I
nte
rna
t
io
na
l J
o
urna
l o
f
P
o
w
er
E
lect
ro
nics
a
nd
Driv
e
Sy
s
t
e
m
s
(
I
J
P
E
DS
)
Vo
l.
1
2
,
No
.
3
,
Sep
tem
b
er
202
1
,
p
p
.
1
9
0
0
~
1
9
1
1
I
SS
N:
2088
-
8694
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
p
ed
s
.
v
1
2
.i
3
.
pp
1
9
0
0
-
1
9
1
1
1900
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
p
e
d
s
.
ia
esco
r
e.
co
m
Ro
tating
blade
faults cla
ss
ificatio
n
o
f
a
roto
r
-
dis
k
-
bl
a
de sy
ste
m
using
artificial n
e
ura
l net
w
o
rk
Abdu
l
la
hi Abu
ba
k
a
r
M
a
s
’
u
d
1
,
Ah
m
a
d J
a
m
a
l
2
,
Su
ra
j
ud
deen
Adew
us
i
3
,
Aru
na
cha
la
m
Su
n
da
ra
m
4
1
,4
De
p
a
rtm
e
n
t
o
f
El
e
c
tri
c
a
l
a
n
d
El
e
c
tro
n
ic E
n
g
i
n
e
e
rin
g
,
Ju
b
a
il
In
d
u
strial
Co
ll
e
g
e
,
Kin
g
d
o
m
o
f
S
a
u
d
i
A
ra
b
ia
1,
2
P
rin
c
e
S
a
u
d
b
i
n
T
h
u
n
a
y
a
n
Re
se
a
rc
h
Ce
n
tre,
Ro
y
a
l
Co
m
m
is
sio
n
f
o
r
Ju
b
a
il
,
Ki
n
g
d
o
m
o
f
S
a
u
d
i
A
ra
b
ia
2,
3
De
p
a
rtm
e
n
t
o
f
M
e
c
h
a
n
ica
l
En
g
in
e
e
rin
g
,
Ju
b
a
il
U
n
iv
e
rsity
Co
ll
e
g
e
,
Kin
g
d
o
m
o
f
S
a
u
d
i
A
ra
b
ia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Ma
y
1
2
,
2
0
21
R
ev
i
s
ed
Ju
l
9
,
2
0
21
A
cc
ep
ted
J
u
l
17
,
2
0
21
In
th
is
p
a
p
e
r,
th
e
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
(
A
N
N)
h
a
s
b
e
e
n
u
ti
li
z
e
d
f
o
r
ro
tatin
g
m
a
c
h
in
e
r
y
f
a
u
lt
s
d
e
tec
t
io
n
a
n
d
c
las
sif
ica
ti
o
n
.
F
irst,
e
x
p
e
rim
e
n
ts
w
e
r
e
p
e
rf
o
r
m
e
d
to
m
e
a
su
re
th
e
late
ra
l
v
ib
ra
ti
o
n
sig
n
a
ls
o
f
lab
o
ra
to
ry
t
e
st
rig
s
f
o
r
ro
to
r
-
d
isk
-
b
lad
e
w
h
e
n
t
h
e
b
la
d
e
s
a
re
d
e
f
e
c
ti
v
e
.
A
ro
to
r
-
d
isk
-
b
l
a
d
e
sy
st
e
m
w
it
h
6
re
g
u
lar
b
lad
e
s
a
n
d
5
b
la
d
e
s
w
it
h
v
a
rio
u
s
d
e
f
e
c
ts
w
a
s
c
o
n
stru
c
ted
.
S
e
c
o
n
d
,
t
h
e
A
NN
wa
s
a
p
p
li
e
d
t
o
c
las
sify
th
e
d
if
f
e
r
e
n
t
x
-
a
n
d
y
-
a
x
is
late
ra
l
v
ib
ra
ti
o
n
s
d
u
e
to
d
if
f
e
r
e
n
t
b
lad
e
f
a
u
lt
s.
T
h
e
re
su
lt
s
b
a
s
e
d
o
n
train
in
g
a
n
d
tes
ti
n
g
w
it
h
d
if
f
e
r
e
n
t
d
a
ta
sa
m
p
le
s
o
f
th
e
f
a
u
lt
ty
p
e
s
in
d
ica
te
th
a
t
t
h
e
A
NN
is
ro
b
u
st
a
n
d
c
a
n
e
ff
e
c
ti
v
e
l
y
id
e
n
ti
fy
a
n
d
d
isti
n
g
u
is
h
d
if
f
e
re
n
t
b
lad
e
fa
u
lt
s
c
a
u
se
d
b
y
late
ra
l
v
ib
ra
ti
o
n
s
in
a
ro
to
r.
A
s
c
o
m
p
a
re
d
to
th
e
li
tera
tu
re
,
th
e
p
re
se
n
t
p
a
p
e
r
p
re
se
n
ts
a
n
o
v
e
l
w
o
rk
o
f
id
e
n
ti
fy
in
g
a
n
d
c
las
si
fy
in
g
v
a
rio
u
s
ro
tatin
g
b
lad
e
f
a
u
lt
s
c
o
m
m
o
n
ly
e
n
c
o
u
n
tere
d
i
n
ro
tatin
g
m
a
c
h
i
n
e
s
u
sin
g
A
N
N.
Ex
p
e
ri
m
e
n
tal
d
a
ta
o
f
late
r
a
l
v
ib
ra
ti
o
n
s
o
f
th
e
ro
to
r
-
d
isk
-
b
la
d
e
s
y
ste
m
in
b
o
t
h
x
-
a
n
d
y
-
d
irec
ti
o
n
s
a
re
u
se
d
f
o
r
t
h
e
train
i
n
g
a
n
d
tes
ti
n
g
o
f
th
e
n
e
tw
o
rk
.
K
ey
w
o
r
d
s
:
A
r
ti
f
icial
n
e
u
r
al
n
et
w
o
r
k
s
B
lad
e
f
au
lts
B
lad
e
s
y
s
te
m
P
y
t
h
o
n
R
o
tatin
g
m
ac
h
in
er
y
T
h
is
is
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
:
A
b
d
u
l
lah
i
A
b
u
b
ak
ar
Ma
s
‟
u
d
Dep
ar
t
m
en
t o
f
E
lectr
ical
an
d
E
lectr
o
n
ic
E
n
g
in
ee
r
i
n
g
J
u
b
ail
I
n
d
u
s
tr
ial
C
o
lleg
e
a
n
d
P
r
in
ce
Sau
d
b
in
T
h
u
n
a
y
a
n
R
e
s
ea
r
ch
C
e
n
tr
e,
J
u
b
ail
P
.
O.
B
o
x
,
1
0
0
9
9
,
Kin
g
d
o
m
o
f
Sau
d
i
A
r
ab
ia
E
m
ail:
m
as
u
d
_
a@
j
ic.
ed
u
.
s
a
1.
I
NT
RO
D
UCT
I
O
N
Fo
r
co
m
p
etiti
v
en
e
s
s
an
d
h
i
g
h
q
u
alit
y
o
p
er
ati
n
g
r
eq
u
ir
e
m
e
n
ts
i
n
i
n
d
u
s
tr
ies,
th
e
n
ee
d
f
o
r
ef
f
ec
ti
v
e
r
o
tatin
g
m
ac
h
i
n
es
h
as
b
ee
n
s
ig
n
i
f
ica
n
t.
T
h
is
ca
n
b
e
ac
h
i
ev
ed
in
p
ar
t
th
r
o
u
g
h
co
n
t
in
u
o
u
s
m
o
n
ito
r
i
n
g
o
f
o
p
er
atin
g
co
n
d
it
io
n
s
to
d
etec
t
an
y
f
a
u
lt
b
ef
o
r
e
t
h
e
r
o
tati
n
g
m
ac
h
in
e
s
b
r
ea
k
d
o
w
n
.
Fa
u
lt
s
i
n
r
o
tatin
g
b
lad
es a
r
e
v
er
y
d
i
f
f
ic
u
lt
to
d
etec
t
b
y
d
ir
ec
t
b
lad
e
v
ib
r
atio
n
m
ea
s
u
r
e
m
en
ts
an
d
e
v
al
u
atio
n
.
T
h
is
i
s
b
ec
au
s
e
b
lad
es
o
f
te
n
o
p
er
ate
in
v
er
y
h
ar
s
h
en
v
ir
o
n
m
en
ts
.
I
n
ad
d
itio
n
,
n
o
s
i
g
n
if
i
ca
n
t
r
esu
lts
h
av
e
b
ee
n
o
b
tain
e
d
in
th
e
q
u
est
f
o
r
a
m
o
r
e
e
f
f
icie
n
t a
r
tif
ic
ial
in
telli
g
en
ce
s
y
s
te
m
f
o
r
r
ea
l ti
m
e
m
o
n
ito
r
in
g
o
f
t
h
ese
f
au
lts
y
et.
T
h
e
p
er
f
o
r
m
a
n
ce
an
d
r
eliab
ili
t
y
o
f
s
p
in
n
i
n
g
m
ac
h
i
n
es
ar
e
m
ai
n
tai
n
ed
b
y
co
n
s
tan
t
l
y
tr
ac
k
in
g
t
h
eir
w
o
r
k
i
n
g
co
n
d
itio
n
s
in
o
r
d
er
to
d
etec
t
an
y
d
ef
ec
t
s
b
ef
o
r
e
ca
u
s
i
n
g
a
s
ig
n
if
ican
t
p
r
o
b
l
em
o
r
f
ail
u
r
e
o
f
th
e
s
y
s
te
m
t
h
at
m
a
y
i
m
p
ac
t
o
u
tp
u
t
tar
g
et
s
.
T
h
er
e
ar
e
r
esear
ch
ar
ticles
o
n
t
h
e
r
o
to
r
v
ib
r
atio
n
r
esp
o
n
s
e
[
1
]
-
[
3
]
,
th
a
t
ar
e
u
s
u
all
y
m
o
n
ito
r
ed
in
t
h
e
i
n
d
u
s
tr
y
a
n
d
a
n
al
y
s
ed
f
o
r
id
en
tif
icat
io
n
o
f
f
au
lts
.
T
h
ese
ef
f
o
r
ts
h
a
v
e
r
e
s
u
l
ted
in
th
e
d
ev
elo
p
m
e
n
t
o
f
k
n
o
w
led
g
e
b
ase
an
d
s
tan
d
ar
d
p
r
o
ce
d
u
r
es
to
id
en
ti
f
y
m
ac
h
i
n
e
f
au
lts
s
u
ch
a
s
s
h
a
f
t
m
is
a
lig
n
m
e
n
t
a
n
d
u
n
b
ala
n
ce
,
r
o
to
r
ec
ce
n
tr
icit
y
,
b
lad
e
cr
ac
k
,
t
w
is
t
in
g
,
a
n
d
r
u
b
b
i
n
g
.
Ma
n
y
in
d
u
s
tr
ial
r
o
tatin
g
m
ac
h
in
e
s
h
av
e
r
o
to
r
s
,
d
is
k
s
,
b
lad
es,
an
d
v
an
e
s
.
He
n
ce
,
it
i
s
i
m
p
o
r
ta
n
t
to
m
o
n
i
to
r
th
e
d
y
n
a
m
ic
r
esp
o
n
s
e
o
f
th
e
m
ac
h
in
e
p
ar
ts
s
u
c
h
as
r
o
to
r
-
d
is
k
-
b
lad
e
s
y
s
te
m
t
h
at
p
o
ten
t
iall
y
ca
n
d
ev
elo
p
m
ac
h
in
e
f
a
u
lts
co
n
tr
ib
u
ted
b
y
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
I
SS
N:
2088
-
8
694
R
o
ta
tin
g
b
l
a
d
e
fa
u
lts
cla
s
s
ific
a
tio
n
o
f
a
r
o
to
r
-
d
is
k
-
b
la
d
e
s
ystem
u
s
in
g
… (
A
b
d
u
lla
h
i A
b
u
b
a
ka
r
Ma
s
’
u
d
)
1901
b
lad
e
d
ef
ec
ts
.
Desp
ite
th
e
f
a
ct
th
at
v
ib
r
atio
n
r
esp
o
n
s
e
o
f
r
o
to
r
ca
n
b
e
ea
s
ily
m
ea
s
u
r
e
d
w
it
h
th
e
h
e
lp
o
f
p
r
o
x
i
m
it
y
p
r
o
b
e
an
d
ac
ce
ler
o
m
e
ter
,
d
ir
ec
t
m
ea
s
u
r
e
m
e
n
t
o
f
v
ib
r
atio
n
r
esp
o
n
s
e
o
f
th
e
b
lad
es
s
o
m
eti
m
e
s
b
ec
o
m
e
s
v
er
y
d
i
f
f
icu
lt
an
d
ex
p
en
s
iv
e
s
in
ce
b
lad
es
s
o
m
eti
m
e
s
o
p
er
ate
at
v
er
y
e
x
tr
e
m
e
co
n
d
itio
n
s
o
f
te
m
p
er
atu
r
e
a
n
d
p
r
ess
u
r
e
en
c
o
u
n
ter
ed
i
n
g
as
a
n
d
s
tea
m
tu
r
b
in
es.
A
l
s
o
,
s
tr
ai
n
g
au
g
es
m
o
u
n
ted
o
n
t
h
e
b
lad
e
s
to
m
ea
s
u
r
e
b
lad
e
d
ef
lectio
n
h
av
e
s
h
o
r
t
o
p
er
atin
g
li
f
e
d
u
e
t
o
th
ese
e
x
tr
e
m
e
co
n
d
itio
n
s
,
b
lad
e
ae
r
o
d
y
n
a
m
ic
s
in
t
er
f
er
e
n
ce
.
As
a
r
esu
lt,
n
o
n
-
co
n
tactin
g
an
d
n
o
n
-
i
n
tr
u
s
i
v
e
t
ec
h
n
iq
u
es
to
m
ea
s
u
r
e
th
e
b
lad
e
v
ib
r
atio
n
s
d
u
e
to
b
lad
e
d
ef
ec
ts
h
av
e
b
ee
n
d
e
v
el
o
p
ed
.
On
e
s
u
c
h
tec
h
n
iq
u
e
is
u
s
ed
b
y
L
a
w
s
o
n
an
d
I
v
e
y
[
4
]
,
in
w
h
ic
h
b
lad
e
tip
ti
m
i
n
g
is
m
ea
s
u
r
ed
u
s
i
n
g
tip
clea
r
an
ce
ca
p
ac
itan
ce
p
r
o
b
es.
