I
nte
rna
t
io
na
l J
o
urna
l o
f
Appl
ied P
o
wer
E
ng
i
neer
ing
(
I
J
AP
E
)
Vo
l.
1
5
,
No
.
3
,
Sep
tem
b
er
20
2
6
,
p
p
.
1
1
5
7
~
1
167
I
SS
N:
2252
-
8
7
9
2
,
DOI
:
1
0
.
1
1
5
9
1
/ijap
e
.
v
1
5
.
i
3
.
pp
1
1
5
7
-
1
1
6
7
1157
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//
ija
p
e.
ia
esco
r
e.
co
m/
Predic
tive
ma
inte
na
nce f
o
r in
ducti
o
n mo
tors
:
a
no
v
e
l sy
nerg
y
of
deep learning
and
ma
chine learni
ng
t
echni
ques
V.
Ra
j
ini
1
,
K
a
runy
a
H
a
ri
k
ri
s
hn
a
n
2
,
K
rism
a
din
a
t
a
3
1
D
e
p
a
r
t
me
n
t
o
f
El
e
c
t
r
i
c
a
l
a
n
d
El
e
c
t
r
o
n
i
c
s E
n
g
i
n
e
e
r
i
n
g
,
S
r
i
S
i
v
a
su
b
r
a
m
a
n
i
y
a
N
a
d
a
r
C
o
l
l
e
g
e
o
f
E
n
g
i
n
e
e
r
i
n
g
,
C
h
e
n
n
a
i
,
I
n
d
i
a
2
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
r
i
n
g
,
S
h
i
v
N
a
d
a
r
U
n
i
v
e
r
s
i
t
y
,
C
h
e
n
n
a
i
,
I
n
d
i
a
3
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
,
F
a
c
u
l
t
y
o
f
E
n
g
i
n
e
e
r
i
n
g
,
U
n
i
v
e
r
si
t
a
s
N
e
g
e
r
i
P
a
d
a
n
g
,
P
a
d
a
n
g
,
I
n
d
o
n
e
si
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Au
g
2
6
,
2
0
2
5
R
ev
is
ed
Ma
r
1
7
,
2
0
2
6
Acc
ep
ted
J
u
n
9
,
2
0
2
6
Co
n
d
it
io
n
m
o
n
it
o
rin
g
o
f
in
d
u
c
ti
o
n
m
o
t
o
rs
is
v
i
tal
fo
r
p
re
v
e
n
t
in
g
u
n
e
x
p
e
c
ted
d
o
wn
t
ime
s
a
n
d
m
in
imiz
in
g
th
e
m
a
in
ten
a
n
c
e
c
o
sts
in
i
n
d
u
stria
l
se
tt
in
g
s.
A
p
re
d
ictiv
e
m
a
in
ten
a
n
c
e
m
o
d
e
l
fo
r
e
a
rly
d
e
tec
ti
o
n
o
f
fa
u
lt
s
is
p
ro
p
o
se
d
.
Th
e
m
o
to
r
c
u
r
re
n
t,
fl
u
x
,
v
i
b
ra
ti
o
n
,
th
e
rm
a
l
a
n
d
a
c
o
u
stic
e
m
issio
n
sig
n
a
tu
re
s
a
re
c
o
m
m
o
n
ly
u
se
d
fo
r
fa
u
lt
d
e
t
e
c
ti
o
n
a
s
th
e
se
sig
n
a
ls
re
v
e
a
l
fa
u
lt
sp
e
c
ifi
c
fre
q
u
e
n
c
y
c
o
m
p
o
n
e
n
ts.
S
i
g
n
a
l
p
r
o
c
e
ss
in
g
tec
h
n
iq
u
e
s
li
k
e
wa
v
e
let
tran
sfo
rm
a
n
d
Hil
b
e
rt
tra
n
sfo
rm
a
re
u
se
d
t
o
id
e
n
ti
fy
fa
u
lt
s
a
t
i
n
c
ip
ien
t
sta
g
e
s
.
M
a
c
h
i
n
e
lea
rn
in
g
a
n
d
d
e
e
p
lea
rn
i
n
g
m
o
d
e
ls
a
re
u
se
d
n
o
wa
d
a
y
s
to
e
x
trac
t
fe
a
tu
re
s
a
n
d
c
las
sify
th
e
fa
u
lt
s
a
c
c
u
ra
tely
.
Th
e
c
u
rre
n
t
a
n
d
flu
x
si
g
n
a
ls
fr
o
m
h
e
a
lt
h
y
a
n
d
fa
u
lt
y
m
o
to
rs
fo
r
in
ter
tu
r
n
fa
u
lt
s
we
re
a
n
a
ly
z
e
d
i
n
t
h
is
w
o
rk
a
n
d
fe
a
tu
re
s
we
re
e
x
trac
ted
th
ro
u
g
h
v
a
rio
u
s
sig
n
a
l
p
ro
c
e
ss
in
g
m
e
th
o
d
s.
T
h
e
s
e
fe
a
tu
re
s
we
re
th
e
n
u
se
d
to
train
m
o
d
e
ls,
in
c
lu
d
i
n
g
d
e
e
p
lea
rn
in
g
a
rc
h
it
e
c
tu
re
s,
to
c
las
sify
m
o
to
r
fa
u
lt
s.
W
h
il
e
t
h
e
c
las
sic
a
l
m
a
c
h
in
e
lea
rn
in
g
m
o
d
e
ls
p
r
o
v
i
d
e
d
a
re
a
so
n
a
b
ly
a
c
c
u
ra
te
fa
u
lt
c
las
sifica
ti
o
n
,
th
e
c
o
n
v
o
l
u
ti
o
n
a
l
n
e
u
ra
l
n
e
two
r
k
p
ro
v
id
e
d
a
v
e
ry
g
o
o
d
c
las
sifica
ti
o
n
a
c
c
u
ra
c
y
.
T
h
e
fi
n
d
i
n
g
s
s
h
o
w
th
a
t
d
e
e
p
lea
rn
in
g
m
o
d
e
ls
e
x
c
e
l
i
n
d
e
tec
ti
n
g
fa
u
lt
s,
e
sp
e
c
ially
u
n
d
e
r
n
o
isy
a
n
d
v
a
ry
i
n
g
o
p
e
ra
ti
o
n
a
l
c
o
n
d
it
i
o
n
s,
o
u
tp
e
rfo
r
m
in
g
t
h
e
trad
it
i
o
n
a
l
m
e
th
o
d
s.
Th
e
se
m
o
d
e
ls
o
ffe
r
a
sc
a
lab
le,
re
a
l
-
ti
m
e
so
l
u
ti
o
n
f
o
r
imp
r
o
v
in
g
t
h
e
re
li
a
b
il
it
y
a
n
d
e
fficie
n
c
y
o
f
in
d
u
c
ti
o
n
m
o
t
o
r
a
n
d
fa
c
il
it
a
te
m
o
re
re
li
a
b
le
a
ss
e
ss
m
e
n
ts
a
n
d
c
o
n
tri
b
u
te t
o
wa
rd
s e
n
e
rg
y
e
fficie
n
c
y
a
n
d
e
x
ten
d
e
d
li
fe
ti
m
e
o
f
e
q
u
i
p
m
e
n
t.
K
ey
w
o
r
d
s
:
C
o
n
d
itio
n
m
o
n
ito
r
in
g
C
u
r
r
en
t sig
n
atu
r
e
Fau
lt d
etec
tio
n
Flu
x
s
ig
n
atu
r
e
I
n
d
u
ctio
n
m
o
to
r
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
:
V
.
R
ajin
i
Dep
ar
tm
en
t o
f
E
lectr
ical
an
d
E
lectr
o
n
ics E
n
g
in
ee
r
i
n
g
,
Sri
S
iv
asu
b
r
am
an
iy
a
Nad
ar
C
o
lleg
e
o
f
E
n
g
in
ee
r
in
g
OM
R
R
o
ad
,
Kala
v
ak
k
am
,
C
h
en
n
ai,
T
am
il
N
ad
u
,
I
n
d
ia
E
m
ail: r
ajin
iv
@
s
s
n
.
ed
u
.
in
1.
I
NT
RO
D
UCT
I
O
N
I
n
m
o
s
t
in
d
u
s
tr
ial
en
v
ir
o
n
m
en
ts
,
m
o
n
ito
r
i
n
g
t
h
e
h
ea
lth
o
f
m
ac
h
in
es
is
im
p
o
r
ta
n
t
to
im
p
r
o
v
e
p
r
o
d
u
ctiv
ity
.
An
ap
p
r
o
p
r
iate
m
ain
ten
an
ce
s
tr
ateg
y
is
im
p
o
r
tan
t
to
a
v
o
id
th
e
f
ailu
r
e
o
f
m
ac
h
in
er
y
s
o
th
at
th
e
co
s
t
o
f
p
r
o
d
u
ctio
n
an
d
tim
e
c
an
b
e
m
in
im
ized
.
T
h
e
m
ain
te
n
an
ce
o
f
m
ac
h
i
n
es
ca
n
b
e
r
esp
o
n
s
iv
e,
p
r
e
v
en
tiv
e,
o
r
p
r
ed
ictiv
e
in
n
atu
r
e.
W
h
i
le
th
e
r
esp
o
n
s
iv
e
m
ain
ten
a
n
ce
is
a
r
ea
ctiv
e
a
p
p
r
o
ac
h
w
h
er
e
f
ix
-
o
n
f
ailu
r
e
h
ap
p
en
s
,
t
h
e
p
r
ev
en
tiv
e
m
ai
n
ten
an
ce
is
a
p
r
e
-
p
lan
n
e
d
a
p
p
r
o
ac
h
.
T
h
e
p
r
ed
ictiv
e
a
p
p
r
o
ac
h
is
b
ased
o
n
f
o
r
ec
asti
n
g
s
tr
ateg
y
.
T
h
is
ap
p
r
o
ac
h
o
f
f
er
s
a
c
o
s
t
an
d
ti
m
e
ad
v
an
ta
g
e
as
co
m
p
ar
e
d
to
r
esp
o
n
s
iv
e
an
d
p
r
ev
en
tiv
e
a
p
p
r
o
ac
h
es
,
r
esp
ec
tiv
ely
.
A
co
m
b
i
n
atio
n
o
f
v
ar
i
o
u
s
in
tellig
en
t
m
o
d
els
ca
n
b
e
u
s
ed
f
o
r
a
r
o
b
u
s
t
f
o
r
ec
ast
an
d
b
etter
p
r
ed
ictio
n
ac
cu
r
ac
y
.
I
t
is
ess
en
tial
to
h
av
e
a
r
o
b
u
s
t
m
eth
o
d
o
lo
g
y
f
o
r
f
a
u
lt
d
etec
tio
n
s
o
th
at
th
ey
ca
n
b
e
id
en
tifie
d
at
an
in
cip
ien
t
s
tag
e
an
d
ca
n
b
e
m
itig
ated
at
th
e
ea
r
lies
t.
Hen
ce
,
d
e
v
elo
p
in
g
a
r
eliab
le
in
tellig
en
t m
o
d
el
f
o
r
m
o
n
ito
r
i
n
g
th
e
h
ea
lth
o
f
th
e
m
o
s
t w
id
ely
u
s
ed
in
d
u
ctio
n
m
o
t
o
r
is
th
e
m
ain
m
o
tiv
atio
n
o
f
th
is
wo
r
k
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
7
9
2
I
n
t J Ap
p
l Po
wer
E
n
g
,
Vo
l.
1
5
,
No
.
3
,
Sep
tem
b
er
20
2
6
:
1
1
5
7
-
1
1
6
7
1158
Var
io
u
s
p
h
y
s
ical
s
ig
n
atu
r
es
li
k
e
v
ib
r
atio
n
,
tem
p
er
atu
r
e,
ac
o
u
s
tic
em
is
s
io
n
,
cu
r
r
e
n
t,
v
o
ltag
e,
p
o
wer
,
an
d
m
ag
n
etic
f
lu
x
ar
e
g
en
er
al
ly
u
s
ed
.
T
h
ese
m
et
h
o
d
s
r
e
q
u
i
r
e
s
en
s
o
r
s
an
d
d
ev
ices
f
o
r
d
at
a
ac
q
u
is
itio
n
.
T
h
e
m
o
to
r
c
u
r
r
e
n
t
b
ased
co
n
d
itio
n
m
o
n
ito
r
in
g
is
a
n
o
n
-
i
n
v
asiv
e
m
eth
o
d
r
e
q
u
ir
in
g
o
n
ly
lo
w
-
c
o
s
t
cu
r
r
en
t
s
en
s
o
r
s
with
a
s
im
p
le
im
p
lem
en
tatio
n
.
I
t
h
as
th
e
ab
ilit
y
to
d
etec
t
e
l
ec
tr
ical
an
d
m
ec
h
an
ical
f
au
lts
,
in
clu
d
in
g
b
ea
r
i
n
g
f
au
lts
.
T
h
e
f
lu
x
b
ased
co
n
d
itio
n
m
o
n
ito
r
in
g
r
e
q
u
ir
es
a
f
lu
x
co
il
to
b
e
in
s
er
ted
in
th
e
m
ac
h
in
e
d
u
r
in
g
d
esig
n
an
d
m
ea
s
u
r
in
g
th
e
in
d
u
ce
d
v
o
ltag
e
ac
r
o
s
s
th
is
co
il.
T
h
is
r
esear
ch
u
s
es
cu
r
r
en
t
a
n
d
f
lu
x
s
ig
n
atu
r
es
f
o
r
f
au
lt
class
if
icatio
n
.
