T
EL
K
O
M
N
I
K
A
,
V
ol
.
15,
N
o
.
3,
S
ept
em
be
r
2017,
pp.
97
7
~
983
I
SSN
:
1
6
9
3
-
6
930
,
ac
c
r
edi
ted
A
b
y
D
IK
T
I
,
D
e
c
r
e
e
N
o
:
58/
D
I
K
T
I
/K
ep/
2013
D
O
I:
10.12928/
T
E
LK
O
M
N
I
K
A
.
v
1
5
i
3
.
5149
977
R
e
c
ei
v
ed
M
ay
2
5
,
201
7
;
R
ev
i
s
ed
Au
g
u
s
t
1
8
,
201
7
;
A
c
c
ept
ed
A
u
gus
t
3
0
,
201
7
T
h
e
C
o
r
r
e
l
a
ti
o
n
o
f
S
t
at
i
st
i
cal
I
m
a
g
e
a
n
d
Pa
rti
a
l
D
i
s
c
h
a
rg
e
Pu
l
s
e
C
o
u
n
t
o
f L
D
P
E
-
N
R
C
o
m
p
o
s
i
te
A
u
l
i
a
*
1
,
Zul
k
urna
i
n
A
bd
u
l
-
M
al
ek
2
,
Y
a
nua
r
Zul
a
rdi
a
n
y
s
a
h
A
ri
e
f
3
,
Ek
a
Pu
t
r
a
W
a
l
d
i
4
1
,4
J
ur
us
an T
ek
ni
k
E
l
ek
t
r
o,
U
n
i
v
er
s
i
t
as
A
nd
al
as
,
P
adan
g,
I
ndon
es
i
a
1,
2
,
3
I
ns
t
i
t
ut
V
ol
t
an da
n A
r
us
T
i
ngg
i
,
U
ni
v
er
s
i
t
i
T
ek
nol
o
gi
M
al
a
y
s
i
a,
J
oh
or
,
M
al
a
y
s
i
a
*
C
or
r
es
po
ndi
n
g aut
hor
,
e
-
ma
i
l
:
aul
i
a.
un
and
@
gm
ai
l
.
c
om
A
b
st
r
a
ct
H
i
gh v
ol
t
a
ge i
n
s
ul
at
i
on
m
us
t
b
e des
i
gn
ed i
n s
uc
h a w
ay
t
hat
i
t
i
s
v
er
y
r
es
i
s
t
ant
t
o ag
ei
n
g i
nc
l
u
di
n
g
t
hat
f
r
om
par
t
i
al
di
s
c
har
g
e (
P
D
)
.
M
any
s
t
udi
es
w
er
e pr
ev
i
ous
l
y
c
ar
r
i
e
d out
on c
o
m
pos
i
t
es
bas
e
d on l
ow
dens
i
t
y
pol
y
et
h
y
l
en
e (
LD
P
E
)
.
H
ow
ev
er
,
t
he u
s
e of
nat
u
r
a
l
r
u
bber
(
N
R
)
an
d nan
os
i
l
i
c
a (
S
i
O
2
)
i
n t
he LD
P
E
-
N
R
bas
ed c
om
pos
i
t
es
i
s
r
el
at
i
v
el
y
new
.
F
ur
t
her
m
or
e,
t
he P
D
r
es
i
s
t
ant
p
er
f
or
m
a
nc
e
of
t
he c
o
m
p
os
i
t
es
i
s
y
et
t
o
be
ex
t
e
n
s
i
v
el
y
r
es
ear
c
h
ed.
T
hi
s
w
or
k
ai
m
s
t
o
an
al
y
z
e
t
he
c
or
r
e
l
at
i
o
n
bet
w
een
P
D
pul
s
e
c
ou
nt
an
d
i
t
s
r
el
at
ed
i
m
a
ge t
o
i
nt
er
pr
et
i
ng t
h
e
ef
f
ec
t
o
f
P
D
s
i
gn
al
s
.
T
he r
es
ul
t
s
s
h
ow
t
her
e
i
s
a
s
t
r
ong c
or
r
el
at
i
o
n
bet
w
een P
D
p
ul
s
e c
ount
and
t
he
s
ta
ti
st
i
ca
l
i
m
ag
e.
T
he r
e
s
ul
t
s
i
n
di
c
at
e t
hat
t
he s
ur
f
ac
e i
m
a
ge s
t
at
i
s
t
i
c
a
l
ana
l
y
s
i
s
c
a
n
b
e
us
e
d
as
a
t
o
ol
t
o
j
us
t
i
f
y
t
h
e
t
ot
al
of
t
he
P
D
pul
s
e
c
ou
nt
on
t
he
s
ur
f
ac
e
f
or
di
f
f
er
ent
s
a
m
p
l
es
of
c
om
pos
i
t
e.