Ma
n
y
r
esear
ch
er
s
h
av
e
e
x
tr
ac
ted
th
e
b
lad
e
d
y
n
a
m
ic
r
esp
o
n
s
e
f
r
o
m
co
u
p
led
r
esp
o
n
s
e
s
o
f
r
o
to
r
-
d
is
k
-
b
lad
e
s
y
s
te
m
s
[
5
]
,
[
6
]
.
Fo
r
th
is
,
th
e
y
estab
li
s
h
ed
a
r
elatio
n
s
h
ip
b
et
w
ee
n
r
o
to
r
to
r
s
io
n
al
v
ib
r
atio
n
an
d
b
lad
e
d
y
n
a
m
ic
r
esp
o
n
s
e.
Ho
w
ev
er
,
o
n
e
s
t
u
d
y
s
h
o
w
s
b
lad
e
f
au
l
t
d
etec
tio
n
f
r
o
m
r
o
to
r
to
r
s
io
n
al
v
ib
r
atio
n
s
.
T
h
er
e
ar
e
f
e
w
s
tu
d
ie
s
t
h
at
e
x
h
ib
it
r
elati
o
n
s
h
ip
b
et
w
ee
n
th
e
r
o
to
r
-
b
lad
e
s
y
s
te
m
an
d
b
lad
e
v
ib
r
atio
n
[
7
]
.
Fe
w
s
tu
d
ie
s
ex
h
ib
it
r
elatio
n
s
h
ip
b
et
w
ee
n
th
e
r
o
to
r
-
b
lad
e
s
y
s
te
m
a
n
d
b
lad
e
v
ib
r
atio
n
[
6
]
,
[
8
]
.
On
e
o
f
t
h
ese
s
t
u
d
ie
s
m
ea
s
u
r
ed
[
8
]
,
an
al
y
ticall
y
a
n
d
ex
p
er
i
m
en
tall
y
,
t
h
e
b
lad
e
v
ib
r
atio
n
s
t
h
r
o
u
g
h
t
h
e
later
al
v
ib
r
atio
n
o
f
th
e
s
u
p
p
o
r
t
o
f
a
r
o
to
r
-
b
lad
e
s
y
s
te
m
.
E
v
en
th
o
u
g
h
s
p
ec
i
f
ic
b
lad
e
f
a
u
lt
s
w
er
e
n
o
t
i
n
v
esti
g
ated
in
th
is
s
t
u
d
y
,
t
h
e
y
p
r
o
v
id
ed
th
e
b
asis
f
o
r
b
lad
e
f
au
lt d
etec
tio
n
f
r
o
m
r
o
to
r
later
al
v
ib
r
atio
n
.
A
S
u
b
s
ta
n
tial
a
m
o
u
n
t
o
f
w
o
r
k
o
n
th
e
d
etec
tio
n
o
f
v
ar
io
u
s
r
o
tatin
g
m
ac
h
i
n
e
f
a
u
lt
s
s
u
ch
as
b
lad
e
d
is
to
r
tio
n
,
s
h
a
f
t
m
is
ali
g
n
m
en
t,
r
o
to
r
u
n
b
alan
ce
,
r
o
t
o
r
cr
ac
k
.
u
s
i
n
g
m
ac
h
i
n
e
lear
n
i
n
g
alg
o
r
ith
m
s
h
as
b
ee
n
u
n
d
er
ta
k
en
.
A
n
u
m
b
er
o
f
r
e
v
ie
w
ar
ticles
ar
e
f
o
u
n
d
o
n
f
au
lt
d
iag
n
o
s
is
o
f
r
o
tatin
g
m
ac
h
in
er
y
u
s
in
g
t
h
e
ar
tif
icial
i
n
tel
lig
e
n
ce
(
A
I
)
.
L
i
u
et
a
l.
[
9
]
p
r
esen
ted
a
co
m
p
r
eh
en
s
i
v
e
r
ev
ie
w
o
f
p
r
ev
io
u
s
r
esear
ch
o
n
s
u
ch
f
au
lts
u
s
i
n
g
n
aï
v
e
B
a
y
e
s
,
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e,
d
ee
p
lear
n
in
g
al
g
o
r
ith
m
s
a
n
d
th
e
k
-
n
ea
r
est
n
ei
g
h
b
o
r
(
KNN)
.
T
h
e
b
en
ef
its
,
co
n
s
tr
ain
ts
,
an
d
p
r
ac
tical
i
m
p
licati
o
n
s
o
f
s
u
c
h
A
I
alg
o
r
it
h
m
s
w
er
e
a
ls
o
d
is
c
u
s
s
ed
.
An
o
th
er
r
ev
ie
w
p
ap
er
b
y
L
ei
et
a
l.
[
1
0
]
p
r
esen
ted
th
e
m
ac
h
in
e
lear
n
in
g
al
g
o
r
ith
m
s
u
til
ize
d
o
v
er
th
e
y
ea
r
s
to
d
etec
t
m
ac
h
i
n
e
f
au
l
ts
.
T
h
e
y
ca
teg
o
r
ized
th
e
a
lg
o
r
it
h
m
s
co
in
ed
as
in
te
lli
g
en
t
f
a
u
lt
d
iag
n
o
s
is
(
I
FD)
i
n
to
(
a)
tr
ad
itio
n
al
m
ac
h
in
e
lear
n
i
n
g
th
eo
r
ies
s
u
ch
as
t
h
e
p
r
o
b
ab
il
is
tic
g
r
ap
h
ica
l
m
et
h
o
d
(
P
GM
)
,
ar
tif
icial
n
e
u
r
al
n
et
w
o
r
k
(
A
NN)
,
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SVM)
,
k
-
n
ea
r
est
n
ei
g
h
b
o
r
(
KNN)
,
(
b
)
co
n
v
o
l
u
tio
n
a
l
n
e
u
r
al
n
et
w
o
r
k
(
C
N
N)
,
R
esNet
a
n
d
d
ee
p
lear
n
in
g
th
eo
r
y
s
u
c
h
a
s
d
ee
p
b
elief
n
et
w
o
r
k
(
DB
N)
,
an
d
(
c)
t
r
an
s
f
er
lear
n
in
g
t
h
eo
r
ies
s
u
ch
a
s
tr
an
s
f
er
co
m
p
o
n
en
t
an
al
y
s
is
(
T
C
A)
an
d
th
e
g
e
n
er
ativ
e
ad
v
er
s
ar
i
al
n
et
w
o
r
k
(
G
A
N)
.
T
h
ey
in
d
icate
d
th
at
al
m
o
s
t
a
ll
th
e
I
FD
co
u
ld
b
e
u
tili
ze
d
to
d
iag
n
o
s
e
r
o
tatin
g
m
a
c
h
i
n
er
y
f
au
lts
.
B
o
th
r
ev
ie
w
p
ap
er
s
f
o
cu
s
o
n
d
i
f
f
er
en
t
m
a
ch
in
e
lear
n
in
g
al
g
o
r
ith
m
s
a
n
d
s
tate
-
of
-
th
e
-
ar
t
tech
n
iq
u
e
s
u
s
e
f
u
l
i
n
d
ia
g
n
o
s
in
g
d
if
f
er
e
n
t
r
o
tatin
g
m
ac
h
i
n
e
f
au
lts
r
ath
er
t
h
an
o
n
t
h
e
f
au
l
ts
o
r
an
y
s
p
ec
i
f
ic
m
ac
h
in
e
f
a
u
lt.
S
án
ch
ez
et
a
l.
[
1
1
]
u
s
ed
th
e
r
a
n
d
o
m
f
o
r
est
a
n
d
KNN
m
ac
h
i
n
e
lear
n
i
n
g
alg
o
r
ith
m
s
a
s
class
if
ier
s
o
f
t
h
e
g
e
ar
b
o
x
an
d
b
ea
r
i
ng
f
au
lts
.
T
h
r
o
u
g
h
th
e
s
e,
a
m
et
h
o
d
o
lo
g
ical
s
tr
u
ct
u
r
e
f
o
r
d
et
ec
tin
g
m
u
ltip
le
f
au
lts
in
r
o
tatin
g
m
ac
h
in
er
y
w
a
s
d
is
cu
s
s
ed
.
Fro
m
t
h
e
v
ib
r
atio
n
s
i
g
n
al,
t
h
e
y
ca
lc
u
lated
t
h
ir
t
y
f
ea
tu
r
es
in
th
e
ti
m
e
d
o
m
ai
n
t
h
r
o
u
g
h
f
u
n
ctio
n
r
an
k
i
n
g
tech
n
iq
u
e
s
s
u
ch
a
s
r
elief
,
in
f
o
r
m
a
tio
n
g
ain
a
n
d
th
e
c
h
i
-
s
q
u
ar
e.
C
o
n
s
id
er
in
g
s
o
m
e
s
p
ec
i
f
ic
r
o
t
atin
g
m
ac
h
i
n
e
f
a
u
lts
,
o
n
e
o
f
t
h
e
f
au
l
ts
i
s
t
h
e
s
h
a
f
t
u
n
b
ala
n
ce
.
Ou
t
o
f
m
an
y
p
ap
er
s
o
n
t
h
i
s
t
y
p
e
o
f
f
au
lt,
a
r
ep
r
esen
tat
iv
e
p
ap
er
b
y
Go
h
ar
i
an
d
E
y
ad
i
[
1
2
]
p
r
ese
n
ted
a
co
m
p
ar
ativ
e
s
tu
d
y
o
f
t
w
o
m
ac
h
i
n
e
lear
n
i
n
g
alg
o
r
ith
m
s
n
a
m
el
y
t
h
e
d
ec
is
io
n
tr
ee
alg
o
r
ith
m
s
a
n
d
th
e
KNN
to
id
en
tify
t
h
e
s
h
a
f
t
u
n
b
ala
n
ce
p
ar
a
m
eter
s
s
u
c
h
as
ec
ce
n
tr
ic
r
ad
iu
s
,
d
is
c
n
u
m
b
er
an
d
th
e
ec
ce
n
tr
ic
m
ass
v
al
u
e.
T
h
ey
co
n
clu
d
ed
o
n
th
e
b
asis
o
f
th
ei
r
r
esu
lt
s
t
h
at
K
NN
e
s
ti
m
ates
t
h
e
f
au
lts
b
etter
th
a
n
t
h
e
Dec
is
io
n
T
r
ee
alg
o
r
ith
m
.
An
o
th
er
f
au
lt
is
t
h
e
b
lad
e
r
u
b
-
i
m
p
ac
t.
P
r
o
s
v
ir
in
et
a
l.
[
1
3
]
id
en
ti
f
ied
th
i
s
f
au
l
t
in
t
u
r
b
in
e
s
u
s
in
g
a
d
ee
p
n
eu
r
a
l
n
et
w
o
r
k
(
DNN)
a
n
d
th
e
Au
t
o
en
co
d
er
-
b
ased
n
o
n
l
in
ea
r
f
u
n
ct
io
n
ap
p
r
o
x
i
m
atio
n
.
T
h
e
y
u
s
ed
a
t
w
o
-
s
tep
p
r
o
ce
s
s
f
o
r
th
e
d
iag
n
o
s
i
s
.
Fir
s
t,
t
h
e
y
u
s
ed
a
d
ee
p
u
n
d
er
-
co
m
p
le
t
e
d
e
-
n
o
i
s
i
n
g
au
to
e
n
co
d
er
f
o
r
d
eter
m
i
n
i
n
g
t
h
e
n
o
n
li
n
ea
r
f
u
n
ctio
n
o
f
th
e
s
y
s
te
m
a
n
d
th
e
n
t
h
e
y
u
s
ed
t
h
e
r
es
u
ltin
g
r
esid
u
al
s
ig
n
als
f
r
o
m
t
h
e
o
r
ig
in
al
s
i
g
n
a
ls
a
s
in
p
u
t to
t
h
e
DNN
to
f
i
n
d
th
e
c
u
r
r
en
t sta
te
o
f
th
e
b
lad
e
-
r
o
to
r
s
y
s
te
m
.
C
las
s
ical
A
N
N
al
g
o
r
ith
m
s
h
a
v
e
b
ee
n
u
ti
lized
b
y
m
a
n
y
r
e
s
ea
r
ch
er
s
i
n
t
h
e
p
ast
to
d
ia
g
n
o
s
e
v
ar
io
u
s
r
o
tatin
g
m
ac
h
i
n
e
f
au
l
ts
.
Z
h
o
n
g
et
a
l.
[
1
4
]
u
tili
ze
d
s
in
g
le
b
ac
k
-
p
r
o
p
ag
atio
n
(
B
P
)
,
s
in
g
le
ellip
s
o
id
b
ac
k
p
r
o
p
ag
atio
n
,
an
d
h
ier
ar
c
h
ical
d
ia
g
n
o
s
tic
ar
ti
f
icia
l
n
e
u
r
al
n
et
w
o
r
k
(
H
D
A
NN)
al
g
o
r
ith
m
s
to
cla
s
s
i
f
y
d
if
f
e
r
e
n
t
r
o
to
r
s
y
s
te
m
f
au
lts
.
T
h
ey
s
h
o
w
ed
th
at
HD
A
N
N
r
eq
u
ir
ed
o
n
l
y
k
n
o
w
n
f
a
u
lt
p
atter
n
s
i
n
s
tead
o
f
tr
ain
i
n
g
t
h
e
e
n
tire
n
e
t
w
o
r
k
.
T
h
e
f
a
u
lt
d
ata
w
er
e
co
llected
e
x
p
er
i
m
e
n
tall
y
t
h
r
o
u
g
h
a
m
o
to
r
an
d
test
r
ig
r
o
to
r
s
y
s
te
m
a
n
d
t
h
en
f
ed
in
to
th
e
t
h
r
ee
al
g
o
r
ith
m
s
f
o
r
f
a
u
lt
id
en
t
if
icatio
n
.
T
h
e
y
co
n
cl
u
d
ed
th
at
HD
A
NN
clas
s
i
f
ied
th
e
f
au
lts
m
o
s
t
ac
cu
r
atel
y
.
N
ah
v
i
an
d
E
s
f
a
h
a
n
ian
[
1
5
]
,
u
s
ed
m
u
l
ti
-
la
y
er
f
ee
d
f
o
r
w
ar
d
n
eu
r
al
n
e
t
w
o
r
k
w
it
h
n
o
n
li
n
ea
r
n
eu
r
o
n
s
(
s
ig
m
o
id
al
ac
tiv
atio
n
f
u
n
ctio
n
)
to
id
en
ti
f
y
4
0
r
o
tatin
g
m
ac
h
i
n
e
f
au
lts
.