Am
id
v
ar
io
u
s
ap
p
r
o
ac
h
es
p
r
o
p
o
s
ed
in
th
e
liter
atu
r
e
,
th
e
p
e
r
f
o
r
m
a
n
ce
o
f
m
o
d
el
-
b
ased
ap
p
r
o
ac
h
is
m
o
s
tly
d
ep
en
d
en
t
o
n
th
e
ex
cl
u
s
iv
e
m
o
to
r
m
o
d
els.
Kn
o
wle
d
g
e
-
b
ased
a
p
p
r
o
ac
h
es
u
s
in
g
s
tatis
tical
an
d
d
ee
p
lear
n
in
g
alg
o
r
ith
m
s
ar
e
g
ain
in
g
in
cr
ea
s
in
g
i
n
ter
est
am
o
n
g
th
e
s
cien
tific
co
m
m
u
n
ity
f
o
r
co
n
d
itio
n
m
o
n
ito
r
in
g
o
f
m
ac
h
in
es.
T
h
e
f
ea
t
u
r
es
d
e
r
iv
ed
f
r
o
m
v
ar
i
o
u
s
ac
q
u
ir
ed
s
ig
n
atu
r
es
ar
e
u
s
ed
in
th
ese
alg
o
r
ith
m
s
.
Fo
u
r
ier
t
r
an
s
f
o
r
m
,
w
av
elet
t
r
an
s
f
o
r
m
(
W
T
)
,
an
d
co
n
v
o
lu
tio
n
al
n
e
u
r
al
n
etwo
r
k
s
(
C
NN)
ar
e
u
s
ed
f
o
r
th
is
p
u
r
p
o
s
e
b
y
s
o
m
e
r
esear
ch
er
s
.
A
h
y
b
r
id
a
p
p
r
o
ac
h
b
ased
o
n
em
p
ir
ical
a
n
d
ad
ap
tiv
e
f
ea
tu
r
es
wer
e
u
s
ed
f
o
r
b
ea
r
in
g
f
au
lt
class
if
icatio
n
o
f
b
y
Xie
e
t
al
.
[
1
]
u
s
in
g
v
ib
r
atio
n
s
ig
n
atu
r
es.
B
az
an
et
al
.
[
2
]
h
av
e
u
s
ed
cu
r
r
en
t sig
n
atu
r
es with
in
f
o
r
m
atio
n
th
e
o
r
etic
m
ea
s
u
r
em
en
ts
an
d
in
tellig
en
t
to
o
ls
s
u
ch
as
m
u
lti
-
lay
er
p
e
r
ce
p
tr
o
n
-
ar
tific
ial
n
eu
r
a
l
n
etwo
r
k
,
k
-
n
ea
r
est n
eig
h
b
o
u
r
s
(
KNN)
,
an
d
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
es
(
SVM)
f
o
r
th
e
d
etec
tio
n
o
f
a
b
r
asiv
e
wea
r
in
th
e
b
ea
r
in
g
s
o
f
in
d
u
ctio
n
m
o
to
r
.
Ho
a
n
g
an
d
Kan
g
[
3
]
h
av
e
u
s
ed
th
e
i
n
f
o
r
m
atio
n
f
u
s
io
n
o
n
c
u
r
r
e
n
t
s
ig
n
atu
r
es.
A
f
au
lt
d
iag
n
o
s
is
f
r
am
ewo
r
k
b
ased
o
n
a
C
NN,
wh
ich
in
teg
r
ates
d
o
m
ain
k
n
o
wled
g
e
an
d
b
r
o
ad
lear
n
in
g
s
y
s
tem
was
p
r
o
p
o
s
ed
b
y
Fen
g
et
al
.
[
4
]
f
o
r
b
ea
r
in
g
f
au
lts
.
T
h
e
ap
p
r
o
ac
h
p
r
o
p
o
s
e
d
b
y
Z
h
a
o
et
a
l.
[
5
]
u
s
ed
a
m
u
lti
s
en
s
o
r
d
ev
ice
f
o
r
co
llectin
g
th
e
tim
e
d
o
m
ai
n
-
b
ased
v
ib
r
atio
n
s
ig
n
als.
A
m
u
ltima
n
if
o
ld
d
ee
p
ex
tr
em
e
lear
n
in
g
m
ac
h
in
e
wa
s
u
s
ed
f
o
r
f
ea
tu
r
e
e
x
tr
ac
tio
n
an
d
f
a
u
lt
d
iag
n
o
s
is
.
W
av
elet
d
ec
o
m
p
o
s
itio
n
an
d
ar
tific
ial
n
eu
r
al
n
etwo
r
k
s
wer
e
u
s
ed
b
y
Sad
eg
h
ia
n
et
al
.
[
6
]
f
o
r
r
o
t
o
r
f
a
u
lt
d
etec
tio
n
o
f
in
d
u
ctio
n
m
o
to
r
s
.
R
ez
en
d
e
et
a
l.
[
7
]
h
av
e
d
em
o
n
s
tr
ated
th
e
v
iab
ilit
y
o
f
th
e
co
m
b
in
ed
u
s
e
o
f
C
NN
tech
n
iq
u
es
with
th
er
m
al
an
aly
s
is
f
o
r
d
etec
tio
n
o
f
f
ailu
r
es
in
elec
tr
ic
m
o
to
r
s
.
A
d
ata
d
r
iv
en
d
ee
p
lear
n
in
g
m
o
d
el
is
in
v
esti
g
ated
b
y
So
h
aib
et
a
l.
[
8
]
a
n
d
T
h
eiss
ler
et
a
l.
[
9
]
.
A
d
etailed
r
e
v
iew
o
f
liter
atu
r
e
i
n
p
r
ed
ictiv
e
m
ain
t
en
an
ce
f
o
r
i
n
d
u
s
tr
y
4
.
0
is
d
o
n
e
[
10
]
-
[
1
3
]
T
h
e
d
ig
ital
twin
c
o
n
ce
p
t
is
p
i
ck
in
g
u
p
m
o
m
en
tu
m
in
in
d
u
s
t
r
ial
s
ec
to
r
.
Falek
as
an
d
Kar
lis
[
1
4
]
h
a
v
e
in
v
esti
g
ated
a
d
etailed
r
ev
iew
in
th
is
ar
ea
.
As
m
o
s
t
o
f
th
e
i
n
d
u
s
tr
ial
d
r
iv
es
u
s
e
p
o
wer
el
ec
tr
o
n
ic
d
e
v
ices,
a
r
ev
iew
o
f
th
e
ap
p
licatio
n
o
f
A
I
tech
n
iq
u
es in
in
d
u
s
tr
ial
d
r
i
v
e
s
was w
r
itten
b
y
Z
h
ao
et
a
l
.
[
1
5
]
.
R
ajin
i
et
al
.
[
1
6
]
h
a
v
e
in
v
esti
g
ated
th
e
class
if
icatio
n
o
f
in
d
u
c
tio
n
m
o
t
o
r
f
au
lts
with
m
ac
h
in
e
lear
n
in
g
an
d
wav
elet
f
ea
tu
r
e
ex
tr
ac
tio
n
tech
n
iq
u
es
.
T
h
e
p
u
r
p
o
s
e
o
f
t
h
is
s
tu
d
y
was
to
d
em
o
n
s
tr
ate
th
e
ap
p
licab
ilit
y
o
f
th
e
p
r
o
p
o
s
ed
C
NN
m
o
d
el
al
o
n
g
with
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
ex
is
tin
g
m
ac
h
in
e
lear
n
in
g
m
o
d
els.
Var
io
u
s
d
ee
p
lear
n
in
g
ar
c
h
itectu
r
es
ap
p
lied
to
in
d
u
ctio
n
m
o
to
r
f
au
lt
d
iag
n
o
s
is
ar
e
g
iv
en
by
Sik
in
y
i
et
a
l
.
[
17
]
.
R
ajap
ak
s
h
a
et
a
l.
[
1
8
]
c
o
m
p
a
r
ed
s
u
p
er
v
is
ed
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
u
s
in
g
f
ea
tu
r
es
d
er
iv
e
d
f
r
o
m
th
e
f
ast
Fo
u
r
ier
tr
an
s
f
o
r
m
(
FF
T
)
o
f
ac
o
u
s
tic
d
ata
f
o
r
co
n
d
itio
n
m
o
n
ito
r
in
g
o
f
in
d
u
ctio
n
m
o
to
r
s
.
A
ca
s
e
s
tu
d
y
wh
e
r
e
t
h
e
FF
T
is
u
s
ed
to
ex
tr
ac
t e
n
er
g
y
f
ea
t
u
r
es,
f
ee
d
in
g
i
n
to
ML
m
o
d
els f
o
r
d
iag
n
o
s
is
is
d
o
n
e
b
y
Kh
alil
an
d
R
o
s
tam
[
19
]
.
A
f
r
am
ewo
r
k
b
ased
o
n
d
ee
p
lear
n
in
g
f
o
r
h
ea
lth
m
o
n
ito
r
in
g
o
f
was
g
iv
en
b
y
Gan
g
u
ly
et
a
l.
[
20
]
.
T
wo
p
o
wer
f
u
l
d
ee
p
lear
n
in
g
ar
ch
itectu
r
es,
C
NN
an
d
L
S
T
M,
to
ex
tr
ac
t
b
o
th
s
p
atial
(
lo
ca
l)
an
d
tem
p
o
r
al
(
s
eq
u
en
ce
)
f
ea
tu
r
es
f
r
o
m
v
ib
r
a
tio
n
s
ig
n
als
f
o
r
f
au
lt d
iag
n
o
s
is
ar
e
d
o
n
e
b
y
Su
n
et
al
.
[
21
].
T
h
e
d
ig
ital
twin
an
d
in
d
u
s
tr
ial
I
o
T
f
o
r
m
o
n
ito
r
in
g
,
a
k
ey
tr
en
d
s
h
o
win
g
th
e
p
r
ac
ti
ca
l
d
ep
lo
y
m
en
t
o
f
co
n
d
itio
n
m
o
n
ito
r
in
g
s
y
s
tem
s
ar
e
d
o
n
e
b
y
San
to
s
e
t
a
l.
[
22
]
.
Fo
r
m
o
n
ito
r
in
g
th
e
elec
tr
ical
d
r
iv
es,
a
s
im
u
latio
n
-
ass
is
ted
n
eu
r
al
n
etwo
r
k
was
u
s
ed
b
y
[
23
]
.
Mo
to
r
cu
r
r
en
t
s
ig
n
atu
r
e
an
aly
s
is
co
m
b
in
ed
with
au
to
r
e
g
r
ess
iv
e
s
p
ec
tr
al
esti
m
atio
n
(
an
alter
n
ativ
e
to
FF
T
)
f
o
r
h
ig
h
-
r
eso
lu
tio
n
f
au
lt
d
iag
n
o
s
is
,
f
o
cu
s
in
g
o
n
n
o
n
-
in
v
asiv
e
s
en
s
in
g
was
d
o
n
e
b
y
[
24
]
.
A
s
tu
d
y
f
o
cu
s
in
g
o
n
im
p
r
o
v
in
g
d
iag
n
o
s
tic
ac
cu
r
ac
y
a
n
d
r
o
b
u
s
tn
ess
b
y
f
u
s
in
g
d
ata
f
r
o
m
m
u
ltip
le
u
s
in
g
an
ad
v
an
ce
d
Dee
p
L
ea
r
n
in
g
f
r
am
ewo
r
k
was g
iv
en
b
y
[
25
].
T
h
is
p
ap
er
aim
s
to
d
em
o
n
s
tr
ate
th
e
v
iab
ilit
y
o
f
a
h
y
b
r
id
ap
p
r
o
ac
h
t
h
at
co
m
b
in
es
FF
T
with
m
ac
h
in
e
lear
n
in
g
(
ML
)
a
n
d
d
ee
p
lear
n
in
g
(
DL
)
m
o
d
els
u
s
in
g
th
e
cu
r
r
en
t
an
d
f
lu
x
s
ig
n
at
u
r
es
o
f
elec
tr
ic
m
o
to
r
s
to
u
n
d
er
s
tan
d
an
d
d
iag
n
o
s
e
f
ailu
r
es.
T
h
is
jo
in
t
a
p
p
r
o
a
ch
is
a
h
ig
h
ly
ef
f
ec
tiv
e,
c
u
ttin
g
-
ed
g
e
m
eth
o
d
i
n
th
e
f
ield
o
f
p
r
ed
ictiv
e
m
ai
n
ten
an
ce
(
P
M)
f
o
r
elec
tr
ic
m
o
to
r
s
.
T
h
is
m
eth
o
d
o
lo
g
y
is
cr
u
cial
f
o
r
I
n
d
u
s
tr
y
4
.
0
ap
p
licatio
n
s
,
en
ab
lin
g
p
r
o
g
n
o
s
tics
an
d
h
ea
lth
m
an
ag
e
m
en
t
(
PHM)
s
y
s
tem
s
th
at
m
ax
im
ize
m
o
t
o
r
life
s
p
an
a
n
d
m
in
im
ize
u
n
p
lan
n
ed
d
o
wn
tim
e.
2.
M
E
T
H
O
D
2
.
1
.
E
x
perim
ent
a
l
s
et
up
T
h
e
d
ataset
was
cu
r
ated
b
y
co
n
d
u
ctin
g
ex
p
e
r
im
en
ts
.
T
h
e
t
est
s
etu
p
co
n
s
is
t
s
o
f
two
m
o
to
r
s
,
o
n
e
is
tak
en
as
th
e
h
ea
lth
y
m
o
to
r
an
d
f
au
lts
ar
e
s
im
u
lated
in
t
h
e
o
th
e
r
m
o
to
r
ca
lled
test
m
o
to
r
.