K
eyw
o
r
d
s
:
p
a
r
t
i
al
d
i
s
c
h
ar
ge,
s
t
at
i
s
t
i
c
al
c
or
r
el
at
i
on,
l
ow
d
ens
i
t
y
pol
y
et
h
y
l
en
e,
nat
ur
al
r
ubb
er
,
s
ur
f
a
c
e
i
m
ag
e
C
o
p
y
r
i
g
h
t
©
2017 U
n
i
ve
r
si
t
a
s A
h
m
ad
D
ah
l
an
.
A
l
l
r
i
g
h
t
s
r
ese
r
ved
.
1
.
I
nt
roduc
t
i
on
D
i
f
f
er
ent
m
et
hod of
t
he P
D
anal
y
s
i
s
p
r
ogr
a
m
i
s
r
ec
ent
l
y
us
ed b
y
t
he ex
per
t
i
s
e i
n P
D
f
ie
l
d
.
F
or
ex
a
m
pl
e,
M
ont
e
C
ar
l
o
m
et
ho
d
w
a
s
us
ed
t
o
ev
al
uat
e
t
he
P
D
di
s
t
r
i
but
i
o
n
t
o
get
t
he
P
D
pul
s
es
p
ha
s
e
and
t
hei
r
r
el
a
t
ed
am
pl
i
t
ud
e
[
1
]
,
t
h
e
G
ene
t
i
c
al
go
r
i
t
hm
w
a
s
us
ed
t
o
r
c
og
ni
z
e
t
he
P
D
pat
t
er
n
i
n
g
a
s
i
ns
ul
at
e
d
s
w
i
t
c
hge
ar
(
G
I
S
)
[
2
]
and
t
he
s
t
at
i
s
t
i
c
al
an
al
y
s
i
s
of
P
D
p
a
t
t
er
n
w
a
s
u
s
e
d
t
o
as
s
e
s
s
t
h
e i
n
s
ul
at
i
o
n
s
y
s
t
em
i
n
hi
gh v
ol
t
ag
e
[
3
]
.
T
he
s
e
m
et
hod
s
c
ont
r
i
but
e di
f
f
e
r
ent
l
y
t
o
enha
nc
e
t
he
P
D
anal
y
s
i
s
a
ppr
o
ac
h
i
n
or
der
un
de
r
s
t
a
nd
t
he
P
D
ph
enom
e
na.
T
he
w
ay
t
o
p
r
es
ent
t
he
P
D
d
at
a
i
s
a
no
t
her
i
m
por
t
a
n
t
par
t
of
t
he
P
D
an
al
y
s
i
s
.
T
he
m
os
t
c
om
m
on
w
ay
t
o
p
r
e
s
ent
t
h
e
P
D
dat
a
i
s
t
he
g
r
aph
of
ϕ
-
q
-
N
pat
t
e
r
n
t
hat
r
epr
es
e
nt
i
ng
t
he
ph
a
s
e
angl
e
of
P
D
o
c
c
ur
r
en
c
e
(
ϕ
)
,
t
he
P
D
appa
r
ent
c
h
ar
ge
(
q
)
and
t
he
p
ul
s
e
s
'
num
ber
of
P
D
(
N
)
[
4
,
5
]
.
T
hi
s
g
r
ap
h i
s
c
al
l
ed ph
a
s
e r
es
ol
v
ed
P
D
(
P
R
P
D
)
pat
t
er
n
t
hat
i
s
c
u
r
r
ent
l
y
us
e
d by
m
a
ny
r
es
ea
r
c
her
s
t
o ev
al
uat
e t
h
e P
D
dat
a.
U
s
i
ng t
hi
s
g
r
a
ph,
t
he pha
s
e angl
e
of
P
D
c
an be
ea
s
i
l
y
r
ec
ogni
z
e
d
a
s
w
el
l
as
t
h
e
appa
r
ent
P
D
c
ha
r
ge
and
n
um
be
r
of
P
D
pul
s
es
i
n
ev
e
r
y
pha
s
e
a
ngl
e.
A
l
t
hough
t
h
e
num
be
r
of
P
D
p
ul
s
e
c
oun
t
i
s
o
ne
p
ar
a
m
et
er
of
P
D
pha
s
e
r
e
s
ol
v
e
an
al
y
s
i
s
,
b
ut
t
he
det
ai
l
anal
y
s
i
s
P
D
pul
s
e
c
ount
and
i
t
s
e
f
f
ec
t
t
o
t
he
phy
s
i
c
al
p
r
op
er
t
i
es
of
i
ns
ul
at
i
on
m
at
er
i
al
w
as
not
di
s
c
u
s
s
ed
i
n
dept
h.