T
h
ey
co
u
ld
tr
ai
n
t
h
e
n
eu
r
al
n
et
w
o
r
k
f
r
o
m
s
elec
ted
d
ata
an
d
th
en
u
tili
ze
it
to
id
en
tify
th
e
f
a
u
lts
.
A
d
e
w
u
s
i
a
n
d
A
lb
ed
o
o
r
[
1
6
]
u
s
ed
m
u
lti
-
la
y
er
f
ee
d
-
f
o
r
w
ar
d
A
N
N
to
d
etec
t
cr
ac
k
in
a
r
o
to
r
.
T
w
o
-
a
n
d
th
r
ee
-
la
y
er
n
et
w
o
r
k
s
w
it
h
d
if
f
er
e
n
t
n
u
m
b
er
o
f
n
e
u
r
o
n
s
i
n
ea
ch
la
y
e
r
w
er
e
tes
ted
to
d
eter
m
i
n
e
an
o
p
ti
m
al
A
NN
co
n
f
i
g
u
r
atio
n
.
Vib
r
atio
n
s
i
g
n
al
s
o
f
a
n
o
r
m
a
l
r
o
to
r
an
d
th
at
w
it
h
a
p
r
o
p
ag
atin
g
cr
ac
k
w
e
r
e
u
s
ed
to
tr
ain
an
d
test
th
e
A
NN
u
s
i
n
g
b
ac
k
-
p
r
o
p
ag
atio
n
(
B
P)
alg
o
r
ith
m
.
T
h
e
r
esu
lts
s
h
o
w
ed
th
at
t
h
e
p
er
f
o
r
m
an
ce
o
f
a
th
r
ee
-
la
y
er
ANN
w
a
s
b
etter
th
an
th
at
o
f
t
w
o
-
la
y
er
ANN
i
n
d
et
ec
tin
g
t
h
e
s
e
v
er
it
y
o
f
t
h
e
p
r
o
p
ag
ati
n
g
cr
ac
k
i
n
t
h
e
r
o
to
r
.
Ng
u
y
e
n
et
a
l.
[
1
7
]
,
o
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2088
-
8
694
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
,
Vo
l.
12
,
No
.
3
,
Sep
tem
b
er
202
1
:
190
0
–
1
9
1
1
1902
th
e
o
th
er
h
a
n
d
,
w
o
r
k
ed
o
n
th
e
g
ea
r
b
o
x
s
y
s
te
m
an
d
d
ev
elo
p
e
d
a
d
ee
p
n
eu
r
al
n
et
w
o
r
k
b
ased
d
iag
n
o
s
i
s
m
o
d
el
to
d
iag
n
o
s
e
an
d
clas
s
i
f
y
m
u
lti
-
l
ev
el
to
o
th
cu
t
g
ea
r
f
a
u
lt
s
o
p
er
atin
g
u
n
d
er
v
ar
iab
le
s
h
a
f
t
s
p
ee
d
s
.
T
h
ey
f
ir
s
t
i
m
p
le
m
en
ted
t
h
e
ad
ap
tiv
e
n
o
is
e
co
n
tr
o
l
ap
p
r
o
ac
h
to
r
e
m
o
v
e
an
y
u
n
w
an
ted
n
o
is
e
in
t
h
e
o
r
ig
in
al
v
ib
r
atio
n
s
ig
n
al
an
d
th
e
n
e
m
p
lo
y
ed
th
e
s
tack
ed
s
p
ar
s
e
au
to
en
co
d
er
b
ased
d
e
ep
n
eu
r
al
n
et
w
o
r
k
to
d
i
ag
n
o
s
e
a
n
d
class
i
f
y
th
e
g
ea
r
b
o
x
f
au
l
ts
.
I
n
ad
d
itio
n
to
th
e
ap
p
licatio
n
o
f
A
N
N
to
d
iag
n
o
s
e
s
p
ec
i
f
ic
r
o
tatin
g
m
ac
h
in
e
f
au
l
ts
,
s
o
m
e
p
r
ev
io
u
s
w
o
r
k
s
ar
e
r
ep
o
r
ted
h
er
e
id
en
t
if
y
in
g
r
o
to
r
-
b
ea
r
in
g
s
y
s
te
m
f
au
lts
u
s
i
n
g
m
u
lt
i
-
la
y
er
B
P
n
e
u
r
al
n
e
t
w
o
r
k
[
1
8
]
,
r
o
llin
g
ele
m
en
t
b
ea
r
in
g
f
a
u
lts
u
s
i
n
g
ti
m
e
-
d
o
m
ai
n
f
ea
t
u
r
es
in
A
NN
[
1
9
]
,
m
in
o
r
f
a
u
lts
s
u
c
h
as
s
cr
atc
h
a
n
d
h
o
le
in
th
e
b
ea
r
in
g
s
o
f
in
d
u
ctio
n
m
o
to
r
s
u
s
in
g
f
i
v
e
m
ac
h
i
n
e
lear
n
in
g
al
g
o
r
ith
m
s
a
n
d
co
n
v
o
lu
tio
n
al
n
e
u
r
al
n
et
w
o
r
k
[
2
0
]
,
an
d
co
m
p
o
u
n
d
b
ea
r
in
g
d
ef
ec
ts
s
u
ch
as
i
n
n
er
r
ac
e,
o
u
ter
r
ac
e
an
d
r
o
ller
d
ef
ec
t
f
o
r
e
m
b
ed
d
ed
s
y
s
te
m
s
u
n
d
er
v
ar
y
i
n
g
r
o
tatio
n
al
s
p
ee
d
s
u
s
in
g
Mo
b
ileNet
-
v
2
;
a
li
g
h
t
s
tate
-
of
-
th
e
-
ar
t
co
n
v
o
l
u
tio
n
al
n
eu
r
al
n
et
w
o
r
k
m
o
d
el
[
2
1
]
.
I
n
ad
d
itio
n
,
tw
o
ar
ticles
r
elate
d
to
r
o
tatin
g
s
h
af
ts
ad
d
r
ess
ed
th
e
th
e
d
etec
tio
n
o
f
b
all
b
ea
r
in
g
f
au
lts
p
r
o
d
u
ci
n
g
v
ib
r
atio
n
s
i
n
R
o
tat
in
g
s
h
a
f
ts
u
s
in
g
m
u
l
ti
-
la
y
er
f
ee
d
f
o
r
w
ar
d
n
e
u
r
al
n
et
w
o
r
k
[
2
2
]
an
d
in
d
u
ctio
n
m
o
to
r
s
h
a
f
t
m
is
a
lig
n
m
e
n
t
s
u
s
i
n
g
m
u
lti
-
s
ca
le
en
tr
o
p
y
(
MSE
)
co
u
p
led
w
it
h
b
ac
k
-
p
r
o
p
ag
atio
n
n
eu
r
a
l
n
et
w
o
r
k
[
2
3
]
.
R
elate
d
to
th
e
p
r
esen
t st
u
d
y
o
f
u
tili
zi
n
g
A
NN
f
o
r
b
lad
e
f
au
lt
s
id
en
ti
f
icatio
n
,
s
o
m
e
p
r
ev
io
u
s
w
o
r
k
ca
n
b
e
f
o
u
n
d
.
R
ao
a
n
d
Srin
i
v
as
[
2
4
]
u
s
ed
s
in
g
le
la
y
er
n
e
u
r
al
n
e
t
w
o
r
k
to
lo
ca
te
t
h
e
o
p
ti
m
al
p
o
s
itio
n
o
f
r
o
tatio
n
al
co
n
s
tr
ain
t
alo
n
g
t
h
e
len
g
t
h
o
f
a
p
r
e
-
t
w
is
ted
b
lad
e
o
n
a
r
o
t
o
r
s
y
s
te
m
i
n
o
r
d
er
to
in
cr
ea
s
e
th
e
lo
w
es
t
n
at
u
r
al
f
r
eq
u
en
c
y
u
n
d
er
g
o
in
g
to
r
s
io
n
al
v
ib
r
atio
n
.
T
w
o
p
ap
er
s
ar
e
f
o
u
n
d
th
at
ad
d
r
ess
t
h
e
b
lad
e
f
au
lts
o
f
v
ar
iab
le
r
o
tatio
n
al
s
p
ee
d
w
in
d
t
u
r
b
in
e
s
.
First
i
s
b
y
C
h
e
n
et
a
l.
[
2
5
]
in
w
h
ich
th
e
y
u
s
ed
t
h
e
d
ee
p
le
ar
n
in
g
ap
p
r
o
ac
h
b
y
e
m
b
ed
d
in
g
atte
n
tio
n
m
ec
h
an
i
s
m
i
n
to
lo
n
g
-
s
h
o
r
t
-
ter
m
m
e
m
o
r
y
n
e
u
r
al
n
et
w
o
r
k
to
id
en
tify
th
e
w
i
n
d
tu
r
b
in
e
b
lad
e
i
m
b
alan
ce
f
a
u
lt
d
u
e
to
th
e
ac
c
u
m
u
latio
n
o
f
ice
o
n
t
h
e
b
lad
es.
Seco
n
d
p
ap
er
b
y
J
o
s
h
u
v
a
et
a
l.
[
2
6
]
p
r
esen
ted
a
s
tu
d
y
i
n
v
e
s
ti
g
ati
n
g
v
ar
io
u
s
w
in
d
t
u
r
b
in
e
b
lad
e
f
au
lts
u
s
i
n
g
f
ir
s
t
v
ar
iatio
n
al
m
o
d
e
d
ec
o
m
p
o
s
itio
n
(
VM
D)
f
o
r
ex
p
er
i
m
e
n
tal
d
ata
s
ig
n
al
p
r
e
-
p
r
o
ce
s
s
i
n
g
an
d
t
h
en
m
u
l
ti
-
la
y
er
p
er
ce
p
tr
o
n
(
ML
P
)
f
o
r
b
lad
e
f
au
l
t
class
i
f
icatio
n
.
T
h
e
ex
p
er
i
m
en
tal
d
ata
o
f
later
al
v
ib
r
at
io
n
s
o
f
t
h
e
s
h
a
f
t
i
n
o
n
l
y
v
er
tica
l
a
x
is
,
i.e
.
,
y
-
a
x
is
w
as
ac
q
u
ir
ed
s
p
ec
if
ic
to
w
in
d
t
u
r
b
in
es.
2.
E
XP
E
R
I
M
E
NT
AND
E
S
T
A
B
L
I
SH
M
E
NT
O
F
T
H
E
S
I
M
UL
A
T
I
O
N
M
O
DE
L
2
.
1
.
E
x
peri
m
e
nta
l set
-
up
a
nd
da
t
a
m
ea
s
ure
m
ent
Fig
u
r
e
1
s
h
o
w
s
th
e
ex
p
er
i
m
en
tal
test
r
ig
,
co
n
s
is
tin
g
o
f
a
v
ar
iab
le
s
p
ee
d
elec
tr
ic
m
o
to
r
,
a
r
o
to
r
,
a
6
-
b
lad
e
d
is
k
an
d
2
b
all
b
ea
r
in
g
s
u
p
p
o
r
ts
.
T
h
e
d
is
k
an
d
th
e
6
n
o
r
m
al
b
lad
es
w
er
e
d
esig
n
ed
an
d
f
ab
r
icate
d
in
a
w
o
r
k
s
h
o
p
.
I
n
ad
d
itio
n
to
th
e
6
n
o
r
m
al
b
lad
es,
m
o
r
e
b
lad
es
w
er
e
d
esi
g
n
ed
an
d
f
ab
r
icate
d
an
d
th
e
f
o
llo
w
i
n
g
d
ef
ec
ts
w
er
e
i
n
tr
o
d
u
ce
d
o
n
t
h
e
b
lad
es:
t
w
i
s
ted
,
b
en
t,
b
u
l
g
ed
,
cr
ac
k
ed
,
s
lo
tted
b
u
lg
ed
,
cr
ac
k
ed
an
d
s
lo
tted
b
lad
es.
So
m
e
o
f
t
h
e
b
lad
es
with
d
e
f
ec
ts
ar
e
s
h
o
w
n
i
n
Fi
g
u
r
e
2
(
a)
.
Fo
u
r
ac
ce
ler
o
m
eter
s
ar
e
attac
h
ed
to
t
h
e
t
w
o
s
u
p
p
o
r
ts
alo
n
g
t
h
e
x
-
an
d
y
-
ax
es
to
m
ea
s
u
r
e
th
e
la
ter
al
v
ib
r
atio
n
o
f
t
h
e
r
o
to
r
-
d
is
k
-
b
lad
e
s
y
s
te
m
.
T
h
e
s
ig
n
al
s
o
f
t
h
e
ac
ce
ler
o
m
eter
s
ar
e
co
n
n
ec
ted
to
a
s
i
x
-
ch
a
n
n
el
B
&
K
d
ata
ac
q
u
i
s
itio
n
s
y
s
te
m
a
n
d
t
h
en
to
a
lap
to
p
w
it
h
P
u
l
s
e
s
o
f
t
w
ar
e
(
F
i
g
u
r
e
2
(
b)
)
.
T
h
e
P
u
ls
e
s
o
f
t
w
ar
e
is
ca
p
ab
le
o
f
m
ea
s
u
r
in
g
,
a
n
a
l
y
s
i
n
g
a
n
d
s
to
r
i
n
g
v
ib
r
atio
n
s
i
g
n
als
f
r
o
m
t
h
e
ac
c
eler
o
m
e
ter
s
an
d
t
h
e
s
i
g
n
al
s
ca
n
b
e
ex
p
o
r
ted
to
ex
ce
l f
ile
f
o
r
f
u
r
t
h
er
p
r
o
ce
s
s
in
g
.
Fig
u
r
e
1
.
E
x
p
er
i
m
en
tal
test
-
r
i
g
f
o
r
r
o
to
r
-
d
is
k
-
b
lad
e
r
esp
o
n
s
es a
t a
co
n
s
ta
n
t r
u
n
n
in
g
s
p
ee
d
First,
th
e
tes
t
-
r
i
g
w
as
o
p
er
ated
b
y
r
u
n
n
i
n
g
th
e
elec
tr
ic
m
o
t
o
r
an
d
r
o
to
r
-
d
is
k
-
b
lad
e
s
y
s
te
m
a
t
a
s
p
ee
d
o
f
1
0
0
0
r
p
m
a
n
d
th
e
v
ib
r
atio
n
s
ig
n
al
s
i
n
ti
m
e
d
o
m
a
in
f
o
r
t
h
e
s
et
-
u
p
w
i
th
6
n
o
r
m
al
b
lad
es
w
er
e
o
b
tain
ed
f
r
o
m
th
e
f
o
u
r
ac
ce
ler
o
m
eter
s
an
d
s
a
v
ed
in
th
e
lap
to
p
.