I
t
also
h
as
a
d
y
n
am
o
m
eter
wh
ich
ca
n
b
e
c
o
u
p
led
to
a
n
y
o
n
e
o
f
th
e
two
id
en
tical
in
d
u
ctio
n
m
o
to
r
s
.
T
wo
id
en
tical
m
o
to
r
s
h
av
e
th
e
f
o
llo
win
g
s
p
ec
if
icati
o
n
:
t
h
r
ee
-
p
h
ase
,
1
HP,
4
1
5
V,
5
0
Hz,
4
p
o
le
,
an
d
1
5
0
0
r
p
m
.
Flu
x
c
o
il
with
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ap
p
l Po
wer
E
n
g
I
SS
N:
2252
-
8
7
9
2
P
r
ed
ictive
ma
in
ten
a
n
ce
fo
r
in
d
u
ctio
n
mo
t
o
r
s
:
a
n
o
ve
l syn
er
g
y
o
f d
ee
p
lea
r
n
in
g
…
(
V
.
R
a
jin
i
)
1159
8
0
tu
r
n
s
is
also
in
s
er
ted
in
b
o
th
th
e
m
o
to
r
s
f
o
r
th
e
f
lu
x
m
ea
s
u
r
em
en
t.
T
h
e
test
m
o
to
r
is
p
r
o
v
id
ed
with
a
n
ex
ter
n
al
tap
p
i
n
g
to
cr
ea
te
s
t
ato
r
in
ter
t
u
r
n
s
h
o
r
t
cir
c
u
it
f
au
lt.
Pro
v
is
io
n
is
g
i
v
en
f
o
r
t
h
e
r
ep
lace
m
e
n
t
o
f
r
o
to
r
/b
ea
r
i
n
g
with
p
u
n
ch
ed
r
o
to
r
/b
ea
r
in
g
to
c
r
ea
te
r
o
to
r
/b
ea
r
in
g
f
a
u
lt
,
r
esp
ec
tiv
ely
.
Dy
n
am
o
m
eter
r
atin
g
is
0
.
7
5
k
W
,
1
5
0
0
r
p
m
,
0
.
5
k
g
-
m
,
ex
citatio
n
o
f
8
5
v
o
lts
.
B
y
a
d
j
u
s
tin
g
th
e
ex
citatio
n
v
o
ltag
e
o
f
th
e
d
y
n
am
o
m
ete
r
s
etu
p
,
th
e
lo
ad
o
f
th
e
i
n
d
u
ctio
n
m
o
to
r
ca
n
b
e
v
ar
ied
.
Data
ac
q
u
is
itio
n
ca
r
d
with
m
ea
s
u
r
e
m
en
t c
o
m
p
u
ti
n
g
USB
1
2
0
8
FS
–
Plu
s
is
u
s
ed
to
ac
q
u
ir
e
th
e
s
tato
r
cu
r
r
en
t
s
ig
n
al
u
n
d
er
h
ea
lth
y
an
d
v
ar
io
u
s
f
au
lty
co
n
d
itio
n
s
.
T
h
e
d
ata
ac
q
u
is
itio
n
was
d
o
n
e
u
s
i
n
g
a
USB
-
1
2
0
8
FS
with
8
c
o
n
f
ig
u
r
a
b
le
an
al
o
g
in
p
u
ts
f
o
r
11
-
b
it
s
in
g
le
-
en
d
e
d
eig
h
t
in
p
u
ts
o
r
f
o
u
r
1
2
-
b
it
d
if
f
er
en
tial
in
p
u
ts
an
d
two
1
2
-
b
i
t
an
alo
g
o
u
tp
u
t
ch
an
n
els.
An
a
lo
g
in
p
u
t
ch
an
n
els
ar
e
u
s
ed
t
o
g
et
th
e
s
ig
n
al
f
r
o
m
th
e
2
:1
r
atio
cu
r
r
en
t
s
en
s
o
r
an
d
th
e
lo
ad
to
r
q
u
e
s
ig
n
al
f
r
o
m
th
e
d
y
n
am
o
m
eter
.
DSO
-
X
3
0
2
4
A
MY
5
4
0
2
0
5
1
8
-
2
0
0
MH
z
is
u
s
ed
f
o
r
ac
q
u
i
r
in
g
th
e
s
tato
r
cu
r
r
en
t
an
d
f
lu
x
s
ig
n
als.
Fig
u
r
e
1
s
h
o
ws
th
e
ex
p
er
im
en
tal
s
etu
p
u
s
ed
f
o
r
th
e
p
r
esen
t
in
v
e
s
tig
atio
n
.
T
h
e
o
v
er
all
ap
p
r
o
ac
h
is
d
ep
icted
in
Fig
u
r
e
2
.
T
h
is
test
b
en
ch
was
m
eticu
lo
u
s
ly
s
et
u
p
to
en
ab
le
th
e
em
u
latio
n
o
f
s
tato
r
win
d
in
g
in
ter
-
tu
r
n
s
h
o
r
t
-
cir
cu
its
.
T
h
is
in
v
o
l
v
ed
p
r
o
v
is
io
n
o
n
th
e
s
tato
r
cir
cu
it
t
o
f
a
cilitate
th
e
in
s
er
tio
n
o
f
s
h
o
r
t
cir
cu
its
at
v
ar
io
u
s
lev
els,
r
an
g
in
g
f
r
o
m
s
u
b
tle
d
ef
ec
ts
to
s
ev
er
e
s
itu
atio
n
s
.
Fig
u
r
e
1
.
E
x
p
er
im
e
n
tal
s
etu
p
Fig
u
r
e
2
.
Ov
e
r
all
b
lo
ck
d
iag
r
a
m
2
.
2
.
T
he
d
a
t
a
s
et
a
nd
da
t
a
e
x
plo
ra
t
io
n
T
h
e
d
ata
s
et
h
as
th
e
f
lu
x
an
d
cu
r
r
en
t
s
ig
n
atu
r
es
f
o
r
n
o
lo
a
d
,
h
alf
lo
a
d
an
d
f
u
ll
lo
ad
ap
p
lied
to
th
e
m
o
to
r
.
T
h
e
d
r
i
v
e
f
r
eq
u
e
n
cy
is
also
ad
ju
s
ted
f
r
o
m
2
5
Hz
to
5
0
Hz
with
5
Hz
in
cr
em
en
ts
.
A
s
am
p
le
cu
r
r
en
t a
n
d
f
lu
x
s
p
ec
tr
a
r
ec
o
r
d
e
d
is
s
h
o
wn
in
Fig
u
r
e
3
.
C
h
1
r
ep
r
esen
ts
th
e
f
lu
x
an
d
C
h
2
,
3
,
an
d
4
ar
e
th
e
th
r
ee
p
h
ase
cu
r
r
en
ts
.
C
o
n
s
eq
u
e
n
tly
,
a
d
ec
is
io
n
was
m
ad
e
to
lev
er
a
g
e
t
h
ese
two
s
ets
o
f
m
ea
s
u
r
em
en
ts
in
d
ep
en
d
en
tly
to
co
n
s
tr
u
ct
d
is
tin
ct
p
r
e
d
ictiv
e
m
o
d
els,
ea
ch
f
o
cu
s
ed
o
n
id
en
tify
in
g
t
h
e
co
r
r
esp
o
n
d
in
g
f
au
lt
class
es.
T
h
is
s
ep
ar
atio
n
o
f
d
ata
was
in
ten
d
e
d
to
ass
ess
th
e
in
d
iv
id
u
al
c
o
r
r
elatio
n
s
b
etwe
en
cu
r
r
e
n
t
v
alu
e
s
an
d
f
au
lt
class
es,
as we
ll a
s
b
etwe
en
f
lu
x
an
d
f
a
u
lt c
lass
es.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
7
9
2
I
n
t J Ap
p
l Po
wer
E
n
g
,
Vo
l.
1
5
,
No
.
3
,
Sep
tem
b
er
20
2
6
:
1
1
5
7
-
1
1
6
7
1160
2
.
3
.
F
e
a
t
ure
e
x
t
r
a
ct
io
n
T
h
e
FFT
was
em
p
lo
y
ed
to
e
x
tr
ac
t
v
iab
le
f
ea
tu
r
es
f
r
o
m
t
h
e
m
ea
s
u
r
ed
c
u
r
r
en
t
a
n
d
f
lu
x
d
ata
an
d
tr
an
s
f
o
r
m
t
h
e
ex
tr
ac
te
d
f
ea
t
u
r
es
as
d
ee
m
ed
ef
f
icien
t
f
o
r
th
e
m
o
d
el
to
lear
n
an
d
r
e
p
r
o
d
u
ce
.
T
h
e
FFT
,
co
m
m
o
n
l
y
ab
b
r
ev
iated
as FFT,
is
a
co
m
m
o
n
y
et
ef
f
ec
tiv
e
tec
h
n
iq
u
e
em
p
lo
y
e
d
in
s
ig
n
al
p
r
o
ce
s
s
in
g
to
an
aly
ze
an
d
r
ep
r
esen
t
s
ig
n
als
in
th
e
f
r
eq
u
en
cy
d
o
m
ain
.
T
h
e
FF
T
ac
co
m
p
lis
h
es
th
is
b
y
d
ec
o
m
p
o
s
i
n
g
a
s
ig
n
al
in
to
it
s
co
n
s
titu
en
t
s
in
u
s
o
id
al
co
m
p
o
n
en
ts
o
f
v
ar
y
in
g
f
r
eq
u
e
n
cies
an
d
am
p
litu
d
es
as
s
h
o
wn
in
Fig
u
r
e
4
.
T
h
er
ef
o
r
e,
it
is
p
ar
ticu
lar
ly
u
s
ef
u
l
f
o
r
id
en
t
if
y
in
g
d
o
m
in
an
t
f
r
eq
u
e
n
cies
an
d
p
er
io
d
ic
co
m
p
o
n
e
n
ts
with
in
a
s
ig
n
al.
FF
T
is
d
ep
lo
y
e
d
f
o
r
th
e
g
i
v
en
f
lu
x
an
d
cu
r
r
e
n
t sig
n
als to
ca
p
tu
r
e
th
e
co
n
s
titu
en
t f
r
eq
u
en
cy
co
m
p
o
n
en
ts
.
T
h
e
d
ataset
in
itially
co
n
s
is
ted
o
f
1
0
0
,
0
0
0
-
tim
e
d
o
m
ai
n
s
am
p
les
co
llected
o
v
er
a
1
0
-
s
ec
o
n
d
in
ter
v
al.
T
o
f
ac
ilit
ate
ef
f
icien
t
p
r
o
ce
s
s
in
g
f
ir
s
tly
,
in
f
i
n
ite
v
alu
es
we
r
e
r
em
o
v
ed
f
r
o
m
th
e
d
ataset.
Su
b
s
eq
u
en
tly
,
FF
T
was u
s
ed
f
o
r
f
r
e
q
u
en
c
y
d
o
m
ai
n
co
n
v
er
s
io
n
.
Fig
u
r
e
3
.
Sam
p
le
c
u
r
r
e
n
t a
n
d
f
lu
x
s
ig
n
atu
r
es
Fig
u
r
e
4
.
T
im
e
an
d
f
r
eq
u
en
cy
d
o
m
ain
s
p
ec
tr
a
T
h
e
F
F
T
o
p
e
r
a
t
i
o
n
g
e
n
e
r
a
t
e
d
1
0
0
,
0
0
0
s
a
m
p
l
e
s
i
n
t
h
e
f
r
e
q
u
e
n
c
y
d
o
m
a
i
n
.
T
o
a
v
o
i
d
r
e
d
u
n
d
a
n
t
d
a
t
a
,
a
d
o
w
n
-
s
a
m
p
l
i
n
g
s
t
e
p
w
as
in
t
r
o
d
u
c
e
d
b
y
t
a
k
i
n
g
a
d
v
a
n
t
a
g
e
o
f
t
h
e
s
y
m
m
e
t
r
y
i
n
s
i
g
n
al
m
a
g
n
i
t
u
d
e
a
r
o
u
n
d
d
i
s
c
o
n
t
i
n
u
it
i
es
.
T
h
i
s
al
l
o
w
e
d
f
o
r
t
h
e
r
e
m
o
v
a
l
o
f
o
n
e
h
a
l
f
o
f
t
h
e
v
a
l
u
e
s
w
h
il
e
r
e
t
a
i
n
i
n
g
t
h
e
r
e
l
e
v
a
n
t
i
n
f
o
r
m
a
t
i
o
n
.
F
r
o
m
t
h
e
1
0
0
,
0
0
0
f
r
e
q
u
e
n
c
y
d
o
m
a
i
n
s
a
m
p
l
es
,
o
n
l
y
5
0
,
0
0
0
s
am
p
l
e
s
w
e
r
e
r
e
t
a
i
n
e
d
f
o
r
f
u
r
t
h
e
r
a
n
a
l
y
s
is
.
Du
r
in
g
o
u
r
ab
latio
n
s
tu
d
ies,
we
id
en
tifie
d
th
at
s
am
p
lin
g
f
r
eq
u
en
cies
at
5
Hz
in
ter
v
als
p
r
o
v
id
ed
th
e
o
p
tim
al
p
er
f
o
r
m
a
n
ce
f
o
r
o
u
r
m
o
d
el.
T
h
is
f
in
d
in
g
was
p
iv
o
t
al
as
it
allo
wed
u
s
to
s
ig
n
if
ica
n
tly
r
ed
u
ce
th
e
d
ata
d
im
en
s
io
n
ality
with
o
u
t
co
m
p
r
o
m
is
in
g
th
e
m
o
d
el'
s
ef
f
icac
y
.
