e
cr
u
cial
p
ar
ts
o
f
th
e
SG.
Ho
wev
er
,
th
e
in
tr
o
d
u
c
tio
n
o
f
s
m
ar
t
d
ev
ic
es
lik
e
PM
Us
in
to
MG
s
n
ec
es
s
itates
s
af
e
d
ata
s
to
r
ag
e
an
d
p
r
o
ce
s
s
in
g
tech
n
iq
u
es
an
d
in
c
r
ea
s
in
g
r
elian
ce
o
n
co
m
m
u
n
icatio
n
lin
k
a
g
es
ac
r
o
s
s
t
h
e
m
an
y
C
PS
lev
els.
C
y
b
er
att
ac
k
r
is
k
is
in
cr
ea
s
ed
b
y
th
is
r
elian
ce
o
n
c
o
m
m
u
n
ic
atio
n
ch
an
n
els
an
d
d
at
a
s
to
r
ag
e
s
y
s
tem
s
,
esp
ec
ially
f
o
r
v
ital
in
f
r
astru
ctu
r
es
lik
e
d
ata
ce
n
ter
s
,
h
o
s
p
itals
,
an
d
m
ilit
ar
y
in
s
tallatio
n
s
.
T
h
e
f
u
n
d
am
en
tal
id
ea
b
e
h
in
d
th
e
SG
i
s
m
o
d
er
n
izatio
n
o
f
elec
tr
ical
n
etwo
r
k
th
r
o
u
g
h
in
te
g
r
atio
n
o
f
a
r
tific
ial
in
tellig
en
ce
,
s
ig
n
al
p
r
o
ce
s
s
in
g
,
i
m
p
r
o
v
ed
au
to
m
atic
co
n
tr
o
l,
co
m
m
u
n
icatio
n
s
,
an
d
in
f
o
r
m
a
tio
n
tech
n
o
lo
g
y
.
Var
io
u
s
lev
e
ls
o
f
th
e
g
r
id
ar
e
m
o
n
ito
r
ed
b
y
th
e
s
m
ar
t
m
eter
,
FDR
,
PMU,
S
C
ADA,
W
A
M
S,
an
d
o
th
er
m
o
n
ito
r
in
g
a
n
d
m
ea
s
u
r
in
g
s
y
s
tem
s
.
As
a
r
esu
lt,
a
s
m
ar
t
g
r
i
d
m
u
s
t
s
to
r
e
an
d
d
is
tr
ib
u
te
en
o
r
m
o
u
s
am
o
u
n
ts
o
f
r
ea
l
-
tim
e
d
ata
am
o
n
g
u
s
er
s
,
co
n
tr
o
l
ce
n
te
r
s
,
an
d
u
tili
ties
.
A
s
a
r
esu
lt
,
d
a
t
a
a
n
a
l
y
ti
c
s
w
il
l
b
e
v
e
r
y
h
e
l
p
f
u
l
f
o
r
p
r
o
c
e
s
s
i
n
g
a
n
d
e
v
a
l
u
at
in
g
t
h
i
s
e
n
o
r
m
o
u
s
v
o
l
u
m
e
o
f
p
o
w
e
r
s
y
s
t
e
m
d
at
a
[
2
]
.
Ma
jo
r
b
lack
o
u
ts
th
at
h
av
e
o
cc
u
r
r
ed
in
s
ev
er
al
p
o
wer
s
y
s
tem
s
th
r
o
u
g
h
o
u
t
th
e
g
lo
b
e
h
av
e
m
ad
e
it
clea
r
h
o
w
v
alu
a
b
le
PMU
d
ata
is
,
a
n
d
in
s
tallin
g
PMUs
o
n
t
h
e
n
e
two
r
k
s
o
f
p
o
wer
t
r
an
s
m
is
s
io
n
th
at
r
elate
to
m
o
s
t
m
ajo
r
p
o
wer
s
y
s
tem
s
h
as
b
ec
o
m
e
a
cr
u
cial
p
r
esen
t
e
n
d
ea
v
o
r
.
T
h
is
ar
ticle
ad
d
r
ess
es
th
e
ap
p
licatio
n
s
o
f
wid
e
-
ar
ea
m
ea
s
u
r
em
en
t
s
y
s
tem
(
W
AM
S)
an
d
PMU
tech
n
o
lo
g
y
f
o
r
b
etter
p
o
wer
n
etwo
r
k
m
o
n
ito
r
in
g
,
p
r
o
tectio
n
,
an
d
co
n
tr
o
l.
I
t
also
o
f
f
er
s
a
b
r
ief
i
n
tr
o
d
u
ctio
n
to
th
ese
tech
n
o
lo
g
ies
[
3
]
.
T
h
e
p
o
ten
tials
o
f
wid
e
ar
ea
tech
n
o
lo
g
ies
i.e
.
wid
e
ar
ea
m
o
n
ito
r
in
g
,
p
r
o
t
ec
tio
n
,
a
n
d
co
n
tr
o
l,
o
r
W
AM
PAC
—
ar
e
d
is
cu
s
s
ed
in
th
is
s
tu
d
y
.
PMUs
m
u
s
t
b
e
p
o
s
itio
n
ed
a
p
p
r
o
p
r
iately
b
ased
o
n
th
e
r
ea
l
-
tim
e
a
p
p
licatio
n
as
W
AM
PA
C
d
ep
lo
y
m
en
t n
ec
ess
itate
s
d
is
tr
ib
u
ted
p
h
aso
r
m
ea
s
u
r
em
e
n
ts
ac
r
o
s
s
th
e
s
y
s
tem
.
T
h
e
p
u
r
p
o
s
e
o
f
p
h
aso
r
m
ea
s
u
r
em
en
t
u
n
it
(
PMU)
te
ch
n
o
lo
g
y
an
d
its
u
s
e
in
th
e
p
o
wer
s
y
s
tem
ar
e
d
i
s
cu
s
s
ed
in
th
is
s
tu
d
y
[
4
]
.
T
h
is
liter
atu
r
e
p
r
esen
ts
th
e
r
esu
lts
o
f
an
ex
p
e
r
im
en
tal
s
tu
d
y
th
at
s
h
o
ws
h
o
w
m
alici
o
u
s
ass
au
lts
af
f
ec
t
th
e
PMUs
in
s
m
ar
t
g
r
id
s
.
A
s
im
u
lated
attac
k
en
v
ir
o
n
m
en
t
ar
ch
itectu
r
e
is
s
u
g
g
est
ed
,
an
d
a
p
h
y
s
ical
test
-
b
ed
eq
u
i
p
p
ed
with
a
n
etwo
r
k
attac
k
e
n
v
ir
o
n
m
en
t,
co
m
p
lete
with
m
ain
s
tr
ea
m
PMUs,
is
estab
lis
h
ed
.
T
ests
f
o
r
c
y
b
er
attac
k
s
a
r
e
co
n
d
u
cted
,
in
clu
d
i
n
g
d
ec
eit
f
u
l
co
m
m
u
n
icatio
n
an
d
in
te
r
f
er
en
ce
with
GPS
s
ig
n
als.
T
h
e
ex
p
er
im
en
tal
s
tu
d
y
f
in
d
in
g
s
h
av
e
s
h
o
wn
PMU'
s
wea
k
n
ess
e
s
an
d
v
u
ln
er
ab
ilit
ies to
m
alicio
u
s
ass
au
lts
.
Ad
d
itio
n
ally
,
th
e
f
o
u
n
d
atio
n
al
r
esear
ch
f
o
r
im
p
r
o
v
i
n
g
c
y
b
er
s
ec
u
r
ity
p
r
o
tectio
n
f
o
r
wid
e
ar
ea
m
ea
s
u
r
em
en
t
s
y
s
tem
s
(
W
AM
S)
in
s
m
ar
t
g
r
i
d
s
h
as
b
ee
n
estab
lis
h
ed
[
5
]
.
T
h
is
wo
r
k
p
r
esen
ts
a
u
n
iq
u
e
d
en
s
ity
-
b
ased
s
p
atial
clu
s
ter
in
g
m
eth
o
d
f
o
r
d
ata
m
an
ip
u
latio
n
ass
au
lts
o
n
PMU
m
ea
s
u
r
em
en
ts
,
in
clu
d
in
g
o
n
lin
e
d
etec
tio
n
,
class
if
icatio
n
,
an
d
d
ata
r
ec
o
v
er
y
.
T
h
e
s
u
g
g
ested
ap
p
r
o
ac
h
is
en
tire
ly
d
ata
-
d
r
i
v
en
an
d
ca
n
h
a
n
d
le
m
an
y
m
ea
s
u
r
em
e
n
t
ass
au
lts
a
t
o
n
c
e
with
o
u
t
ad
d
in
g
m
o
r
e
h
ar
d
war
e
to
th
e
cu
r
r
en
t
s
etu
p
.
Ad
d
itio
n
ally
,
th
e
s
u
g
g
ested
m
eth
o
d
d
o
es n
o
t r
el
y
o
n
tr
ad
itio
n
al
s
tate
esti
m
ate
[
6
]
.
Usi
n
g
t
h
e
s
u
b
s
p
ac
e
id
en
tif
icatio
n
ap
p
r
o
ac
h
,
a
d
ata
-
g
u
id
e
d
d
esig
n
s
ch
em
e
o
f
u
n
tr
ac
ea
b
le
f
ak
e
d
ata
-
in
cu
lca
tio
n
attac
k
s
to
cy
b
er
-
p
h
y
s
ical
s
y
s
tem
s
i
s
in
itially
p
r
esen
ted
in
t
h
is
wo
r
k
.
Ne
x
t,
b
y
s
o
lv
in
g
a
r
estricte
d
o
p
tim
i
za
tio
n
p
r
o
b
lem
a
n
d
co
n
s
id
er
in
g
th
e
lim
itatio
n
s
o
f
en
er
g
y
lim
itatio
n
an
d
u
n
d
et
ec
tab
ilit
y
,
th
e
ef
f
ec
ts
o
f
u
n
n
o
ticea
b
le
b
ad
d
ata
-
in
d
u
cin
g
ass
au
lts
as
s
es
s
ed
.
Fu
r
th
er
m
o
r
e
,
co
d
in
g
th
eo
r
y
is
u
s
ed
to
s
tu
d
y
th
e
d
etec
tio
n
o
f
p
lan
n
ed
d
ata
-
d
r
i
v
en
f
ak
e
d
at
a
-
in
d
u
cin
g
attac
k
s
.
Ultim
ately
,
s
im
u
latio
n
s
co
n
d
u
cted
o
n
a
f
ly
i
n
g
v
e
h
icle
m
o
d
el
ar
e
s
h
o
wn
to
co
n
f
ir
m
th
e
e
f
f
ic
ac
y
o
f
th
e
s
u
g
g
ested
tech
n
iq
u
es
[
7
]
.
Attack
er
s
m
ig
h
t
u
s
e
th
e
c
o
m
m
u
n
icatio
n
f
la
w
in
wid
e
-
ar
ea
m
o
n
ito
r
in
g
s
y
s
tem
s
(
W
AM
S)
to
tar
g
et
W
AM
S
r
ec
o
r
d
s
with
m
alicio
u
s
d
ata
in
teg
r
ity
ass
au
lts
,
wh
ich
c
o
u
ld
h
av
e
d
is
astro
u
s
r
esu
lts
.
I
n
r
esp
o
n
s
e
to
th
e
cy
b
er
s
ec
u
r
ity
is
s
u
es
b
r
o
u
g
h
t
to
lig
h
t
b
y
W
AM
S,
s
p
ec
i
f
ic
m
ac
h
in
e
lear
n
in
g
-
b
ased
m
e
th
o
d
s
h
av
e
r
ec
e
n
tly
b
ee
n
cr
ea
ted
to
v
er
if
y
th
e
s
o
u
r
ce
in
f
o
r
m
atio
n
o
f
W
AM
S
d
ata.
Mo
s
t
o
f
th
e
m
eth
o
d
s
o
f
s
o
u
r
ce
au
th
e
n
ticatio
n
n
o
w
in
u
s
e
aim
to
v
er
if
y
W
AM
S
d
ata
f
r
o
m
a
lim
ited
q
u
an
tity
o
f
s
ites
d
is
tr
ib
u
ted
ac
r
o
s
s
an
e
x
p
an
s
iv
e
g
eo
g
r
ap
h
ic
r
eg
io
n
,
wh
ich
co
u
l
d
n
o
t
f
u
ll
y
r
ef
lect
W
AM
S
'
s
o
p
er
atin
g
co
n
d
itio
n
i
n
r
ea
l
-
wo
r
ld
n
et
wo
r
k
s
.
T
h
is
s
tu
d
y
'
s
o
b
jectiv
e
is
to
ascer
tain
if
m
a
ch
in
e
lear
n
in
g
-
b
ased
m
eth
o
d
s
ca
n
b
e
u
s
ed
to
r
ea
l
-
wo
r
ld
p
o
w
er
g
r
id
s
in
o
r
d
er
to
d
ev
elo
p
r
eliab
le
s
o
u
r
ce
au
th
en
ticatio
n
o
f
W
AM
S
d
ata.
Fo
u
r
m
ac
h
i
n
e
lear
n
i
n
g
-
b
ase
d
"state
-
of
-
th
e
-
a
r
t"
tech
n
iq
u
es
th
at
co
m
b
in
e
s
h
all
o
w
an
d
d
ee
p
lear
n
in
g
[
8
]
.
T
o
m
ak
e
th
e
s
m
ar
t
g
r
id
co
m
p
letely
v
is
ib
le,
th
e
n
o
t
m
an
y
PMUs
ar
e
ar
r
an
g
e
d
i
n
p
h
ases
in
lin
e
with
in
ten
d
ed
r
ein
f
o
r
ce
m
e
n
t
-
b
ased
lear
n
in
g
m
eth
o
d
.
Mo
s
t
s
u
s
ce
p
tib
le
b
u
s
e
s
th
at
h
av
e
t
h
e
ab
ilit
y
o
f
g
ettin
g
c
o
m
p
r
o
m
is
ed
b
y
ad
ju
s
tin
g
th
e
f
ewe
s
t
am
o
u
n
t
o
f
m
ea
s
u
r
es
ar
e
id
en
tifie
d
u
s
in
g
a
m
u
ltis
tag
e
o
p
tim
u
m
PMU
p
lace
m
e
n
t
m
eth
o
d
th
at
co
m
b
i
n
es
a
least
-
ef
f
o
r
t
attac
k
m
o
d
el
with
a
r
ein
f
o
r
ce
m
en
t
lear
n
in
g
tech
n
iq
u
e
[
9
]
.
T
h
e
liter
atu
r
e
o
f
f
e
r
s
a
n
in
n
o
v
ativ
e
s
tr
ateg
y
f
o
r
cr
ea
ti
n
g
a
n
d
i
d
en
tify
in
g
th
r
ea
ts
to
d
ata
in
teg
r
ity
in
s
m
ar
t
g
r
id
s
.
Ad
d
itio
n
ally
,
it
o
f
f
er
s
a
way
to
o
p
tim
ize
th
e
cr
ea
ti
o
n
o
f
FDI
A
ag
ain
s
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Po
w
E
lec
&
Dr
i Sy
s
t
I
SS
N:
2088
-
8
6
9
4
E
n
s
emb
le
lea
r
n
in
g
b
a
s
ed
fa
u
lt
d
etec
tio
n
u
s
in
g
P
MU d
a
ta
in
imb
a
la
n
ce
d
…
(
K
ir
u
th
ika
K
r
is
h
n
a
n
)
853
th
e
co
n
tr
o
l
ce
n
ter
'
s
s
tate
esti
m
atio
n
tech
n
iq
u
es.
T
h
e
tech
n
i
q
u
e
f
o
r
g
e
n
er
atin
g
AC
s
ta
te
e
s
tim
ate
attac
k
s
with
b
o
th
e
n
tire
an
d
p
ar
tial
i
n
f
o
r
m
a
tio
n
is
p
r
esen
ted
to
g
eth
er
with
DC
s
tate
esti
m
atio
n
ass
au
lts
.
I
t
also
r
ec
o
m
m
e
n
d
s
in
co
r
p
o
r
atin
g
m
eth
o
d
s
f
o
r
t
h
e
v
o
tin
g
-
b
ased
e
n
s
em
b
le
lear
n
in
g
ap
p
r
o
ac
h
(
MV
C
C
)
to
d
et
ec
t
FDI
A
in
s
m
ar
t
g
r
id
s
.
Nex
t,
a
3
9
b
u
s
New
E
n
g
lan
d
s
y
s
tem
an
d
an
I
E
E
E
2
4
b
u
s
s
y
s
tem
ar
e
em
p
lo
y
ed
as
t
est
s
y
s
tem
s
f
o
r
th
e
m
o
d
el,
a
n
d
f
ictitio
u
s
d
ata
i
n
j
ec
tio
n
attac
k
s
ar
e
cr
ea
ted
an
d
d
etec
ted
.
T
h
e
d
etec
tio
n
ap
p
r
o
ac
h
is
co
m
p
ar
ed
ag
ain
s
t
en
s
em
b
le
m
eth
o
d
s
,
cl
ass
ical
weig
h
ted
least
s
q
u
ar
es,
an
d
m
o
s
t
m
o
d
er
n
m
ac
h
in
e
l
ea
r
n
in
g
alg
o
r
ith
m
s
cu
r
r
en
tly
i
n
u
s
e
[
1
0
]
.
A
r
ea
l
-
tim
e
s
eq
u
en
tial
ap
p
r
o
ac
h
f
o
r
d
etec
tin
g
an
d
class
if
y
in
g
f
au
lty
d
ata
was
p
r
esen
t
ed
in
th
e
liter
atu
r
e.