On
e
o
f
th
e
6
n
o
r
m
al
b
lad
es
w
as
t
h
e
n
r
ep
lace
d
w
ith
o
n
e
b
lad
e
w
it
h
a
d
ef
ec
t
an
d
t
h
e
s
et
-
u
p
was
o
p
er
ated
ag
ain
at
1
0
0
r
p
m
t
o
o
b
tain
th
e
v
ib
r
atio
n
s
i
g
n
al
o
f
t
h
e
s
y
s
te
m
w
it
h
a
d
ef
ec
tiv
e
b
lad
e.
T
h
e
d
ef
ec
ti
v
e
b
lad
e
w
as
later
r
e
m
o
v
ed
a
n
d
s
u
b
s
eq
u
en
t
l
y
r
ep
lace
d
w
it
h
o
t
h
er
d
ef
ec
ti
v
e
b
lad
es
o
n
e
af
ter
th
e
o
th
er
to
o
b
tain
th
e
v
ib
r
atio
n
s
i
g
n
als
f
o
r
all
th
e
d
ef
ec
tiv
e
b
lad
es.
A
l
l
th
e
ti
m
e
d
o
m
ain
v
ib
r
atio
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
I
SS
N:
2088
-
8
694
R
o
ta
tin
g
b
l
a
d
e
fa
u
lts
cla
s
s
ific
a
tio
n
o
f
a
r
o
to
r
-
d
is
k
-
b
la
d
e
s
ystem
u
s
in
g
… (
A
b
d
u
lla
h
i A
b
u
b
a
ka
r
Ma
s
’
u
d
)
1903
s
ig
n
al
s
o
f
th
e
r
o
to
r
-
d
is
k
-
b
lad
e
s
y
s
te
m
w
er
e
e
x
p
o
r
ted
to
ex
ce
l
f
i
le
f
o
r
an
al
y
s
i
s
.
T
h
e
v
ib
r
ati
o
n
s
i
g
n
als
f
r
o
m
t
h
e
x
-
a
n
d
y
-
a
x
i
s
ac
ce
ler
o
m
e
ter
s
attac
h
ed
to
th
e
s
u
p
p
o
r
t
th
at
is
clo
s
e
to
th
e
d
is
k
-
b
lad
e
s
y
s
te
m
ar
e
co
n
s
id
er
ed
f
o
r
f
u
r
t
h
er
an
al
y
s
is
.
(
a)
(
b
)
Fig
u
r
e
2
.
(
a)
t
h
r
ee
d
ef
ec
tiv
e
b
l
ad
es (
b
en
t,
t
w
i
s
ted
an
d
b
r
o
k
en
b
lad
es);
(
b
)
P
C
an
d
B
&
K
p
u
ls
e
s
y
s
t
e
m
an
d
v
ib
r
atio
n
d
ata
m
ea
s
u
r
e
m
en
t a
n
d
an
al
y
s
i
s
ac
ce
s
s
o
r
ies.
2
.
2
.
Da
t
a
co
llect
io
n pro
ce
du
re
E
x
p
er
i
m
e
n
ts
ar
e
co
n
d
u
cted
o
n
th
e
d
if
f
er
en
t
b
lad
es
o
f
a
w
i
n
d
tu
r
b
in
e
b
y
v
ar
y
in
g
th
e
f
r
e
q
u
en
c
y
(
F)
f
r
o
m
0
Hz
to
2
0
0
0
Hz.
T
h
e
x
-
an
d
y
-
co
o
r
d
in
ates
f
o
r
d
if
f
er
en
t
s
ets
o
f
b
lad
es
„
n
a
m
el
y
‟
th
e
g
o
o
d
b
lad
es
w
i
th
o
u
t
an
y
d
ef
ec
ts
,
t
w
is
ted
b
lad
es,
b
en
t
b
lad
es,
b
u
lg
ed
b
lad
es,
cr
ac
k
ed
b
lad
es,
an
d
s
lo
tted
b
lad
es;
all
o
f
w
h
ic
h
ar
e
r
ec
o
r
d
e
d
in
an
ex
ce
l
f
ile.
T
h
e
d
ataset
s
o
p
r
ep
a
r
ed
h
as
1
3
attr
ib
u
tes
an
d
2
1
3
4
o
b
s
er
v
atio
n
s
.
T
h
e
attr
ib
u
tes
ar
e
th
e
f
r
eq
u
e
n
c
y
,
x
-
a
n
d
y
-
a
x
is
v
ib
r
atio
n
m
ea
s
u
r
e
m
en
t
s
o
f
t
h
e
g
o
o
d
b
lad
es
w
ith
o
u
t
an
y
d
ef
e
cts,
t
w
is
ted
b
lad
es,
b
en
t b
lad
es,
b
u
lg
ed
b
lad
es,
cr
ac
k
ed
b
lad
es,
an
d
s
lo
tted
b
lad
e
s
.
2
.
3
.
Si
m
ula
t
io
n
m
o
del
P
y
t
h
o
n
i
s
a
h
i
g
h
-
le
v
el
p
r
o
g
r
a
m
m
in
g
la
n
g
u
a
g
e.
P
y
th
o
n
is
a
p
r
o
g
r
am
m
i
n
g
la
n
g
u
a
g
e
t
h
at
ca
n
b
e
u
tili
ze
d
to
cr
ea
te
d
esk
to
p
g
r
ap
h
ical
u
s
er
in
ter
f
ac
e
(
GUI
)
p
r
o
g
r
a
m
s
[
2
8
]
.
Fu
r
th
er
m
o
r
e,
s
in
c
e
P
y
t
h
o
n
is
a
h
i
g
h
-
lev
el
p
r
o
g
r
a
m
m
i
n
g
la
n
g
u
a
g
e,
it
h
elp
s
y
o
u
to
co
n
ce
n
tr
ate
o
n
th
e
ap
p
licatio
n
's
co
r
e
f
ea
t
u
r
es
w
h
ile
i
t
h
a
n
d
les
th
e
m
u
n
d
an
e
p
r
o
g
r
a
m
m
i
n
g
ta
s
k
s
.
P
y
t
h
o
n
w
as
u
s
ed
in
t
h
i
s
s
tu
d
y
to
d
ev
elo
p
th
e
ANN
f
o
r
class
i
f
y
in
g
v
ar
io
u
s
r
o
tatin
g
b
lad
e
f
au
lts
co
m
m
o
n
l
y
e
n
co
u
n
ter
ed
in
r
o
tati
n
g
m
ac
h
in
e
s
.
C
o
m
p
ar
ed
to
o
th
er
s
o
f
t
w
ar
e
‟
s
,
p
y
th
o
n
is
a
g
en
er
al
-
p
u
r
p
o
s
e
p
r
o
g
r
a
m
m
in
g
lan
g
u
ag
e.
P
y
t
h
o
n
g
i
v
es
m
o
r
e
co
n
tr
o
l
o
v
er
o
n
e's
co
d
e
o
r
g
an
izatio
n
a
n
d
b
etter
m
an
a
g
e
m
e
n
t
o
f
n
a
m
e
s
p
ac
e.
T
h
er
e
ar
e
th
r
ee
m
ai
n
s
tep
s
to
d
ev
elo
p
th
e
A
NN
i
n
p
y
t
h
o
n
[
2
9
]
,
i.e
.
1
)
u
s
in
g
t
h
e
p
r
o
v
id
ed
d
ata
as
in
p
u
t
2
)
p
r
ed
ictin
g
th
e
o
u
tp
u
t
3
)
co
m
p
ar
is
o
n
o
f
th
e
p
r
ed
ictio
n
w
it
h
th
e
ac
tu
al
o
u
tp
u
t
an
d
4
)
ch
an
g
i
n
g
i
ts
i
n
ter
n
al
s
tate
i
n
o
r
d
er
to
c
o
r
r
ec
tly
p
r
ed
ict
th
e
n
e
x
t ti
m
e.
3.
ARTI
F
I
CI
AL
N
E
URA
L
NE
T
WO
RK
AN
N
r
ef
er
s
to
a
co
m
p
u
tatio
n
al
m
o
d
el
m
i
m
ic
k
i
n
g
t
h
e
h
u
m
an
b
r
ain
an
d
its
in
f
o
r
m
at
io
n
p
r
o
ce
s
s
in
g
ca
p
ab
ilit
ies
[
3
0
]
,
[
3
1
]
.
A
n
ANN
is
a
n
e
n
s
e
m
b
le
o
f
h
i
g
h
l
y
i
n
ter
co
n
n
ec
ted
p
r
o
ce
s
s
in
g
u
n
i
ts
k
n
o
w
n
a
s
t
h
e
n
eu
r
o
n
s
[
3
2
]
.
T
h
er
e
ar
e
a
n
u
m
b
er
o
f
n
u
m
er
ical
v
al
u
es
b
et
w
ee
n
t
h
e
n
e
u
r
o
n
s
t
h
at
s
h
o
w
s
th
e
s
tr
e
n
g
t
h
o
f
th
e
co
n
n
ec
tio
n
s
k
n
o
w
n
as
t
h
e
w
e
ig
h
ts
[
2
3
]
,
[
3
3
]
.
T
h
e
A
NN
is
ca
p
ab
le
o
f
s
elf
-
o
r
g
an
izat
io
n
an
d
ac
q
u
is
it
io
n
o
f
in
f
o
r
m
atio
n
,
i.e
.
,
lear
n
i
n
g
[
3
4
]
.
Fig
u
r
e
3
s
h
o
w
s
a
t
y
p
ical
t
h
r
e
e
f
ee
d
f
o
r
w
ar
d
n
e
u
r
al
n
et
w
o
r
k
,
co
m
p
r
is
i
n
g
o
f
th
e
i
n
p
u
t
la
y
er
,
h
id
d
en
la
y
er
a
n
d
o
u
tp
u
t la
y
er
.
I
n
th
is
n
et
w
o
r
k
,
t
h
e
m
id
d
le
la
y
er
n
o
d
es o
n
l
y
r
ec
ei
v
e
i
n
f
o
r
m
atio
n
f
r
o
m
t
h
e
n
o
d
es o
f
t
h
e
p
r
ev
io
u
s
la
y
er
an
d
f
o
r
w
ar
d
t
h
eir
o
u
tp
u
ts
to
co
n
s
ec
u
ti
v
e
la
y
e
r
n
o
d
es o
n
l
y
.
T
h
e
h
id
d
en
la
y
er
s
ar
e
n
o
t c
o
n
n
ec
ted
w
it
h
t
h
e
o
u
t
s
id
e
w
o
r
ld
d
ir
ec
t
l
y
.
T
h
e
co
m
p
u
tatio
n
al
ele
m
e
n
ts
o
r
n
o
d
es
u
s
ed
in
n
e
u
r
al
n
et
w
o
r
k
m
o
d
el
s
ar
e
n
o
r
m
all
y
n
o
n
lin
ea
r
an
d
g
e
n
er
all
y
an
alo
g
.
As
s
h
o
w
n
in
Fi
g
u
r
e
3
,
th
e
s
i
m
p
le
s
t
n
o
d
e
s
u
m
s
t
h
e
w
e
ig
h
ted
in
p
u
ts
an
d
tr
an
s
f
er
s
t
h
e
r
es
u
lt
t
h
r
o
u
g
h
n
o
n
l
in
ea
r
it
y
.
T
h
e
m
aj
o
r
p
a
r
t
o
f
th
e
A
N
N
is
it
s
lear
n
i
n
g
alg
o
r
ith
m
.
T
h
e
B
P
,
w
h
ic
h
is
a
s
u
p
er
v
i
s
ed
lear
n
in
g
alg
o
r
ith
m
,
is
t
h
e
m
o
s
t
w
id
el
y
u
s
ed
f
o
r
v
ar
io
u
s
ap
p
licatio
n
s
[
3
5
]
,
[
3
6
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2088
-
8
694
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
,
Vo
l.
12
,
No
.
3
,
Sep
tem
b
er
202
1
:
190
0
–
1
9
1
1
1904
Fig
u
r
e
3
.
T
y
p
ical
f
ee
d
-
f
o
r
w
ar
d
A
NN
3
.
1
.
B
a
ck
-
pro
pa
g
a
t
io
n a
lg
o
rit
hm
A
t
y
p
e
o
f
s
u
p
er
v
is
ed
lear
n
in
g
is
t
h
e
B
P
alg
o
r
ith
m
[
3
7
]
,
[
3
8
]
.
I
t
tak
es
p
lace
i
n
t
w
o
s
ta
g
e
s
,
„
n
a
m
el
y
‟
f
o
r
w
ar
d
p
r
o
p
ag
atio
n
o
f
i
n
f
o
r
m
atio
n
an
d
b
ac
k
w
ar
d
p
r
o
p
ag
atio
n
o
f
er
r
o
r
s
[
1
4
]
.
I
t le
ar
n
s
b
y
p
r
esen
tin
g
th
e
in
p
u
t
an
d
o
u
tp
u
t
d
ata
p
atter
n
s
r
ep
ea
ted
l
y
an
d
p
r
o
p
ag
atin
g
t
h
e
er
r
o
r
b
ac
k
w
ar
d
ea
ch
ti
m
e
an
d
ad
j
u
s
tin
g
th
e
w
ei
g
h
t
s
u
n
t
il
th
e
er
r
o
r
m
i
n
i
m
ize
s
to
a
r
ea
s
o
n
ab
le
lo
w
v
al
u
e.
So
m
e
o
f
th
e
B
P
alg
o
r
ith
m
'
s
f
u
n
d
a
m
en
tal
w
ea
k
n
e
s
s
es
in
cl
u
d
e
lo
n
g
er
co
n
v
er
g
en
ce
p
er
io
d
an
d
s
u
s
ce
p
tib
ilit
y
to
f
ailu
r
e
o
f
tr
ain
i
n
g
.
On
e
o
f
th
e
m
o
d
if
icatio
n
s
in
tr
o
d
u
ce
d
to
r
eso
lv
e
t
h
is
i
s
s
u
e
in
cl
u
d
e
ad
d
in
g
a
m
o
m
e
n
t
u
m
ter
m
f
o
r
f
a
s
ter
tr
ain
in
g
,
t
h
e
r
eb
y
co
m
p
r
o
m
is
i
n
g
ad
d
itio
n
al
m
e
m
o
r
y
s
p
ac
e.