I
n
itially
,
o
u
r
d
ataset
c
o
n
s
is
ted
o
f
5
0
,
0
0
0
s
am
p
les,
ea
ch
r
ep
r
esen
tin
g
a
co
m
p
r
eh
e
n
s
iv
e
r
an
g
e
o
f
f
r
eq
u
e
n
cies.
T
o
f
u
r
th
e
r
s
tr
ea
m
lin
e
th
e
d
ata
s
et
an
ad
d
itio
n
al
r
ed
u
ctio
n
s
tep
was
im
p
lem
en
ted
.
Sp
ec
if
ica
lly
,
ev
er
y
1
0
th
s
am
p
le
f
r
o
m
th
e
o
r
ig
in
al
s
et
o
f
5
0
,
0
0
0
s
am
p
les
was
s
elec
ted
.
T
h
is
s
tr
ateg
ic
d
o
wn
-
s
a
m
p
lin
g
r
esu
lted
in
a
f
in
al
d
ataset
wh
er
e
ea
ch
d
ata
p
o
in
t
co
n
s
is
ted
o
f
5
,
0
0
0
f
ea
t
u
r
es
in
th
e
f
r
eq
u
en
c
y
d
o
m
ai
n
.
B
y
ap
p
ly
in
g
th
ese
two
k
ey
s
tep
s
,
s
am
p
lin
g
at
5
Hz
in
ter
v
als
an
d
th
e
n
r
ed
u
cin
g
th
e
to
tal
n
u
m
b
e
r
o
f
s
am
p
les b
y
a
f
ac
to
r
o
f
1
0
,
th
e
cr
itical
f
r
eq
u
en
c
y
in
f
o
r
m
atio
n
n
ec
ess
ar
y
f
o
r
th
e
m
o
d
els’
p
er
f
o
r
m
an
ce
was
r
etain
ed
wh
ile
s
ig
n
if
ican
tly
r
ed
u
cin
g
th
e
d
ataset
s
ize.
A
s
am
p
le
FF
T
f
o
r
cu
r
r
en
t a
n
d
f
l
u
x
s
ig
n
als is
g
iv
en
in
Fig
u
r
e
5
.
Fig
u
r
e
5
.
Sam
p
le
FF
T
s
p
ec
tr
a
f
o
r
cu
r
r
en
t a
n
d
f
lu
x
s
ig
n
als
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ap
p
l Po
wer
E
n
g
I
SS
N:
2252
-
8
7
9
2
P
r
ed
ictive
ma
in
ten
a
n
ce
fo
r
in
d
u
ctio
n
mo
t
o
r
s
:
a
n
o
ve
l syn
er
g
y
o
f d
ee
p
lea
r
n
in
g
…
(
V
.
R
a
jin
i
)
1161
T
o
f
u
r
th
er
u
n
d
er
s
tan
d
th
e
p
atter
n
o
f
f
au
lts
in
t
h
e
d
ata,
s
p
ec
t
r
o
g
r
am
s
o
f
ea
ch
ch
an
n
el
wer
e
g
en
er
ate
d
s
ep
ar
ately
an
d
s
u
b
s
eq
u
en
tly
u
s
ed
to
p
r
ed
ict
f
a
u
lts
in
an
i
n
d
u
ctio
n
m
o
to
r
.
Fo
r
ea
ch
ch
a
n
n
el,
a
s
p
ec
tr
o
g
r
a
m
was
co
m
p
u
ted
u
s
in
g
th
e
s
h
o
r
t
-
tim
e
Fo
u
r
ier
tr
an
s
f
o
r
m
(
ST
FT)
with
a
s
eg
m
en
t
len
g
th
o
f
1
0
2
4
s
am
p
les,
an
o
v
er
lap
o
f
5
1
2
s
am
p
les,
an
d
1
0
2
4
FF
T
p
o
in
ts
.
T
h
e
s
p
ec
tr
o
g
r
am
,
wh
ich
co
n
v
er
te
d
th
e
tim
e
-
s
er
ies d
ata
in
to
th
e
f
r
eq
u
e
n
cy
d
o
m
ain
,
was
th
en
r
escaled
to
a
0
-
2
5
5
r
an
g
e
to
f
ac
ilit
ate
im
ag
e
r
ep
r
esen
tatio
n
.
T
h
ese
r
escaled
s
p
ec
tr
o
g
r
am
s
wer
e
s
av
ed
as
PNG
im
ag
es
in
d
esig
n
ated
d
ir
ec
to
r
ies,
allo
win
g
f
o
r
v
is
u
al
an
aly
s
is
an
d
f
u
r
th
er
p
r
o
ce
s
s
in
g
in
f
a
u
lt p
r
ed
ictio
n
m
o
d
els.
3.
F
AULT
C
L
A
SS
I
F
I
CA
T
I
O
N
M
E
T
H
O
DS A
ND
R
E
SU
L
T
S
T
h
e
p
ar
a
m
etr
ic
class
if
ier
s
,
s
u
ch
as
m
ax
im
u
m
lik
elih
o
o
d
clas
s
if
icatio
n
(
ML
C
)
p
r
o
v
id
e
h
ig
h
ac
cu
r
ac
y
wh
en
th
e
d
ata
h
as
a
n
o
r
m
al
d
i
s
tr
ib
u
tio
n
.
T
h
eir
p
er
f
o
r
m
a
n
ce
d
r
asti
ca
lly
f
alls
wh
en
th
e
d
ata
h
as
a
non
–
n
o
r
m
al
d
is
tr
ib
u
tio
n
.
No
n
p
a
r
am
etr
ic
alter
n
ativ
es
lik
e
d
ec
is
io
n
tr
ee
,
r
an
d
o
m
f
o
r
est,
a
n
d
ar
tifi
cial
n
eu
r
al
n
etwo
r
k
(
ANN)
class
if
ier
s
ar
e
b
ec
o
m
in
g
p
o
p
u
lar
f
o
r
d
ata
w
h
ich
d
o
n
o
t h
av
e
n
o
r
m
al
d
is
tr
ib
u
tio
n
s
.
3
.
1
.
Dec
is
io
n t
re
e
A
d
ec
is
io
n
tr
ee
(
DT
)
is
a
f
lo
wch
ar
t
-
s
ty
le
m
o
d
el
i
n
wh
ich
ea
ch
in
ter
n
al
n
o
d
e
test
s
an
attr
ib
u
te,
ea
ch
ed
g
e
co
r
r
esp
o
n
d
s
to
an
o
u
tc
o
m
e
o
f
th
at
test
,
an
d
ea
ch
leaf
s
to
r
es
a
cla
s
s
lab
el.
T
o
cl
ass
if
y
an
u
n
lab
eled
in
s
tan
ce
XXX,
th
e
attr
ib
u
te
test
s
ar
e
ev
alu
ated
f
r
o
m
t
h
e
r
o
o
t
to
a
s
in
g
le
leaf
;
th
e
lab
el
at
th
at
ter
m
in
al
n
o
d
e
is
r
etu
r
n
ed
as
th
e
p
r
ed
ictio
n
.
D
T
s
ar
e
h
ig
h
ly
i
n
ter
p
r
etab
le
a
n
d
o
f
te
n
ac
h
iev
e
c
o
m
p
etitiv
e
ac
cu
r
ac
y
.
I
t
h
as
an
in
h
er
en
t
f
ea
tu
r
e
o
f
h
an
d
lin
g
m
ix
ed
d
ata
t
y
p
es
with
an
ab
ilit
y
o
f
ac
c
o
m
m
o
d
atin
g
m
is
s
in
g
f
e
atu
r
es
.
Ho
wev
er
,
a
s
m
all
ch
an
g
e
in
th
e
tr
ain
i
n
g
d
ata
ca
n
r
esu
lt
in
a
d
if
f
er
en
t
p
r
ed
ictio
n
.
DT
s
s
u
p
p
o
r
t
m
u
lti
-
class
clas
s
if
icatio
n
n
ativ
ely
an
d
ca
n
also
b
e
c
o
n
f
ig
u
r
ed
f
o
r
r
eg
r
ess
io
n
tas
k
s
.
Af
ter
tr
ain
in
g
,
in
f
er
en
ce
is
co
m
p
u
tatio
n
ally
in
ex
p
en
s
iv
e.
A
d
ec
is
io
n
tr
ee
class
if
ier
lear
n
s
a
h
ier
ar
c
h
y
o
f
ax
is
-
alig
n
ed
r
u
les
th
at
p
ar
titi
o
n
th
e
in
p
u
t
s
p
ac
e
in
to
r
eg
io
n
s
with
(
id
ea
lly
)
h
o
m
o
g
en
eo
u
s
class
lab
els.
T
h
e
m
o
d
el
is
n
o
n
-
p
a
r
am
etr
ic:
it
d
o
es
n
o
t
ass
u
m
e
a
f
ix
e
d
f
u
n
ctio
n
al
f
o
r
m
o
r
r
e
q
u
ir
e
iter
ativ
e
weig
h
t u
p
d
ates a
s
in
p
ar
am
etr
ic
m
o
d
els.
I
n
itially
,
u
p
o
n
d
ep
l
o
y
in
g
t
h
e
alg
o
r
ith
m
o
n
th
e
cu
r
r
en
t
d
ataset,
two
k
ey
o
b
s
er
v
atio
n
s
em
er
g
ed
.
Firstl
y
,
th
e
d
ec
is
io
n
b
o
u
n
d
ar
i
es
ar
e
v
er
y
s
h
allo
w,
in
d
icatin
g
th
at
o
n
l
y
a
f
ew
f
r
eq
u
e
n
cy
b
in
s
ar
e
g
en
u
in
ely
u
s
ef
u
l
f
o
r
p
r
ed
ictio
n
s
.
Seco
n
d
ly
,
m
o
s
t
o
f
th
ese
s
ig
n
if
ican
t
b
in
s
ar
e
in
th
e
h
ig
h
-
f
r
eq
u
en
cy
r
e
g
io
n
.
T
h
is
s
u
g
g
ests
th
at
th
e
cr
u
cial
f
lu
ct
u
atio
n
s
n
ec
ess
ar
y
f
o
r
ef
f
ec
tiv
e
f
au
lt
p
r
e
d
ictio
n
ar
e
p
r
im
ar
i
ly
f
o
u
n
d
in
h
ig
h
e
r
f
r
eq
u
e
n
cies,
th
u
s
r
e
q
u
ir
in
g
a
co
n
ce
n
tr
ated
f
o
c
u
s
o
n
th
ese
r
eg
io
n
s
.
Fig
u
r
e
s
6
(
a
)
an
d
6
(
b
)
s
h
o
w
th
e
co
n
f
u
s
io
n
m
atr
ix
o
f
DT
with
c
u
r
r
e
n
t a
n
d
f
lu
x
,
r
esp
ec
tiv
ely
.
(
a)
(
b
)
Fig
u
r
e
6
.
C
o
n
f
u
s
io
n
m
atr
i
x
o
f
d
ec
is
io
n
t
r
ee
with
(
a
)
c
u
r
r
e
n
t
a
n
d
(
b
)
f
lu
x
s
ig
n
atu
r
e
Giv
en
th
ese
o
b
s
er
v
atio
n
s
,
it
w
as
clea
r
th
at
m
o
r
e
s
o
p
h
is
ticate
d
tech
n
i
q
u
es
wer
e
n
ec
ess
ar
y
to
im
p
r
o
v
e
th
e
m
o
d
el'
s
p
er
f
o
r
m
an
ce
an
d
h
an
d
le
th
e
in
h
er
e
n
t
n
o
n
-
lin
ea
r
ity
in
th
e
d
ata.
T
h
is
r
ea
lizatio
n
led
u
s
to
ex
p
lo
r
e
en
s
em
b
le
lear
n
in
g
tech
n
iq
u
es,
p
ar
ticu
lar
ly
r
an
d
o
m
f
o
r
est
,
w
h
ich
o
f
f
er
ed
a
m
o
r
e
r
o
b
u
s
t
an
d
f
lex
ib
le
ap
p
r
o
ac
h
to
ca
p
tu
r
in
g
th
e
co
m
p
lex
ity
o
f
th
e
d
ata.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
7
9
2
I
n
t J Ap
p
l Po
wer
E
n
g
,
Vo
l.
1
5
,
No
.
3
,
Sep
tem
b
er
20
2
6
:
1
1
5
7
-
1
1
6
7
1162
3
.
2
.
Ra
nd
o
m
f
o
re
s
t
cl
a
s
s
if
ier
T
h
e
R
F
class
if
ier
is
an
e
n
s
em
b
le
-
b
ased
a
p
p
r
o
ac
h
u
s
in
g
m
u
ltip
le
d
ec
is
io
n
tr
ee
class
if
icatio
n
an
d
r
eg
r
ess
io
n
tr
ee
s
.
R
an
d
o
m
m
u
l
tip
le
tr
ee
s
an
d
r
esam
p
led
tr
ain
in
g
d
ata
ar
e
u
s
ed
to
im
p
r
o
v
e
th
e
p
er
f
o
r
m
a
n
ce
.
I
n
d
iv
id
u
al
tr
ee
p
r
ed
ictio
n
s
ar
e
u
s
ed
f
o
r
f
in
al
m
o
d
el
p
r
ed
icti
o
n
.
I
m
p
lem
en
tin
g
th
e
r
an
d
o
m
f
o
r
est
class
if
ier
d
id
n
o
t
lead
t
o
s
ig
n
if
ican
t
im
p
r
o
v
em
en
ts
in
th
e
f
au
lt
d
etec
tio
n
.