I
n
itially
,
th
e
Ha
n
k
el
-
m
atr
ix
'
s
lo
w
r
an
k
ch
ar
a
cter
is
tic
to
q
u
ick
ly
id
en
tif
y
er
r
o
n
eo
u
s
d
ata
is
im
p
lem
en
ted
.
Seco
n
d
,
s
tep
is
t
o
ca
teg
o
r
ize
m
alicio
u
s
d
ata
in
t
o
two
g
r
o
u
p
s
:
r
ea
l
-
wo
r
l
d
o
cc
u
r
r
en
ce
s
an
d
o
n
lin
e
attac
k
s
.
T
h
e
m
eth
o
d
u
tili
ze
s
th
e
m
u
lti
-
ch
a
n
n
el
Han
k
el
-
m
atr
ix
'
s
lo
w
r
an
k
ap
p
r
o
x
im
atio
n
e
r
r
o
r
b
ef
o
r
e
to
a
n
d
later
r
an
d
o
m
co
lu
m
n
p
er
m
u
tat
io
n
s
o
n
g
o
in
g
p
h
y
s
ical
ev
en
ts
.
I
n
th
e
im
p
r
o
b
ab
le
ev
en
t
th
at
co
m
p
r
o
m
is
ed
d
ata
is
d
is
co
v
er
ed
to
b
e
th
e
r
esu
lt
o
f
a
cy
b
er
attac
k
,
o
u
r
s
u
g
g
ested
m
eth
o
d
th
e
n
m
o
v
es
o
n
to
id
en
tify
th
e
s
o
u
r
ce
o
f
attac
k
s
.
T
wo
p
o
ten
tial
c
y
b
er
at
tack
r
o
u
tes
ex
a
m
in
ed
ar
e
GP
S
s
p
o
o
f
in
g
an
d
f
ak
e
d
ata
in
je
ctio
n
attac
k
s
(
FDI
A
an
d
GSA)
.
T
o
d
if
f
e
r
en
tiate
b
etwe
en
th
em
,
th
e
ap
p
r
o
ac
h
lev
er
ag
es
th
e
r
an
k
-
1
clo
s
en
ess
er
r
o
r
o
f
th
e
s
in
g
le
-
ch
an
n
el
Han
k
el
m
atr
i
x
with
u
n
wr
ap
p
ed
p
h
ase
a
n
g
le
d
ata
[
1
1
]
.
Ap
p
ly
in
g
m
ac
h
in
e
lear
n
in
g
-
b
ased
a
p
p
r
o
ac
h
es
to
PMU
d
ata
is
o
n
e
o
f
th
e
m
o
s
t
cr
u
cial
attac
k
d
etec
tio
n
m
ea
s
u
r
es.
An
aly
zin
g
th
e
r
esid
u
e
o
f
th
e
o
b
s
er
v
er
s
a
n
d
esti
m
ato
r
s
is
an
o
th
er
m
et
h
o
d
.
Usi
n
g
PMU
d
ata,
th
is
s
tu
d
y
at
tem
p
ts
to
d
etec
t
ass
au
lts
o
n
p
o
wer
s
y
s
tem
s
u
s
in
g
b
o
th
tech
n
iq
u
es.
T
h
e
s
ty
le
o
f
a
ttack
,
s
u
ch
as
m
an
-
in
-
th
e
-
m
id
d
le
(
MitM
)
o
r
a
p
o
te
n
tial
d
en
i
al
-
of
-
s
er
v
ice
(
D
o
S),
is
id
en
tifie
d
u
s
in
g
an
alg
o
r
ith
m
.
L
astl
y
,
th
e
s
u
g
g
ested
p
r
o
c
ed
u
r
e
is
r
ep
licated
u
s
in
g
an
e
x
am
p
le
I
E
E
E
p
o
we
r
s
y
s
tem
,
an
d
en
co
u
r
ag
in
g
o
u
tco
m
es th
at
co
n
f
ir
m
th
e
ap
p
r
o
ac
h
'
s
ef
f
ec
tiv
en
ess
is
ex
p
lain
ed
in
d
e
tail [
1
2
]
.
A
n
ew
v
o
tin
g
-
b
ased
tech
n
iq
u
e
f
o
r
d
etec
tio
n
f
o
r
s
y
s
tem
ic
c
y
b
er
i
n
tr
u
s
io
n
s
is
d
ev
el
o
p
ed
with
in
th
is
wo
r
k
.
T
h
e
attac
k
er
in
t
h
e
cy
b
e
r
attac
k
u
n
d
er
e
x
am
in
atio
n
in
t
r
o
d
u
ce
s
a
s
p
an
o
f
f
alse
d
ata
in
PMU
in
an
attem
p
t,
r
ep
licate
f
ictitio
u
s
s
h
o
r
t
cir
cu
it
o
cc
u
r
r
en
ce
s
in
th
e
s
y
s
tem
.
T
h
e
s
u
g
g
ested
s
p
o
ttin
g
p
r
o
ce
d
u
r
e
m
ak
es
u
s
e
o
f
a
v
ar
iety
o
f
m
ac
h
in
e
lear
n
in
g
(
ML
)
tech
n
iq
u
es,
s
u
ch
as
en
s
e
m
b
le
lear
n
i
n
g
,
r
ec
u
r
r
en
t
n
eu
r
al
n
etwo
r
k
(
R
NN)
,
f
ee
d
f
o
r
war
d
n
e
u
r
al
n
etwo
r
k
(
FNN)
,
d
ec
is
io
n
tr
ee
s
,
d
is
cr
im
in
an
t
an
aly
s
is
,
k
-
n
ea
r
est
n
eig
h
b
o
r
s
(
KNN)
class
if
icatio
n
,
s
u
p
p
o
r
t v
ec
to
r
m
ac
h
in
e
(
SVM)
,
an
d
n
aiv
e
b
a
y
es.
B
y
d
eter
m
in
in
g
t
h
e
av
er
a
g
e
o
u
tp
u
t d
ep
e
n
d
in
g
o
n
d
etec
to
r
p
e
r
f
o
r
m
an
ce
,
t
h
e
v
o
tin
g
-
b
ased
tech
n
i
q
u
e
m
a
y
b
e
ab
le
to
d
if
f
e
r
en
tiate
b
etwe
en
FDI
attac
k
s
an
d
ac
tu
al
s
h
o
r
t
cir
cu
it
f
a
ilu
r
es.
T
o
m
in
im
ize
r
ed
u
n
d
a
n
cy
an
d
e
n
h
an
ce
r
ele
v
an
ce
,
th
e
m
e
ch
a
n
ical
an
d
elec
tr
ical
co
m
p
o
n
en
ts
o
f
t
h
e
s
y
s
tem
ar
e
o
p
tim
ally
s
elec
ted
f
o
r
tr
ain
in
g
o
b
jectiv
es
[
1
3
]
.
L
iter
at
u
r
e
p
r
e
s
en
ts
an
en
s
em
b
le
b
ag
g
ed
tr
ee
f
o
r
r
elativ
ely
ac
cu
r
ate
r
ea
l
-
tim
e
attac
k
an
d
d
ef
ec
t d
etec
tio
n
.
T
h
is
s
u
g
g
ested
s
tr
u
ctu
r
e
is
p
r
ed
icate
d
o
n
d
ata
f
r
o
m
th
e
p
h
aso
r
m
ea
s
u
r
em
en
t
u
n
it
(
PMU)
an
d
r
elay
s
d
u
r
in
g
n
o
r
m
al,
cy
b
e
r
attac
k
,
an
d
f
ailu
r
e
co
n
d
itio
n
s
.
T
h
is
s
tu
d
y
co
m
p
ar
es th
e
ef
f
ec
tiv
en
ess
o
f
th
e
r
ec
o
m
m
en
d
ed
m
eth
o
d
ag
ain
s
t sev
er
al
m
ac
h
i
n
e
lear
n
i
n
g
m
eth
o
d
s
an
d
v
alid
ates
it
in
a
MA
T
L
A
B
/Si
m
u
lin
k
test
in
g
en
v
ir
o
n
m
e
n
t
[
1
4
]
.
L
iter
atu
r
e
h
as
co
n
d
u
ct
ed
a
co
m
p
r
eh
e
n
s
iv
e
in
v
esti
g
atio
n
o
f
b
ig
d
ata
a
n
aly
tics
ap
p
licatio
n
s
,
cu
r
r
en
t iss
u
e
s
,
an
d
s
o
lu
tio
n
s
[
1
5
]
.
I
n
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
s
(
I
DS)
ar
e
cr
itical
to
o
v
e
r
s
ee
th
e
s
ec
u
r
ity
o
f
cy
b
er
-
p
h
y
s
ical
en
er
g
y
a
n
d
p
o
wer
s
y
s
tem
s
p
r
esen
t
in
SG
with
in
cr
ea
s
in
g
m
ac
h
i
n
e
-
to
-
m
ac
h
in
e
c
o
n
n
ec
tio
n
s
.
Sti
ll,
I
DS
is
f
in
d
in
g
it
v
e
r
y
ch
allen
g
in
g
to
r
eliab
ly
d
i
f
f
er
e
n
tiate
b
etwe
en
b
e
n
ig
n
an
d
m
alev
o
len
t
ev
e
n
ts
d
u
e
to
th
e
m
an
y
-
s
o
u
r
ce
d
,
lar
g
e,
lin
k
ed
,
an
d
o
f
ten
n
o
is
e
-
co
n
tai
n
in
g
u
n
wa
n
ted
d
ata
th
at
s
av
es a
r
an
g
e
o
f
c
o
n
cu
r
r
en
t c
y
b
er
an
d
p
h
y
s
ical
ac
tiv
ity
.
T
o
d
ea
l
with
th
ese
an
d
s
im
ilar
is
s
u
es,
h
er
e,
a
r
o
b
u
s
t
s
tar
t
-
to
-
f
in
is
h
f
r
am
ewo
r
k
in
lin
e
with
th
e
en
s
em
b
le
m
ac
h
in
e
lear
n
in
g
a
n
d
s
tack
e
d
d
e
n
o
is
in
g
au
to
e
n
co
d
e
r
(
SDAE
)
to
ex
tr
ac
t
n
ew
ch
ar
ac
te
r
is
tic
s
ets
in
f
o
r
m
ed
b
y
attac
k
s
a
n
d
n
o
is
e
f
r
o
m
cy
b
er
-
p
h
y
s
ical
s
y
s
tem
d
ata
an
d
in
teg
r
ate
m
u
ltip
le
in
f
o
r
m
atio
n
s
o
u
r
ce
s
f
o
r
au
t
h
en
tic
ev
e
n
t
ca
teg
o
r
izatio
n
.
T
h
e
p
u
t
f
o
r
war
d
ed
m
eth
o
d
o
lo
g
y
in
f
lu
en
ce
s
s
to
ch
asti
c
d
if
f
e
r
en
ce
o
f
an
o
m
aly
ex
tr
ac
tio
n
(
SDAE
)
to
f
ir
s
t
p
r
o
v
id
e
s
m
aller
-
d
im
en
s
io
n
al
attr
ib
u
tes
th
at
p
er
m
it
th
e
r
esto
r
atio
n
o
f
a
clu
tter
-
f
r
ee
in
p
u
t
f
r
o
m
clu
tter
-
d
am
ag
ed
p
er
t
u
r
b
atio
n
s
.
No
v
el
ch
ar
ac
ter
is
tics
th
at
will
m
ain
t
ain
an
d
u
p
d
ate
i
n
f
o
r
m
atio
n
as
n
o
r
m
al,
f
a
u
lt,
a
n
d
attac
k
ev
en
ts
ag
ain
s
t a
r
an
g
e
o
f
s
y
n
th
etic
attac
k
d
ata,
with
th
e
g
o
al
o
f
im
p
r
o
v
in
g
ca
teg
o
r
izatio
n
b
y
in
teg
r
atin
g
attac
k
an
d
n
o
is
y
in
p
u
ts
.
I
n
ad
d
itio
n
,
e
n
s
em
b
le
lear
n
i
n
g
-
b
ased
m
u
lti
-
class
if
ier
class
if
icatio
n
,
c
o
n
s
id
er
in
g
th
e
h
eter
o
g
en
e
o
u
s
n
at
u
r
e
o
f
t
h
e
i
n
p
u
ts
s
u
ch
as
PMU
m
ea
s
u
r
e
m
en
ts
,
s
y
s
tem
lo
g
s
,
an
d
I
DS
aler
ts
,
an
d
class
if
y
in
g
th
e
s
p
ec
im
en
s
b
ased
o
n
th
e
S
DAE
-
ex
tr
ac
ted
ch
ar
ac
ter
is
tics
u
s
in
g
th
e
e
x
tr
em
e
g
r
ad
ie
n
t
b
o
o
s
tin
g
(
XGBo
o
s
t)
tech
n
iq
u
e
,
is
d
e
v
elo
p
ed
.
Mo
r
e
o
v
er
,
n
o
r
m
aliza
tio
n
a
n
d
o
v
e
r
s
am
p
lin
g
we
r
e
u
s
ed
t
o
en
h
an
ce
th
e
d
ata'
s
b
alan
ce
an
d
h
o
m
o
g
en
eity
.
T
h
e
p
r
esen
t
SDAE
+X
GB
o
o
s
t a
p
p
r
o
ac
h
at
tain
s
m
o
r
e
th
a
n
9
0
% c
lass
if
icatio
n
co
r
r
ec
tn
ess
o
n
a
p
r
ac
tical
d
ataset
co
m
p
r
is
in
g
3
7
s
u
b
-
t
y
p
es
o
f
n
o
r
m
al,
f
a
u
lt,
an
d
attac
k
o
b
tain
ed
v
ia
co
-
s
im
u
latio
n
s
o
n
a
h
ar
d
war
e
-
in
-
th
e
-
lo
o
p
(
HI
L
)
te
s
tb
ed
s
ec
u
r
ity
test
b
ed
[
1
6
]
.
Sp
ec
if
ically
,
in
th
e
c
o
n
tex
t
o
f
cy
b
er
th
r
ea
ts
,
a
u
n
iq
u
e
"g
r
ee
d
y
"
m
eth
o
d
f
o
r
PMU
p
la
ce
m
en
t
is
d
ev
elo
p
ed
.
Acc
o
r
d
i
n
g
to
t
h
is
r
esear
ch
,
c
y
b
er
r
is
k
m
ay
g
r
ea
tly
r
aise
a
p
o
wer
s
y
s
tem
'
s
u
n
o
b
s
er
v
ab
ilit
y
r
is
k
,
n
ec
ess
itatin
g
th
e
in
clu
s
io
n
o
f
PMU
allo
ca
tio
n
s
[
1
7
]
.
T
o
d
ev
elo
p
a
co
m
p
lete
ar
c
h
itectu
r
e
th
at
is
r
esis
tan
t
to
cy
b
er
attac
k
s
an
d
u
s
es
s
tr
ateg
ically
p
o
s
itio
n
ed
p
h
aso
r
m
ea
s
u
r
em
en
t
u
n
its
(
P
MU
s
)
to
co
u
n
ter
ac
t
s
tr
u
ctu
r
al
wea
k
n
ess
es
in
s
m
ar
t
g
r
id
s
,
a
b
r
an
d
-
n
ew
h
y
b
r
id
b
etwe
en
n
ess
ce
n
tr
ality
(
HB
C
)
m
etr
ic
is
p
u
t
f
o
r
th
th
at
s
u
cc
ess
f
u
lly
p
in
p
o
in
ts
a
s
y
s
tem
's
m
o
s
t
cr
u
cial
lin
es.
A
d
is
tin
ct
o
b
jectiv
e
f
u
n
ctio
n
is
c
r
ea
ted
with
th
e
p
u
r
p
o
s
e
o
f
d
eli
b
er
ately
in
s
er
tin
g
PMUs
in
to
th
e
s
y
s
tem
in
o
r
d
er
to
s
tr
en
g
th
en
its
d
ef
en
s
es
ag
ain
s
t
an
y
attac
k
s
b
y
b
o
g
u
s
d
ata
in
jectio
n
o
n
th
ese
s
u
s
ce
p
tib
le
lin
es.
Fin
d
in
g
th
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
6
9
4
I
n
t J Po
w
E
lec
&
Dr
i Sy
s
t
,
Vo
l.
16
,
No
.
2
,
J
u
n
e
20
25
:
8
5
1
-
8
6
3
854
b
est PMU
lo
ca
tio
n
r
esu
lts
in
t
h
e
f
ewe
s
t set
s
o
f
m
ea
s
u
r
em
en
ts
r
eq
u
ir
ed
to
d
ef
en
d
th
e
s
tate
v
ar
iab
les ag
ain
s
t a
ll
k
in
d
s
o
f
attac
k
s
o
n
d
ata
in
teg
r
ity
.
T
h
is
d
esig
n
'
s
ef
f
ec
tiv
en
es
s
is
i
llu
s
tr
ated
with
th
e
I
E
E
E
1
4
-
b
u
s
s
y
s
tem
[
1
8
]
.
Data
-
d
r
iv
en
h
ac
k
i
n
g
tech
n
iq
u
es
lik
e
th
e
f
ak
e
d
ata
in
jectio
n
attac
k
(
FDI
A)
s
er
io
u
s
ly
jeo
p
ar
d
ize
th
e
s
tate
s
o
f
th
e
g
r
id
.
L
iter
atu
r
e
[
1
9
]
p
r
esen
ts
an
ef
f
ec
tiv
e
f
o
r
m
u
latio
n
m
et
h
o
d
f
o
r
b
lin
d
FDI
A
th
at
r
eq
u
ir
es
ex
ac
t
m
ea
s
u
r
em
en
t
s
u
b
s
p
ac
e
in
f
o
r
m
atio
n
.