T
h
e
f
o
llo
w
in
g
eq
u
a
tio
n
g
i
v
es t
h
e
o
u
tp
u
t o
f
t
h
e
n
e
u
r
o
n
[
3
7
]
:
(
)
∑
(
1
)
W
h
er
e
w
1
j
an
d
x
j
ar
e
t
h
e
ass
o
c
iated
w
ei
g
h
ts
a
n
d
th
e
i
n
p
u
ts
o
f
t
h
e
A
NN,
r
esp
ec
ti
v
el
y
.
Fig
u
r
e
4
s
h
o
w
h
o
w
t
h
e
B
P
alg
o
r
it
h
m
is
i
m
p
le
m
e
n
ted
to
g
et
h
er
w
i
th
t
h
e
p
r
o
ce
s
s
in
g
ele
m
e
n
ts
.
I
n
p
u
t
s
(
X
p
1
..........
...
X
pNo
)
ar
e
in
itial
l
y
p
r
o
p
ag
ated
an
d
t
h
e
r
e
s
p
o
n
s
e
i
s
co
m
p
u
ted
v
ia
t
h
e
n
e
t
w
o
r
k
.
T
h
en
th
e
o
u
tp
u
t
er
r
o
r
(
T
p
1
……
….
T
pNM
)
i
s
p
r
o
p
ag
ated
b
ac
k
w
ar
d
th
r
o
u
g
h
th
e
n
et
w
o
r
k
an
d
th
e
w
ei
g
h
ts
ar
e
m
o
d
i
f
ied
ac
co
r
d
in
g
to
th
e
alg
o
r
ith
m
b
ased
o
n
th
e
g
r
ad
ien
t
d
esce
n
t
[
3
0
]
.
T
h
is
m
et
h
o
d
co
n
tin
u
es
u
n
t
il
th
e
av
er
ag
e
m
ea
n
s
q
u
ar
e
er
r
o
r
at
th
e
o
u
tp
u
t
r
ea
ch
es
an
accep
t
ab
le
m
i
n
i
m
u
m
v
a
lu
e
.
F
o
r
th
is
w
o
r
k
,
th
e
B
P
alg
o
r
ith
m
w
ill
b
e
i
m
p
le
m
e
n
ted
in
t
h
e
A
N
N
to
class
if
y
th
e
d
if
f
er
e
n
t b
lad
e
f
au
lts
.
Fig
u
r
e
4
.
T
y
p
ical
i
m
p
le
m
en
tat
io
n
o
f
th
e
B
P
alg
o
r
ith
m
4.
DATA
E
XP
L
O
RA
T
I
O
N
A
N
D
P
RE
-
P
RO
CE
S
SI
N
G
T
h
er
e
ar
e
s
ev
er
al
i
m
p
o
r
tan
t
s
tep
s
i
n
d
ata
e
x
p
lo
r
atio
n
a
n
d
p
r
e
-
p
r
o
ce
s
s
in
g
.
T
h
is
s
ec
tio
n
g
i
v
es
an
ac
co
u
n
t o
f
t
h
e
s
tep
s
in
v
o
lv
ed
i
n
d
ata
ex
p
lo
r
atio
n
an
d
p
r
e
-
p
r
o
ce
s
s
i
n
g
f
o
r
v
ib
r
atio
n
f
au
l
ts
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
I
SS
N:
2088
-
8
694
R
o
ta
tin
g
b
l
a
d
e
fa
u
lts
cla
s
s
ific
a
tio
n
o
f
a
r
o
to
r
-
d
is
k
-
b
la
d
e
s
ystem
u
s
in
g
… (
A
b
d
u
lla
h
i A
b
u
b
a
ka
r
Ma
s
’
u
d
)
190
5
4
.
1
.
E
x
plo
ra
t
o
ry
da
t
a
a
na
ly
s
is
T
h
e
n
ex
t
i
m
p
o
r
ta
n
t
s
tep
a
f
ter
d
ata
co
llectio
n
i
s
t
h
e
e
x
p
lo
r
ato
r
y
d
ata
an
al
y
s
is
.
T
h
e
d
ata
is
an
al
y
s
ed
b
y
s
u
m
m
ar
izin
g
t
h
e
b
asic
s
tatis
t
ical
d
escr
ip
tio
n
o
f
th
e
d
ata
f
o
r
f
in
d
in
g
an
o
m
alies,
m
is
s
i
n
g
v
al
u
es,
d
u
p
licate
v
alu
e
s
,
an
d
p
atter
n
s
o
r
r
elatio
n
s
h
ip
b
et
w
ee
n
t
h
e
attr
ib
u
te
s
.
Statis
t
ical
d
escr
ip
tio
n
o
f
t
h
e
d
ata
r
ev
ea
ls
t
h
at
all
attr
ib
u
tes
ar
e
f
lo
ati
n
g
p
o
in
t
n
u
m
b
er
s
an
d
th
er
e
ar
e
n
o
m
i
s
s
i
n
g
v
al
u
es
i
n
th
e
d
ataset.
T
h
e
r
a
n
g
e
o
f
t
h
e
attr
ib
u
te
f
r
eq
u
en
c
y
i
s
f
r
o
m
0
Hz
to
2
0
0
0
Hz.
Fo
r
all
o
th
er
attr
ib
u
tes
,
th
e
m
i
n
i
m
u
m
v
al
u
e
is
a
v
er
y
s
m
all
co
n
ti
n
u
o
u
s
f
lo
ati
n
g
p
o
in
t
v
a
lu
e
t
h
at
h
a
s
a
p
o
w
er
o
f
1
e
-
9
.
Fu
r
t
h
er
an
al
y
s
is
o
f
t
h
e
d
ata
u
s
in
g
d
u
p
licate
d
in
p
an
d
as,
a
d
ata
an
al
y
s
is
s
o
f
t
w
ar
e,
r
ev
ea
l
t
h
er
e
ar
e
n
o
d
u
p
licate
r
o
w
s
in
t
h
e
d
ataset.
T
o
f
u
r
th
er
an
a
l
y
s
e
t
h
e
d
ataset
f
o
r
o
u
tlier
s
,
s
k
e
w
n
es
s
an
d
k
u
r
to
s
is
f
u
n
ctio
n
s
i
n
p
an
d
as
h
a
v
e
b
ee
n
u
s
ed
a
n
d
th
e
r
esu
l
ts
r
ev
ea
l
h
ig
h
p
o
s
i
tiv
e
v
al
u
e
f
o
r
s
k
e
w
an
d
k
u
r
to
s
i
s
f
o
r
all
attr
ib
u
tes
ex
ce
p
t
t
h
e
f
r
eq
u
e
n
c
y
.
Hi
g
h
p
o
s
itiv
e
v
al
u
e
f
o
r
s
k
e
w
i
n
d
i
ca
tes
all
a
ttrib
u
te
s
ex
ce
p
t
f
r
eq
u
en
c
y
ar
e
r
i
g
h
t
s
k
e
w
ed
w
h
i
c
h
i
n
d
icate
s
t
h
e
d
ata
d
is
tr
ib
u
tio
n
is
n
o
t
s
y
m
m
etr
ical
an
d
h
i
g
h
l
y
s
k
e
w
ed
w
it
h
lo
n
g
er
tail
s
to
t
h
e
r
i
g
h
t
s
i
d
e.
Hig
h
p
o
s
iti
v
e
v
alu
e
s
f
o
r
k
u
r
to
s
is
r
ev
ea
l
n
u
m
er
o
u
s
o
u
tlie
r
s
in
th
e
d
ataset
f
o
r
all
attr
ib
u
tes
ex
ce
p
t
f
r
eq
u
e
n
c
y
.
T
h
e
r
esu
lt
s
s
h
o
w
n
in
T
ab
le
1
in
d
icat
e
s
i
g
n
if
ican
t
r
ed
u
cti
o
n
in
s
k
e
w
n
e
s
s
a
n
d
k
u
r
to
s
is
v
al
u
es
w
h
ich
i
n
d
icat
es
t
h
e
tr
a
n
s
f
o
r
m
ed
d
atase
t
i
s
n
o
w
clo
s
er
to
n
o
r
m
al
d
is
tr
ib
u
tio
n
w
it
h
r
ed
u
ce
d
o
u
tlier
s
.
E
x
p
lo
r
ato
r
y
d
ata
an
al
y
s
i
s
r
ev
ea
ls
t
h
at
th
e
d
ataset
h
as
n
u
m
er
o
u
s
o
u
tlier
s
an
d
is
p
o
s
itiv
el
y
h
i
g
h
l
y
sk
e
w
ed
.
T
h
er
ef
o
r
e,
t
h
e
tr
ai
n
i
n
g
o
f
t
h
e
ANN
w
o
u
ld
n
o
t
b
e
id
ea
l
an
d
w
i
th
t
h
e
h
elp
o
f
a
s
u
i
t
ab
le
tr
an
s
f
o
r
m
at
io
n
alg
o
r
ith
m
,
t
h
e
d
ataset
h
a
s
to
b
e
tr
an
s
f
o
r
m
ed
to
a
n
e
w
d
ata
s
et
d
ev
o
id
o
f
o
u
tlier
s
an
d
s
k
e
w
n
e
s
s
.
T
ab
le
1
.
A
n
al
y
s
i
s
o
f
th
e
at
tr
ib
u
tes i
n
t
h
e
d
ataset
B
e
f
o
r
e
p
r
o
c
e
ssi
n
g
A
f
t
e
r
p
r
o
c
e
ssi
n
g
S
k
e
w
n
e
ss
K
u
r
t
o
si
s
S
k
e
w
n
e
ss
K
u
r
t
o
si
s
F
0
-
1
.
2
-
1
.
9
7
4
8
3
5
.
5
3
2
3
6
G
o
o
d
X
1
4
.
9
4
0
3
9
2
5
4
.
9
3
6
8
-
0
.
9
0
6
3
9
0
.
6
4
9
1
1
4
G
o
o
d
Y
4
5
.
2
5
4
2
6
2
0
7
4
.
9
6
3
-
1
.
4
2
5
9
3
4
.
2
9
1
2
9
0
Tw
i
st
e
d
X
1
4
.
8
7
6
5
6
3
2
4
.
2
8
2
4
-
0
.
7
4
6
9
4
-
0
.
1
3
5
7
1
Tw
i
st
e
d
Y
4
3
.
7
7
0
0
1
1
9
8
1
.
8
8
6
-
1
.
1
2
6
7
7
2
.
4
7
3
1
8
2
B
e
n
t
X
1
1
.
3
3
7
5
6
1
6
2
.
3
0
0
6
-
0
.
8
4
8
7
5
0
.
2
8
6
3
3
9
B
e
n
t
Y
4
4
.
9
9
9
0
5
2
0
5
8
.
7
4
2
-
1
.
3
6
6
1
2
4
.
1
8
8
4
8
7
B
u
l
g
e
d
X
2
4
.
8
4
0
2
5
7
5
9
.
1
2
7
1
-
0
.
9
0
4
4
9
0
.
6
1
0
6
4
8
B
u
l
g
e
d
Y
4
6
.
1
2
2
0
1
2
1
2
9
.
4
6
1
-
1
.
5
5
1
3
1
5
.
3
7
2
1
0
9
C
r
a
c
k
e
d
X
2
0
.
2
1
3
3
6
5
4
6
.
9
6
3
5
-
0
.
8
8
5
5
1
0
.
4
8
0
5
5
3
C
r
a
c
k
e
d
Y
4
5
.
8
9
0
5
2
2
1
1
5
.
0
4
3
-
1
.
4
9
0
2
4
.
6
5
6
8
0
1
S
l
o
t
t
e
d
X
1
8
.
8
8
7
8
7
4
6
3
.
5
6
9
8
-
0
.
8
2
6
6
5
0
.
0
7
2
9
2
7
S
l
o
t
t
e
d
Y
4
6
.
1
6
1
9
7
2
1
3
1
.
9
4
2
-
1
.
4
7
5
8
3
5
.
1
5
8
6
2
3
4
.
2
.
Da
t
a
t
ra
ns
f
o
rm
a
t
io
n
E
v
en
m
an
y
tr
a
n
s
f
o
r
m
at
io
n
s
ar
e
av
ailab
le,
lo
g
tr
a
n
s
f
o
r
m
a
tio
n
i
s
ap
p
lied
to
al
l
attr
ib
u
t
es
o
f
t
h
is
d
ataset
to
r
ed
u
ce
th
e
s
k
e
w
n
e
s
s
to
m
a
k
e
th
e
d
ata
as
n
o
r
m
al
as
p
o
s
s
ib
le.
Fo
r
th
e
attr
ib
u
te
f
r
eq
u
en
c
y
,
th
e
lo
g
tr
an
s
f
o
r
m
atio
n
is
ap
p
lied
to
all
v
alu
e
s
ex
ce
p
t t
h
e
f
ir
s
t v
a
lu
e
w
h
ic
h
is
ze
r
o
.
Fo
r
all
o
th
er
v
al
u
es,
t
h
e
ap
p
licatio
n
o
f
th
e
L
o
g
tr
an
s
f
o
r
m
atio
n
r
e
s
u
lts
in
a
n
e
g
ati
v
e
v
alu
e
a
n
d
h
en
ce
,
af
ter
ap
p
l
y
in
g
L
o
g
tr
an
s
f
o
r
m
atio
n
,
th
e
s
i
g
n
o
f
th
e
d
ata
is
ch
an
g
ed
f
r
o
m
n
eg
ati
v
e
to
p
o
s
itiv
e.
T
h
e
b
asic
s
tatis
tical
d
etails
o
f
t
h
e
tr
an
s
f
o
r
m
ed
d
ataset
ar
e
s
u
m
m
ar
ized
in
T
ab
le
2.
T
ab
le
2
.
Statis
tical
d
escr
ip
tio
n
o
f
th
e
d
ata
af
ter
tr
an
s
f
o
r
m
atio
n
C
o
u
n
t
M
e
a
n
S
t
a
n
d
a
r
d
d
e
v
i
a
t
i
o
n
M
i
n
i
m
u
m
2
5
%
5
0
%
7
5
%
M
a
x
i
m
u
m
F
r
e
q
u
e
n
c
y
2
1
3
4
2
.
8
6
6
2
9
0
.
4
3
5
0
7
-
0
.
0
2
8
0
3
2
.
6
9
8
9
0
2
.
9
9
9
9
3
3
.