Fig
u
r
e
s
7
(
a)
an
d
7
(
b
)
s
h
o
w
t
h
e
co
n
f
u
s
io
n
m
atr
i
x
o
f
th
e
r
a
n
d
o
m
f
o
r
est alg
o
r
ith
m
with
th
e
cu
r
r
e
n
t a
n
d
f
lu
x
s
ig
n
atu
r
es
,
r
esp
ec
tiv
ely
.
(
a)
(
b
)
Fig
u
r
e
7
.
C
o
n
f
u
s
io
n
m
atr
i
x
o
f
r
an
d
o
m
f
o
r
est
:
(
a
)
f
lu
x
a
n
d
(
b
)
cu
r
r
en
t sig
n
at
u
r
e
T
h
e
p
er
f
o
r
m
a
n
ce
m
etr
ics
lik
e
F1
s
co
r
e,
p
r
ec
is
io
n
,
r
ec
all
,
a
n
d
f
a
u
lt
ac
cu
r
ac
y
f
o
r
s
tato
r
c
u
r
r
en
t
a
n
d
f
lu
x
ar
e
co
m
p
a
r
ed
in
F
ig
u
r
e
8
.
W
h
ile
b
o
th
d
ec
is
io
n
tr
ee
s
an
d
r
an
d
o
m
f
o
r
est
class
if
ier
s
d
em
o
n
s
tr
ated
d
ec
e
n
t
p
er
f
o
r
m
an
ce
,
ac
h
ie
v
in
g
an
a
v
er
ag
e
o
f
7
1
%
ac
cu
r
ac
y
,
f
o
r
cu
r
r
en
t
s
ig
n
atu
r
es,
we
f
u
r
t
h
er
in
v
esti
g
ated
d
ee
p
lear
n
in
g
tech
n
iq
u
es
to
ac
h
iev
e
b
etter
p
e
r
f
o
r
m
an
ce
with
c
u
r
r
en
t
s
ig
n
atu
r
es.
As
a
r
esu
lt,
a
n
ex
p
e
r
im
en
t
with
a
s
im
p
le
n
eu
r
al
n
etwo
r
k
was b
u
i
lt.
Fig
u
r
e
8
.
E
v
alu
atio
n
s
co
r
es o
f
f
au
lt d
etec
tio
n
3
.
3
.
Art
if
ici
a
l
neura
l net
wo
rk
An
ANN,
r
ep
r
esen
ts
a
f
u
n
d
am
en
tal
class
o
f
m
ac
h
in
e
lear
n
in
g
m
o
d
els
in
s
p
ir
ed
b
y
th
e
h
u
m
an
b
r
ain
.
ANNs
co
n
s
is
t
o
f
in
ter
co
n
n
ec
t
ed
lay
er
s
o
f
ar
tific
ial
n
e
u
r
o
n
s
th
at
p
r
o
ce
s
s
an
d
tr
a
n
s
f
o
r
m
in
p
u
t
d
ata
to
ca
p
tu
r
e
th
e
u
n
d
er
ly
i
n
g
n
o
n
lin
ea
r
ity
.
A
p
er
ce
p
tr
o
n
,
th
e
f
u
n
d
am
en
t
al
u
n
it
o
f
ANN,
ac
ce
p
ts
m
u
ltip
le
in
p
u
ts
,
an
d
th
e
weig
h
ted
s
u
m
o
f
ea
c
h
in
p
u
t
is
ad
d
ed
u
p
.
T
h
e
weig
h
ts
en
h
an
ce
o
r
in
h
ib
it
a
s
ig
n
al
.
A
b
ias
v
alu
e
is
th
en
ad
d
ed
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ap
p
l Po
wer
E
n
g
I
SS
N:
2252
-
8
7
9
2
P
r
ed
ictive
ma
in
ten
a
n
ce
fo
r
in
d
u
ctio
n
mo
t
o
r
s
:
a
n
o
ve
l syn
er
g
y
o
f d
ee
p
lea
r
n
in
g
…
(
V
.
R
a
jin
i
)
1163
to
th
e
weig
h
te
d
s
u
m
b
ef
o
r
e
i
t
is
p
ass
ed
to
an
ac
tiv
atio
n
f
u
n
ctio
n
t
o
in
tr
o
d
u
ce
n
o
n
-
lin
e
ar
ity
to
th
e
o
u
tp
u
ts
.
Hen
ce
,
th
e
m
o
d
el
lear
n
s
co
m
p
lex
an
d
n
o
n
lin
ea
r
p
atter
n
s
[9
]
,
[
1
0
]
.
Per
ce
p
tr
o
n
s
a
r
e
s
tack
ed
to
g
eth
er
i
n
p
a
r
allel
(
in
a
la
y
er
)
an
d
s
eq
u
en
tially
o
v
er
m
u
ltip
le
lay
e
r
s
,
wh
er
e
i
n
p
u
t
d
ata
g
o
es
th
r
o
u
g
h
r
e
p
etitiv
e
tr
an
s
f
o
r
m
atio
n
s
,
u
ltima
tely
p
r
o
d
u
cin
g
a
n
o
u
tp
u
t
o
r
p
r
ed
ictio
n
.
Fig
u
r
e
9
d
e
p
icts
a
s
in
g
le
p
er
ce
p
tr
o
n
.
T
h
e
eq
u
atio
n
f
o
r
a
s
in
g
l
e
p
er
ce
p
tr
o
n
is
g
iv
en
i
n
(
1
)
.
=
(
+
)
(
1
)
Fig
u
r
e
9
.
Sin
g
le
p
er
ce
p
tr
o
n
T
h
e
d
en
s
ely
co
n
n
ec
ted
AN
N
with
8
lay
e
r
s
is
u
s
ed
f
o
r
th
is
s
tu
d
y
.
E
ac
h
lay
e
r
em
p
lo
y
s
R
eL
U
ac
tiv
atio
n
f
u
n
ctio
n
s
.
T
h
e
o
u
tp
u
t
lay
er
u
s
es
a
So
f
tm
ax
ac
tiv
atio
n
f
u
n
ctio
n
.
T
o
p
r
ev
e
n
t
o
v
er
f
itti
n
g
,
a
d
r
o
p
o
u
t
r
ate
o
f
0
.
2
is
ap
p
lied
to
ea
c
h
l
ay
er
.
T
h
e
ANN
p
er
f
o
r
m
e
d
c
o
m
p
ar
ativ
ely
well,
ac
h
iev
in
g
a
n
ac
cu
r
ac
y
o
f
8
2
%.
Fig
u
r
e
1
0
s
h
o
ws
th
e
co
n
f
u
s
io
n
m
atr
ix
o
f
ANN
with
cu
r
r
en
t
s
ig
n
atu
r
es.
I
t
is
ev
id
en
t
f
r
o
m
Fig
u
r
e
1
0
th
at
th
e
ANN
s
lig
h
tly
o
u
tp
er
f
o
r
m
e
d
th
e
r
an
d
o
m
f
o
r
est
class
if
ier
d
u
e
to
its
ca
p
ab
ilit
y
to
ca
p
tu
r
e
th
e
non
-
lin
ea
r
ities
an
d
h
ig
h
er
-
o
r
d
e
r
in
ter
ac
tio
n
s
with
i
n
th
e
d
ata,
w
h
ich
ar
e
c
r
u
cial
f
o
r
ac
cu
r
ate
f
lu
x
p
r
ed
ictio
n
.
Fig
u
r
e
1
0
.
C
o
n
f
u
s
io
n
m
atr
ix
o
f
ANN
f
o
r
cu
r
r
en
t sig
n
atu
r
e
Giv
en
th
e
c
o
m
p
lex
ity
o
f
th
e
cu
r
r
e
n
t
d
ata,
a
d
ec
is
io
n
was
m
ad
e
t
o
f
u
r
th
er
ex
p
l
o
r
e
d
ee
p
lear
n
in
g
tech
n
iq
u
es f
o
r
f
au
lt c
lass
if
icat
io
n
u
s
in
g
th
e
th
r
ee
d
if
f
er
en
t c
u
r
r
en
t c
h
a
n
n
els.
T
o
lev
er
a
g
e
th
e
tem
p
o
r
al
p
atter
n
s
an
d
s
eq
u
en
tial
n
at
u
r
e
o
f
th
e
d
ata,
we
d
ec
id
ed
t
o
im
p
lem
e
n
t
a
1
D
C
NN.
B
o
th
th
e
s
p
ec
tr
o
g
r
am
s
o
f
a
s
in
g
le
ch
an
n
el
o
f
cu
r
r
en
t a
n
d
th
e
r
aw
s
ig
n
al
wer
e
in
p
u
t sep
a
r
ately
f
o
r
co
m
p
ar
is
o
n
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
7
9
2
I
n
t J Ap
p
l Po
wer
E
n
g
,
Vo
l.
1
5
,
No
.
3
,
Sep
tem
b
er
20
2
6
:
1
1
5
7
-
1
1
6
7
1164
3
.
4
.
T
he
pro
po
s
ed
CNN
A
s
tr
aig
h
tf
o
r
war
d
1
D
C
NN
was
im
p
lem
en
ted
in
Py
T
o
r
ch
,
s
p
ec
if
ically
d
esig
n
ed
f
o
r
class
if
icatio
n
task
s
.
T
h
e
ar
ch
itectu
r
e
co
n
s
is
ts
o
f
a
s
er
ies
o
f
co
n
v
o
lu
tio
n
al
lay
er
s
,
ea
ch
f
o
llo
wed
b
y
th
e
G
au
s
s
ian
er
r
o
r
lin
ea
r
u
n
it
(
GE
L
U)
ac
tiv
atio
n
f
u
n
cti
o
n
,
an
d
cu
lm
in
ates
in
a
lin
ea
r
lay
er
f
o
r
f
in
al
class
if
icatio
n
.
A
s
et
o
f
f
ilter
s
to
th
e
in
p
u
t
d
ata
ar
e
ap
p
lied
b
y
ea
ch
lay
er
an
d
f
ea
t
u
r
e
m
ap
s
c
ap
tu
r
in
g
v
ar
io
u
s
asp
ec
ts
o
f
th
e
in
p
u
t
s
ig
n
al
ar
e
ca
p
tu
r
ed
.
T
h
e
GE
L
U
ac
tiv
atio
n
f
u
n
ctio
n
is
a
p
p
lied
a
f
ter
ea
c
h
co
n
v
o
lu
tio
n
al
lay
er
.
T
h
e
d
ec
is
io
n
to
u
s
e
GE
L
U
o
v
er
o
th
er
ac
tiv
atio
n
f
u
n
ctio
n
s
,
s
u
ch
as
R
eL
U
o
r
s
ig
m
o
id
,
w
as
d
r
iv
en
b
y
its
s
u
p
er
io
r
p
er
f
o
r
m
an
ce
i
n
h
a
n
d
lin
g
co
m
p
lex
d
ata
d
is
tr
ib
u
tio
n
s
an
d
its
ab
ilit
y
to
p
r
o
v
id
e
a
s
m
o
o
th
tr
an
s
itio
n
f
o
r
n
eg
ativ
e
in
p
u
ts
.
T
h
is
co
n
tr
ib
u
tes
to
im
p
r
o
v
ed
g
r
ad
ie
n
t
f
lo
w
a
n
d
m
o
r
e
s
tab
le
tr
ain
i
n
g
.
GE
L
U
co
m
b
in
es
th
e
ad
v
an
tag
es
o
f
R
eL
U
with
th
e
s
m
o
o
th
n
ess
o
f
a
Gau
s
s
ian
d
is
tr
ib
u
tio
n
,
e
n
ab
lin
g
it
to
b
etter
ap
p
r
o
x
im
ate
th
e
in
p
u
t
d
is
tr
ib
u
tio
n
an
d
en
h
an
ce
th
e
m
o
d
el'
s
lear
n
in
g
ca
p
ab
ilit
y
.
T
h
e
f
in
al
lin
ea
r
lay
er
m
ap
s
th
e
h
ig
h
-
lev
el
f
ea
t
u
r
e
s
ex
tr
ac
ted
b
y
th
e
co
n
v
o
l
u
tio
n
al
lay
er
s
to
t
h
e
d
es
ir
ed
o
u
tp
u
t c
lass
es,
wh
ich
in
th
is
ca
s
e
is
s
et
to
s
ev
en
.
Un
d
er
s
tan
d
in
g
th
e
r
ec
ep
tiv
e
f
ield
o
f
co
n
v
o
l
u
tio
n
al
la
y
er
s
is
ess
en
tial
f
o
r
in
ter
p
r
etin
g
th
e
s
p
atial
co
n
tex
t
ca
p
tu
r
ed
b
y
th
e
n
etwo
r
k
.
T
h
e
r
ec
ep
tiv
e
f
ield
d
eter
m
in
es
th
e
ex
te
n
t
o
f
th
e
in
p
u
t
s
p
ac
e
th
at
af
f
ec
ts
th
e
o
u
tp
u
t
o
f
a
n
e
u
r
o
n
.
I
n
th
e
co
n
tex
t
o
f
C
NN
,
a
lar
g
e
r
r
ec
ep
t
iv
e
f
ield
allo
ws
ea
ch
n
e
u
r
o
n
to
in
co
r
p
o
r
ate
m
o
r
e
g
lo
b
al
in
f
o
r
m
atio
n
,
w
h
ile
a
s
m
aller
r
ec
ep
tiv
e
f
ield
f
o
cu
s
es
o
n
m
o
r
e
lo
ca
lized
p
atter
n
s
.
Acc
u
r
ate
ca
lcu
latio
n
o
f
th
e
r
ec
ep
tiv
e
f
ield
h
elp
s
in
d
esig
n
in
g
n
etwo
r
k
s
th
at
ca
n
ef
f
ec
tiv
ely
ca
p
t
u
r
e
th
e
n
ec
e
s
s
ar
y
f
ea
tu
r
es
f
o
r
a
g
iv
en
task
.