T
he u
s
e of
s
t
at
i
s
t
i
c
al
an
al
y
s
i
s
t
o an
al
y
z
e t
he P
D
d
at
a i
s
v
er
y
u
s
ef
ul
t
o qu
an
t
i
f
y
t
he
par
t
i
al
di
s
c
ha
r
ge pa
r
am
et
e
r
s
t
o enh
an
c
e t
he P
D
pat
t
er
n
[
5
-
7]
.
A
n anal
y
s
i
s
of
t
he k
u
r
t
os
i
s
a
nd
t
h
e
s
k
ew
ne
s
s
c
o
nf
i
r
m
e
d
t
h
at
po
s
i
t
i
v
e
an
d
ne
gat
i
v
e
po
l
ar
i
t
y
had
c
ou
nt
er
c
l
oc
k
w
i
s
e
and
c
l
oc
k
w
i
s
e
c
l
u
s
t
er
i
n
g
di
s
t
r
i
but
i
on,
r
e
s
pec
t
i
v
el
y
.
T
h
r
ee P
D
pa
r
a
m
et
er
s
w
hi
c
h
ar
e
t
he pul
s
e
du
r
at
i
on,
p
ul
s
e
m
agni
t
ud
e
an
d
pul
s
e
p
ha
s
e
angl
e
w
er
e
al
s
o
i
nt
r
od
uc
ed
f
or
s
t
at
i
s
t
i
c
al
ev
al
uat
i
o
n
of
P
D
dat
a
[
8
]
.
T
he pul
s
e d
ur
at
i
on i
s
t
o c
or
r
el
at
e t
o t
he di
s
c
h
ar
g
e
m
ec
ha
ni
s
m
w
i
t
h t
he c
o
n
di
t
i
on of
s
ur
f
ac
e
r
es
i
s
t
an
c
e.
T
he pul
s
e m
a
gni
t
ude i
s
t
o
c
or
r
el
at
e t
he
v
ol
t
age s
t
r
es
s
i
n t
he di
s
c
h
ar
gi
n
g c
av
i
t
y
a
s
t
he i
ndi
c
at
i
o
n
of
t
he
f
or
m
a
t
i
on of
s
u
r
f
ac
e c
ha
r
ge
s
an
d s
ur
f
a
c
e
c
o
ndu
c
t
i
v
i
t
y
.
T
he pul
s
e
p
has
e
angl
e i
s
t
o c
or
r
el
at
e t
he
det
er
i
o
r
at
i
o
n ene
r
gy
w
i
t
h t
he nat
ur
e
of
phy
s
i
c
al
c
han
ge
s
i
n
t
he
di
s
c
ha
r
ge.
I
ns
ul
at
i
o
n
det
er
i
o
r
at
i
on
i
s
a
l
ong
-
t
e
r
m
pr
oc
e
s
s
t
hat
oc
c
ur
s
ov
e
r
a
pr
ol
on
ged
o
pe
r
at
i
ona
l
c
on
di
t
i
on.
S
o
i
t
i
s
not
un
c
o
m
m
on
t
h
at
i
f
t
he
i
n
s
ul
at
i
o
n
s
y
s
t
em
ha
s
a
s
t
abl
e
P
D
a
c
t
i
v
i
t
y
(
ev
en
w
i
t
h
a hi
gh P
D
l
e
v
el
)
,
i
ns
ul
at
i
o
n f
ai
l
ur
e m
ay
not
oc
c
ur
e
a
r
l
y
.
I
f
a s
i
gni
f
i
c
ant
i
n
c
r
e
as
e i
n P
D
ac
t
i
v
i
t
y
w
i
t
hi
n a
c
er
t
ai
n pe
r
i
o
d o
f
t
i
m
e unde
r
t
he s
am
e t
es
t
c
ondi
t
i
o
n
s
i
s
det
e
c
t
e
d,
i
t
w
a
r
r
a
nt
ed
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN
:
1
6
9
3
-
6
930
T
E
L
K
O
M
NI
K
A
Vo
l
.
15
,
N
o
.
3,
S
ept
em
ber
20
17
:
977
–
9
83
978
i
nv
es
t
i
gat
i
o
n
and
i
s
a
w
ar
n
i
ng
s
i
gn
of
p
ot
ent
i
al
i
ni
t
i
at
i
on
of
i
n
s
ul
at
i
on
f
ai
l
ur
e.