I
n
o
r
d
e
r
to
allay
th
is
wo
r
r
y
,
ef
f
icien
t
im
p
lem
en
tatio
n
o
f
n
ew,
r
o
b
u
s
t,
n
o
n
lin
ea
r
d
ee
p
lear
n
in
g
m
o
d
els
th
at
ca
n
,
in
ad
d
itio
n
to
ef
f
ec
tiv
ely
d
e
tectin
g
th
e
ex
is
ten
ce
o
f
b
lin
d
attac
k
in
tr
u
s
io
n
s
in
r
ea
l
tim
e,
p
in
p
o
in
t
th
eir
p
r
ec
is
e
l
o
ca
tio
n
s
is
to
b
e
im
p
lem
en
te
d
.
T
h
ese
v
er
s
io
n
s
ca
n
wo
r
k
i
n
co
n
ju
n
ctio
n
with
co
n
v
en
tio
n
al
b
ad
d
ata
d
etec
to
r
s
to
o
f
f
er
a
p
r
ac
tical
an
d
af
f
o
r
d
ab
le
s
o
lu
tio
n
.
B
y
id
en
tify
in
g
t
h
e
d
is
cr
ep
an
cy
with
th
e
co
-
o
cc
u
r
r
en
ce
d
e
p
en
d
e
n
cy
o
f
th
e
attac
k
v
ec
to
r
s
ad
d
e
d
to
th
e
r
aw
d
ata,
th
ese
n
eu
r
al
n
etwo
r
k
m
o
d
els
also
d
em
o
n
s
tr
ate
a
m
u
ltil
ab
el
cla
s
s
if
icatio
n
tech
n
iq
u
e.
Mo
r
eo
v
er
,
it
is
d
em
o
n
s
tr
ated
th
at
t
h
ese
d
ee
p
lear
n
in
g
s
tr
u
ctu
r
es
ar
e
m
o
d
el
-
f
r
e
e,
s
u
g
g
esti
n
g
th
at
ass
au
lts
m
ig
h
t
b
e
id
en
tifie
d
with
o
u
t
r
e
q
u
ir
in
g
s
t
atis
tical
k
n
o
wled
g
e
o
f
th
e
g
r
id
.
On
th
e
s
tan
d
ar
d
I
E
E
E
test
b
en
ch
,
th
e
s
u
g
g
ested
f
r
am
ewo
r
k
'
s
p
er
f
o
r
m
a
n
ce
is
ass
ess
ed
u
n
d
er
a
r
an
g
e
o
f
ass
au
lt a
n
d
n
o
is
e
s
ce
n
ar
io
s
[
1
9
]
.
On
e
o
f
th
e
b
ig
g
est
r
is
k
s
to
th
e
s
af
ety
,
d
ep
en
d
ab
ilit
y
,
an
d
c
o
s
t
-
ef
f
ec
tiv
e
o
p
e
r
atio
n
o
f
p
o
w
er
s
y
s
tem
s
n
o
wad
ay
s
is
cy
b
er
attac
k
s
.
I
t is ch
allen
g
in
g
to
id
e
n
tify
an
d
class
if
y
v
ar
io
u
s
cy
b
er
attac
k
s
wh
ile
m
ain
tain
in
g
th
e
s
tab
ilit
y
an
d
s
ec
u
r
ity
o
f
th
e
p
o
wer
in
f
r
astru
ct
u
r
e.
An
au
to
m
ate
d
tech
n
iq
u
e
b
ased
o
n
t
h
e
c
o
n
v
o
lu
ti
o
n
al
n
e
u
r
al
n
etwo
r
k
f
o
r
th
e
r
ec
o
g
n
itio
n
an
d
ca
teg
o
r
izatio
n
o
f
v
ar
i
o
u
s
cy
b
er
attac
k
s
to
ad
d
r
ess
th
is
p
r
o
b
lem
.
T
h
e
co
n
v
o
l
u
tio
n
al
n
eu
r
al
n
etwo
r
k
co
llects
tem
p
o
r
al
in
f
o
r
m
atio
n
an
d
s
p
atial
i
n
ter
ac
tio
n
s
b
etwe
en
v
ar
i
o
u
s
n
o
d
es
f
r
o
m
th
e
p
r
io
r
o
p
er
atio
n
al
s
tate
o
f
th
e
s
en
t
d
ata
p
ac
k
ets.
T
h
e
s
u
g
g
ested
s
tr
u
ctu
r
e'
s
ca
p
s
u
les
h
av
e
s
ig
n
if
ican
t
ef
f
ec
ts
o
n
p
r
eser
v
i
n
g
th
e
m
ea
s
u
r
em
en
t m
atr
ix
'
s
to
p
o
lo
g
ical
c
o
n
s
is
ten
cy
.
Ad
d
itio
n
ally
,
th
e
s
u
g
g
ested
ap
p
r
o
ac
h
elim
in
ates
th
e
in
f
lu
e
n
ce
o
f
u
n
ce
r
tain
ty
in
s
y
s
tem
ch
ar
ac
ter
is
tics
o
n
d
etec
tio
n
p
e
r
f
o
r
m
an
ce
an
d
is
m
o
d
el
-
f
r
ee
.
I
n
th
is
s
tu
d
y
,
m
a
n
y
ty
p
es
o
f
c
o
m
m
o
n
cy
b
er
attac
k
s
ar
e
e
x
am
i
n
ed
a
n
d
m
o
d
eled
,
s
u
ch
as
r
e
p
lay
,
d
e
n
ial
o
f
s
er
v
ice,
b
o
g
u
s
d
ata
in
jectio
n
,
tim
e
-
d
ela
y
,
an
d
d
ec
e
p
tio
n
ass
au
lts
.
T
h
e
s
u
g
g
ested
s
o
lu
tio
n
m
ay
ac
h
iev
e
9
9
.
9
7
%
d
etec
tio
n
ac
cu
r
ac
y
o
n
a
s
in
g
le
c
y
b
er
at
tack
an
d
9
6
.
2
5
%
d
etec
tio
n
a
cc
u
r
ac
y
o
n
m
u
ltip
le
c
y
b
er
att
ac
k
s
,
ac
co
r
d
in
g
to
n
u
m
er
ical
f
in
d
in
g
s
o
n
th
e
I
E
E
E
3
9
-
b
u
s
test
s
y
s
tem
.
T
h
e
r
esu
l
ts
o
f
co
m
p
ar
is
o
n
s
h
o
w
th
at
th
e
s
u
g
g
ested
ap
p
r
o
ac
h
p
er
f
o
r
m
s
b
etter
th
an
co
n
v
en
ti
o
n
al
n
eu
r
al
n
etwo
r
k
s
.
T
h
e
is
s
u
e
o
f
m
u
ltip
le
attac
k
s
d
etec
tio
n
an
d
ca
teg
o
r
izatio
n
is
r
eso
lv
ed
b
y
th
is
tech
n
iq
u
e
[
2
0
]
.
L
iter
atu
r
e
aim
s
t
o
ca
r
ef
u
lly
ex
p
lain
s
ev
er
al
tech
n
iq
u
e
s
an
d
p
r
o
ce
s
s
es
f
o
r
cy
b
er
-
s
ec
u
r
ity
in
en
e
r
g
y
s
y
s
tem
s
an
d
ex
am
in
e
r
elev
an
t
s
o
lu
tio
n
ap
p
r
o
ac
h
es.
Ad
d
itio
n
ally
,
a
tech
n
ical
ex
am
in
atio
n
an
d
d
e
b
ate
o
f
th
e
tr
aits
an
d
r
elev
a
n
ce
o
f
s
ev
e
r
al
cy
b
e
r
-
attac
k
m
o
d
els
is
ca
r
r
ied
o
u
t.
T
h
e
m
o
s
t
r
ec
en
t
r
esear
c
h
to
p
ics
ar
e
d
is
cu
s
s
ed
,
alo
n
g
with
cu
ttin
g
-
e
d
g
e
cy
b
e
r
s
ec
u
r
ity
m
eth
o
d
s
f
o
r
p
o
wer
s
y
s
tem
s
an
d
s
u
p
er
g
r
id
s
,
s
u
ch
b
lo
c
k
ch
ain
an
d
q
u
an
tu
m
co
m
p
u
tin
g
.
T
h
e
talk
co
v
er
s
ess
en
tial
p
r
o
tectio
n
m
ec
h
an
is
m
s
an
d
p
r
o
b
lem
-
s
o
lv
in
g
s
tr
ateg
ies.
Fi
n
ally
,
s
o
m
e
th
o
u
g
h
ts
o
n
SGs
'
c
y
b
er
-
s
ec
u
r
ity
in
th
e
f
u
t
u
r
e
ar
e
ex
p
r
ess
ed
[
2
1
]
.
T
h
e
n
ew
p
o
wer
s
y
s
tem
will
f
ac
e
s
i
g
n
if
ican
t
r
is
k
an
d
s
ec
u
r
ity
co
n
ce
r
n
s
b
ec
au
s
e
o
f
th
e
ex
ten
s
iv
e
in
teg
r
atio
n
o
f
cy
b
er
an
d
p
h
y
s
ical
s
y
s
tem
s
.
T
o
tack
le
th
is
is
s
u
e,
a
g
am
e
-
th
eo
r
etic
o
p
tim
u
m
d
e
f
en
s
e
r
eso
u
r
ce
all
o
ca
tio
n
s
tr
ateg
y
is
p
u
t
f
o
r
th
to
p
r
o
ac
tiv
ely
g
u
ar
d
ag
ain
s
t
p
o
s
s
ib
le
cy
b
e
r
attac
k
s
o
n
s
m
ar
t
g
r
id
s
.
Usi
n
g
th
is
tech
n
iq
u
e,
an
id
ea
l
r
eso
u
r
ce
allo
ca
tio
n
f
o
r
a
2
-
lay
er
g
am
e
m
o
d
el
is
p
r
o
d
u
ce
d
.
W
h
ile
th
e
o
th
er
tier
in
v
o
l
v
es
s
ev
er
al
d
ef
en
s
e
n
o
d
es
in
a
n
o
n
c
o
o
p
er
ativ
e
g
am
e,
th
e
f
ir
s
t
lay
er
in
v
o
lv
es
attac
k
er
s
a
n
d
d
ef
e
n
s
iv
e
n
o
d
es
in
an
e
v
o
lu
tio
n
ar
y
g
am
e
.
Af
ter
an
aly
zin
g
th
e
o
f
f
en
s
iv
e
a
n
d
d
ef
en
s
iv
e
ev
o
l
u
tio
n
o
u
tco
m
es
o
f
ev
er
y
s
ce
n
ar
io
,
a
s
o
lu
tio
n
to
th
e
m
u
lti
-
n
o
d
e
r
eso
u
r
ce
allo
ca
tio
n
p
r
o
b
lem
is
g
en
er
ated
.
I
n
co
n
tr
ast
to
ea
r
lier
r
esear
ch
,
th
e
attac
k
er
'
s
co
n
s
tr
ain
ed
r
atio
n
ality
is
co
n
s
id
er
ed
b
ased
o
n
th
e
in
teg
r
ity
,
u
s
ab
ilit
y
,
a
n
d
c
o
n
f
id
en
ti
ality
in
d
ices.
T
o
m
ea
s
u
r
e
p
la
y
er
g
ain
s
,
q
u
a
n
tu
m
r
esp
o
n
s
e
eq
u
aliza
tio
n
is
ad
d
ed
in
th
e
in
ter
im
.
L
astl
y
,
alg
o
r
ith
m
s
ar
e
u
s
ed
to
s
h
o
w
th
at
th
e
ap
p
r
o
ac
h
s
u
g
g
ested
in
th
is
r
esear
ch
is
b
o
th
p
r
ac
tic
al
an
d
ef
f
icien
t
[
2
2
]
.
Dis
t
r
ib
u
ted
d
en
ial
-
of
-
s
er
v
ice
(
DDo
S)
ass
au
lts
ar
e
o
n
e
k
in
d
o
f
cy
b
er
attac
k
th
at
f
r
e
q
u
en
tly
tar
g
ets
s
m
ar
t
g
r
id
n
etwo
r
k
s
.
Fu
r
th
e
r
m
o
r
e
,
s
y
n
ch
r
o
p
h
aso
r
tech
n
o
lo
g
y
s
h
ield
s
th
e
wid
e
-
ar
ea
m
ea
s
u
r
em
en
t
s
y
s
tem
(
W
AM
S)
f
r
o
m
c
o
m
p
licated
d
if
f
icu
lt
s
itu
atio
n
s
b
y
m
an
ag
i
n
g
s
ev
e
r
al
co
n
ce
r
n
s
in
a
g
r
id
.
B
ec
au
s
e
o
f
co
m
m
u
n
icatio
n
p
r
o
to
co
ls
,
v
en
d
o
r
r
estrictio
n
s
,
an
d
th
e
co
m
p
le
x
ity
o
f
th
e
a
s
s
au
lt,
d
etec
tin
g
DDo
S
attac
k
s
is
d
if
f
icu
lt.
Fo
r
m
ea
s
u
r
em
en
t
PMU
d
ata,
attac
k
er
s
tar
g
et
th
e
p
h
aso
r
d
ata
co
n
ce
n
tr
ato
r
(
PDC
)
d
atab
ase
in
W
AM
S.
Nev
er
th
eless
,
d
esig
n
m
ak
es
s
u
r
e
th
at
th
e
en
d
ap
p
licatio
n
m
ak
es
u
s
e
o
f
th
e
r
eg
u
lar
PDC
d
ata
s
tr
ea
m
ev
en
d
u
r
in
g
th
e
in
tr
u
s
io
n
.
PMU
-
g
en
er
ated
d
ata
in
W
AM
S
is
q
u
ick
ly
v
er
if
ied
u
s
in
g
th
e
s
u
g
g
ested
attac
k
d
etec
tio
n
tech
n
iq
u
e.
Var
io
u
s
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
a
r
e
em
p
lo
y
e
d
to
id
e
n
tify
DDo
S
ass
au
lts
;
y
et
th
e
m
o
s
t
ef
f
ec
tiv
e
d
etec
tio
n
m
o
d
el
r
em
ain
s
u
n
claim
ed
.
T
h
is
s
tu
d
y
aim
s
to
d
eter
m
i
n
e
(
a)
th
e
b
est
m
ac
h
in
e
lear
n
in
g
m
eth
o
d
f
o
r
d
etec
tin
g
DDo
S
attac
k
s
an
d
(
b
)
th
e
ac
cu
r
ac
y
o
f
th
e
alg
o
r
ith
m
s
th
at
a
r
e
tau
g
h
t.
T
h
is
s
tu
d
y
o
f
f
er
s
a
h
y
b
r
i
d
ap
p
r
o
ac
h
b
ased
o
n
m
ac
h
in
e
lear
n
in
g
th
at
y
ield
s
8
3
.
2
3
%
ac
c
u
r
ac
y
.
T
h
e
s
u
g
g
ested
m
o
d
el
is
c
r
ea
ted
u
s
in
g
th
e
P
y
th
o
n
co
m
p
iler
,
a
n
d
th
e
o
u
tc
o
m
e
d
em
o
n
s
tr
ates
h
o
w
ef
f
ec
tiv
ely
th
e
s
u
g
g
ested
d
etec
tio
n
tech
n
i
q
u
e
r
aises
th
e
ac
cu
r
ac
y
o
f
DDo
S
ass
au
lt
d
etec
tio
n
[
2
3
]
.
T
h
e
r
es
ea
r
ch
s
u
g
g
ests
a
n
o
v
el
a
p
p
r
o
a
ch
co
m
b
in
in
g
d
ata
m
o
n
ito
r
in
g
an
d
f
u
zz
y
m
ac
h
in
e
lear
n
in
g
m
o
d
el
class
if
icatio
n
t
o
id
en
tify
s
m
ar
t
g
r
id
p
r
o
b
lem
s
.
Her
e,
d
ata
f
r
o
m
th
e
s
m
ar
t
g
r
i
d
h
as
b
ee
n
tr
ac
k
ed
u
s
in
g
im
p
r
o
v
e
d
s
m
ar
t
s
en
s
o
r
m
eter
in
g
,
wh
ich
r
u
n
s
in
t
h
e
c
lo
u
d
at
th
e
ed
g
e
o
f
th
e
n
etwo
r
k
.
Af
te
r
war
d
s
,
th
e
m
o
n
ito
r
ed
d
ata
was
ca
teg
o
r
i
ze
d
u
s
in
g
a
n
a
d
v
er
s
ar
ial
n
e
u
r
al
n
etwo
r
k
with
f
u
zz
y
r
ei
n
f
o
r
ce
m
en
t
e
n
co
d
e
r
.
T
h
r
o
u
g
h
p
u
t,
m
ea
n
av
e
r
ag
e
p
r
ec
is
io
n
,
ac
cu
r
ac
y
,
s
ca
lab
ilit
y
,
an
d
d
e
p
en
d
a
b
ilit
y
ar
e
all
tak
e
n
in
to
c
o
n
s
id
e
r
atio
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Po
w
E
lec
&
Dr
i Sy
s
t
I
SS
N:
2088
-
8
6
9
4
E
n
s
emb
le
lea
r
n
in
g
b
a
s
ed
fa
u
lt
d
etec
tio
n
u
s
in
g
P
MU d
a
ta
in
imb
a
la
n
ce
d
…
(
K
ir
u
th
ika
K
r
is
h
n
a
n
)
855
wh
ile
d
o
in
g
ex
p
er
im
en
tal
s
tu
d
y
.
I
m
p
r
o
v
ed
m
o
n
ito
r
in
g
an
d
p
r
ed
ictio
n
a
p
p
r
o
ac
h
es
ca
n
b
o
o
s
t
th
e
ex
is
tin
g
g
r
id
'
s
p
o
ten
tial
u
tili
za
tio
n
wh
ile
r
ed
u
cin
g
f
a
u
lt
f
r
eq
u
e
n
cy
.