T
h
e
r
ec
ep
tiv
e
f
i
eld
ca
lcu
lato
r
f
u
n
ctio
n
c
o
m
p
u
tes
th
e
r
ec
ep
tiv
e
f
ield
f
o
r
th
e
en
tire
n
etwo
r
k
ar
ch
itectu
r
e,
lay
e
r
b
y
lay
er
.
T
h
is
f
u
n
ctio
n
iter
ates
th
r
o
u
g
h
t
h
e
lis
t
o
f
lay
er
s
,
c
alcu
latin
g
a
n
d
ac
cu
m
u
latin
g
t
h
e
r
ec
ep
tiv
e
f
iel
d
s
ize
at
ea
ch
s
t
ep
b
ased
o
n
th
e
co
r
r
esp
o
n
d
in
g
lay
er
s
'
k
er
n
el
s
ize,
s
tr
id
e,
an
d
d
ilatio
n
.
E
a
ch
s
u
cc
ess
iv
e
lay
er
ex
p
an
d
s
th
e
r
ec
ep
tiv
e
f
ield
ac
co
r
d
in
g
to
its
p
ar
am
eter
s
.
T
h
e
r
ec
ep
tiv
e
f
iel
d
R
out
af
ter
a
co
n
v
o
lu
tio
n
al
lay
e
r
ca
n
b
e
co
m
p
u
ted
iter
ativ
ely
.
Fo
r
a
1
D
co
n
v
o
l
u
tio
n
al
lay
er
with
in
p
u
t
r
ec
ep
tiv
e
f
ield
s
ize
R
in
,
k
er
n
el
s
ize
K,
s
tr
id
e
S,
an
d
d
ilatio
n
D,
th
e
r
ec
ep
tiv
e
f
ield
R
out
is
g
iv
en
in
(
2
).
=
+
(
−
1
)
×
(
2
)
T
h
e
C
NN
p
er
f
o
r
m
e
d
b
etter
w
ith
s
p
ec
tr
o
g
r
am
i
n
p
u
ts
with
7
lay
er
s
o
f
co
n
v
o
lu
tio
n
with
a
n
ac
cu
r
ac
y
o
f
8
6
%
.
T
h
e
co
n
f
u
s
io
n
m
at
r
ices
g
iv
en
in
Fig
u
r
e
11
ar
e
a
test
am
en
t
to
th
is
f
ac
t.
F
FT
o
f
th
e
r
aw
s
ig
n
als
p
r
o
v
id
e
a
tim
e
-
f
r
e
q
u
en
cy
r
ep
r
esen
tatio
n
o
f
th
e
in
p
u
t
s
ig
n
a
ls
,
allo
win
g
th
e
C
NNs
to
ca
p
tu
r
e
tem
p
o
r
al
an
d
f
r
eq
u
e
n
cy
d
o
m
ain
f
ea
tu
r
es
s
im
u
ltan
eo
u
s
ly
th
u
s
allo
win
g
f
o
r
an
ef
f
icien
t
class
if
icatio
n
o
f
th
e
d
ata.
T
h
e
ev
alu
atio
n
s
co
r
es o
f
f
au
lt d
ete
ctio
n
f
o
r
C
NN
is
g
iv
en
in
Fig
u
r
e
12.
Fig
u
r
e
1
1
.
C
o
n
f
u
s
io
n
m
atr
ix
o
f
C
NN
f
o
r
cu
r
r
e
n
t
s
ig
n
atu
r
e
Fig
u
r
e
1
2
.
E
v
alu
atio
n
s
co
r
es f
o
r
C
C
N
4.
I
NF
E
R
E
NC
E
S AN
D
DIS
CU
SS
I
O
N
T
h
e
p
er
f
o
r
m
an
ce
o
f
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
f
o
r
f
au
lt
class
if
icat
io
n
o
f
in
d
u
ctio
n
m
o
to
r
s
ar
e
ev
alu
ated
in
th
is
wo
r
k
.
T
h
e
ex
p
er
im
en
tal
s
etu
p
in
v
o
l
v
ed
co
llectin
g
cu
r
r
en
t
an
d
f
lu
x
s
ig
n
atu
r
es
f
r
o
m
b
o
th
h
ea
lth
y
an
d
f
au
lty
m
o
to
r
s
.
I
n
i
tially
,
th
e
d
ataset
co
n
s
i
s
ted
o
f
1
0
0
,
0
0
0
tim
e
-
d
o
m
ai
n
s
am
p
les
r
ec
o
r
d
ed
o
v
er
a
10
-
s
ec
o
n
d
in
ter
v
al.
T
o
f
ac
ilit
ate
an
aly
s
is
,
th
ese
s
ig
n
als
wer
e
co
n
v
er
ted
to
th
e
f
r
eq
u
en
cy
d
o
m
ain
u
s
in
g
FF
T
,
r
ed
u
cin
g
th
e
d
ata
to
5
0
,
0
0
0
s
am
p
les
af
ter
a
s
tr
ateg
ic
d
o
wn
-
s
am
p
lin
g
p
r
o
ce
s
s
th
at
elim
in
ated
r
ed
u
n
d
a
n
t
in
f
o
r
m
atio
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ap
p
l Po
wer
E
n
g
I
SS
N:
2252
-
8
7
9
2
P
r
ed
ictive
ma
in
ten
a
n
ce
fo
r
in
d
u
ctio
n
mo
t
o
r
s
:
a
n
o
ve
l syn
er
g
y
o
f d
ee
p
lea
r
n
in
g
…
(
V
.
R
a
jin
i
)
1165
T
h
e
f
ir
s
t
ap
p
r
o
ac
h
e
m
p
lo
y
ed
was
a
d
ec
is
io
n
tr
ee
c
lass
if
ier
.
Ho
wev
er
,
th
e
r
es
u
lts
wer
e
u
n
d
er
wh
elm
i
n
g
.
T
h
e
d
ec
is
io
n
tr
ee
m
o
d
el
id
en
tifie
d
o
n
l
y
a
f
e
w
h
ig
h
-
f
r
eq
u
en
cy
b
in
s
as
s
ig
n
if
ican
t,
r
esu
ltin
g
in
s
h
allo
w
d
ec
is
io
n
b
o
u
n
d
ar
ies
a
n
d
an
ac
c
u
r
ac
y
r
ate
o
f
a
p
p
r
o
x
im
ately
7
1
%.
T
h
is
p
er
f
o
r
m
a
n
ce
was
in
s
u
f
f
icien
t
f
o
r
r
eliab
le
f
au
lt
d
etec
tio
n
,
in
d
icatin
g
th
at
th
e
d
ec
is
io
n
tr
ee
s
tr
u
g
g
led
to
ca
p
tu
r
e
th
e
co
m
p
lex
r
elatio
n
s
h
ip
s
with
in
th
e
d
ata.
T
o
im
p
r
o
v
e
u
p
o
n
th
ese
r
esu
lts
,
a
r
an
d
o
m
f
o
r
est
class
if
ier
wa
s
im
p
lem
en
ted
.
T
h
is
en
s
em
b
le
lear
n
in
g
tech
n
iq
u
e
aim
ed
t
o
en
h
an
ce
r
o
b
u
s
tn
ess
an
d
f
lex
ib
ilit
y
b
y
c
o
m
b
in
in
g
m
u
ltip
le
d
ec
is
io
n
tr
ee
s
.
Desp
ite
th
ese
en
h
an
ce
m
e
n
ts
,
th
e
r
an
d
o
m
f
o
r
est
class
if
ier
d
id
n
o
t
ac
h
iev
e
s
ig
n
if
ican
t
im
p
r
o
v
em
en
t
s
,
m
ain
tain
in
g
an
ac
cu
r
ac
y
s
im
ilar
to
th
e
d
ec
is
io
n
tr
ee
,
ar
o
u
n
d
7
1
%.
W
h
ile
it
o
f
f
er
ed
b
etter
h
a
n
d
lin
g
o
f
d
ata
co
m
p
lex
ity
,
t
h
e
g
ain
s
wer
e
m
ar
g
in
al.
Seek
in
g
f
u
r
th
er
im
p
r
o
v
em
en
t
,
an
ANN
with
eig
h
t
la
y
er
s
was
in
tr
o
d
u
ce
d
.
T
h
e
ANN
in
co
r
p
o
r
ated
R
eL
U
ac
tiv
atio
n
f
u
n
ctio
n
s
an
d
d
r
o
p
o
u
t
lay
er
s
f
o
r
r
e
g
u
lar
iz
atio
n
,
wh
ich
r
esu
lted
in
a
n
o
ti
ce
ab
le
p
er
f
o
r
m
an
ce
b
o
o
s
t.
T
h
e
ANN
ac
h
iev
ed
an
ac
cu
r
ac
y
o
f
8
2
%,
o
u
tp
er
f
o
r
m
i
n
g
th
e
p
r
ev
io
u
s
m
o
d
els.
T
h
is
im
p
r
o
v
em
e
n
t
was
p
r
im
ar
ily
d
u
e
to
t
h
e
ANN’
s
ab
ilit
y
to
ca
p
tu
r
e
co
m
p
lex
n
o
n
lin
ea
r
r
elatio
n
s
h
ip
s
an
d
h
ig
h
e
r
-
o
r
d
e
r
in
ter
ac
tio
n
s
with
in
th
e
d
ata,
m
a
k
in
g
it m
o
r
e
s
u
itab
le
f
o
r
f
a
u
lt c
lass
if
icati
o
n
.
T
h
e
f
in
al
m
o
d
el
test
ed
was
a
1
D
C
NN
d
ev
elo
p
ed
u
s
in
g
Py
T
o
r
ch
.
T
h
e
C
NN
em
p
lo
y
ed
co
n
v
o
lu
tio
n
al
lay
er
s
with
GE
L
U
ac
tiv
atio
n
f
u
n
ctio
n
s
,
d
esig
n
ed
to
p
r
o
ce
s
s
s
p
ec
tr
o
g
r
am
in
p
u
ts
as
well
as
r
aw
s
ig
n
als.
T
h
e
C
NN
o
u
tp
er
f
o
r
m
ed
all
o
th
e
r
m
o
d
els,
ac
h
iev
in
g
a
n
ac
cu
r
ac
y
o
f
9
6
%
wh
en
u
s
in
g
s
p
ec
tr
o
g
r
am
in
p
u
ts
.
T
h
is
s
u
p
er
io
r
p
e
r
f
o
r
m
an
ce
ca
n
b
e
attr
ib
u
ted
to
t
h
e
C
NN’
s
ca
p
ab
ilit
y
to
s
im
u
ltan
eo
u
s
ly
ca
p
tu
r
e
tem
p
o
r
al
an
d
f
r
eq
u
e
n
cy
d
o
m
ain
f
ea
tu
r
es,
w
h
ich
ar
e
cr
u
cial
f
o
r
ac
cu
r
ate
class
if
icatio
n
.
T
h
e
an
aly
s
is
also
h
ig
h
lig
h
ted
th
e
d
is
tin
ctio
n
b
etwe
en
u
s
in
g
f
lu
x
an
d
cu
r
r
e
n
t
d
ata
f
o
r
f
au
lt
d
etec
tio
n
.
W
h
ile
f
lu
x
d
ata
g
en
er
ally
y
ield
ed
b
etter
class
if
icatio
n
r
esu
lt
s
,
th
e
p
r
ac
tical
co
n
s
id
er
atio
n
s
o
f
co
s
t
a
n
d
s
en
s
o
r
av
ailab
ilit
y
m
ak
e
cu
r
r
en
t
d
ata
a
m
o
r
e
ec
o
n
o
m
ical
ch
o
ice,
d
esp
ite
th
e
s
lig
h
t c
o
m
p
r
o
m
is
e
in
ac
cu
r
ac
y
.
5.
CO
NCLU
SI
O
N
T
h
is
p
ap
er
aim
e
d
to
ad
d
r
ess
th
e
co
m
p
lex
ities
o
f
f
a
u
lt
cl
ass
if
icatio
n
o
f
in
d
u
ctio
n
m
o
t
o
r
f
au
lts
.
A
r
ea
l tim
e
d
ata
ca
p
tu
r
in
g
is
d
o
n
e
with
th
e
h
elp
o
f
a
n
ex
p
er
i
m
en
tal
s
etu
p
f
o
r
in
ter
tu
r
n
f
au
l
ts
.
T
h
e
d
ata
in
clu
d
e
th
e
cu
r
r
e
n
t
an
d
th
e
f
lu
x
s
ig
n
als.
A
r
ea
s
o
n
ab
l
y
g
o
o
d
ac
c
u
r
ac
y
a
n
d
r
ec
all
was
p
r
o
v
id
ed
b
y
th
e
class
ical
m
ac
h
in
e
lear
n
in
g
m
o
d
els.
ANN
an
d
C
NN
p
r
o
v
id
ed
im
p
r
o
v
ed
ac
cu
r
ac
y
.
T
h
e
1
D
C
NN
o
u
tp
er
f
o
r
m
ed
ANN
with
9
6
% a
cc
u
r
ac
y
m
ak
in
g
it
h
ig
h
ly
ef
f
ec
tiv
e
f
o
r
f
au
lt c
lass
if
icatio
n
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
Au
th
o
r
s
s
tate
n
o
f
u
n
d
in
g
in
v
o
lv
ed
.