1
7
6
0
2
3
.
3
0
0
9
6
G
o
o
d
X
2
1
3
4
5
.
6
9
4
5
5
0
.
7
8
2
8
6
2
.
4
8
5
9
6
5
.
1
5
3
3
5
5
.
8
9
3
0
8
6
.
2
9
7
5
5
8
.
7
3
8
4
5
G
o
o
d
Y
2
1
3
4
5
.
7
2
4
4
2
0
.
5
7
9
2
9
1
.
5
9
3
4
1
5
.
4
4
5
1
1
5
.
8
5
1
1
2
6
.
1
0
7
6
6
8
.
5
2
2
1
6
Tw
i
st
e
d
X
2
1
3
4
5
.
6
0
9
8
4
0
.
8
6
2
8
6
2
.
4
0
9
9
7
4
.
9
2
5
3
6
5
.
9
0
1
6
4
6
.
3
0
9
4
2
8
.
6
7
3
1
2
Tw
i
st
e
d
Y
2
1
3
4
5
.
6
7
1
3
8
0
.
5
9
9
8
9
1
.
9
8
1
6
5
5
.
3
3
3
7
7
5
.
8
0
1
1
6
6
.
0
8
9
5
5
8
.
8
1
8
6
1
B
e
n
t
X
2
1
3
4
5
.
6
7
8
7
1
0
.
7
9
3
9
5
2
.
7
5
3
2
7
5
.
1
6
9
2
4
5
.
8
9
2
1
9
6
.
2
9
5
0
9
8
.
6
8
1
6
3
B
e
n
t
Y
2
1
3
4
5
.
7
1
6
8
6
0
.
6
0
3
2
7
1
.
4
7
9
7
7
5
.
4
2
9
5
4
5
.
8
4
3
2
6
6
.
1
2
0
8
8
8
.
9
5
6
8
7
B
u
l
g
e
d
X
2
1
3
4
5
.
7
2
4
0
2
0
.
7
7
2
8
0
2
.
2
5
6
8
8
5
.
2
0
4
7
5
5
.
9
3
1
4
9
6
.
3
1
4
9
3
8
.
6
6
3
9
3
B
u
l
g
e
d
Y
2
1
3
4
5
.
7
6
0
8
6
0
.
6
0
5
5
0
0
.
8
9
4
3
8
5
.
4
7
0
3
0
5
.
8
9
3
5
8
6
.
1
5
1
7
3
8
.
9
6
4
1
9
C
r
a
c
k
e
d
X
2
1
3
4
5
.
7
1
2
4
3
0
.
7
8
8
0
9
2
.
3
4
2
5
2
5
.
2
0
4
5
3
5
.
9
3
7
6
8
6
.
3
1
6
2
6
8
.
3
9
5
6
7
C
r
a
c
k
e
d
Y
2
1
3
4
5
.
7
4
4
7
8
0
.
6
0
0
0
7
1
.
2
9
3
4
6
5
.
4
6
0
9
8
5
.
8
7
8
5
0
6
.
1
3
8
0
1
8
.
7
6
3
3
7
S
l
o
t
t
e
d
X
2
1
3
4
5
.
6
7
2
2
0
0
.
8
1
9
1
2
2
.
4
6
9
7
5
5
.
0
8
9
0
5
5
.
9
2
9
1
7
6
.
3
0
3
0
8
8
.
3
8
4
8
2
S
l
o
t
t
e
d
Y
2
1
3
4
5
.
6
9
8
6
9
0
.
6
0
1
3
0
0
.
7
3
0
8
9
5
.
3
9
3
0
8
5
.
8
3
6
2
1
6
.
1
0
1
7
9
8
.
7
6
5
7
9
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2088
-
8
694
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
,
Vo
l.
12
,
No
.
3
,
Sep
tem
b
er
202
1
:
190
0
–
1
9
1
1
1906
5.
DE
S
I
G
N
O
F
T
H
E
ANN
A
N
D
T
H
E
P
R
E
DIC
T
E
D
RE
SU
L
T
S
5
.
1
.
Desig
n o
f
t
he
ANN
T
h
is
s
ec
tio
n
d
escr
ib
es
t
h
e
ar
c
h
itect
u
r
e
o
f
t
h
e
ANN
d
es
ig
n
e
d
to
p
r
ed
ict
th
e
1
0
f
a
u
lt
s
(
5
d
ef
ec
ts
in
ea
ch
x
-
a
n
d
y
-
d
ir
ec
tio
n
s
)
i
n
t
h
e
b
lad
es
o
f
t
h
e
r
o
tatin
g
m
ac
h
in
e
s
.
T
h
e
in
p
u
t
la
y
er
o
f
A
N
N
co
n
s
i
s
ts
o
f
th
r
ee
n
eu
r
o
n
s
„
n
a
m
el
y
‟
th
e
f
r
eq
u
e
n
c
y
,
g
o
o
d
x
a
n
d
g
o
o
d
y
attr
ib
u
tes.
T
h
e
o
u
tp
u
t
la
y
er
co
n
s
is
t
s
o
f
1
0
n
eu
r
o
n
s
n
a
m
e
l
y
,
x
-
a
n
d
y
-
a
x
i
s
v
ib
r
atio
n
m
ea
s
u
r
e
m
e
n
t
s
o
f
t
w
is
t
ed
b
l
ad
es,
b
en
t
b
lad
es,
b
u
l
g
ed
b
lad
es,
cr
ac
k
ed
b
lad
es
an
d
s
lo
tted
b
lad
es.
Af
ter
tr
ial
an
d
er
r
o
r
w
i
th
a
n
u
m
b
er
o
f
n
e
u
r
o
n
s
,
a
s
in
g
le
h
id
d
en
la
y
er
with
8
0
n
e
u
r
o
n
s
w
it
h
lin
ea
r
ac
tiv
a
tio
n
f
u
n
ctio
n
is
co
n
s
id
er
ed
b
et
w
ee
n
t
h
e
i
n
p
u
t
an
d
o
u
tp
u
t
la
y
er
as
s
h
o
w
n
i
n
Fi
g
u
r
e
5
.
B
ef
o
r
e
ch
o
o
s
in
g
t
h
e
li
n
ea
r
ac
ti
v
atio
n
,
co
r
r
elatio
n
b
et
w
ee
n
th
e
a
ttri
b
u
tes
w
er
e
co
n
s
id
er
ed
an
d
a
p
o
s
itiv
e
co
r
r
elatio
n
b
et
w
ee
n
th
e
i
n
p
u
t a
n
d
o
u
tp
u
t
attr
ib
u
tes
w
a
s
f
o
u
n
d
as s
h
o
w
n
in
th
e
h
ea
t
m
ap
(
s
ee
Fig
u
r
e
6
)
.
Fig
u
r
e
5
.
T
h
e
ar
ch
itectu
r
es o
f
th
e
A
NN
to
f
in
d
t
h
e
f
a
u
lt
s
in
t
h
e
r
o
tatin
g
m
ac
h
i
n
e
Fig
u
r
e
6
.
Hea
t
m
ap
o
f
th
e
co
r
r
elatio
n
b
et
w
ee
n
t
h
e
attr
ib
u
te
s
af
ter
d
ata
tr
an
s
f
o
r
m
atio
n
5
.
2
.
Select
io
n o
f
t
he
t
ra
ini
ng
a
nd
t
esting
da
t
a
s
et
Scik
it
-
lear
n
m
ac
h
in
e
lear
n
i
n
g
lib
r
ar
y
a
v
ailab
le
f
o
r
m
ac
h
i
n
e
l
ea
r
n
in
g
f
o
r
p
y
t
h
o
n
p
r
o
g
r
a
m
m
i
n
g
[
2
1
]
is
u
s
ed
to
s
p
lit
t
h
e
tr
a
n
s
f
o
r
m
ed
d
ataset
i
n
to
tr
ai
n
in
g
a
n
d
te
s
ti
n
g
d
ataset.
Ni
n
et
y
p
er
ce
n
t
o
f
t
h
e
d
ataset
i
s
u
s
ed
f
o
r
tr
ain
i
n
g
an
d
te
n
p
er
ce
n
t
o
f
th
e
d
ataset
is
u
s
ed
f
o
r
test
in
g
.
1
9
2
0
o
b
s
er
v
atio
n
s
ar
e
r
an
d
o
m
l
y
s
p
lit
as
tr
ain
i
n
g
d
ata
an
d
2
1
4
o
b
s
er
v
atio
n
s
a
r
e
s
p
lit as test
in
g
d
ata.
5
.
3
.
T
ra
ini
ng
o
f
t
he
ANN
T
en
s
o
r
Flo
w
is
a
m
ac
h
in
e
lea
r
n
in
g
p
lat
f
o
r
m
u
s
ed
f
o
r
tr
ain
in
g
n
e
u
r
al
n
et
w
o
r
k
s
a
n
d
it
is
an
o
p
en
s
o
u
r
ce
lib
r
ar
y
w
it
h
p
len
t
y
o
f
f
ast
n
u
m
er
ical
co
m
p
u
ti
n
g
al
g
o
r
ith
m
s
av
ailab
le
to
cr
ea
te
ANN
m
o
d
els
an
d
to
tr
ain
t
h
e
m
.
Ker
a
s
is
a
n
A
P
I
th
at
r
u
n
s
o
n
to
p
o
f
T
en
s
o
r
Flo
w
an
d
en
ab
les
f
a
s
t
ex
p
er
i
m
en
tati
o
n
to
f
i
n
e
t
u
n
e
a
n
d
tr
ain
A
NN
m
o
d
els.
T
r
ain
in
g
o
f
A
NN
in
th
i
s
r
esear
ch
is
ca
r
r
i
ed
o
u
t
u
s
in
g
Ker
as
a
n
d
T
en
s
o
r
Flo
w
a
s
it
h
elp
s
to
s
eq
u
en
tiall
y
b
u
ild
an
d
tes
t n
e
u
r
al
n
et
w
o
r
k
v
er
y
q
u
ic
k
l
y
w
it
h
m
i
n
i
m
al
li
n
es o
f
co
d
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
I
SS
N:
2088
-
8
694
R
o
ta
tin
g
b
l
a
d
e
fa
u
lts
cla
s
s
ific
a
tio
n
o
f
a
r
o
to
r
-
d
is
k
-
b
la
d
e
s
ystem
u
s
in
g
… (
A
b
d
u
lla
h
i A
b
u
b
a
ka
r
Ma
s
’
u
d
)
1907
A
d
a
m
o
p
ti
m
izer
(
a
d
ap
tiv
e
m
o
m
en
t
est
i
m
a
to
r
)
is
u
s
ed
f
o
r
u
p
g
r
ad
in
g
t
h
e
w
ei
g
h
ts
o
f
t
h
e
d
e
n
s
e
la
y
er
s
o
f
th
e
A
N
N
d
u
r
in
g
t
h
e
tr
ain
i
n
g
.
A
d
a
m
u
s
es
ad
ap
tiv
e
lear
n
in
g
r
ate
d
u
r
i
n
g
tr
ai
n
i
n
g
p
h
a
s
e
an
d
ch
a
n
g
es
t
h
e
in
d
iv
id
u
al
lear
n
in
g
r
ate
s
f
o
r
ea
ch
p
ar
a
m
eter
b
ased
o
n
s
q
u
ar
ed
g
r
ad
ien
ts
a
n
d
m
o
m
e
n
tu
m
to
e
n
ab
le
f
a
s
ter
co
n
v
er
g
e
n
ce
.
A
d
a
m
is
s
u
p
er
io
r
to
s
to
ch
asti
c
g
r
ad
ien
t
d
e
s
c
en
t
o
p
ti
m
izer
w
h
ic
h
u
s
e
s
a
le
ar
n
in
g
r
ate
t
h
at
d
o
es
n
o
t v
ar
y
d
u
r
in
g
th
e
i
ter
atio
n
s
an
d
h
en
ce
p
r
ef
er
r
ed
f
o
r
tr
ain
i
n
g
th
e
ANN.
Du
r
in
g
tr
ai
n
in
g
t
h
e
m
ea
n
s
q
u
ar
ed
er
r
o
r
(
MSE
)
lo
s
s
f
u
n
ctio
n
is
m
in
i
m
ized
.
T
h
e
m
ax
i
m
u
m
ep
o
ch
is
s
p
ec
if
ied
as
8
0
.
Ver
y
co
m
m
o
n
is
s
u
e
w
it
h
tr
ai
n
i
n
g
n
e
u
r
al
n
et
w
o
r
k
m
o
d
el
i
s
t
h
e
p
r
o
b
le
m
o
f
o
v
er
f
itti
n
g
.
T
o
o
v
er
co
m
e
t
h
is
p
r
o
b
lem
E
ar
l
y
S
to
p
p
in
g
f
ea
tu
r
e
av
aila
b
le
in
k
e
r
as
to
s
to
p
tr
ain
in
g
w
h
e
n
a
m
o
n
ito
r
ed
m
etr
ic
h
a
s
s
to
p
p
ed
i
m
p
r
o
v
in
g
h
as
b
ee
n
u
s
ed
.
T
h
e
p
atien
ce
ar
g
u
m
en
t
av
ailab
le
i
n
E
ar
l
y
Sto
p
p
in
g
f
u
n
ctio
n
is
s
et
to
f
i
v
e
w
h
ic
h
s
to
p
s
t
h
e
tr
ai
n
in
g
w
h
e
n
th
e
lo
s
s
f
u
n
ctio
n
h
as
n
o
t
s
ig
n
if
ica
n
tl
y
i
m
p
r
o
v
ed
o
v
er
5
iter
at
io
n
s
.
T
h
e
2
0
p
er
ce
n
t
o
f
t
h
e
tr
ain
i
n
g
d
ata
w
as
u
s
ed
f
o
r
v
alid
atio
n
to
te
s
t
t
h
e
m
o
d
el
d
u
r
in
g
tr
ain
i
n
g
.
T
h
e
c
o
n
v
er
g
e
n
ce
o
f
th
e
m
o
d
el
is
p
r
o
v
id
ed
in
T
ab
le
3
.
I
n
T
ab
le
3
,
MSE
,
M
A
E
,
V
AL
_
MSE
an
d
V
AL
_
M
A
E
r
ef
er
to
m
ea
n
s
q
u
ar
ed
er
r
o
r
d
u
r
in
g
tr
ai
n
i
n
g
,
m
ea
n
a
b
s
o
lu
te
er
r
o
r
d
u
r
in
g
tr
ai
n
i
n
g
,
m
ea
n
s
q
u
ar
ed
er
r
o
r
d
u
r
in
g
v
a
lid
atio
n
,
an
d
m
ea
n
ab
s
o
lu
te
er
r
o
r
d
u
r
in
g
v
alid
ati
o
n
r
esp
ec
tiv
e
l
y
.