T
h
e
s
u
g
g
ested
m
eth
o
d
ac
h
ie
v
ed
9
3
%
ac
cu
r
ac
y
,
9
4
%
th
r
o
u
g
h
p
u
t,
8
1
%
d
ep
en
d
ab
ilit
y
,
8
9
%
m
ea
n
a
v
er
ag
e
p
r
ec
is
i
o
n
,
a
n
d
9
4
%
s
ca
lab
ilit
y
[
2
4
]
.
T
h
e
liter
atu
r
e
[
2
5
]
o
f
f
er
s
an
ag
g
r
eg
ated
in
teg
er
lin
ea
r
p
r
o
g
r
am
m
in
g
m
eth
o
d
u
t
ilizin
g
m
icr
o
-
s
y
n
ch
r
o
p
h
aso
r
u
n
it
p
lace
m
en
t
f
o
r
m
ac
h
in
e
lear
n
in
g
to
ac
co
m
p
lis
h
co
m
p
lete
o
b
s
er
v
a
b
ilit
y
o
f
th
e
au
to
m
ated
s
m
ar
t
g
r
id
s
,
k
ee
p
i
n
g
in
m
in
d
th
at
th
e
d
is
tr
ib
u
tio
n
s
y
s
tem
s
ar
e
r
ec
o
n
f
ig
u
r
ab
le.
T
h
e
s
u
g
g
ested
s
to
ch
asti
c
ap
p
r
o
ac
h
p
r
o
v
i
d
es a
m
u
lti
-
s
tag
e
m
ec
h
an
is
m
f
o
r
m
icr
o
-
s
y
n
c
h
r
o
p
h
aso
r
u
n
it
p
lacin
g
d
ep
en
d
in
g
o
n
th
e
d
em
an
d
a
n
d
l
o
ad
s
ize
o
f
th
e
s
y
s
tem
,
in
ad
d
itio
n
to
p
r
e
-
p
lan
n
ed
s
ec
tio
n
alizin
g
a
n
d
tie
s
witch
es.
T
h
is
m
eth
o
d
m
ay
a
ls
o
b
e
u
s
ed
to
a
p
p
ly
t
h
e
n
o
-
in
j
ec
tio
n
lim
its
o
f
th
e
r
ep
r
esen
tatio
n
t
o
n
ar
r
o
w
f
in
d
in
g
s
p
ac
e
f
o
r
th
e
is
s
u
e.
I
n
ad
d
itio
n
,
a
n
ew
a
p
p
r
o
ac
h
d
er
iv
ed
f
r
o
m
t
h
e
wh
al
e
o
p
tim
izatio
n
tech
n
i
q
u
e
(
W
OM
)
is
p
r
esen
ted
f
o
r
f
in
d
i
n
g
b
est
d
esig
n
f
o
r
e
v
er
y
p
h
ase
wh
ile
co
n
cu
r
r
e
n
tly
in
cr
ea
s
in
g
th
e
r
eliab
ilit
y
in
d
ic
es
an
d
l
o
wer
in
g
th
e
ex
p
e
n
s
es
ass
o
ciate
d
with
cu
s
to
m
er
d
is
r
u
p
tio
n
s
a
n
d
p
o
wer
o
u
tag
es.
T
h
e
W
OM
h
e
u
r
is
tic
s
er
v
es
as
th
e
f
o
u
n
d
atio
n
f
o
r
th
e
r
estru
ctu
r
in
g
p
r
o
ce
s
s
,
an
d
an
in
teg
e
r
lin
e
ar
p
r
o
g
r
a
m
m
in
g
f
r
am
ewo
r
k
is
u
s
ed
to
ar
r
an
g
e
th
e
m
icr
o
-
s
y
n
c
h
r
o
p
h
aso
r
s
.
T
o
m
a
n
ag
e
th
e
u
n
ce
r
tain
ty
ef
f
ec
ts
,
a
s
to
ch
asti
c
f
r
am
ewo
r
k
d
er
i
v
ed
f
r
o
m
o
n
p
o
in
t
esti
m
atio
n
is
co
n
s
tr
u
cted
,
ac
co
u
n
tin
g
f
o
r
p
r
ed
ictio
n
er
r
o
r
o
r
m
eter
in
g
d
e
v
ice
u
n
ce
r
tain
ty
.
T
h
e
s
u
g
g
ested
ap
p
r
o
ac
h
en
s
u
r
es
d
is
tin
ctn
ess
o
f
th
e
d
is
tr
ib
u
tio
n
n
etwo
r
k
b
o
th
b
ef
o
r
e
an
d
af
ter
r
ea
r
r
an
g
e
m
en
t
with
in
allo
tted
tim
e
lim
it,
as
co
n
f
ir
m
ed
b
y
s
im
u
latio
n
an
d
n
u
m
e
r
ical
r
esu
lts
o
n
an
ac
tu
al
s
y
s
tem
.
Ad
d
itio
n
ally
,
th
e
r
esu
lts
d
em
o
n
s
tr
ate
th
at
ev
en
wh
en
th
e
s
y
s
tem
is
s
u
b
jecte
d
to
v
ar
io
u
s
r
e
co
n
f
ig
u
r
atio
n
s
an
d
to
p
o
lo
g
ies,
s
y
s
tem
o
b
s
er
v
ab
il
ity
m
ay
b
e
ass
u
r
ed
at
v
a
r
y
in
g
lo
ad
le
v
els
[
2
5
]
.
T
h
e
a
n
o
m
aly
d
etec
tio
n
a
n
d
id
en
tific
atio
n
m
o
d
u
le
(
ADI
M)
,
a
u
n
iq
u
e
co
m
p
o
n
e
n
t
p
r
o
p
o
s
e
d
in
th
e
s
tu
d
y
,
is
d
esig
n
ed
t
o
i
d
en
tify
an
o
m
alies
or
er
r
o
n
eo
u
s
d
ata
p
o
in
ts
b
ef
o
r
e
th
e
s
tate
esti
m
atio
n
p
r
o
ce
s
s
b
eg
in
s
.
W
e
p
r
o
v
id
e
a
d
ee
p
lea
r
n
in
g
tec
h
n
iq
u
e
th
at
d
em
o
n
s
tr
ates
ex
ce
p
tio
n
al
p
r
ec
is
io
n
in
d
etec
tin
g
ir
r
eg
u
lar
itie
s
with
in
an
o
n
g
o
in
g
d
ata
s
tr
ea
m
.
C
o
m
p
r
eh
en
s
iv
e
test
in
g
o
n
a
r
an
g
e
o
f
test
s
ce
n
ar
io
s
th
at
ef
f
ec
tiv
ely
c
o
v
er
a
v
ar
iety
o
f
n
etwo
r
k
to
p
o
l
o
g
ies,
tr
an
s
f
o
r
m
e
r
ty
p
es,
an
d
lo
ad
co
n
d
itio
n
s
d
em
o
n
s
tr
a
tes
th
e
ca
p
ab
ilit
y
o
f
ADI
M.
I
t
is
d
em
o
n
s
tr
ated
th
at
an
o
m
alie
s
m
ay
b
e
ef
f
icien
tly
f
o
u
n
d
an
d
r
ec
o
g
n
ized
with
A
DI
M,
h
en
ce
lo
wer
in
g
th
e
r
e
q
u
ir
em
en
t
f
o
r
t
h
e
co
m
p
o
n
en
t
t
h
a
t
d
etec
ts
f
au
lty
d
ata
an
d
en
h
a
n
cin
g
s
tate
esti
m
atio
n
'
s
o
v
er
all
r
esp
o
n
s
iv
en
ess
.
Ou
r
s
tu
d
y
estab
lis
h
es th
e
g
r
o
u
n
d
wo
r
k
f
o
r
c
r
ea
tin
g
an
an
o
m
aly
s
y
s
tem
f
o
r
p
o
wer
s
y
s
tem
m
ea
s
u
r
em
en
ts
th
at
is
b
a
s
ed
o
n
d
etec
tio
n
an
d
id
e
n
tific
atio
n
[
2
6
]
.
T
h
e
d
ata
s
o
u
r
ce
s
an
d
SG
ar
c
h
itectu
r
e
ar
e
b
r
ief
ly
s
u
m
m
a
r
ized
in
t
h
e
p
ap
e
r
'
s
f
ir
s
t
p
ar
t.
Fu
r
th
er
m
o
r
e,
e
x
am
p
les
o
f
f
r
au
d
u
le
n
t
d
ata
attac
k
s
an
d
d
at
a
s
ec
u
r
ity
r
eq
u
ir
em
e
n
ts
ar
e
s
h
o
wn
.
T
h
e
m
o
s
t
r
ec
en
t
ML
-
b
ase
d
d
etec
tio
n
m
eth
o
d
s
ar
e
th
en
s
u
m
m
ar
ize
d
u
s
in
g
th
e
th
r
ee
p
r
im
ar
y
d
etec
ti
o
n
s
ce
n
ar
io
s
:
n
o
n
-
tech
n
ical
lo
s
s
es,
s
tate
esti
m
atio
n
,
an
d
lo
ad
f
o
r
ec
asti
n
g
.
Fin
ally
,
c
o
n
s
id
er
in
g
th
e
lim
itatio
n
s
o
f
th
e
cu
r
r
en
t
m
ac
h
in
e
lear
n
in
g
-
b
as
ed
tech
n
iq
u
es,
we
in
v
esti
g
ate
ad
d
itio
n
al
r
esear
ch
o
p
p
o
r
tu
n
ities
at
th
e
co
n
clu
s
io
n
o
f
th
e
p
r
o
ject.
W
e
s
p
ec
if
i
ca
lly
co
v
er
in
t
r
u
s
io
n
d
etec
tio
n
ag
ain
s
t
h
o
s
tile
ass
a
u
lts
,
a
co
o
p
e
r
ativ
e
an
d
d
ec
e
n
tr
alize
d
d
etec
tio
n
f
r
am
ewo
r
k
,
p
r
iv
ac
y
-
p
r
eser
v
i
n
g
d
etec
tio
n
,
an
d
a
f
ew
p
o
s
s
ib
le
cu
ttin
g
-
ed
g
e
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
[
2
7
]
.
A
h
y
b
r
id
d
e
ep
lear
n
in
g
s
y
s
tem
th
at
tar
g
ets
d
en
ial
-
of
-
s
er
v
ice
ass
au
lts
o
n
th
e
Sm
ar
t
Gr
id
's
c
o
m
m
u
n
icatio
n
n
etwo
r
k
.
T
h
e
g
ated
r
ec
u
r
r
e
n
t
u
n
it
an
d
co
n
v
o
lu
tio
n
al
n
e
u
r
al
n
etwo
r
k
ap
p
r
o
ac
h
es
h
y
b
r
id
ize
th
e
s
u
g
g
ested
ap
p
r
o
ac
h
.
T
h
e
b
en
ch
m
ar
k
cy
b
er
s
ec
u
r
ity
d
ataset
f
r
o
m
th
e
C
an
ad
ian
I
n
s
titu
te
o
f
C
y
b
er
s
ec
u
r
ity
I
n
tr
u
s
i
o
n
Dete
ctio
n
Sy
s
tem
is
u
s
ed
i
n
s
im
u
latio
n
s
.
W
ith
a
co
m
p
r
eh
en
s
iv
e
ac
cu
r
ac
y
r
ate
o
f
9
9
.
7
%,
th
e
s
im
u
latio
n
r
esu
lts
s
h
o
w
th
at
th
e
s
u
g
g
ested
ap
p
r
o
ac
h
w
o
r
k
s
b
etter
th
an
th
e
ex
is
tin
g
in
t
r
u
s
io
n
d
et
ec
tio
n
s
y
s
tem
s
[
2
8
]
.
T
h
er
e
ar
e
s
ev
er
al
o
b
s
tacle
s
to
t
h
e
cy
b
er
s
ec
u
r
ity
o
f
th
e
s
m
ar
t
g
r
id
,
m
o
s
t
o
f
th
em
s
tem
f
r
o
m
m
alev
o
len
t
ass
au
lts
o
n
d
ev
ices
co
n
n
ec
ted
to
th
e
s
y
s
tem
.
T
h
ese
attac
k
s
co
u
ld
tr
y
to
je
o
p
ar
d
ize
th
e
p
r
iv
ac
y
o
f
PMU
an
d
I
P
ca
m
er
a
s
en
s
o
r
d
ata,
o
r
th
ey
m
ig
h
t
tr
y
to
u
n
d
er
m
in
e
th
e
p
o
w
er
s
u
p
p
ly
to
s
p
ec
if
ic
c
u
s
to
m
er
s
.
T
h
er
ef
o
r
e
,
u
s
in
g
s
ec
u
r
ity
m
ea
s
u
r
es
lik
e
n
etw
o
r
k
in
tr
u
s
io
n
p
r
ev
en
tio
n
s
y
s
tem
s
(
NI
PS
)
an
d
f
ir
ewa
lls
e
q
u
ip
p
e
d
with
ea
ch
d
is
tr
ib
u
ted
s
y
s
tem
lin
k
e
d
in
to
th
e
g
r
id
is
th
e
ea
s
iest
m
eth
o
d
to
p
r
ev
e
n
t
s
u
ch
cy
b
er
-
attac
k
is
s
u
es.
T
h
e
g
r
id
is
n
ea
r
ly
im
p
e
n
etr
ab
ly
p
r
o
tect
ed
f
r
o
m
h
ac
k
er
s
th
an
k
s
to
s
ev
er
al
attac
k
an
d
d
e
f
en
s
e
m
ec
h
a
n
is
m
s
.
T
h
u
s
,
em
p
lo
y
in
g
FDI
an
d
MitM
attac
k
s
ce
n
ar
i
o
s
,
th
e
aim
o
f
th
is
s
tu
d
y
is
to
g
iv
e
an
a
n
aly
s
is
o
f
PMU
an
d
I
P
ca
m
er
a
s
en
s
o
r
ass
au
lts
at
th
e
co
m
p
o
n
e
n
t le
v
e
l w
h
en
co
u
p
led
t
o
an
I
E
E
E
1
3
-
n
o
d
e
s
y
s
tem
.
Gr
id
Attack
An
aly
ze
r
,
a
s
m
ar
t g
r
id
attac
k
an
aly
s
is
to
o
l,
is
u
s
ed
t
o
co
llect
d
ata
an
d
s
im
u
late
t
h
e
r
esear
ch
[
2
9
]
.
Su
ch
attac
k
s
ar
e
co
n
ce
aled
b
y
co
v
er
tly
m
o
d
if
y
i
n
g
th
e
SC
A
DA
an
d
PMU
m
etr
ics.
T
h
is
p
ap
er
in
v
esti
g
ates
cy
b
er
-
p
h
y
s
ical
ass
au
lts
th
at
ar
e
co
v
er
t
a
n
d
tar
g
et
p
o
wer
s
y
s
tem
s
.
B
y
f
lick
in
g
th
e
co
r
r
esp
o
n
d
in
g
s
witch
es
o
r
cir
c
u
it
b
r
ea
k
e
r
s
,
o
n
e
o
r
m
an
y
lin
es
an
d
b
u
s
es
ca
n
b
e
in
ter
r
u
p
ted
d
u
r
in
g
o
n
e
o
f
th
ese
ass
au
lts
.
T
h
e
r
esear
ch
d
e
v
elo
p
s
a
f
r
am
ewo
r
k
b
ased
o
n
th
e
non
-
lin
ea
r
p
o
wer
f
lo
w
m
o
d
el
to
ch
ar
ac
ter
ize
s
u
ch
attac
k
s
an
d
s
u
g
g
ests
a
m
eth
o
d
f
o
r
s
p
o
ttin
g
s
u
ch
cy
b
er
-
p
h
y
s
ical
attac
k
s
u
s
in
g
s
witch
in
g
tr
a
n
s
ien
ts
.
T
h
e
d
etec
tio
n
tech
n
iq
u
e
tak
es
u
s
e
o
f
th
e
f
a
ct
th
at
a
p
h
y
s
ical
s
ep
ar
atio
n
will
r
esu
lt
in
a
h
ig
h
tr
an
s
ien
t
r
ate
f
o
r
th
e
s
y
s
tem
.
T
h
ese
PMU
-
o
b
s
er
v
ed
tr
an
s
ien
t
c
o
m
p
o
n
en
ts
ar
e
u
tili
ze
d
to
d
etec
t
co
v
er
t
lin
e
d
is
co
n
n
ec
tio
n
an
d
b
u
s
b
lack
o
u
t
ass
au
lts
,
a
s
well
a
s
to
v
alid
ate
th
e
co
r
r
ec
tn
ess
o
f
th
e
s
tead
y
-
s
tate
v
al
u
es
o
f
SC
ADA
an
d
PMU
m
ea
s
u
r
em
en
t
s
.
T
h
e
s
u
g
g
ested
m
eth
o
d
m
a
y
b
e
ab
le
t
o
i
d
en
tify
cy
b
er
-
p
h
y
s
ical
ass
au
lts
th
at
m
ask
f
r
eq
u
en
t
lin
e
d
is
co
n
n
ec
tio
n
s
an
d
b
u
s
f
ailu
r
es
u
n
d
er
v
ar
i
o
u
s
lo
ad
s
ce
n
ar
io
s
,
ac
co
r
d
in
g
to
ex
p
er
im
en
ts
co
n
d
u
cted
o
n
th
e
I
E
E
E
3
0
b
u
s
s
y
s
tem
[
3
0
]
.