AUTHO
R
CO
NT
RI
B
UT
I
O
NS ST
A
T
E
M
E
N
T
T
h
is
jo
u
r
n
al
u
s
es
th
e
C
o
n
tr
i
b
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT
)
to
r
ec
o
g
n
ize
in
d
iv
id
u
a
l
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
th
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
llab
o
r
atio
n
.
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
V.
R
ajin
i
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
Kar
u
n
y
a
Har
ik
r
is
h
n
an
✓
✓
✓
✓
✓
✓
✓
Kr
is
m
ad
in
ata
✓
✓
✓
✓
✓
✓
✓
✓
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
s
i
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
T
h
e
au
t
h
o
r
s
d
ec
lar
e
th
at
th
e
y
h
av
e
n
o
k
n
o
wn
c
o
m
p
etin
g
f
in
an
cial
in
ter
ests
o
r
p
er
s
o
n
al
r
el
atio
n
s
h
ip
s
th
at
co
u
ld
h
av
e
ap
p
ea
r
ed
t
o
in
f
lu
en
ce
th
e
wo
r
k
r
e
p
o
r
te
d
in
t
h
is
p
ap
er
.
Au
th
o
r
s
s
tate
n
o
co
n
f
lict o
f
in
ter
est.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
7
9
2
I
n
t J Ap
p
l Po
wer
E
n
g
,
Vo
l.
1
5
,
No
.
3
,
Sep
tem
b
er
20
2
6
:
1
1
5
7
-
1
1
6
7
1166
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
d
ata
th
at
s
u
p
p
o
r
t
th
e
f
in
d
in
g
s
o
f
th
is
s
tu
d
y
ar
e
av
ailab
le
f
r
o
m
th
e
co
r
r
esp
o
n
d
in
g
au
t
h
o
r
,
[
VR
]
,
u
p
o
n
r
ea
s
o
n
ab
le
r
eq
u
est.
RE
F
E
R
E
NC
E
S
[
1
]
J.
X
i
e
,
Z.
L
i
,
Z
.
Z
h
o
u
,
a
n
d
S
.
L
i
u
,
“
A
n
o
v
e
l
b
e
a
r
i
n
g
f
a
u
l
t
c
l
a
ssi
f
i
c
a
t
i
o
n
met
h
o
d
b
a
s
e
d
o
n
X
G
B
o
o
s
t
:
t
h
e
f
u
si
o
n
o
f
d
e
e
p
l
e
a
r
n
i
n
g
-
b
a
s
e
d
f
e
a
t
u
r
e
s
a
n
d
e
m
p
i
r
i
c
a
l
f
e
a
t
u
r
e
s,”
I
EE
E
T
ra
n
s
a
c
t
i
o
n
s
o
n
I
n
st
r
u
m
e
n
t
a
t
i
o
n
a
n
d
M
e
a
s
u
r
e
m
e
n
t
,
v
o
l
.
7
0
,
p
p
.
1
–
9
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
TI
M
.
2
0
2
0
.
3
0
4
2
3
1
5
.
[
2
]
G
.
H
.
B
a
z
a
n
,
P
.
R
.
S
c
a
l
a
ss
a
r
a
,
W
.
E
n
d
o
,
a
n
d
A
.
G
o
e
d
t
e
l
,
“
I
n
f
o
r
ma
t
i
o
n
t
h
e
o
r
e
t
i
c
a
l
m
e
a
s
u
r
e
m
e
n
t
s
f
r
o
m
i
n
d
u
c
t
i
o
n
m
o
t
o
r
s
u
n
d
e
r
sev
e
r
a
l
l
o
a
d
a
n
d
v
o
l
t
a
g
e
c
o
n
d
i
t
i
o
n
s
f
o
r
b
e
a
r
i
n
g
f
a
u
l
t
s
c
l
a
ssi
f
i
c
a
t
i
o
n
,
”
I
EEE
T
ra
n
s
a
c
t
i
o
n
s
o
n
I
n
d
u
s
t
ri
a
l
I
n
f
o
rm
a
t
i
c
s
,
v
o
l
.
1
6
,
n
o
.
6
,
p
p
.
3
6
4
0
–
3
6
5
0
,
Ju
n
.
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
TI
I
.
2
0
1
9
.
2
9
3
9
6
7
8
.
[
3
]
D
.
T.
H
o
a
n
g
a
n
d
H
.
J.
K
a
n
g
,
“
A
m
o
t
o
r
c
u
r
r
e
n
t
s
i
g
n
a
l
-
b
a
se
d
b
e
a
r
i
n
g
f
a
u
l
t
d
i
a
g
n
o
si
s
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
a
n
d
i
n
f
o
r
mat
i
o
n
f
u
s
i
o
n
,
”
I
EEE
T
ra
n
s
a
c
t
i
o
n
s
o
n
I
n
s
t
ru
m
e
n
t
a
t
i
o
n
a
n
d
Me
a
su
r
e
m
e
n
t
,
v
o
l
.
6
9
,
n
o
.
6
,
p
p
.
3
3
2
5
–
3
3
3
3
,
Ju
n
.
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
TI
M
.
2
0
1
9
.
2
9
3
3
1
1
9
.
[
4
]
J.
F
e
n
g
,
Y
.
Y
a
o
,
S
.
L
u
,
a
n
d
Y
.
Li
u
,
“
D
o
m
a
i
n
k
n
o
w
l
e
d
g
e
-
b
a
se
d
d
e
e
p
-
b
r
o
a
d
l
e
a
r
n
i
n
g
f
r
a
mew
o
r
k
f
o
r
f
a
u
l
t
d
i
a
g
n
o
si
s,
”
I
E
E
E
T
ra
n
s
a
c
t
i
o
n
s
o
n
I
n
d
u
st
r
i
a
l
E
l
e
c
t
r
o
n
i
c
s
,
v
o
l
.
6
8
,
n
o
.
4
,
p
p
.
3
4
5
4
–
3
4
6
4
,
A
p
r
.
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
TI
E.
2
0
2
0
.
2
9
8
2
0
8
5
.
[
5
]
X
.
Z
h
a
o
,
M
.
Ji
a
,
P
.
D
i
n
g
,
C
.
Y
a
n
g
,
D
.
S
h
e
,
a
n
d
Z.
Li
u
,
“
I
n
t
e
l
l
i
g
e
n
t
f
a
u
l
t
d
i
a
g
n
o
si
s
o
f
mu
l
t
i
c
h
a
n
n
e
l
mo
t
o
r
-
r
o
t
o
r
sy
st
e
m
b
a
s
e
d
o
n
mu
l
t
i
m
a
n
i
f
o
l
d
d
e
e
p
e
x
t
r
e
m
e
l
e
a
r
n
i
n
g
mac
h
i
n
e
,
”
I
E
EE/
AS
ME
T
r
a
n
s
a
c
t
i
o
n
s
o
n
M
e
c
h
a
t
r
o
n
i
c
s
,
v
o
l
.
2
5
,
n
o
.
5
,
p
p
.
2
1
7
7
–
2
1
8
7
,
O
c
t
.
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
T
M
E
C
H
.
2
0
2
0
.
3
0
0
4
5
8
9
.
[
6
]
A
.
S
a
d
e
g
h
i
a
n
,
Z.
Y
e
,
a
n
d
B
.
W
u
,
“
O
n
l
i
n
e
d
e
t
e
c
t
i
o
n
o
f
b
r
o
k
e
n
r
o
t
o
r
b
a
r
s
i
n
i
n
d
u
c
t
i
o
n
mo
t
o
r
s b
y
w
a
v
e
l
e
t
p
a
c
k
e
t
d
e
c
o
m
p
o
si
t
i
o
n
a
n
d
a
r
t
i
f
i
c
i
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s,”
I
EE
E
T
r
a
n
sa
c
t
i
o
n
s
o
n
I
n
s
t
ru
m
e
n
t
a
t
i
o
n
a
n
d
Me
a
su
r
e
m
e
n
t
,
v
o
l
.
5
8
,
n
o
.
7
,
p
p
.
2
2
5
3
–
2
2
6
3
,
Ju
l
.
2
0
0
9
,
d
o
i
:
1
0
.
1
1
0
9
/
TI
M
.
2
0
0
9
.
2
0
1
3
7
4
3
.
[
7
]
S
.
W
.
F
.
d
e
R
e
z
e
n
d
e
,
P
.
E.
C
.
P
e
r
e
i
r
a
,
B
.
P
.
B
a
r
e
l
l
a
,
A
.
L
.
Li
ma,
R
.
M
.
F
.
N
e
t
o
,
a
n
d
J.
d
o
s
R
.
V
.
d
e
M
o
u
r
a
,
“
A
p
p
l
i
c
a
t
i
o
n
o
f
d
e
e
p
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s
i
n
t
h
e
d
e
v
e
l
o
p
m
e
n
t
o
f
p
r
e
d
i
c
t
i
v
e
mai
n
t
e
n
a
n
c
e
a
n
d
f
a
u
l
t
d
e
t
e
c
t
i
o
n
i
n
e
l
e
c
t
r
i
c
m
o
t
o
r
s,”
Ar
c
h
i
v
e
s
o
f
C
u
rre
n
t
Re
se
a
rc
h
I
n
t
e
rn
a
t
i
o
n
a
l
,
v
o
l
.
2
3
,
n
o
.
3
,
p
p
.
4
2
–
5
2
,
2
0
2
3
,
d
o
i
:
1
0
.
9
7
3
4
/
a
c
r
i
/
2
0
2
3
/
v
2
3
i
3
5
6
4
.
[
8
]
M
.
S
o
h
a
i
b
,
S
.
M
u
s
h
t
a
q
,
a
n
d
J
.
U
d
d
i
n
,
“
D
e
e
p
l
e
a
r
n
i
n
g
f
o
r
d
a
t
a
-
d
r
i
v
e
n
p
r
e
d
i
c
t
i
v
e
ma
i
n
t
e
n
a
n
c
e
,
”
i
n
I
n
t
e
l
l
i
g
e
n
t
S
y
st
e
m
s
Re
f
e
r
e
n
c
e
L
i
b
r
a
ry
,
v
o
l
.
2
0
7
,
i
n
I
n
t
e
l
l
i
g
e
n
t
S
y
s
t
e
ms
R
e
f
e
r
e
n
c
e
L
i
b
r
a
r
y
,
v
o
l
.
2
0
7
,
C
h
a
m:
S
p
r
i
n
g
e
r
,
2
0
2
1
,
p
p
.
7
1
–
95
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
030
-
7
5
4
9
0
-
7
_
3
.
[
9
]
A
.
T
h
e
i
ssl
e
r
,
J.
P
é
r
e
z
-
V
e
l
á
z
q
u
e
z
,
M
.
K
e
t
t
e
l
g
e
r
d
e
s,
a
n
d
G
.
El
g
e
r
,
“
P
r
e
d
i
c
t
i
v
e
m
a
i
n
t
e
n
a
n
c
e
e
n
a
b
l
e
d
b
y
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
:
u
se
c
a
ses
a
n
d
c
h
a
l
l
e
n
g
e
s
i
n
t
h
e
a
u
t
o
m
o
t
i
v
e
i
n
d
u
s
t
r
y
,
”
Re
l
i
a
b
i
l
i
t
y
En
g
i
n
e
e
ri
n
g
a
n
d
S
y
s
t
e
m
S
a
f
e
t
y
,
v
o
l
.
2
1
5
,
p
.
1
0
7
8
6
4
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
r
e
ss.
2
0
2
1
.
1
0
7
8
6
4
.
[
1
0
]
M
.
D
r
a
k
a
k
i
,
Y
.
L.
K
a
r
n
a
v
a
s,
I
.
A
.
T
z
i
a
f
e
t
t
a
s,
V
.
L
i
n
a
r
d
o
s,
a
n
d
P
.
T
z
i
o
n
a
s,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
n
d
d
e
e
p
l
e
a
r
n
i
n
g
b
a
s
e
d
m
e
t
h
o
d
s
t
o
w
a
r
d
i
n
d
u
s
t
r
y
4
.
0
p
r
e
d
i
c
t
i
v
e
mai
n
t
e
n
a
n
c
e
i
n
i
n
d
u
c
t
i
o
n
m
o
t
o
r
s:
α
s
t
a
t
e
o
f
t
h
e
a
r
t
su
r
v
e
y
,
”
J
o
u
r
n
a
l
o
f
I
n
d
u
st
r
i
a
l
En
g
i
n
e
e
ri
n
g
a
n
d
Ma
n
a
g
e
m
e
n
t
,
v
o
l
.
1
5
,
n
o
.
1
,
p
p
.
3
1
–
5
7
,
2
0
2
2
,
d
o
i
:
1
0
.
3
9
2
6
/
j
i
e
m.
3
5
9
7
.
[1
1
]
O
.
S
e
r
r
a
d
i
l
l
a
,
E
.
Z
u
g
a
st
i
,
J.
R
o
d
r
i
g
u
e
z
,
a
n
d
U
.
Z
u
r
u
t
u
z
a
,
“
D
e
e
p
l
e
a
r
n
i
n
g
mo
d
e
l
s
f
o
r
p
r
e
d
i
c
t
i
v
e
m
a
i
n
t
e
n
a
n
c
e
:
a
su
r
v
e
y
,
c
o
m
p
a
r
i
so
n
,
c
h
a
l
l
e
n
g
e
s
a
n
d
p
r
o
s
p
e
c
t
s
,
”
Ap
p
l
i
e
d
I
n
t
e
l
l
i
g
e
n
c
e
,
v
o
l
.
5
2
,
n
o
.
1
0
,
p
p
.
1
0
9
3
4
–
1
0
9
6
4
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
0
4
8
9
-
021
-
0
3
0
0
4
-
y.
[1
2
]
E.
I
r
g
a
t
,
E
.