As
s
ee
n
f
r
o
m
T
ab
le
3
th
e
m
o
d
el
co
n
v
er
g
ed
in
1
7
iter
atio
n
s
w
h
e
n
th
er
e
i
s
n
o
s
i
g
n
if
ican
t
i
m
p
r
o
v
e
m
e
n
t
in
m
s
e
d
u
r
i
n
g
tr
ain
i
n
g
.
T
h
e
ea
r
l
y
s
to
p
p
i
n
g
f
ea
t
u
r
e
u
s
ed
in
th
i
s
w
o
r
k
en
s
u
r
es
t
h
e
tr
ai
n
i
n
g
co
n
ti
n
u
es
u
n
t
il
t
h
e
er
r
o
r
s
ig
n
i
f
ica
n
tl
y
d
e
cr
ea
s
es
an
d
er
r
o
r
f
o
r
tr
ai
n
in
g
an
d
v
al
id
atio
n
d
ata
is
al
m
o
s
t
s
i
m
ilar
to
av
o
id
o
v
er
f
itti
n
g
.
T
h
e
m
s
e
d
u
r
i
n
g
tr
ain
i
n
g
a
n
d
v
alid
atio
n
i
s
al
m
o
s
t
t
h
e
s
a
m
e
at
th
e
en
d
o
f
th
e
1
7
th
i
ter
atio
n
s
w
h
ich
i
n
d
i
ca
tes
t
h
e
m
o
d
el
w
i
ll
n
o
t
o
v
er
f
it
f
o
r
u
n
s
ee
n
d
ata.
T
h
e
p
l
o
t
th
e
co
n
v
er
g
e
n
ce
ch
ar
ac
ter
is
tic
s
ar
e
s
h
o
w
n
i
n
Fi
g
u
r
e
7
(
a)
an
d
Fig
u
r
e
7
(
b
)
.
Fro
m
Fi
g
u
r
e
7
(
a)
an
d
Fi
g
u
r
e
7
(
b
)
,
it
is
e
v
id
e
n
t
th
at
d
u
r
i
n
g
tr
ai
n
i
n
g
o
f
A
NN,
t
h
e
MSE
an
d
M
A
E
r
ed
u
ce
s
to
t
h
e
lo
w
e
s
t p
o
s
s
ib
le
v
alu
e.
T
ab
le
3
.
C
o
n
v
er
g
e
n
ce
c
h
ar
ac
t
er
is
tics
o
f
t
h
e
m
o
d
el
Ep
o
c
h
M
S
E
M
A
E
V
A
L
_
M
S
E
V
A
L
_
M
A
E
0
1
8
.
7
1
1
6
8
3
2
.
6
4
9
9
9
3
0
.
3
5
8
5
4
9
0
.
4
3
9
0
9
7
1
0
.
2
3
1
8
9
1
0
.
3
6
9
9
2
7
0
.
1
7
1
9
9
6
0
.
3
2
2
5
3
8
2
0
.
1
4
3
5
3
7
0
.
2
9
0
1
9
9
0
.
1
2
9
4
3
4
0
.
2
7
5
0
0
2
3
0
.
1
2
9
2
6
4
0
.
2
7
4
9
0
2
0
.
1
2
7
0
6
0
.
2
7
0
7
3
5
-
-
-
-
-
-
-
-
-
-
13
0
.
0
6
4
6
4
7
0
.
1
8
9
0
0
5
0
.
0
5
4
9
0
6
0
.
1
7
2
8
1
14
0
.
0
6
0
7
3
9
0
.
1
8
2
1
8
4
0
.
0
4
8
8
2
2
0
.
1
6
3
7
7
3
15
0
.
0
6
6
7
0
6
0
.
1
9
3
2
1
8
0
.
0
6
8
1
9
0
.
2
0
8
1
5
1
16
0
.
0
6
0
9
8
4
0
.
1
8
2
1
8
1
0
.
0
5
7
5
8
0
.
1
8
2
9
2
5
17
0
.
0
6
4
3
2
9
0
.
1
8
9
4
0
6
0
.
0
6
1
7
1
2
0
.
1
9
5
2
6
3
(
a)
MSE
(
b
)
MA
E
Fig
u
r
e
7
.
P
lo
t o
f
(
a)
MSE
,
an
d
(
b
)
MA
E
d
u
r
in
g
tr
ai
n
i
n
g
o
f
ANN
5
.
4
.
ANN
predict
io
n r
es
ults
I
n
th
is
s
ec
tio
n
,
th
e
p
lo
t
o
f
th
e
ac
tu
al
v
al
u
e
an
d
th
e
p
r
ed
icted
v
alu
e
f
o
r
all
1
0
f
au
lts
ar
e
s
h
o
w
n
.
I
n
Fig
u
r
e
s
8
(
a)
-
(
j
)
,
th
e
tr
ain
ed
m
o
d
el
s
ar
e
ev
alu
ated
o
n
th
e
test
d
ata
an
d
th
e
p
lo
ts
f
o
r
th
e
1
0
attr
ib
u
tes
ar
e
s
h
o
w
n
.
T
h
e
p
r
ed
icted
v
alu
es
f
o
r
ea
ch
f
au
lt
ar
e
s
h
o
w
n
i
n
th
e
y
-
a
x
is
,
w
h
ile
th
e
ac
t
u
al
v
al
u
es
ar
e
s
h
o
w
n
i
n
x
-
ax
is
.
T
h
e
r
es
u
lt
s
h
o
w
s
t
h
at
ANN
ca
n
p
r
ed
ict
t
h
e
x
-
an
d
y
-
ax
es
la
ter
al
v
ib
r
atio
n
s
o
f
m
o
s
t
o
f
t
h
e
f
au
l
ts
v
er
y
clo
s
el
y
.
A
4
5
°
li
n
e
in
d
icate
s
a
p
r
ed
ictio
n
w
it
h
m
i
n
i
m
al
er
r
o
r
s
an
d
t
h
e
d
ata
p
o
in
ts
a
w
a
y
f
r
o
m
t
h
e
4
5
°
l
in
e
in
d
icate
s
th
e
er
r
o
r
in
p
r
ed
ictio
n
b
y
t
h
e
A
N
N.
Ge
n
er
all
y
,
th
e
r
esu
lt
s
i
m
p
l
y
t
h
at
t
h
e
ac
tu
al
a
n
d
p
r
ed
icted
v
al
u
es
f
o
r
m
o
s
t o
f
t
h
e
f
au
l
ts
ar
e
clo
s
e
l
y
s
i
m
ilar
w
it
h
w
id
er
v
ar
iatio
n
s
in
f
e
w
f
a
u
lt e
x
a
m
p
les.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2088
-
8
694
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
,
Vo
l.
12
,
No
.
3
,
Sep
tem
b
er
202
1
:
190
0
–
1
9
1
1
1908
(
a)
(
b
)
(
c)
(
d
)
(
e)
(
f
)
(
g
)
(
h
)
(
i)
(
j
)
Fig
u
r
e
8
.
P
lo
t o
f
ac
tu
al
an
d
p
r
ed
icted
v
alu
es
f
o
r
1
0
f
au
lt
s
,
(
a)
t
w
i
s
ted
x
,
(
b
)
t
w
i
s
ted
y
,
(
c)
b
en
tx
,
(
d
)
b
en
t
y
,
(
e)
b
u
lg
ed
x
,
(
f
)
b
u
lg
ed
y
,
(
g
)
cr
ac
k
ed
x
,
(
h
)
cr
ac
k
ed
y
,
(
i)
s
lo
tted
x
,
an
d
(
j
)
s
lo
tted
y
6.
CO
NCLU
SI
O
N
I
n
t
h
is
p
ap
er
,
th
e
ANN
h
as
b
ee
n
ap
p
lied
to
cla
s
s
i
f
y
d
i
f
f
er
e
n
t
f
a
u
lts
o
f
r
o
tatin
g
m
ac
h
i
n
es
.
First,
th
e
later
al
v
ib
r
atio
n
s
o
f
a
r
o
to
r
-
d
is
k
-
b
lad
e
d
ev
ice
co
n
s
is
ti
n
g
o
f
6
h
ea
lt
h
y
b
lad
es
w
er
e
an
al
y
ze
d
ex
p
er
i
m
e
n
tall
y
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
I
SS
N:
2088
-
8
694
R
o
ta
tin
g
b
l
a
d
e
fa
u
lts
cla
s
s
ific
a
tio
n
o
f
a
r
o
to
r
-
d
is
k
-
b
la
d
e
s
ystem
u
s
in
g
… (
A
b
d
u
lla
h
i A
b
u
b
a
ka
r
Ma
s
’
u
d
)
1909
an
d
th
r
o
u
g
h
f
in
i
te
ele
m
en
t
m
o
d
al
an
al
y
s
is
.
T
h
e
s
y
s
te
m
r
es
p
o
n
s
es
i
n
t
w
o
d
ir
ec
tio
n
s
w
er
e
o
b
tain
ed
at
1
0
0
0
r
p
m
.
Fi
v
e
f
au
lt
y
b
lad
es (
t
w
i
s
te
d
,
b
en
t,
b
u
l
g
ed
,
cr
ac
k
ed
,
an
d
s
lo
tted
)
w
er
e
r
ep
lace
d
i
n
t
u
r
n
b
y
o
n
e
o
f
th
e
6
s
a
f
e
b
lad
es.
C
o
m
p
ar
is
o
n
s
o
f
a
t
w
i
s
ted
b
lad
e's
r
esp
o
n
s
e
m
ag
n
it
u
d
es
w
it
h
t
h
o
s
e
o
f
h
ea
lt
h
y
b
l
ad
es
in
d
icate
th
a
t
t
h
e
v
ib
r
atio
n
a
m
p
li
tu
d
es
i
n
th
e
ac
ce
ler
atio
n
r
esp
o
n
s
e
r
an
g
e
o
f
t
h
e
f
ir
s
t
t
w
o
r
eso
n
an
t
f
r
eq
u
e
n
c
ies
d
ec
r
ea
s
ed
,
w
h
i
le
th
e
a
m
p
li
tu
d
e
n
ea
r
er
to
th
e
r
u
n
n
i
n
g
s
p
ee
d
in
cr
ea
s
ed
.
Af
ter
o
b
tain
i
n
g
t
h
e
d
ata
o
n
th
e
af
o
r
e
m
en
tio
n
ed
r
o
tatio
n
al
m
ac
h
in
e
f
au
lts
,
t
h
e
ANN
w
as
u
tili
ze
d
to
b
e
ab
le
to
class
i
f
y
an
d
d
i
s
cr
i
m
i
n
ate
t
h
e
x
-
an
d
y
-
ax
is
later
al
v
ib
r
atio
n
s
o
f
th
e
r
o
to
r
.
T
h
e
ANN
w
a
s
co
d
ed
in
p
y
t
h
o
n
s
o
f
t
w
ar
e.
I
n
itiall
y
,
co
r
r
elatio
n
b
et
w
ee
n
t
h
e
attr
ib
u
te
s
w
a
s
co
n
s
id
er
ed
an
d
it
w
as
s
h
o
w
n
th
at
th
er
e
is
a
p
o
s
itiv
e
co
r
r
elatio
n
b
et
w
ee
n
t
h
e
i
n
p
u
t
an
d
o
u
tp
u
t
attr
ib
u
te
s
.
Fo
r
tr
ain
i
n
g
t
h
e
ANN,
9
0
%
o
f
t
h
e
d
ata
s
et
i
s
u
s
ed
an
d
1
0
%
o
f
t
h
e
d
atase
t
is
u
s
e
d
f
o
r
test
i
n
g
.
T
h
e
r
esu
lt
s
h
o
ws
th
a
t
A
NN
ca
n
v
er
y
ac
cu
r
ate
l
y
p
r
ed
ict
th
e
t
w
o
-
d
i
m
en
s
io
n
al
(
x
-
a
n
d
y
-
a
x
e
s
)
l
ater
al
v
ib
r
atio
n
s
o
f
m
o
s
t
o
f
t
h
e
f
a
u
lts
.
T
h
e
r
es
u
lts
i
m
p
l
y
t
h
at
t
h
e
A
NN
ca
n
ef
f
ec
tiv
e
l
y
r
ec
o
g
n
ize
an
d
d
is
cr
i
m
i
n
ate
d
if
f
er
en
t
later
al
v
ib
r
a
tio
n
in
a
r
o
to
r
d
ep
en
d
in
g
o
n
tr
ain
in
g
an
d
test
i
n
g
w
it
h
d
i
f
f
er
en
t
d
ata
s
a
m
p
les
o
f
th
e
f
a
u
lts
t
y
p
e
s
.
F
u
r
th
er
r
ese
ar
ch
co
n
ce
n
tr
ates
o
n
t
h
e
ap
p
li
ca
tio
n
o
f
th
e
A
NN
f
o
r
p
r
ac
tical
f
au
lt d
etec
tio
n
an
d
class
if
icatio
n
.
ACK
NO
WL
E
D
G
E
M
E
NT
S
A
c
k
n
o
w
led
g
e
m
en
t
s
ar
e
d
u
e
to
J
u
b
ail
I
n
d
u
s
tr
ial
C
o
lle
g
e
an
d
J
u
b
ail
Un
iv
er
s
it
y
C
o
lle
g
e,
R
o
y
al
C
o
m
m
is
s
io
n
f
o
r
J
u
b
ail
.
RE
F
E
R
E
NC
E
S
[1
]
S
.
A
.
A
d
e
w
u
si
a
n
d
B.
O.
A
l
-
Be
d
o
o
r,
“
W
a
v
e
let
a
n
a
l
y
sis
o
f
v
ib
ra
ti
o
n
sig
n
a
ls
o
f
a
n
o
v
e
rh
a
n
g
ro
to
r
w
it
h
a
p
ro
p
a
g
a
ti
n
g
tran
sv
e
rse
c
ra
c
k
,
”
J
.
S
o
u
n
d
Vi
b
.
,
v
o
l
.
2
4
6
,
n
o
.
5
,
p
p
.
7
7
7
-
7
9
3
,
2
0
0
1
,
d
o
i:
1
0
.
1
0
0
6
/
jsv
i.
2
0
0
0
.
3
6
1
1
.
[2
]
J.