T
h
e
liter
atu
r
e
d
is
cu
s
s
es
th
e
ch
allen
g
e
o
f
g
u
ar
a
n
teein
g
p
r
ec
is
e
o
n
lin
e
tr
an
s
ien
t
s
tab
ilit
y
p
r
e
d
ictio
n
in
co
n
tem
p
o
r
ar
y
p
o
wer
s
y
s
tem
s
,
wh
ich
r
ely
in
cr
ea
s
in
g
ly
h
ea
v
ily
o
n
s
m
ar
t
g
r
id
tech
n
o
lo
g
ies
an
d
ar
e
th
u
s
m
o
r
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
6
9
4
I
n
t J Po
w
E
lec
&
Dr
i Sy
s
t
,
Vo
l.
16
,
No
.
2
,
J
u
n
e
20
25
:
8
5
1
-
8
6
3
856
v
u
ln
er
ab
le
to
c
y
b
er
attac
k
s
.
De
s
p
ite
th
e
r
ap
id
g
r
o
wth
o
f
tech
n
o
lo
g
y
,
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
f
o
r
s
tab
ilit
y
p
r
ed
ictio
n
p
r
esen
tly
lack
th
e
r
esil
ien
ce
n
ee
d
ed
to
ef
f
ec
tiv
ely
r
esis
t
th
e
co
m
p
lex
a
n
d
ev
e
r
-
ev
o
lv
in
g
n
atu
r
e
o
f
cy
b
er
attac
k
s
.
T
h
e
s
tu
d
y
also
ev
alu
ates
th
e
im
p
ac
t
o
f
to
p
o
lo
g
ical
m
o
d
if
icatio
n
s
an
d
t
h
e
in
c
o
r
p
o
r
atio
n
o
f
r
en
ewa
b
le
en
er
g
y
o
n
th
ese
m
ac
h
in
e
lear
n
in
g
-
b
ased
tech
n
iq
u
es,
as
well
a
s
cy
b
er
attac
k
s
.
T
r
an
s
ien
t
s
tab
ilit
y
p
r
ed
ictio
n
tech
n
iq
u
es
u
s
ed
o
n
lin
e
ar
e
ess
en
tial
f
o
r
tr
ac
k
i
n
g
a
n
d
p
r
ed
ictin
g
p
o
wer
s
y
s
tem
b
eh
av
io
r
in
r
ea
l
tim
e
d
u
r
in
g
d
is
r
u
p
tio
n
s
.
T
o
ass
ess
th
e
r
o
b
u
s
tn
ess
o
f
th
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
s
in
th
e
co
n
te
x
t
o
f
th
e
p
o
ten
tial
f
o
r
attac
k
er
s
to
d
is
r
u
p
t
co
m
m
u
n
ic
atio
n
an
d
h
en
ce
af
f
ec
t
th
e
p
o
wer
s
y
s
tem
,
th
e
s
tu
d
y
r
ep
r
o
d
u
ce
s
m
an
y
s
ce
n
ar
io
s
.
T
h
e
r
esu
lts
s
h
o
w
th
at
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
p
er
f
o
r
m
wo
r
s
e
d
u
r
in
g
cy
b
er
attac
k
s
,
r
esu
l
tin
g
in
a
lar
g
e
d
r
o
p
in
th
e
co
r
r
ec
tn
ess
o
f
tr
a
n
s
ien
t
s
tab
ilit
y
f
o
r
ec
asts
wh
e
n
co
m
p
ar
ed
to
n
o
r
m
al
wo
r
k
in
g
s
ettin
g
s
.
T
h
is
d
em
o
n
s
tr
ates
h
o
w
v
ital
it
is
to
h
a
v
e
c
u
ttin
g
-
ed
g
e
c
y
b
er
s
ec
u
r
ity
d
ef
e
n
s
es
to
m
ain
tain
p
o
wer
s
y
s
tem
s
'
ca
p
ac
ity
f
o
r
p
r
ed
ictio
n
[
3
1
]
.
T
o
d
ay
'
s
cy
b
er
-
p
h
y
s
ical
g
r
id
s
ar
e
h
ea
v
ily
in
teg
r
atin
g
co
s
t
-
ef
f
ec
tiv
e
co
m
m
u
n
icatio
n
n
etwo
r
k
s
an
d
th
e
in
ter
n
et
o
f
th
i
n
g
s
(
I
o
T
)
,
wh
i
ch
h
as
led
to
s
er
io
u
s
s
ec
u
r
it
y
p
r
o
b
lem
s
.
Mo
r
e
s
p
ec
if
icall
y
,
n
etwo
r
k
s
ec
u
r
ity
b
ec
o
m
es
m
o
r
e
v
u
ln
er
a
b
le
d
u
e
to
wir
eless
co
m
m
u
n
icatio
n
tech
n
o
lo
g
y
.
I
n
ad
d
itio
n
to
t
h
e
well
-
r
esear
ch
ed
cy
b
er
s
ec
u
r
ity
ch
allen
g
es,
we
als
o
n
ee
d
to
co
n
s
id
er
p
h
y
s
ical
lay
er
s
ec
u
r
ity
.
As
a
r
esu
lt,
a
lo
t
o
f
wo
r
k
h
as
g
o
n
e
in
to
cr
ea
tin
g
a
s
o
lu
tio
n
to
co
p
e
with
cy
b
er
s
ec
u
r
ity
p
r
o
b
lem
s
.
B
u
t
th
er
e
h
asn
'
t
b
ee
n
m
u
ch
wo
r
k
d
o
n
e
o
n
cr
ea
tin
g
en
cr
o
ac
h
m
e
n
t
f
in
d
in
g
s
y
s
tem
s
f
o
r
p
h
y
s
ical
s
ec
u
r
ity
.
T
h
is
wo
r
k
p
r
o
v
i
d
es
a
s
h
ar
p
m
o
d
el
th
at
d
etec
ts
an
d
class
if
ies
a
s
s
au
lts
u
s
in
g
a
co
m
b
in
atio
n
o
f
m
ac
h
in
e
lear
n
i
n
g
t
ec
h
n
iq
u
es,
in
cl
u
d
in
g
id
en
tify
i
n
g
th
e
k
in
d
o
f
attac
k
at
th
e
p
h
y
s
ical
b
a
n
d
.
Ad
d
itio
n
ally
,
th
e
s
u
g
g
ested
m
eth
o
d
lo
c
alize
s
th
e
attac
k
o
r
v
u
ln
e
r
ab
ilit
y
to
ce
r
tain
s
y
s
te
m
p
ar
am
eter
s
o
r
attr
ib
u
tes,
wh
ich
ca
n
ass
is
t
cy
b
er
s
ec
u
r
ity
s
p
ec
ialis
ts
in
r
ed
u
cin
g
th
e
ef
f
ec
t
o
f
attac
k
s
o
n
co
m
m
u
n
icatio
n
g
r
id
s
.
T
h
e
p
r
o
p
o
s
ed
m
o
d
el
is
co
m
p
a
r
ed
with
co
n
v
en
tio
n
al
m
ac
h
in
e
lear
n
in
g
class
if
ier
s
u
s
in
g
an
SG d
ataset
th
at
is
s
im
u
lated
at
Oak
R
id
g
e
Natio
n
al
L
ab
o
r
ato
r
ies.
B
y
d
iv
id
in
g
th
e
d
ata
an
d
ca
lcu
lati
n
g
th
e
co
m
p
a
r
is
o
n
b
etwe
en
th
e
c
o
n
f
in
e
d
m
etr
ics
g
en
er
ate
d
b
y
th
e
s
u
g
g
es
ted
m
o
d
el,
th
e
co
n
f
in
em
en
t
o
f
er
r
o
r
s
an
d
attac
k
s
is
v
er
if
ied
.
W
h
en
co
n
tr
ast
to
p
ee
r
m
eth
o
d
o
lo
g
ies,
th
e
r
esu
lts
s
h
o
w
h
o
w
g
o
o
d
th
is
m
eth
o
d
is
at
class
if
y
in
g
th
r
ea
ts
an
d
co
n
f
i
n
in
g
th
em
[
3
2
]
.
T
h
e
latest
wav
e
o
f
r
esil
ien
t
f
ak
e
d
ata
in
jectio
n
attac
k
m
eth
o
d
s
n
ec
e
s
s
itates
a
d
ee
p
co
m
p
r
eh
e
n
s
io
n
o
f
th
e
ass
o
ciate
d
p
o
wer
g
r
id
n
etwo
r
k
'
s
s
tr
u
ctu
r
e.
T
h
is
s
tu
d
y
s
u
g
g
ests
th
r
ee
m
eth
o
d
s
th
at
ar
e
in
d
ep
en
d
en
t
o
f
n
etwo
r
k
ar
c
h
itectu
r
e
f
o
r
in
tr
o
d
u
cin
g
f
ictitio
u
s
d
ata
in
to
th
e
s
m
ar
t
g
r
id
.
T
h
e
s
e
m
eth
o
d
s
in
clu
d
e
d
elta
th
r
esh
o
ld
s
,
lin
e
ar
r
eg
r
ess
io
n
,
a
n
d
li
n
ea
r
r
eg
r
ess
io
n
with
tim
estam
p
.
I
t
is
in
te
n
d
ed
to
clo
s
e
th
e
g
a
p
s
in
r
ea
l
-
tim
e
d
ata
m
ea
s
u
r
em
en
ts
,
h
e
n
c
e
in
cr
ea
s
in
g
th
e
p
r
o
b
a
b
ilit
y
th
a
t
tam
p
er
ed
d
ata
w
o
n
'
t
b
e
d
etec
ted
.
Mo
d
er
n
d
ef
en
s
e
s
tr
ateg
ies
lik
e
A
C
s
tate
e
s
tim
a
tio
n
,
tem
p
o
r
al
b
eh
a
v
i
o
r
s
b
ased
on
f
alse
d
ata
d
etec
tio
n
,
b
ad
d
ata
d
etec
tio
n
,
an
d
SVM
d
em
o
n
s
tr
ate
th
e
r
esil
ien
ce
o
f
th
e
s
u
g
g
ested
attac
k
ap
p
r
o
ac
h
es [
3
3
]
.
B
ased
o
n
th
e
a
tta
ck
e
r
’
s
p
e
r
s
p
e
c
tiv
e
a
n
d
u
s
i
n
g
t
h
e
P
MU
as
t
h
e
a
ttac
k
-
d
e
f
e
n
s
e
ta
r
g
et
in
t
h
e
p
o
w
er
s
y
s
t
em
,
a
m
u
lti
-
l
ev
el
g
am
e
m
o
d
el
f
o
r
FD
I
A
is
s
u
g
g
est
ed
.
I
n
a
m
u
lti
-
s
ta
g
e
g
a
m
e
,
s
p
e
ci
al
att
e
n
ti
o
n
is
p
ai
d
t
o
d
ata
m
a
n
i
p
u
lat
io
n
,
s
tr
ate
g
i
c
m
o
d
if
i
ca
ti
o
n
s
,
a
n
d
m
u
lti
-
p
a
th
at
ta
ck
s
.
T
h
e
P
MU
s
et
u
p
f
r
o
m
t
h
e
i
n
t
r
u
d
e
r
’
s
v
iew
is
u
s
ed
to
g
en
er
ate
f
a
k
e
d
ata
,
t
h
e
s
t
r
i
k
e
r
a
n
g
e
is
o
p
t
im
ize
d
,
an
d
th
e
a
ttac
k
r
ep
er
c
u
s
s
i
o
n
s
a
r
e
ass
ess
e
d
.
Se
co
n
d
,
t
h
e
Nas
h
eq
u
il
ib
r
i
u
m
p
o
i
n
t
m
ay
b
e
es
ta
b
lis
h
ed
b
y
u
s
i
n
g
th
e
id
ea
s
o
f
tw
o
-
p
l
a
y
e
r
ze
r
o
-
s
u
m
g
a
m
e
t
h
e
o
r
y
a
n
d
a
cc
o
u
n
ti
n
g
f
o
r
b
o
t
h
t
o
t
al
i
n
c
o
m
e
a
n
d
m
u
lti
-
p
a
th
at
tac
k
s
i
n
m
u
lt
i
-
s
ta
g
e
g
a
m
es
t
o
d
ete
r
m
in
e
t
h
e
b
est
a
tta
ck
-
d
e
f
e
n
s
e
c
o
m
b
in
ati
o
n
.
L
astl
y
,
a
d
is
c
u
s
s
i
o
n
o
f
t
h
e
e
x
p
er
i
m
e
n
t
f
i
n
d
in
g
s
f
o
r
b
o
t
h
s
in
g
l
e
-
a
n
d
m
u
lti
-
s
t
ag
e
g
am
es
f
o
ll
o
ws.
T
h
e
s
i
m
u
l
ati
o
n
'
s
f
i
n
d
i
n
g
s
s
h
o
w
t
h
at
at
ta
ck
er
s
c
a
n
m
o
r
e
e
f
f
e
cti
v
el
y
a
n
d
ef
f
i
cie
n
t
ly
e
m
p
l
o
y
t
h
e
s
u
g
g
est
e
d
m
u
lti
-
s
ta
g
e
g
a
m
e
a
p
p
r
o
ac
h
[
3
4
]
.
L
it
er
a
t
u
r
e
p
r
es
en
ts
a
r
ea
lis
tic
b
i
-
le
v
e
l
m
i
x
e
d
-
i
n
te
g
e
r
l
in
e
a
r
p
r
o
g
r
a
m
m
i
n
g
(
B
MI
L
P
)
m
o
d
el
t
o
r
ep
lic
at
e
f
r
a
u
d
u
l
en
t
d
ata
i
n
j
ec
t
io
n
s
(
F
DI
s
)
,
w
h
i
ch
a
tte
m
p
t
t
o
o
v
e
r
l
o
a
d
s
e
v
e
r
a
l
li
n
es
o
f
tr
a
n
s
m
is
s
i
o
n
a
n
d
c
r
e
ate
a
c
o
ll
a
p
s
e
o
f
p
o
w
er
s
u
p
p
l
y
i
n
v
er
y
b
i
g
g
r
i
d
s
.
I
n
co
n
t
r
as
t
t
o
p
r
ev
io
u
s
s
t
u
d
ies
,
t
h
is
m
o
d
el
ac
c
o
u
n
ts
f
o
r
th
e
p
o
s
s
ib
ili
ty
t
h
a
t
att
ac
k
e
r
s
m
a
y
o
n
l
y
h
a
v
e
r
est
r
icte
d
ac
ce
s
s
to
m
ea
s
u
r
em
e
n
t
b
u
s
es
a
n
d
s
im
u
l
at
es
p
r
o
b
l
em
s
o
n
ce
r
t
ai
n
li
n
es
t
h
a
t
ar
e
o
v
e
r
l
o
o
k
e
d
b
y
c
u
r
r
en
t
DC
s
ta
te
es
ti
m
ati
o
n
.
F
u
r
t
h
e
r
m
o
r
e
,
it
is
d
e
m
o
n
s
tr
ate
d
th
a
t
s
te
alt
h
y
FD
I
s
m
ay
b
e
id
en
tifi
e
d
u
s
in
g
an
o
b
s
e
r
v
ati
o
n
f
r
a
m
ew
o
r
k
b
ase
d
o
n
r
e
cu
r
s
i
v
e
we
ig
h
te
d
l
ea
s
t
-
s
q
u
a
r
es
(
W
L
S
)
s
tat
e
est
im
ati
o
n
,
b
u
t
cl
ass
ic
al
W
L
S
est
im
ati
o
n
is
u
n
a
b
le
to
d
e
tec
t
FD
I
s
.
T
h
is
h
e
lp
s
p
r
o
t
ec
t
t
h
e
s
y
s
te
m
ag
ai
n
s
t
m
a
n
y
k
in
d
s
o
f
t
h
r
ea
ts
.
T
wo
b
en
c
h
m
ar
k
s
ar
e
u
til
ize
d
t
o
v
e
r
i
f
y
th
e
e
f
f
ic
ac
y
o
f
t
h
e
s
u
g
g
est
e
d
at
tac
k
m
o
d
el
an
d
d
ete
cti
o
n
s
y
s
t
em
o
n
t
h
e
r
e
al
g
r
i
d
:
th
e
I
E
E
E
1
1
8
-
b
u
s
b
e
n
ch
m
a
r
k
an
d
a
2
0
0
0
-
b
u
s
a
r
t
if
ici
al
g
r
i
d
t
h
at
m
i
m
ics
th
e
T
e
x
as
,
USA
,
el
ec
tr
i
ca
l n
et
wo
r
k
[
3
5
]
.
L
it
e
r
at
u
r
e
[
3
6
]
p
r
o
p
o
s
es
a
m
ac
h
i
n
e
le
ar
n
i
n
g
m
et
h
o
d
th
at
u
s
es
p
o
in
t o
f
co
m
m
o
n
co
u
p
li
n
g
(
PC
C
)
s
e
n
s
o
r
s
a
lo
n
e
t
o
id
en
ti
f
y
c
y
b
e
r
a
tta
ck
s
i
n
p
h
o
to
v
o
lt
aic
(
PV)
f
a
r
m
s
.
Fi
r
s
t
,
a
t
h
o
r
o
u
g
h
cy
b
e
r
-
at
tac
k
m
o
d
el
f
o
r
a
p
h
o
t
o
v
o
lta
ic
f
ar
m
is
cr
ea
t
ed
,
c
o
n
s
id
e
r
i
n
g
t
h
e
v
a
r
ie
ty
o
f
o
p
er
ati
o
n
al
cir
c
u
m
s
t
a
n
ce
s
.
T
w
o
c
y
b
e
r
a
tta
c
k
k
i
n
d
s
t
h
at
a
r
e
o
f
t
e
n
h
ar
d
e
r
t
o
i
d
en
ti
f
y
a
r
e
p
a
r
ti
c
u
lar
ly
i
n
c
lu
d
ed
i
n
th
e
att
ac
k
m
o
d
el
.