A
n
i
nc
r
ea
s
e
of
P
D
ac
t
i
v
i
t
y
r
at
he
r
t
han
an
a
bs
ol
ut
e
P
D
m
a
gni
t
ude
c
an
i
ndi
c
at
e
s
ev
er
e
pr
oba
bl
y
i
n
s
ul
at
i
o
n
f
ai
l
u
r
e.
A
s
t
abl
e
P
D
a
c
t
i
v
i
t
y
ov
er
t
i
m
e i
ndi
c
at
e
s
t
h
at
i
ns
ul
at
i
on
det
er
i
o
r
at
i
o
n has
not
pr
ogr
es
s
ed
m
u
c
h.
A
s
i
gni
f
i
c
ant
i
n
c
r
e
as
e i
n P
D
a
c
t
i
v
i
t
y
w
i
t
hi
n a r
el
at
i
v
el
y
s
ho
r
t
pe
r
i
od of
t
i
m
e (
e.
g.
s
i
x
m
on
t
hs
)
i
ndi
c
at
es
t
h
a
t
t
he r
at
e of
i
ns
ul
at
i
on d
et
er
i
o
r
at
i
on
a
nd t
he l
i
k
el
i
h
ood of
m
a
c
h
i
ne f
ai
l
ur
e h
as
i
nc
r
ea
s
ed
[
9,
10]
.
A
n
i
m
age
h
i
s
t
ogr
am
i
s
t
h
e
gr
a
phi
c
al
r
ep
r
e
s
ent
at
i
on
o
f
t
h
e
i
nt
ens
i
t
y
di
s
t
r
i
b
ut
i
on
of
a
d
i
g
i
t
a
l ima
g
e
.
I
t
pl
ot
s
t
he nu
m
be
r
o
f
p
i
x
e
ls
f
or
e
ac
h i
nt
en
s
i
t
y
v
al
ue.
T
he
hor
i
z
ont
al
ax
i
s
o
f
t
h
e
gr
ap
h
r
e
p
r
es
ent
s
t
h
e
t
onal
v
ar
i
at
i
on
s
,
w
hi
l
e
t
he
v
er
t
i
c
al
ax
i
s
r
ep
r
es
ent
s
t
h
e
n
um
be
r
of
pi
x
el
s
i
n
t
hat
par
t
i
c
ul
ar
t
on
e.
T
h
e
l
ef
t
s
i
de
of
t
he
hor
i
z
ont
al
ax
i
s
r
ep
r
e
s
e
nt
s
t
he
bl
a
c
k
and
da
r
k
ar
e
as
,
t
he m
i
ddl
e
r
e
pr
e
s
ent
s
m
e
di
um
g
r
ey
an
d t
he
r
i
ght
h
and
s
i
d
e r
ep
r
es
ent
s
l
i
ght
and pur
e w
hi
t
e
ar
ea
s
.
T
he v
er
t
i
c
al
ax
i
s
r
e
pr
e
s
ent
s
t
he
s
i
z
e of
t
h
e a
r
ea t
hat
i
s
c
a
pt
ur
ed
i
n
ea
c
h on
e of
t
h
e
s
e
z
on
es
.
T
h
us
,
t
he
hi
s
t
og
r
am
f
or
a
v
er
y
da
r
k
i
m
a
ge
w
i
l
l
hav
e
t
he
m
aj
or
i
t
y
of
i
t
s
dat
a
poi
nt
s
on
t
h
e
l
ef
t
s
i
de an
d c
ent
e
r
of
t
he
gr
ap
h.
C
onv
er
s
el
y
,
t
he hi
s
t
ogr
am
f
or
a v
er
y
br
i
g
ht
i
m
age w
i
t
h f
e
w
dar
k
ar
ea
s
a
nd/
or
s
had
ow
s
w
i
l
l
hav
e
m
os
t
of
i
t
s
dat
a
poi
nt
s
o
n
t
he
r
i
ght
s
i
de
and
c
ent
er
of
t
he
gr
ap
h.
Li
m
i
t
e
d
r
ef
er
e
nc
es
w
hi
c
h
di
s
c
u
s
s
ed
t
he
i
m
ag
e
du
e
t
o
t
he
P
D
ev
e
nt
s
c
a
n
be
f
ou
nd
i
n
[
11]
.
T
he
i
m
age
s
w
er
e
t
ak
e
n
u
s
i
ng
t
he
opt
i
c
al
c
am
e
r
a.