Ç
i
n
a
r
,
A
.
Ü
n
s
a
l
,
a
n
d
A
.
Y
a
z
ı
c
ı
,
“
A
n
I
o
T
-
b
a
se
d
m
o
n
i
t
o
r
i
n
g
s
y
st
e
m
f
o
r
i
n
d
u
c
t
i
o
n
m
o
t
o
r
f
a
u
l
t
s
u
t
i
l
i
z
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
mo
d
e
l
s
,
”
J
o
u
r
n
a
l
o
f
V
i
b
r
a
t
i
o
n
E
n
g
i
n
e
e
ri
n
g
a
n
d
T
e
c
h
n
o
l
o
g
i
e
s
,
v
o
l
.
1
1
,
n
o
.
7
,
p
p
.
3
5
7
9
–
3
5
8
9
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s4
2
4
1
7
-
022
-
0
0
7
6
9
-
5.
[1
3
]
R
.
M
.
S
o
u
z
a
,
E.
G
.
S
.
N
a
sci
me
n
t
o
,
U
.
A
.
M
i
r
a
n
d
a
,
W
.
J
.
D
.
S
i
l
v
a
,
a
n
d
H
.
A
.
L
e
p
i
k
so
n
,
“
D
e
e
p
l
e
a
r
n
i
n
g
f
o
r
d
i
a
g
n
o
s
i
s
a
n
d
c
l
a
ss
i
f
i
c
a
t
i
o
n
o
f
f
a
u
l
t
s
i
n
i
n
d
u
st
r
i
a
l
r
o
t
a
t
i
n
g
m
a
c
h
i
n
e
r
y
,
”
C
o
m
p
u
t
e
rs
&
I
n
d
u
s
t
ri
a
l
E
n
g
i
n
e
e
r
i
n
g
,
v
o
l
.
1
5
3
,
p
.
1
0
7
0
6
0
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
i
e
.
2
0
2
0
.
1
0
7
0
6
0
.
[1
4
]
G
.
F
a
l
e
k
a
s
a
n
d
A
.
K
a
r
l
i
s,
“
D
i
g
i
t
a
l
t
w
i
n
i
n
e
l
e
c
t
r
i
c
a
l
m
a
c
h
i
n
e
c
o
n
t
r
o
l
a
n
d
p
r
e
d
i
c
t
i
v
e
mai
n
t
e
n
a
n
c
e
:
s
t
a
t
e
-
of
-
t
h
e
-
a
r
t
a
n
d
f
u
t
u
r
e
p
r
o
s
p
e
c
t
s
,
”
En
e
rg
i
e
s
,
v
o
l
.
1
4
,
n
o
.
1
8
,
p
.
5
9
3
3
,
2
0
2
1
,
d
o
i
:
1
0
.
3
3
9
0
/
e
n
1
4
1
8
5
9
3
3
.
[1
5
]
S
.
Z
h
a
o
,
F
.
B
l
a
a
b
j
e
r
g
,
a
n
d
H
.
W
a
n
g
,
“
A
n
o
v
e
r
v
i
e
w
o
f
a
r
t
i
f
i
c
i
a
l
i
n
t
e
l
l
i
g
e
n
c
e
a
p
p
l
i
c
a
t
i
o
n
s
f
o
r
p
o
w
e
r
e
l
e
c
t
r
o
n
i
c
s,
”
I
E
E
E
T
ra
n
s
a
c
t
i
o
n
s
o
n
P
o
w
e
r E
l
e
c
t
r
o
n
i
c
s
,
v
o
l
.
3
6
,
n
o
.
4
,
p
p
.
4
6
3
3
–
4
6
5
8
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
TP
EL.
2
0
2
0
.
3
0
2
4
9
1
4
.
[
1
6
]
V
.
R
a
j
i
n
i
,
K
.
B
.
S
u
n
d
h
a
r
a
k
u
mar,
V
.
S
.
N
a
g
a
r
a
j
a
n
,
H
.
K
a
r
u
n
y
a
,
H
.
B
.
M
a
n
o
g
a
r
a
n
,
a
n
d
W
.
A
.
M
e
ma
l
a
,
“
A
c
l
a
ss
i
f
i
c
a
t
i
o
n
a
p
p
r
o
a
c
h
f
o
r
i
n
d
u
c
t
i
o
n
m
o
t
o
r
f
a
u
l
t
s
b
a
s
e
d
o
n
e
mp
i
r
i
c
a
l
m
o
d
e
d
e
c
o
m
p
o
si
t
i
o
n
a
n
d
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
ms
,
”
El
e
c
t
ri
c
a
,
v
o
l
.
2
4
,
n
o
.
2
,
p
p
.
5
1
5
–
5
2
4
,
2
0
2
4
,
d
o
i
:
1
0
.
5
1
5
2
/
e
l
e
c
t
r
i
c
a
.
2
0
2
4
.
2
3
0
3
8
.
[
1
7
]
L.
A
.
S
i
k
i
n
y
i
,
C
.
M
.
M
a
i
n
a
,
a
n
d
L
.
N
g
o
o
,
“
C
o
n
d
i
t
i
o
n
m
o
n
i
t
o
r
i
n
g
a
n
d
f
a
u
l
t
d
i
a
g
n
o
si
s
o
f
i
n
d
u
c
t
i
o
n
m
o
t
o
r
f
a
u
l
t
s
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
:
a
r
e
v
i
e
w
,
”
i
n
I
E
EE
AF
RI
C
O
N
C
o
n
f
e
re
n
c
e
,
K
o
l
k
a
t
a
,
I
n
d
i
a
,
2
0
2
3
,
p
p
.
4
9
3
–
498
,
d
o
i
:
1
0
.
1
1
0
9
/
A
F
R
I
C
O
N
5
5
9
1
0
.
2
0
2
3
.
1
0
2
9
3
5
7
8
.
[
1
8
]
N
.
R
a
j
a
p
a
k
s
h
a
,
S
.
J
a
y
a
si
n
g
h
e
,
H
.
En
sh
a
e
i
,
a
n
d
N
.
Ja
y
a
r
a
t
h
n
e
,
“
S
u
p
e
r
v
i
se
d
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
m
sel
e
c
t
i
o
n
f
o
r
c
o
n
d
i
t
i
o
n
mo
n
i
t
o
r
i
n
g
o
f
i
n
d
u
c
t
i
o
n
m
o
t
o
r
s,
”
i
n
2
0
2
1
I
E
EE
S
o
u
t
h
e
rn
P
o
w
e
r
E
l
e
c
t
r
o
n
i
c
s
C
o
n
f
e
re
n
c
e
,
S
P
EC
2
0
2
1
,
K
i
g
a
l
i
,
R
w
a
n
d
a
,
2
0
2
1
,
p
p
.
3
5
1
–
356
,
d
o
i
:
1
0
.
1
1
0
9
/
S
P
EC
5
2
8
2
7
.
2
0
2
1
.
9
7
0
9
4
3
6
.
[
1
9
]
A
.
F
.
K
h
a
l
i
l
a
n
d
S
.
R
o
st
a
m,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
-
b
a
se
d
p
r
e
d
i
c
t
i
v
e
ma
i
n
t
e
n
a
n
c
e
f
o
r
f
a
u
l
t
d
e
t
e
c
t
i
o
n
i
n
r
o
t
a
t
i
n
g
mac
h
i
n
e
r
y
:
a
c
a
s
e
st
u
d
y
,
”
E
n
g
i
n
e
e
r
i
n
g
,
T
e
c
h
n
o
l
o
g
y
a
n
d
Ap
p
l
i
e
d
S
c
i
e
n
c
e
Re
s
e
a
rc
h
,
v
o
l
.
1
4
,
n
o
.
2
,
p
p
.
1
3
1
8
1
–
1
3
1
8
9
,
2
0
2
4
,
d
o
i
:
1
0
.
4
8
0
8
4
/
e
t
a
sr
.
6
8
1
3
.
[
2
0
]
B
.
G
a
n
g
u
l
y
,
R
.
K
.
R
a
y
,
A
.
C
h
a
t
t
e
r
j
e
e
,
a
n
d
S
.
P
a
u
l
,
“
A
d
e
e
p
l
e
a
r
n
i
n
g
a
i
d
e
d
i
n
t
e
l
l
i
g
e
n
t
f
r
a
mew
o
r
k
f
o
r
c
o
n
d
i
t
i
o
n
mo
n
i
t
o
r
i
n
g
o
f
e
l
e
c
t
r
i
c
a
l
ma
c
h
i
n
e
r
y
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
5
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
2
0
2
3
D
e
v
i
c
e
s
f
o
r
I
n
t
e
g
ra
t
e
d
C
i
rc
u
i
t
,
D
e
v
I
C
2
0
2
3
,
K
a
l
y
a
n
i
,
I
n
d
i
a
,
2
0
2
3
,
p
p
.
8
2
–
86
,
d
o
i
:
1
0
.
1
1
0
9
/
D
e
v
I
C
5
7
7
5
8
.
2
0
2
3
.
1
0
1
3
4
8
3
2
.
[
2
1
]
G
.
S
u
n
,
X
.
L
i
,
a
n
d
F
.
H
u
a
n
g
,
“
H
y
b
r
i
d
C
N
N
-
LST
M
f
a
u
l
t
d
i
a
g
n
o
si
s
m
o
d
e
l
f
o
r
r
o
l
l
i
n
g
b
e
a
r
i
n
g
b
a
s
e
d
o
n
m
u
l
t
i
-
d
o
ma
i
n
f
e
a
t
u
r
e
f
u
si
o
n
,
”
I
E
EE
Ac
c
e
ss
,
v
o
l
.
1
1
,
p
p
.
1
1
7
7
8
9
–
1
1
7
8
0
0
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
3
.
3
3
2
7
6
6
6
.
[
2
2
]
J.
F
.
D
.
S
a
n
t
o
s
e
t
a
l
.
,
“
D
i
g
i
t
a
l
t
w
i
n
-
b
a
se
d
m
o
n
i
t
o
r
i
n
g
sy
s
t
e
m
o
f
i
n
d
u
c
t
i
o
n
mo
t
o
r
s
u
s
i
n
g
I
o
T
s
e
n
s
o
r
s
a
n
d
t
h
e
r
mo
-
mag
n
e
t
i
c
f
i
n
i
t
e
e
l
e
m
e
n
t
a
n
a
l
y
si
s,”
I
EEE
Ac
c
e
ss
,
v
o
l
.
1
1
,
n
o
.
3
,
p
p
.
1
6
8
2
–
1
6
9
3
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
2
.
3
2
3
2
0
6
3
.
[
2
3
]
F
.
S
c
h
mi
d
,
S
.
M
a
s
o
u
d
i
a
n
,
E.
M
a
r
t
h
,
P
.
Z
o
r
n
,
K
.
K
o
u
t
i
n
i
,
a
n
d
G
.
W
i
d
m
e
r
,
“
S
i
mu
l
a
t
i
o
n
-
a
ssi
s
t
e
d
t
r
a
i
n
i
n
g
o
f
n
e
u
r
a
l
n
e
t
w
o
r
k
s
f
o
r
c
o
n
d
i
t
i
o
n
mo
n
i
t
o
r
i
n
g
o
f
e
l
e
c
t
r
i
c
a
l
d
r
i
v
e
s:
e
n
h
a
n
c
e
d
d
o
ma
i
n
a
d
a
p
t
a
t
i
o
n
m
e
t
h
o
d
s,
”
i
n
I
K
MT
2
0
2
2
;
1
3
.
G
MM
/
ETG
-
S
y
m
p
o
s
i
u
m
,
Li
n
z
,
A
u
s
t
r
i
a
,
2
0
2
2
,
p
p
.
1
–
7.
[
2
4
]
R
.
D
i
v
e
r
s
i
,
A
.
L
e
n
z
i
,
N
.
S
p
e
c
i
a
l
e
,
a
n
d
M
.
B
a
r
b
i
e
r
i
,
“
A
n
a
u
t
o
r
e
g
r
e
ssi
v
e
-
b
a
sed
m
o
t
o
r
c
u
r
r
e
n
t
si
g
n
a
t
u
r
e
a
n
a
l
y
s
i
s
a
p
p
r
o
a
c
h
f
o
r
f
a
u
l
t
d
i
a
g
n
o
si
s
o
f
e
l
e
c
t
r
i
c
mo
t
o
r
-
d
r
i
v
e
n
m
e
c
h
a
n
i
sms
,
”
S
e
n
so
r
s
,
v
o
l
.
2
5
,
n
o
.
4
,
p
.
1
1
3
0
,
F
e
b
.
2
0
2
5
,
d
o
i
:
1
0
.
3
3
9
0
/
s2
5
0
4
1
1
3
0
.
[
2
5
]
T.
L
i
n
,
Z.
R
e
n
,
K
.
H
u
a
n
g
,
Y
.
Z
h
u
,
a
n
d
H
.
R
.
K
a
r
i
mi
,
“
A
n
o
v
e
l
mu
l
t
i
-
se
n
so
r
i
n
f
o
r
ma
t
i
o
n
f
u
s
i
o
n
me
t
h
o
d
f
o
r
f
a
u
l
t
d
i
a
g
n
o
si
s
o
f
r
o
t
a
t
i
n
g
m
a
c
h
i
n
e
r
y
w
i
t
h
mi
ss
i
n
g
si
g
n
a
l
s,”
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
ri
n
g
I
n
f
o
rm
a
t
i
c
s
,
v
o
l
.
6
8
,
2
0
2
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
a
e
i
.
2
0
2
5
.
1
0
3
5
9
5
.
Evaluation Warning : The document was created with Spire.PDF for Python.