W
e
p
r
es
en
t a
n
d
c
o
n
t
r
ast
a
co
n
v
o
l
u
ti
o
n
al
n
eu
r
a
l
n
et
wo
r
k
(
C
N
N)
u
s
i
n
g
m
ic
r
o
-
p
h
ase
m
ea
s
u
r
e
m
e
n
t
u
n
its
(
μ
PMU
)
a
n
d
r
aw
ele
ct
r
ic
wa
v
ef
o
r
m
wit
h
f
i
g
u
r
es
o
f
m
e
r
it
in
r
elat
io
n
to
e
x
is
ti
n
g
m
ac
h
i
n
e
le
a
r
n
in
g
te
ch
n
i
q
u
es
.
I
n
th
e
e
n
d
,
a
d
is
t
r
i
b
u
te
d
g
r
i
d
c
o
n
s
is
ti
n
g
o
f
I
E
E
E
3
7
b
u
s
es
a
n
d
s
o
la
r
f
ar
m
s
is
esta
b
l
is
h
ed
as
a
test
b
ed
f
o
r
c
y
b
er
-
p
h
y
s
i
ca
l
s
ec
u
r
it
y
.
A
f
r
a
m
ew
o
r
k
f
o
r
r
e
al
-
ti
m
e
s
i
m
u
la
ti
o
n
,
d
et
e
cti
o
n
,
a
n
d
v
is
u
ali
za
t
io
n
is
c
r
e
at
ed
t
o
s
h
o
w
h
o
w
t
h
e
s
u
g
g
este
d
a
p
p
r
o
a
c
h
w
o
r
k
s
i
n
an
a
ct
u
al
s
et
ti
n
g
.
T
h
e
f
in
d
i
n
g
s
d
em
o
n
s
t
r
at
e
th
at
t
h
e
s
u
g
g
est
e
d
m
a
ch
in
e
le
ar
n
i
n
g
tec
h
n
i
q
u
es
ar
e
ca
p
a
b
l
e
o
f
a
c
h
i
ev
i
n
g
s
u
f
f
ici
e
n
t
r
es
ili
en
ce
a
n
d
d
et
ec
ti
o
n
a
cc
u
r
ac
y
in
a
r
a
n
g
e
o
f
ass
au
lt
s
it
u
at
io
n
s
[
3
6
]
.
L
i
te
r
at
u
r
e
[
3
7
]
b
u
il
d
s
a
n
I
E
E
E
-
1
1
8
-
n
o
d
e
p
o
w
er
n
etw
o
r
k
a
n
d
a
2
0
0
-
n
o
d
e
s
c
ale
-
f
r
ee
in
f
o
r
m
at
io
n
n
e
tw
o
r
k
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Po
w
E
lec
&
Dr
i Sy
s
t
I
SS
N:
2088
-
8
6
9
4
E
n
s
emb
le
lea
r
n
in
g
b
a
s
ed
fa
u
lt
d
etec
tio
n
u
s
in
g
P
MU d
a
ta
in
imb
a
la
n
ce
d
…
(
K
ir
u
th
ika
K
r
is
h
n
a
n
)
857
u
s
i
n
g
t
h
e
c
o
n
c
e
p
t
o
f
a
n
i
n
t
r
ic
ate
g
r
i
d
t
o
c
r
e
ate
a
s
e
lf
-
s
u
s
ta
i
n
i
n
g
m
o
d
el
.
F
o
ll
o
w
in
g
t
h
e
p
r
o
c
ess
o
f
i
n
f
o
r
m
at
io
n
ca
t
al
o
g
u
i
n
g
f
o
r
d
ete
cti
n
g
a
n
d
r
an
k
i
n
g
th
e
n
etw
o
r
k
'
s
j
u
n
cti
o
n
s
,
a
v
u
ln
e
r
a
b
il
it
y
an
al
y
s
is
is
c
o
n
d
u
ct
e
d
,
a
n
d
a
s
t
r
e
am
m
o
d
el
o
f
t
h
e
p
o
w
er
c
y
b
e
r
-
p
h
y
s
ical
s
y
s
te
m
is
b
u
i
lt
b
as
e
d
o
n
t
h
e
n
o
d
e
c
ar
r
y
i
n
g
c
ap
ac
it
y
.
Sim
u
lati
o
n
d
e
m
o
n
s
tr
ates
th
a
t
i
n
t
e
n
ti
o
n
al
ass
au
lts
ca
u
s
e
t
h
e
s
y
s
te
m
s
t
r
u
ct
u
r
e
t
o
b
r
ea
k
d
o
wn
m
o
r
e
q
u
ic
k
l
y
t
h
a
n
r
a
n
d
o
m
atta
c
k
s
d
o
,
a
n
d
t
h
a
t
f
ai
lu
r
es
m
a
y
b
e
s
u
cc
ess
f
u
ll
y
s
to
p
p
e
d
f
r
o
m
s
p
r
ea
d
i
n
g
b
y
r
ais
in
g
t
h
e
n
o
d
e
th
r
es
h
o
l
d
.
[
3
7
]
.
L
ite
r
at
u
r
e
[
3
8
]
p
u
t
f
o
r
w
ar
d
a
“
q
u
a
n
t
u
m
d
wa
r
f
m
o
n
g
o
o
s
e
o
p
t
im
iza
ti
o
n
wi
th
e
n
s
em
b
le
d
e
ep
le
a
r
n
in
g
b
ase
d
i
n
t
r
u
s
io
n
d
et
ec
ti
o
n
(
QDM
O
-
E
D
L
I
D
)
te
c
h
n
iq
u
e
“
i
n
th
e
co
n
t
e
x
t
o
f
C
PS
.
T
h
e
QD
MO
-
E
DL
I
D
m
et
h
o
d
t
h
a
t
is
b
e
in
g
d
is
cu
s
s
e
d
u
s
es
en
s
em
b
l
e
le
ar
n
i
n
g
a
n
d
f
e
at
u
r
e
s
ele
cti
o
n
(
FS
)
t
o
i
d
e
n
ti
f
y
in
tr
u
s
io
n
s
.
T
h
e
QDM
O
al
g
o
r
it
h
m
is
u
s
e
d
b
y
t
h
e
QDM
O
-
E
DL
I
D
m
et
h
o
d
to
c
h
o
o
s
e
f
e
a
tu
r
e
s
u
b
s
e
ts
.
A
d
d
iti
o
n
al
ly
,
a
m
i
x
o
f
d
ee
p
b
eli
ef
n
etw
o
r
k
s
(
DB
N)
,
c
o
n
v
o
lu
ti
o
n
r
esi
d
u
a
l
n
etw
o
r
k
s
(
C
R
N
)
,
an
d
d
ee
p
a
u
t
o
en
c
o
d
er
(
D
AE
)
m
o
d
els
is
u
ti
liz
ed
to
ca
t
e
g
o
r
i
ze
i
n
t
r
u
s
i
o
n
s
.
Usi
n
g
b
e
n
c
h
m
a
r
k
i
n
t
r
u
s
i
o
n
d
a
tase
ts
,
th
e
QD
MO
-
E
D
L
I
D
te
ch
n
i
q
u
e'
s
e
x
p
e
r
im
e
n
ta
l
r
es
u
lts
w
er
e
e
v
alu
at
ed
.
T
h
e
s
i
m
u
l
a
ti
o
n
'
s
o
u
t
c
o
m
es
d
em
o
n
s
t
r
a
t
e
d
t
h
e
QD
M
O
-
E
D
L
I
D
a
p
p
r
o
a
c
h
'
s
i
n
c
r
e
as
e
d
e
f
f
e
c
ti
v
e
n
e
s
s
i
n
r
e
l
a
t
i
o
n
t
o
s
e
v
e
r
a
l
p
e
r
f
o
r
m
a
n
c
e
m
e
t
r
i
c
s
[
3
8
]
.
I
t
i
s
e
v
i
d
e
n
t
f
r
o
m
t
h
e
p
r
e
c
e
d
i
n
g
p
a
p
e
r
t
h
a
t
a
l
a
r
g
e
b
o
d
y
o
f
r
e
s
e
a
r
c
h
h
a
s
d
e
m
o
n
s
t
r
a
t
e
d
h
o
w
t
h
e
S
MO
T
E
a
n
d
E
n
s
e
m
b
l
e
l
e
a
r
n
i
n
g
al
g
o
r
i
t
h
m
s
e
n
h
a
n
c
e
c
at
e
g
o
r
i
z
a
tio
n
.
T
h
e
s
u
b
j
e
ct
o
f
c
y
b
e
r
a
t
t
a
c
k
d
e
te
c
t
i
o
n
a
n
d
cl
ass
i
f
i
c
at
i
o
n
is
a
d
d
r
es
s
e
d
i
n
t
h
is
p
a
p
e
r
b
y
t
h
e
a
p
p
l
i
c
at
i
o
n
o
f
t
h
e
e
n
s
e
m
b
l
e
l
e
a
r
n
i
n
g
m
e
t
h
o
d
a
n
d
t
h
e
S
MO
T
E
s
a
m
p
l
i
n
g
s
t
r
at
e
g
y
.
T
h
e
e
n
s
e
m
b
l
e
le
a
r
n
i
n
g
a
l
g
o
r
i
t
h
m
is
u
s
e
d
t
o
co
m
p
a
r
e
a
v
a
r
i
e
t
y
o
f
m
e
t
h
o
d
s
,
i
n
c
l
u
d
i
n
g
l
o
g
is
t
i
c
r
e
g
r
e
s
s
i
o
n
,
m
u
l
t
i
l
a
y
e
r
p
e
r
ce
p
t
r
o
n
m
o
d
e
l
i
n
g
,
s
u
p
p
o
r
t
v
e
c
t
o
r
c
l
as
s
i
f
i
e
r
s
,
a
n
d
d
e
c
is
i
o
n
t
r
e
e
c
l
a
s
s
i
f
i
e
r
s
.
T
h
e
t
e
c
h
n
i
q
u
e
o
f
e
n
s
e
m
b
l
e
l
e
a
r
n
i
n
g
u
s
e
s
SV
C
,
M
L
P
,
l
o
g
i
s
ti
c
r
e
g
r
e
s
s
i
o
n
,
a
n
d
d
e
c
i
s
i
o
n
t
r
e
e
a
l
g
o
r
i
t
h
m
s
as
c
o
r
e
c
l
a
s
s
i
f
i
e
r
s
,
w
i
t
h
l
o
g
is
ti
c
r
e
g
r
es
s
i
o
n
s
e
r
v
i
n
g
a
s
t
h
e
m
e
t
a
cl
a
s
s
i
f
i
e
r
.
S
e
v
e
r
al
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
m
s
a
r
e
i
m
p
l
e
m
e
n
t
e
d
a
n
d
c
o
n
t
r
a
s
t
e
d
w
i
t
h
t
h
e
s
ta
c
k
e
n
s
e
m
b
l
e
l
e
a
r
n
i
n
g
s
t
r
a
t
e
g
y
.
A
n
e
n
s
e
m
b
l
e
l
e
a
r
n
i
n
g
-
b
a
s
e
d
t
e
c
h
n
i
q
u
e
f
o
r
i
d
e
n
t
i
f
y
i
n
g
P
M
U
c
y
b
e
r
a
t
t
a
c
k
s
is
p
u
t
f
o
r
wa
r
d
i
n
t
h
i
s
p
a
p
e
r
.
T
h
e
s
p
ec
i
f
i
e
d
e
n
s
e
m
b
l
e
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
i
s
c
o
n
t
r
as
t
e
d
w
it
h
t
h
e
o
u
t
c
o
m
e
s
o
f
m
a
n
y
s
i
n
g
l
e
l
e
a
r
n
in
g
t
e
c
h
n
i
q
u
e
s
,
s
u
c
h
as
l
o
g
is
ti
c
r
e
g
r
e
s
s
i
o
n
,
d
e
c
is
i
o
n
t
r
e
e
s
,
M
L
P
c
la
s
s
i
f
ie
r
s
,
a
n
d
s
u
p
p
o
r
t
v
e
c
t
o
r
m
a
c
h
i
n
e
s
.
P
M
U
in
f
o
r
m
a
t
i
o
n
o
n
t
h
e
i
s
s
u
e
a
n
d
n
o
r
m
a
l
o
p
e
r
a
t
i
o
n
s
i
s
c
o
l
l
e
c
t
e
d
.
P
r
e
p
r
o
c
e
s
s
i
n
g
is
d
o
n
e
o
n
t
h
e
d
a
t
a
s
et
t
o
r
e
m
o
v
e
m
i
s
s
i
n
g
v
a
l
u
e
s
a
n
d
o
u
t
l
i
e
r
s
.
T
h
e
e
n
s
e
m
b
l
e
le
a
r
n
i
n
g
a
l
g
o
r
i
t
h
m
i
s
c
o
m
p
a
r
e
d
i
n
a
n
u
n
b
a
l
a
n
c
e
d
e
n
v
i
r
o
n
m
e
n
t
w
i
t
h
o
t
h
e
r
s
i
n
g
l
e
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
es,
i
n
c
l
u
d
i
n
g
l
o
g
i
s
t
i
c
r
e
g
r
e
s
s
i
o
n
,
d
e
c
i
s
i
o
n
t
r
e
e
s
,
M
L
P
c
l
as
s
i
f
i
e
r
s
,
a
n
d
s
u
p
p
o
r
t
v
e
c
t
o
r
m
a
c
h
i
n
e
s
,
i
n
o
r
d
e
r
t
o
d
e
t
ec
t
PM
U
c
y
b
e
r
a
t
t
ac
k
s
.
2.
E
NS
E
M
B
L
E
L
E
A
RNING
-
B
AS
E
D
P
M
U
CYB
E
R
A
T
T
A
CK
DE
T
E
C
T
I
O
N
US
I
NG
SM
O
T
E
SAM
P
L
I
NG
A
co
llectio
n
o
f
d
ata
ca
lled
"PMU
cy
b
er
-
attac
k
Dete
ctio
n
"
was
cr
ea
ted
s
p
ec
if
ically
to
h
e
lp
m
ac
h
in
e
lear
n
in
g
m
o
d
els
id
e
n
tify
alleg
atio
n
s
o
f
cy
b
er
attac
k
s
.
A
v
a
r
ie
ty
o
f
PMU
d
ata
o
b
s
er
v
e
d
at
v
ar
io
u
s
p
lace
s
in
th
e
p
o
wer
n
etwo
r
k
ar
e
in
clu
d
ed
i
n
th
e
d
ataset.
I
n
ad
d
itio
n
,
o
n
l
y
f
ew
r
e
-
s
ea
r
ch
es
h
av
e
test
ed
th
e
is
o
latio
n
an
d
lo
ca
tio
n
o
f
ass
au
lts
,
wh
ich
is
c
r
u
cial
f
o
r
d
ef
en
d
er
s
to
im
p
lem
en
t
th
e
ap
p
r
o
p
r
iate
co
u
n
ter
m
e
asu
r
es
to
e
n
s
u
r
e
t
h
e
s
y
s
tem
co
n
tin
u
es
to
f
u
n
ctio
n
n
o
r
m
ally
ev
e
n
in
th
e
f
ac
e
o
f
cy
b
er
attac
k
s
s
h
o
u
ld
b
e
b
o
th
f
a
u
lt
-
an
d
attac
k
-
to
le
r
an
t,
s
o
th
at
ev
en
in
th
e
wo
r
s
t
-
ca
s
e
s
ce
n
ar
io
s
,
it
ca
n
co
n
tin
u
e
t
o
f
u
n
ctio
n
as
in
ten
d
ed
.
T
h
er
e
f
o
r
e,
to
in
cr
e
ase
th
e
r
esil
ien
ce
o
f
s
m
ar
t
g
r
i
d
s
,
th
e
d
ef
en
d
e
r
s
m
u
s
t
lev
er
ag
e
f
au
lt
-
to
ler
an
t
co
n
tr
o
l.
T
h
u
s
,
a
d
ev
i
ce
th
at
ca
n
r
eliab
l
y
d
etec
t
o
r
an
ticip
ate
an
attac
k
i
s
d
esp
er
ately
n
ee
d
e
d
.
Fig
u
r
e
1
(
s
ee
Ap
p
en
d
i
x
)
[
3
9
]
c
o
n
tain
s
g
en
er
ato
r
s
G1
an
d
G2
ar
e
u
s
ed
.
in
tellig
en
t
elec
tr
o
n
ic
d
ev
ices
(
I
E
Ds)
R
1
to
R
4
ca
n
tu
r
n
o
n
an
d
o
f
f
th
e
b
r
ea
k
er
s
.
T
h
e
lab
els
o
n
th
ese
b
r
ea
k
er
s
a
r
e
B
R
1
to
B
R
4
.
T
h
e
r
e
ar
e
d
o
u
b
le
lin
es
as
well.
L
in
es
1
an
d
2
ex
ten
d
f
r
o
m
B
R
1
to
B
R
2
an
d
B
R
3
to
B
R
4
,
r
esp
ec
tiv
ely
.
On
e
b
r
ea
k
e
r
is
au
to
m
atica
lly
c
o
n
tr
o
lled
b
y
ea
ch
I
E
D.
C
o
n
s
e
q
u
en
tly
,
R
1
g
o
v
e
r
n
s
B
R
1
,
R
2
g
o
v
er
n
s
B
R
2
,
an
d
s
o
f
o
r
th
.
T
h
e
I
E
Ds
r
elay
o
n
a
f
a
r
awa
y
p
r
o
tectio
n
tec
h
n
iq
u
e
th
at
tr
ip
s
th
e
b
r
ea
k
er
o
n
d
etec
ted
f
au
lts
s
in
ce
th
ey
la
ck
in
n
er
v
alid
atio
n
to
d
is
ce
r
n
b
etwe
en
r
ea
l
an
d
c
o
u
n
te
r
f
eit
f
au
lts
.