De
r
J
e
n
g
,
L
.
Hs
u
,
C.
W
.
Hu
n
,
a
n
d
C.
Y.
Ch
o
u
,
“
Id
e
n
t
if
ica
ti
o
n
f
o
r
b
i
f
u
rc
a
ti
o
n
a
n
d
re
sp
o
n
se
s
o
f
ru
b
-
im
p
a
c
ti
n
g
ro
to
r
sy
ste
m
,
”
Pro
c
e
d
ia
En
g
.
,
v
o
l
.
7
9
,
p
p
.
3
6
9
-
3
7
7
,
2
0
1
4
,
d
o
i:
1
0
.
1
0
1
6
/j
.
p
ro
e
n
g
.
2
0
1
4
.
0
6
.
3
5
7
.
[3
]
S
.
A
.
A
d
e
w
u
si
a
n
d
B.
O.
A
l
-
Be
d
o
o
r,
“
Ex
p
e
rim
e
n
tal
stu
d
y
o
n
th
e
v
ib
ra
ti
o
n
o
f
a
n
o
v
e
rh
u
n
g
ro
to
r
w
it
h
a
p
ro
p
a
g
a
ti
n
g
tran
sv
e
rse
c
ra
c
k
,
”
S
h
o
c
k
Vi
b
.
,
v
o
l
.
9
,
p
p
.
9
1
-
1
0
4
,
2
0
0
2
,
d
o
i:
1
0
.
1
1
5
5
/2
0
0
2
/4
0
5
9
2
8
.
[4
]
C.
P
.
L
a
w
so
n
a
n
d
P
.
C
.
Iv
e
y
,
“
T
u
b
o
m
a
c
h
in
e
r
y
b
lad
e
v
ib
ra
ti
o
n
a
m
p
li
tu
d
e
m
e
a
su
re
m
e
n
t
th
ro
u
g
h
ti
p
ti
m
in
g
w
it
h
c
a
p
a
c
it
a
n
c
e
ti
p
c
lea
ra
n
c
e
p
ro
b
e
s,”
S
e
n
so
rs
Actu
a
to
rs
,
A
P
h
y
s.
,
v
o
l.
1
1
8
,
n
o
.
1
,
p
p
.
1
4
-
2
4
,
2
0
0
5
,
d
o
i:
1
0
.
1
0
1
6
/
S
0
9
2
4
-
4
2
4
7
(0
4
)0
0
4
8
2
-
0.
[5
]
B.
O.
A
l
-
Be
d
o
o
r,
L
.
G
h
o
u
ti
,
S
.
A
.
A
d
e
w
u
si,
Y.
A
l
-
N
a
ss
a
r,
a
n
d
M
.
A
b
d
lsa
m
a
d
,
“
Ex
p
e
rime
n
ts
o
n
th
e
e
x
trac
ti
o
n
o
f
b
lad
e
v
ib
ra
ti
o
n
sig
n
a
tu
re
f
ro
m
th
e
sh
a
f
t
to
rsio
n
a
l
v
ib
ra
ti
o
n
sig
n
a
ls
,
”
J
.
Qu
a
l.
M
a
in
t
.
E
n
g
.
,
v
o
l.
9
,
n
o
.
2
,
p
p
.
1
4
4
-
1
5
9
,
2
0
0
3
,
d
o
i:
1
0
.
1
1
0
8
/
1
3
5
5
2
5
1
0
3
1
0
4
8
2
3
9
8
.
[6
]
B.
O.
A
l
-
Be
d
o
o
r,
Y.
A
l
-
Na
ss
a
r,
L
.
G
h
o
u
ti
,
S
.
A
.
A
d
e
w
u
si,
a
n
d
M
.
A
b
d
lsa
m
a
d
,
“
S
h
a
f
t
late
ra
l
a
n
d
to
rsio
n
a
l
v
ib
ra
ti
o
n
re
sp
o
n
se
s t
o
b
lad
e
(s) ran
d
o
m
v
ib
ra
ti
o
n
e
x
c
it
a
ti
o
n
,
”
Ar
a
b
.
J
.
S
c
i.
En
g
.
,
v
o
l
.
2
9
,
n
o
.
1
C
,
p
p
.
3
9
-
6
7
,
2
0
0
4
.
[7
]
A
.
A
.
G
u
b
ra
n
a
n
d
J.
K.
S
i
n
h
a
,
“
S
h
a
f
t
in
sta
n
tan
e
o
u
s
a
n
g
u
lar
sp
e
e
d
f
o
r
b
lad
e
v
ib
ra
ti
o
n
i
n
r
o
tatin
g
m
a
c
h
in
e
,
”
M
e
c
h
.
S
y
st.
S
i
g
n
a
l
Pr
o
c
e
ss
.
,
v
o
l.
4
4
,
n
o
.
1
-
2
,
p
p
.
4
7
-
5
9
,
2
0
1
4
,
d
o
i:
1
0
.
1
0
1
6
/j
.
y
m
ss
p
.
2
0
1
3
.
0
2
.
0
0
5
.
[8
]
I.
F
.
S
a
n
t
o
s,
C.
M
.
S
a
ra
c
h
o
,
J.
T
.
S
m
it
h
,
a
n
d
J.
Ei
lan
d
,
“
Co
n
tri
b
u
ti
o
n
to
e
x
p
e
r
im
e
n
tal
v
a
li
d
a
ti
o
n
o
f
l
in
e
a
r
a
n
d
n
o
n
-
li
n
e
a
r
d
y
n
a
m
ic
m
o
d
e
ls
f
o
r
re
p
re
s
e
n
ti
n
g
ro
t
o
r
-
b
lad
e
p
a
ra
m
e
tri
c
c
o
u
p
led
v
ib
ra
ti
o
n
s,”
J
.
S
o
u
n
d
Vi
b
.
,
v
o
l.
2
7
1
,
n
o
.
3
-
5
,
p
p
.
8
8
3
-
9
0
4
,
2
0
0
4
,
d
o
i:
1
0
.
1
0
1
6
/
S
0
0
2
2
-
4
6
0
X
(0
3
)0
0
7
5
8
-
2.
[9
]
R.
L
iu
,
B.
Ya
n
g
,
E.
Zi
o
,
a
n
d
X
.
Ch
e
n
,
“
A
rti
f
icia
l
in
telli
g
e
n
c
e
f
o
r
f
a
u
lt
d
iag
n
o
sis
o
f
ro
tatin
g
m
a
c
h
in
e
r
y
:
A
re
v
i
e
w
,
”
M
e
c
h
.
S
y
st.
S
i
g
n
a
l
Pr
o
c
e
ss
.
,
v
o
l.
1
0
8
,
p
p
.
3
3
-
4
7
,
2
0
1
8
,
d
o
i:
1
0
.
1
0
1
6
/j
.
y
m
ss
p
.
2
0
1
8
.
0
2
.
0
1
6
.
[1
0
]
Y.
L
e
i,
B.
Ya
n
g
,
X
.
Jia
n
g
,
F
.
Ji
a
,
N.
L
i,
a
n
d
A
.
K.
Na
n
d
i,
“
A
p
p
li
c
a
ti
o
n
s
o
f
m
a
c
h
in
e
lea
rn
in
g
t
o
m
a
c
h
in
e
fa
u
lt
d
iag
n
o
sis:
A
re
v
ie
w
a
n
d
ro
a
d
m
a
p
,
”
M
e
c
h
.
S
y
st.
S
ig
n
a
l
Pro
c
e
ss
.
,
v
o
l.
1
3
8
,
1
0
6
5
8
7
,
2
0
2
0
,
d
o
i:
1
0
.
1
0
1
6
/
j.
y
m
ss
p
.
2
0
1
9
.
1
0
6
5
8
7
.
[1
1
]
R.
V
.
S
á
n
c
h
e
z
,
P
.
L
u
c
e
ro
,
R.
E.
V
á
sq
u
e
z
,
M
.
Ce
rra
d
a
,
J.
C.
M
a
c
a
n
c
e
la,
a
n
d
D.
Ca
b
re
ra
,
“
F
e
a
tu
re
ra
n
k
in
g
f
o
r
m
u
lt
i
-
f
a
u
lt
d
iag
n
o
sis
o
f
ro
tati
n
g
m
a
c
h
in
e
ry
b
y
u
sin
g
ra
n
d
o
m
f
o
re
st
a
n
d
KN
N,”
J
.
in
tell
.
Fu
zz
y
S
y
st
.
,
v
o
l
.
3
4
,
p
p
.
3
4
6
3
-
3
4
7
3
,
2
0
1
8
,
d
o
i:
1
0
.
3
2
3
3
/JIF
S
-
1
6
9
5
2
6
.
[1
2
]
M
.
G
o
h
a
ri
a
n
d
A
.
M
.
Ey
d
i,
“
M
o
d
e
ll
in
g
o
f
sh
a
f
t
u
n
b
a
lan
c
e
:
M
o
d
e
ll
i
n
g
a
m
u
lt
i
d
isc
s ro
to
r
u
sin
g
K
-
Ne
a
re
st Ne
ig
h
b
o
r
a
n
d
De
c
isio
n
T
re
e
A
lg
o
rit
h
m
s,”
M
e
a
su
re
me
n
t
,
v
o
l.
1
5
1
,
1
0
7
2
5
3
,
2
0
2
0
,
d
o
i:
1
0
.
1
0
1
6
/
j.
m
e
a
su
re
m
e
n
t.
2
0
1
9
.
1
0
7
2
5
3
.
[1
3
]
A
.
E.
P
r
o
sv
iri
n
,
F
.
P
i
lt
a
n
,
a
n
d
J.
M
.
Kim
,
“
Blad
e
ru
b
-
im
p
a
c
t
f
a
u
lt
id
e
n
ti
f
ica
ti
o
n
u
sin
g
a
u
t
o
e
n
c
o
d
e
r
-
b
a
se
d
n
o
n
li
n
e
a
r
f
u
n
c
ti
o
n
a
p
p
r
o
x
im
a
ti
o
n
a
n
d
a
d
e
e
p
n
e
u
ra
l
n
e
tw
o
rk
,
”
S
e
n
so
rs
,
v
o
l
.
2
0
,
6
2
6
5
,
2
0
2
0
,
d
o
i:
1
0
.
3
3
9
0
/s2
0
2
1
6
2
6
5
.
[1
4
]
B.
Zh
o
n
g
,
J.
M
a
c
In
ty
re
,
Y.
He
,
a
n
d
J.
T
a
it
,
“
Hig
h
o
rd
e
r
n
e
u
ra
l
n
e
tw
o
rk
s
f
o
r
si
m
u
lt
a
n
e
o
u
s
d
iag
n
o
sis
o
f
m
u
lt
ip
le
f
a
u
lt
s in
ro
tati
n
g
m
a
c
h
in
e
s,”
Ne
u
r
a
l
Co
m
p
u
t
.
A
p
p
l.
,
v
o
l.
8
,
p
p
.
1
8
9
-
1
9
5
,
1
9
9
9
,
d
o
i:
1
0
.
1
0
0
7
/s
0
0
5
2
1
0
0
5
0
0
2
1
.
[1
5
]
H.
Na
h
v
i
a
n
d
M
.
Esf
a
h
a
n
ian
,
“
F
a
u
lt
id
e
n
ti
f
ica
ti
o
n
i
n
ro
tati
n
g
m
a
c
h
in
e
ry
u
sin
g
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
s,”
Pro
c
.
In
st.
M
e
c
h
.
E
n
g
.
Pa
rt C
J
.
M
e
c
h
.
En
g
.
S
c
i.
,
v
o
l.
2
1
9
,
n
o
.
2
,
p
p
.
1
4
1
-
1
5
8
,
2
0
0
5
,
d
o
i:
1
0
.
1
2
4
3
/
0
9
5
4
4
0
6
0
5
X8
4
6
9
.
[1
6
]
S
.
A
.
A
d
e
w
u
si
a
n
d
B
.
O.
A
l
-
Be
d
o
o
r,
“
De
tec
ti
o
n
o
f
p
r
o
p
a
g
a
ti
n
g
c
ra
c
k
s
in
r
o
to
rs
u
sin
g
n
e
u
ra
l
n
e
t
w
o
rk
s,”
Pro
c
.
AS
M
E
Pi
p
e
C
o
mp
.
An
lys
.
Dia
g
.
Co
n
f
.
,
v
o
l.
4
4
7
,
n
o
.
P
VP2
0
0
2
-
1
5
1
8
,
p
p
.
7
1
-
7
8
,
2
0
0
2
,
d
o
i:
1
0
.
1
1
1
5
/
P
VP2
0
0
2
-
1
5
1
8
.
[1
7
]
C.
D.
Ng
u
y
e
n
,
A
.
E.
P
r
o
sv
iri
n
,
C
.
H.
Kim
,
a
n
d
J.
M
.
Kim
,
“
Co
n
str
u
c
ti
o
n
o
f
a
se
n
siti
v
e
a
n
d
sp
e
e
d
in
v
a
rian
t
g
e
a
rb
o
x
f
a
u
lt
d
iag
n
o
s
is
m
o
d
e
l
u
sin
g
a
n
in
c
o
rp
o
ra
ted
u
ti
li
z
in
g
a
d
a
p
t
iv
e
n
o
is
e
c
o
n
tro
l
a
n
d
a
sta
c
k
e
d
sp
a
rse
a
u
to
e
n
c
o
d
e
r
-
b
a
se
d
d
e
e
p
n
e
u
ra
l
n
e
tw
o
rk
,
”
S
e
n
so
rs
,
v
o
l.
2
1
,
n
o
.
1
,
1
4
2
4
-
8
2
2
0
,
2
0
2
1
,
d
o
i:
1
0
.
3
3
9
0
/s2
1
0
1
0
0
1
8
.
[1
8
]
N.
S
.
V
y
a
s an
d
D.
S
a
ti
sh
k
u
m
a
r,
“
A
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
d
e
sig
n
fo
r
f
a
u
lt
id
e
n
ti
f
ica
ti
o
n
in
a
r
o
to
r
-
b
e
a
rin
g
s
y
ste
m
,
”
M
e
c
h
.
M
a
c
h
.
T
h
e
o
ry
,
v
o
l.
3
6
,
n
o
.
2
,
p
p
.
1
5
7
-
1
7
5
,
2
0
0
1
,
d
o
i:
1
0
.
1
0
1
6
/S
0
0
9
4
-
1
1
4
X
(
0
0
)
0
0
0
3
4
-
3.
Evaluation Warning : The document was created with Spire.PDF for Python.