Op
er
ato
r
s
ca
n
m
an
u
ally
in
s
tr
u
ct
t
h
e
I
E
Ds
R
1
th
r
o
u
g
h
R
4
in
ad
d
itio
n
to
m
an
u
ally
tr
ip
p
in
g
t
h
e
b
r
ea
k
er
s
B
R
1
th
r
o
u
g
h
B
R
4
.
W
h
en
d
o
in
g
r
ep
air
s
o
n
th
e
lin
es o
r
o
th
er
s
y
s
tem
p
a
r
ts
,
th
e
m
an
u
al
o
v
e
r
r
id
e
is
em
p
lo
y
ed
.
3.
M
E
T
H
O
DO
L
O
G
Y
T
h
e
d
ataset
u
s
ed
f
o
r
class
if
ic
atio
n
an
d
th
e
d
if
f
er
en
t
class
es
ar
e
as
g
iv
en
in
th
e
f
o
llo
win
g
:
T
ab
le
1
p
r
o
v
id
es
a
n
ex
p
lan
atio
n
o
f
t
h
e
1
2
8
ch
ar
ac
ter
is
tics
.
E
v
er
y
p
h
aso
r
m
ea
s
u
r
em
e
n
t
u
n
it
(
PMU)
h
as
2
9
d
if
f
e
r
en
t
k
in
d
s
o
f
m
ea
s
u
r
e
m
en
ts
.
A
PMU
o
r
s
y
n
ch
r
o
p
h
aso
r
is
a
d
e
v
ice
th
at
co
m
p
u
tes
th
e
elec
tr
ical
wav
es
o
n
an
en
er
g
y
g
r
id
b
y
s
y
n
c
h
r
o
n
izin
g
with
a
u
s
u
al
tim
e
s
o
u
r
ce
.
Ou
r
s
y
s
tem
h
as
f
o
u
r
PMUs
m
ea
s
u
r
in
g
2
9
attr
ib
u
tes,
o
r
1
1
6
PMU
m
ea
s
u
r
em
en
t c
o
lu
m
n
s
i
n
to
tal.
T
h
e
ev
e
n
ts
th
at
n
ee
d
to
b
e
p
r
e
d
icted
u
s
in
g
m
ac
h
in
e
lea
r
n
in
g
alg
o
r
ith
m
s
ar
e
n
atu
r
al
ev
e
n
ts
th
at
o
cc
u
r
in
th
e
p
o
wer
s
y
s
tem
,
wh
ich
ar
e
s
h
o
w
n
in
T
ab
le
2
,
lik
e
th
e
s
in
g
le
lin
e
to
g
r
o
u
n
d
(
SLG)
f
au
lt
an
d
lin
e
m
ain
ten
an
ce
.
T
h
e
n
o
r
m
al
o
p
er
atio
n
,
as
s
h
o
wn
in
T
ab
le
3
,
is
d
u
e
to
th
e
lo
ad
ch
a
n
g
es
in
th
e
p
o
wer
s
y
s
tem
.
T
ab
le
4
d
etai
ls
th
e
d
if
f
er
en
t
attac
k
ev
en
t
s
ce
n
ar
io
s
in
th
e
p
o
wer
s
y
s
tem
d
u
e
to
a
cy
b
er
-
attac
k
,
wh
ich
in
clu
d
e
d
ata
in
jectio
n
to
tr
ip
th
e
r
elay
an
d
r
em
o
te
tr
ip
p
in
g
o
f
th
e
r
elay
.
T
h
ese
d
ata
ar
e
u
s
ed
to
tr
ain
t
h
e
m
ac
h
in
e
lear
n
in
g
m
o
d
els f
o
r
f
au
lt a
n
d
n
o
n
-
f
a
u
lt
co
n
d
itio
n
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
6
9
4
I
n
t J Po
w
E
lec
&
Dr
i Sy
s
t
,
Vo
l.
16
,
No
.
2
,
J
u
n
e
20
25
:
8
5
1
-
8
6
3
858
T
ab
le
1
.
PMU
d
ata
s
p
ec
if
icati
o
n
s
F
e
a
t
u
r
e
D
e
scri
p
t
i
o
n
P
A
1
:
V
H
-
P
A
3
:
V
H
V
o
l
t
a
g
e
p
h
a
se
a
n
g
l
e
f
o
r
A
,
B
,
a
n
d
C
p
h
a
s
e
s
P
M
1
:
V
-
P
M
3
:
V
V
o
l
t
a
g
e
m
a
g
n
i
t
u
d
e
f
o
r
A
,
B
,
a
n
d
C
p
h
a
ses
P
A
4
:
I
H
-
P
A
6
:
I
H
C
u
r
r
e
n
t
p
h
a
s
e
A
n
g
l
e
f
o
r
A
,
B
,
a
n
d
C
p
h
a
ses
P
M
4
:
I
-
P
M
6
:
I
C
u
r
r
e
n
t
ma
g
n
i
t
u
d
e
f
o
r
A
,
B
,
a
n
d
C
p
h
a
ses
P
A
7
:
V
H
-
P
A
9
:
V
H
V
o
l
t
a
g
e
p
h
a
se
a
n
g
l
e
f
o
r
p
o
si
t
i
v
e
,
n
e
g
a
t
i
v
e
,
a
n
d
z
e
r
o
s
e
q
u
e
n
c
e
c
o
mp
o
n
e
n
t
s
P
M
7
:
V
-
P
M
9
:
V
V
o
l
t
a
g
e
p
h
a
se
ma
g
n
i
t
u
d
e
f
o
r
p
o
si
t
i
v
e
,
n
e
g
a
t
i
v
e
,
a
n
d
z
e
r
o
se
q
u
e
n
c
e
c
o
m
p
o
n
e
n
t
s
P
A
1
0
:
V
H
-
P
A
1
2
:
V
H
C
u
r
r
e
n
t
p
h
a
s
e
a
n
g
l
e
f
o
r
p
o
s
i
t
i
v
e
,
n
e
g
a
t
i
v
e
,
a
n
d
z
e
r
o
s
e
q
u
e
n
c
e
c
o
mp
o
n
e
n
t
s
P
M
1
0
:
V
-
P
M
1
2
:
V
C
u
r
r
e
n
t
p
h
a
s
e
m
a
g
n
i
t
u
d
e
f
o
r
p
o
s
i
t
i
v
e
,
n
e
g
a
t
i
v
e
,
a
n
d
z
e
r
o
se
q
u
e
n
c
e
c
o
m
p
o
n
e
n
t
s
F
R
e
l
a
y
f
r
e
q
u
e
n
c
y
D
F
R
e
l
a
y
f
r
e
q
u
e
n
c
y
d
e
l
t
a
(
d
F
/
d
t
)
P
A
:
Z
R
e
l
a
y
a
p
p
e
a
r
a
n
c
e
i
m
p
e
d
a
n
c
e
P
A
:
ZH
R
e
l
a
y
a
n
g
l
e
o
f
a
p
p
e
a
r
a
n
c
e
i
mp
e
d
a
n
c
e
S
R
e
l
a
y
st
a
t
u
s f
l
a
g
T
ab
le
2
.
Natu
r
al
e
v
en
t scen
ar
i
o
s
S
c
e
n
a
r
i
o
N
a
t
u
r
a
l
e
v
e
n
t
s
(
S
LG
f
a
u
l
t
s)
1
Li
n
e
1
S
LG
f
a
u
l
t
f
r
o
m
1
0
-
1
9
%
2
Li
n
e
1
S
LG
f
a
u
l
t
f
r
o
m
2
0
-
7
9
%
3
Li
n
e
1
S
LG
f
a
u
l
t
f
r
o
m
8
0
-
9
0
%
4
Li
n
e
2
S
LG
f
a
u
l
t
f
r
o
m
1
0
-
1
9
%
5
Li
n
e
2
S
LG
f
a
u
l
t
f
r
o
m
2
0
-
7
9
%
6
Li
n
e
2
S
LG
f
a
u
l
t
f
r
o
m
8
0
-
9
0
%
N
a
t
u
r
a
l
e
v
e
n
t
s
(
l
i
n
e
ma
i
n
t
e
n
a
n
c
e
)
13
Li
n
e
1
ma
i
n
t
e
n
a
n
c
e
14
Li
n
e
2
ma
i
n
t
e
n
a
n
c
e
T
ab
le
3
.
No
e
v
en
t scen
ar
io
s
S
c
e
n
a
r
i
o
N
o
e
v
e
n
t
s (n
o
r
ma
l
o
p
e
r
a
t
i
o
n
)
41
N
o
r
mal
o
p
e
r
a
t
i
o
n
d
u
e
t
o
l
o
a
d
c
h
a
n
g
e
s
Data
was
en
ter
ed
in
to
th
e
AR
FF
f
o
r
m
at
f
o
r
th
e
in
itial
m
u
lticlas
s
d
atase
t,
wh
ich
in
clu
d
ed
f
if
tee
n
d
atasets
wi
th
ar
o
u
n
d
5
,
0
0
0
d
a
ta
item
s
p
er
.
T
h
e
1
2
8
asp
ec
ts
o
r
v
ar
iab
les
th
at
m
ak
e
u
p
th
ese
d
ata
ar
e
m
o
s
tly
d
er
iv
ed
f
r
o
m
s
y
n
ch
r
o
p
h
aso
r
s
o
r
p
h
aso
r
m
ea
s
u
r
in
g
u
n
i
ts
(
PMUs).
T
h
e
d
ata
was
ev
alu
ate
d
at
1
2
0
s
am
p
les
p
er
s
ec
o
n
d
,
an
d
ea
c
h
s
ch
em
e
was
s
im
u
lated
f
o
r
1
7
s
ec
o
n
d
s
[
3
9
]
.
Du
e
to
v
ar
io
u
s
ca
p
ac
ity
p
r
o
b
le
m
s
th
at
ar
e
s
p
ec
if
ic
to
ea
ch
ap
p
r
o
ac
h
,
g
e
n
er
aliza
tio
n
s
u
f
f
e
r
s
in
in
d
ep
e
n
d
en
t
m
ac
h
in
e
lear
n
in
g
tech
n
iq
u
es.
T
h
e
g
e
n
er
aliza
tio
n
p
r
o
b
lem
is
m
o
s
tly
r
eso
lv
ed
b
y
an
alg
o
r
ith
m
th
at
ca
n
c
o
m
b
in
e
th
e
b
en
ef
its
o
f
m
an
y
m
ac
h
in
e
l
ea
r
n
in
g
tech
n
iq
u
es.
An
ef
f
ec
tiv
e
an
o
m
aly
d
etec
tio
n
tech
n
i
q
u
e
is
th
e
is
o
latio
n
f
o
r
est
alg
o
r
ith
m
.
An
i
m
p
r
o
v
ed
im
p
lem
en
tatio
n
o
f
th
e
ex
ten
d
ed
is
o
latio
n
f
o
r
es
t
tech
n
iq
u
e
f
o
r
an
o
m
aly
id
e
n
tific
atio
n
is
in
clu
d
ed
.
T
h
e
g
e
n
er
al
ex
e
cu
tio
n
p
ar
ticu
lar
s
o
f
s
u
g
g
ested
ap
p
r
o
ac
h
ar
e
s
h
o
wn
in
Fig
u
r
e
2
.
T
h
er
e
a
r
e
f
o
u
r
m
ain
co
m
p
o
n
en
ts
to
th
e
im
p
l
em
en
tatio
n
o
f
f
au
lt
p
r
ed
ictio
n
.
d
ata
p
r
e
p
r
o
ce
s
s
in
g
au
to
m
atio
n
,
o
u
tlier
d
etec
tio
n
an
d
f
ea
tu
r
e
e
n
g
in
ee
r
in
g
,
t
r
ain
in
g
an
d
test
in
g
,
m
o
d
el
ev
alu
atio
n
.
Fig
u
r
e
2
illu
s
tr
ates
h
o
w
t
h
e
wh
o
le
d
ata
p
r
ep
ar
atio
n
p
r
o
ce
s
s
is
au
to
m
ated
,
lea
d
in
g
t
o
a
m
ac
h
in
e
lear
n
in
g
p
ar
ad
ig
m
with
n
o
n
ee
d
f
o
r
h
u
m
an
p
a
r
ticip
atio
n
.
T
h
e
SVM
lear
n
in
g
m
et
h
o
d
i
n
cr
ea
s
es
th
e
g
en
e
r
ality
o
f
th
e
lear
n
in
g
p
r
o
ce
s
s
.
An
o
m
aly
d
etec
tio
n
,
d
ata
clea
n
in
g
,
an
d
d
ata
class
if
icat
io
n
in
to
b
alan
ce
d
an
d
u
n
b
ala
n
ce
d
d
at
a
ar
e
all
in
clu
d
ed
in
th
e
au
t
o
m
atio
n
o
f
d
ata
p
r
ep
ar
atio
n
.
T
h
is
estab
li
s
h
es
th
e
s
am
p
le
p
la
n
f
o
r
th
e
p
r
o
p
o
s
ed
im
p
lem
en
tatio
n
.
Usi
n
g
th
e
m
e
an
v
alu
e
as
a
s
tan
d
-
in
,
au
to
m
at
ic
im
p
u
r
ity
clea
n
in
g
an
d
m
is
s
in
g
v
alu
e
im
p
u
tatio
n
ar
e
p
er
f
o
r
m
ed
o
n
th
e
ca
teg
o
r
ized
d
ata.
As
th
e
m
o
s
t
s
ig
n
if
ican
t
elec
tr
ical
ap
p
licatio
n
s
ar
e
f
au
lt
p
r
ed
ictio
n
alg
o
r
ith
m
s
,
th
e
p
r
im
ar
y
f
ac
to
r
in
f
lu
en
cin
g
th
e
r
esear
ch
d
o
n
e
in
t
h
is
way
is
th
e
p
r
ed
ictio
n
alg
o
r
ith
m
'
s
r
eliab
ilit
y
.
Fo
r
a
m
eth
o
d
to
h
a
n
d
le
m
illi
o
n
s
o
f
d
ata
p
o
in
ts
,
it
m
u
s
t
h
av
e
h
ig
h
er
g
en
er
ality
an
d
h
i
g
h
ly
o
r
th
o
g
o
n
al
in
p
u
t.
I
n
o
r
d
er
to
i
n
cr
ea
s
e
p
r
ed
ictio
n
p
e
r
f
o
r
m
a
n
ce
,
m
o
r
e
s
o
p
h
is
ticated
f
ea
tu
r
e
en
g
i
n
ee
r
in
g
tec
h
n
iq
u
e
s
co
u
ld
b
e
r
eq
u
ir
e
d
if
th
er
e
is
a
lar
g
er
co
n
n
ec
tio
n
b
etwe
en
th
e
s
am
p
le
d
ata.
I
t
is
ch
allen
g
in
g
t
o
ac
q
u
ir
e
th
e
d
ata
-
a
war
e
p
r
ep
r
o
ce
s
s
in
g
p
r
o
g
r
a
m
as
th
e
r
ec
o
m
m
en
d
ed
aim
.
T
h
e
b
lo
ck
d
iag
r
am
r
e
q
u
ir
es
th
at
th
e
ch
o
s
en
f
au
lt
d
iag
n
o
s
tic
p
r
o
b
lem
b
e
s
u
b
jecte
d
to
an
ex
ten
d
ed
is
o
la
tio
n
p
r
o
ce
d
u
r
e.
W
h
en
t
h
e
ex
t
en
d
ed
is
o
latio
n
f
o
r
est
is
u
s
ed
f
o
r
f
au
lt
d
ia
g
n
o
s
is
,
th
e
o
u
tlier
id
en
tific
atio
n
s
h
o
w
s
im
p
r
o
v
e
d
o
r
th
o
g
o
n
ality
.
T
h
e
en
tire
d
ataset
is
d
iv
i
d
ed
in
to
tr
ain
in
g
an
d
test
in
g
d
atasets
,
ea
ch
in
clu
d
in
g
th
e
af
o
r
em
en
tio
n
e
d
C
SV
f
iles
.
PM
U
cy
b
er
-
attac
k
tr
ai
n
in
g
d
ata
h
as
th
e
in
f
o
r
m
atio
n
o
n
wh
eth
er
th
er
e
is
attac
k
o
r
n
o
t w
h
ich
ca
n
b
e
u
s
ed
as
th
e
tar
g
e
t
v
ar
iab
le.
T
h
e
tar
g
et
attr
ib
u
te
is
f
ix
ed
to
wh
eth
er
th
e
m
ea
s
u
r
ed
PMU
d
ata
is
cy
b
er
-
attac
k
o
r
n
o
t sin
ce
th
e
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
h
as to
p
r
ed
ict
th
e
s
am
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Po
w
E
lec
&
Dr
i Sy
s
t
I
SS
N:
2088
-
8
6
9
4
E
n
s
emb
le
lea
r
n
in
g
b
a
s
ed
fa
u
lt
d
etec
tio
n
u
s
in
g
P
MU d
a
ta
in
imb
a
la
n
ce
d
…
(
K
ir
u
th
ika
K
r
is
h
n
a
n
)
859
T
ab
le
4
.
Attack
ev
e
n
t scen
ar
io
s
S
c
e
n
a
r
i
o
A
t
t
a
c
k
t
y
p
e
D
a
t
a
i
n
j
e
c
t
i
o
n
A
t
t
a
c
k
S
u
b
-
t
y
p
e
(
S
LG
f
a
u
l
t
r
e
p
l
a
y
)
7
Li
n
e
1
w
i
t
h
t
r
i
p
p
i
n
g
c
o
mm
a
n
d
f
r
o
m
1
0
-
1
9
%
8
Li
n
e
1
w
i
t
h
t
r
i
p
p
i
n
g
c
o
mm
a
n
d
f
r
o
m
2
0
-
7
9
%
9
Li
n
e
1
w
i
t
h
t
r
i
p
p
i
n
g
c
o
mm
a
n
d
f
r
o
m
80
-
9
0
%
10
Li
n
e
2
w
i
t
h
t
r
i
p
p
i
n
g
c
o
mm
a
n
d
f
r
o
m
1
0
-
1
9
%
11
Li
n
e
2
w
i
t
h
t
r
i
p
p
i
n
g
c
o
mm
a
n
d
f
r
o
m
2
0
-
7
9
%
12
Li
n
e
2
w
i
t
h
t
r
i
p
p
i
n
g
c
o
mm
a
n
d
f
r
o
m
8
0
-
9
0
%
R
e
m
o
t
e
t
r
i
p
p
i
n
g
c
o
m
ma
n
d
I
n
j
e
c
t
i
o
n
A
t
t
a
c
k
su
b
-
t
y
p
e
(
c
o
mm
a
n
d
i
n
j
e
c
t
i
o
n
a
g
a
i
n
s
t
s
i
n
g
l
e
r
e
l
a
y
)
15
R
1
r
e
l
a
y
c
o
mm
a
n
d
i
n
j
e
c
t
i
o
n
16
R
2
r
e
l
a
y
c
o
mm
a
n
d
i
n
j
e
c
t
i
o
n
17
R
3
r
e
l
a
y
c
o
mm
a
n
d
i
n
j
e
c
t
i
o
n
18
R
4
r
e
l
a
y
c
o
mm
a
n
d
i
n
j
e
c
t
i
o
n
A
t
t
a
c
k
su
b
t
-
t
y
p
e
(
c
o
mm
a
n
d
i
n
j
e
c
t
i
o
n
a
g
a
i
n
st
si
n
g
l
e
r
e
l
a
y
)
19
R
e
l
a
y
1
a
n
d
R
e
l
a
y
2
c
o
mm
a
n
d
i
n
j
e
c
t
i
o
n
20
R
e
l
a
y
3
a
n
d
R
e
l
a
y
4
c
o
mm
a
n
d
i
n
j
e
c
t
i
o
n
R
e
l
a
y
se
t
t
i
n
g
c
h
a
n
g
e
A
t
t
a
c
k
su
b
-
t
y
p
e
(
d
i
s
a
b
l
i
n
g
r
e
l
a
y
f
u
n
c
t
i
o
n
-
si
n
g
l
e
r
e
l
a
y
d
i
sa
b
l
e
d
a
n
d
f
a
u
l
t
21
Li
n
e
1
S
LG
f
a
u
l
t
f
r
o
m
1
0
-
1
9
%
R
1
d
i
s
a
b
l
e
d
a
n
d
f
a
u
l
t
Li
n
e
1
S
LG
f
a
u
l
t
f
r
o
m
2
0
-
7
9
%
R
1
d
i
s
a
b
l
e
d
a
n
d
f
a
u
l
t
Li
n
e
1
S
LG
f
a
u
l
t
f
r
o
m
1
0
-
4
9
%
R
2
d
i
s
a
b
l
e
d
a
n
d
f
a
u
l
t
Li
n
e
1
S
LG
f
a
u
l
t
f
r
o
m
5
0
-
7
9
%
R
2
d
i
s
a
b
l
e
d
a
n
d
f
a
u
l
t
Li
n
e
1
S
LG
f
a
u
l
t
f
r
o
m
8
0
-
9
0
%
R
2
d
i
s
a
b
l
e
d
a
n
d
f
a
u
l
t
Li
n
e
2
S
LG
f
a
u
l
t
f
r
o
m
1
0
-
1
9
%
R
3
d
i
s
a
b
l
e
d
a
n
d
f
a
u
l
t
Li
n
e
2
S
LG
f
a
u
l
t
f
r
o
m
2
0
-
7
9
%
R
3
d
i
s
a
b
l
e
d
a
n
d
f
a
u
l
t
Li
n
e
2
S
LG
f
a
u
l
t
f
r
o
m
10
-
4
9
%
R
3
d
i
s
a
b
l
e
d
a
n
d
f
a
u
l
t
Li
n
e
2
S
LG
f
a
u
l
t
f
r
o
m
5
0
-
7
9
%
R
4
d
i
s
a
b
l
e
d
a
n
d
f
a
u
l
t
Li
n
e
2
S
LG
f
a
u
l
t
f
r
o
m
8
0
-
9
0
%
R
4
d
i
s
a
b
l
e
d
a
n
d
f
a
u
l
t
Fig
u
r
e
2
.
Ov
e
r
all
b
lo
ck
d
iag
r
a
m
o
f
d
ata
p
r
ep
r
o
ce
s
s
in
g
au
to
m
ated
SVM
lear
n
in
g
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
Py
th
o
n
is
u
s
ed
in
co
n
j
u
n
cti
o
n
with
th
e
Scik
it
-
lear
n
(
s
k
l
ea
r
n
)
,
s
y
n
th
etic
m
i
n
o
r
ity
o
v
er
s
am
p
lin
g
tech
n
iq
u
e
(
SMOT
E
)
,
an
d
P
an
d
as
to
o
lb
o
x
es to
cr
ea
te
cy
b
er
-
attac
k
p
r
ed
ictio
n
.
Usi
n
g
ML
P,
SVM,
an
d
d
ec
is
io
n
tr
ee
alg
o
r
ith
m
s
as th
e
b
ase
cla
s
s
if
ier
s
an
d
lo
g
is
tic
r
eg
r
ess
io
n
as th
e
m
eta
c
lass
if
ier
,
an
en
s
e
m
b
le
lear
n
in
g
co
d
e
is
cr
ea
ted
to
f
in
d
cy
b
e
r
attac
k
s
.
Fig
u
r
e
3
d
is
p
lay
s
th
e
g
r
a
p
h
th
at
was
cr
ea
ted
to
illu
s
tr
a
te
th
e
v
ar
io
u
s
in
p
u
t
v
ar
iab
les.
Sin
ce
t
h
er
e
is
a
n
o
t
icea
b
le
d
is
p
ar
ity
b
etwe
en
th
e
am
o
u
n
t
o
f
cy
b
e
r
attac
k
d
ata
a
n
d
r
eg
u
lar
d
ata,
th
e
d
ata
ap
p
ea
r
s
to
b
e
u
n
b
alan
ce
d
.
T
h
e
in
d
iv
id
u
al
an
d
en
s
em
b
le
l
ea
r
n
in
g
p
ar
a
d
ig
m
s
ar
e
u
s
ed
to
d
esig
n
th
e
ca
te
g
o
r
izatio
n
is
s
u
e.
Data
is
r
esam
p
led
,
an
d
d
ata
im
b
alan
c
e
is
v
er
if
ied
.
Nu
m
er
o
u
s
attr
ib
u
tes
th
at
ar
e
n
o
t
n
ec
ess
ar
y
f
o
r
ca
teg
o
r
izatio
n
a
r
e
r
em
o
v
ed
f
r
o
m
th
e
d
ataset
p
r
io
r
to
r
esam
p
lin
g
.
T
h
e
s
ig
n
if
ica
n
t c
h
ar
ac
ter
is
tics
f
r
o
m
th
e
d
ataset
r
em
ain
af
ter
th
e
in
s
ig
n
if
ican
t
f
ea
tu
r
es
h
av
e
b
e
en
elim
in
ated
.
T
ab
le
5
lis
ts
th
e
p
ar
am
eter
s
f
o
r
th
e
v
ar
i
o
u
s
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
u
tili
ze
d
in
t
h
e
en
s
e
m
b
le
lear
n
i
n
g
tec
h
n
iq
u
e
.
An
ML
ap
p
r
o
ac
h
ca
lled
e
n
s
em
b
le
lear
n
in
g
co
m
b
in
es
m
an
y
b
ase
class
if
ier
s
to
p
r
o
d
u
ce
a
m
o
r
e
p
o
wer
f
u
l
p
r
ed
ictio
n
m
o
d
el.
T
h
e
d
ec
is
io
n
tr
ee
alg
o
r
ith
m
,
m
u
lti
-
lay
er
p
er
ce
p
tr
o
n
(
ML
P),
an
d
s
u
p
p
o
r
t
v
ec
to
r
class
if
ier
s
(
SVC
)
ar
e
t
h
e
b
ase
class
if
ier
s
in
th
is
in
s
tan
ce
.
T
r
ain
i
n
g
ea
c
h
b
ase
class
if
ier
s
e
p
ar
ately
o
n
th
e
tr
ain
in
g
s
et
o
f
d
ata
is
th
e
f
i
r
s
t
s
tag
e
in
th
e
e
n
s
em
b
le
lear
n
in
g
p
r
o
ce
s
s
.
Fo
r
th
e
test
d
ata,
ev
er
y
b
ase
class
if
ier
will
p
r
o
v
id
e
a
u
n
i
q
u
e
s
et
o
f
p
r
ed
ictio
n
s
.
A
m
eta
-
class
if
ier
will
b
e
u
s
ed
to
ag
g
r
eg
ate
th
ese
p
r
e
d
ictio
n
s
an
d
p
r
o
v
id
e
a
f
in
al
p
r
e
d
ictio
n
.
I
n
th
is
in
s
tan
ce
,
lo
g
is
tic
r
eg
r
ess
io
n
s
er
v
es
as
th
e
m
eta
class
if
ier
.
I
t
g
en
er
ates
a
f
in
al
p
r
ed
ictio
n
b
y
u
s
in
g
th
e
i
n
p
u
t
p
r
e
d
ictio
n
s
f
r
o
m
ea
ch
o
f
th
e
b
ase
class
if
ier
s
.
I
n
o
r
d
er
to
g
en
e
r
ate
th
e
m
o
s
t a
cc
u
r
ate
p
r
e
d
ictio
n
p
o
s
s
ib
le,
th
e
m
eta
-
class
if
ie
r
in
teg
r
ates th
e
k
n
o
wled
g
e
g
ain
e
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
6
9
4
I
n
t J Po
w
E
lec
&
Dr
i Sy
s
t
,
Vo
l.
16
,
No
.
2
,
J
u
n
e
20
25
:
8
5
1
-
8
6
3
860
f
r
o
m
th
e
b
asis
class
if
ier
s
'
p
r
e
d
ictio
n
s
.
C
o
m
b
in
in
g
th
e
m
eta
-
class
if
ier
'
s
p
r
ed
ictio
n
s
with
t
h
e
b
asic
class
if
ier
s
'
p
r
ed
ictio
n
s
y
ield
s
th
e
f
in
al
en
s
em
b
le
m
o
d
el.
T
h
e
way
th
is
c
o
m
b
in
atio
n
is
u
s
ed
o
p
tim
izes
th
e
f
o
r
ec
ast
ac
cu
r
ac
y
.
A
n
en
s
em
b
le
lear
n
in
g
ap
p
r
o
ac
h
ca
n
b
o
o
s
t
th
e
m
o
d
el'
s
ac
cu
r
ac
y
b
y
co
m
b
in
in
g
th
e
b
e
n
ef
its
o
f
m
an
y
b
asic
class
if
ier
s
.
T
h
e
d
r
awb
ac
k
s
o
f
s
ep
a
r
ate
class
if
ier
s
ca
n
b
e
m
itig
ated
b
y
co
m
b
in
in
g
th
e
b
e
n
ef
its
an
d
d
r
a
wb
ac
k
s
o
f
ea
ch
f
u
n
d
am
en
tal
cl
ass
if
ier
.
All
th
in
g
s
co
n
s
id
er
ed
,
th
e
e
n
s
em
b
le
lea
r
n
in
g
alg
o
r
ith
m
is
an
im
p
o
r
t
an
t
m
ac
h
in
e
lear
n
i
n
g
tech
n
i
q
u
e
th
at
m
ay
y
ield
ex
tr
ao
r
d
i
n
ar
ily
ac
cu
r
ate
p
r
ed
i
ctio
n
s
.
I
t
em
p
lo
y
s
lo
g
is
tic
r
eg
r
ess
io
n
as
a
m
eta
-
class
if
ier
a
n
d
SVC
,
ML
P,
an
d
d
ec
is
io
n
tr
ee
alg
o
r
ith
m
s
as
b
a
s
e
class
if
ier
s
.
Fi
g
u
r
e
3
d
is
p
lay
s
th
e
g
r
ap
h
s
o
f
v
ar
i
o
u
s
in
p
u
t
v
ar
iab
les
r
ep
r
esen
ts
Fig
u
r
e
3
(
a
)
v
o
ltag
e
p
h
ase
an
g
le
f
o
r
C
p
h
ase,
Fig
u
r
e
3
(
b
)
v
o
ltag
e
m
ag
n
itu
d
e
f
o
r
B
p
h
ase,
Fig
u
r
e
3
(
c
)
v
o
ltag
e
p
h
ase
an
g
le
f
o
r
B
p
h
ase,
a
n
d
Fig
u
r
e
3
(
d
)
v
o
ltag
e
m
ag
n
itu
d
e
f
o
r
A
p
h
ase.
Fig
u
r
e
4
r
e
p
r
esen
ts
an
AUC
-
R
OC
cu
r
v
e
th
at
was
p
r
o
d
u
ce
d
b
y
t
h
e
in
d
iv
id
u
al
an
d
g
r
o
u
p
lear
n
i
n
g
(
s
tack
d
is
tr
ib
u
tio
n
)
tech
n
iq
u
e.
On
co
m
p
ar
is
o
n
with
o
th
er
d
is
tin
ct
m
ac
h
in
e
le
ar
n
in
g
tech
n
iq
u
es,
th
e
s
tack
d
i
s
tr
ib
u
tio
n
ap
p
r
o
ac
h
y
ield
s
th
e
h
ig
h
est
AUC
-
R
O
C
cu
r
v
e
p
er
f
o
r
m
an
ce
.
Ou
t
o
f
al
l
th
e
alg
o
r
ith
m
s
,
lo
g
is
tic
r
eg
r
ess
io
n
d
em
o
n
s
tr
ated
g
o
o
d
p
e
r
f
o
r
m
a
n
ce
.
All
th
e
s
ep
ar
ate
m
eth
o
d
s
ar
e
o
u
t
p
er
f
o
r
m
ed
b
y
th
e
e
n
s
em
b
le
lear
n
in
g
alg
o
r
ith
m
.
Giv
en
th
at
th
e
SM
OT
E
alg
o
r
ith
m
was
u
tili
ze
d
as
th
e
s
am
p
lin
g
tech
n
iq
u
e,
it
is
co
n
clu
d
ed
th
at
th
e
im
p
lem
en
tatio
n
'
s
ac
cu
r
ac
y
is
g
o
o
d
.
Sin
ce
th
er
e
ar
e
s
ev
er
al
SMOT
E
m
eth
o
d
s
av
ai
lab
le,
th
e
d
i
v
er
s
ity
o
f
SMOT
E
im
p
lem
en
tatio
n
o
n
up
-
s
am
p
lin
g
m
a
y
b
e
u
s
ed
t
o
en
h
an
ce
th
e
ac
cu
r
ac
y
an
d
AUC
-
R
O
C
cu
r
v
e
f
o
r
th
e
up
-
s
am
p
led
in
p
u
t c
h
ar
ac
ter
is
tics
.
(
a)
(
b
)
(
c)
(
d
)
F
i
g
u
r
e
3
.
I
n
p
u
t
v
a
r
i
a
b
l
e
s
f
r
o
m
th
e
P
M
U
(
4
a
m
o
n
g
t
h
e
1
2
8
v
a
r
ia
b
l
e
s
)
:
(
a
)
v
o
lt
a
g
e
p
h
a
s
e
a
n
g
le
f
o
r
C
p
h
a
s
e
,
(
b
)
v
o
l
t
a
g
e
m
a
g
n
i
t
u
d
e
f
o
r
B
p
h
a
s
e
,
(
c
)
v
o
l
t
a
g
e
p
h
as
e
a
n
g
l
e
f
o
r
B
p
h
a
s
e
,
a
n
d
(
d
)
v
o
l
t
a
g
e
m
a
g
n
i
t
u
d
e
f
o
r
A
p
h
a
s
e
T
ab
le
5
.
E
n
s
em
b
le
lear
n
in
g
p
a
r
am
eter
s
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
m
o
d
e
l
P
a
r
a
me
t
e
r
s
M
LP
c
l
a
ssi
f
i
e
r
A
c
t
i
v
a
t
i
o
n
=
"
r
e
l
u
"
,
a
l
p
h
a
=
0
.
1
,
h
i
d
d
e
n
_
l
a
y
e
r
_
s
i
z
e
s
=
3
0
,
l
e
a
r
n
i
n
g
_
r
a
t
e
=
"
i
n
v
sc
a
l
i
n
g
"
,
ma
x
_
i
t
e
r
=
5
0
0
0
0
,
r
a
n
d
o
m
_
st
a
t
e
=
1
0
0
0
D
e
c
i
s
i
o
n
t
r
e
e
c
l
a
ss
i
f
i
e
r
max
_
d
e
p
t
h
=
5,
m
a
x
_
f
e
a
t
u
r
e
s
=
"
a
u
t
o
"
,
mi
n
_
s
a
m
p
l
e
s_
l
e
a
f
=
0
.
0
0
5
,
m
i
n
_
sa
mp
l
e
s_
s
p
l
i
t
=
0
.
0
0
5
,
r
a
n
d
o
m
_
s
t
a
t
e
=
2
0
0
0
S
V
C
C
=
8
5
,
d
e
g
r
e
e
=
1
5
,
g
a
mm
a
=
.
8
,
k
e
r
n
e
l
=
"
r
b
f
"
,
p
r
o
b
a
b
i
l
i
t
y
=
Tr
u
e
Lo
g
i
s
t
i
c
r
e
g
r
e
ssi
o
n
r
a
n
d
o
m
_
s
t
a
t
e
=
42
Evaluation Warning : The document was created with Spire.PDF for Python.