T
he
c
om
par
i
s
o
n
bet
w
ee
n
t
he
s
ur
f
a
c
e
i
m
age
of
neat
L
D
P
E
and L
D
P
E
nano
s
i
l
i
c
a
c
om
po
s
i
t
e du
e t
o P
D
ev
e
nt
al
s
o
w
a
s
s
ho
w
n i
n
[
1
2]
,
b
u
t
w
i
t
hout
a
de
e
per
an
al
y
s
i
s
.
T
he
di
s
t
r
i
but
i
on
of
c
ol
our
s
i
n
an
i
m
ag
e
p
r
oc
es
s
i
n
g
t
ec
hni
qu
e
i
s
r
ep
r
es
ent
ed
by
a
c
ol
o
ur
hi
s
t
og
r
am
t
h
at
c
ov
e
r
al
l
pos
s
i
bl
e t
he
i
m
age
's
c
ol
our
s
pac
e.
F
or
di
gi
t
al
i
m
age
s
,
a
c
o
l
our
hi
s
t
og
r
am
r
e
pr
e
s
ent
s
t
h
e
num
be
r
of
pi
x
el
s
t
hat
hav
e c
ol
ou
r
s
i
n
eac
h of
a f
i
x
ed l
i
s
t
of
c
ol
our
r
ang
es
.
I
f
t
he
s
et
of
p
os
s
i
bl
e c
ol
ou
r
v
al
ues
i
s
s
uf
f
i
c
i
ent
l
y
s
m
al
l
,
e
ac
h
of
t
hos
e
c
ol
o
ur
s
m
ay
be
pl
ac
ed on a r
ang
e
by
i
t
s
e
l
f
;
t
hen t
he
h
i
s
t
ogr
am
i
s
m
er
el
y
t
h
e
c
ount
of
pi
x
el
s
t
hat
hav
e
eac
h
pos
s
i
bl
e c
ol
o
ur
.
T
he i
m
ag
e hi
s
t
o
gr
a
m
m
ay
al
s
o
be
r
ep
r
e
s
ent
e
d
and di
s
pl
ay
ed a
s
a
s
m
o
ot
h
f
unc
t
i
on def
i
n
ed ov
er
t
he c
ol
our
s
pa
c
e t
hat
appr
ox
i
m
at
es
t
he pi
x
el
c
ount
s
.
Li
k
e
ot
her
k
i
n
ds
of
hi
s
t
og
r
am
s
,
t
he
c
ol
o
ur
hi
s
t
ogr
am
i
s
a
s
t
at
i
s
t
i
c
t
hat
c
an
be
v
i
ew
e
d a
s
a
n a
ppr
ox
i
m
at
i
on of
an
unde
r
l
y
i
ng
c
o
nt
i
nuou
s
di
s
t
r
i
but
i
on
of
c
ol
our
s
v
al
ue
s
.
A
n
i
m
age
hi
s
t
ogr
am
i
s
a
g
r
aphi
c
al
r
ep
r
es
e
nt
at
i
on
of
t
he
num
be
r
of
pi
x
el
s
i
n
a
n
i
m
age,
w
ho
s
e X
-
ax
i
s
r
ep
r
e
s
ent
s
t
he t
ona
l
s
c
al
e (
bl
ac
k
at
t
he l
ef
t
an
d w
hi
t
e at
t
he
r
i
ght
)
,
an
d Y
-
ax
i
s
r
ep
r
e
s
ent
s
t
he
num
be
r
of
pi
x
el
s
i
n
an
i
m
age
i
n
a
pa
r
t
i
c
ul
a
r
ar
ea
of
t
he
t
onal
s
c
al
e.
I
f
t
her
e
ar
e
m
or
e
pi
x
el
s
,
t
he
p
ea
k
at
t
h
e
c
er
t
ai
n
i
nt
e
ns
i
t
y
l
ev
el
i
s
hi
ghe
r
.
A
c
ol
o
ur
hi
s
t
og
r
am
of
an
i
m
age
al
s
o
r
ep
r
e
s
ent
s
t
h
e
di
s
t
r
i
but
i
on
of
t
he
c
om
po
s
i
t
i
on
of
c
ol
o
ur
s
i
n
t
he
i
m
a
ge.
I
t
s
h
o
w
s
d
i
f
f
e
r
e
n
t
t
y
p
e
s
o
f
c
ol
o
ur
s
a
ppe
ar
ed
an
d t
he
num
b
er
of
pi
x
el
s
i
n
ea
c
h t
y
pe of
t
he
c
o
l
our
s
a
ppe
ar
ed.
T
he
r
el
at
i
o
n
bet
w
e
en
a
c
o
l
our
hi
s
t
og
r
a
m
and
a
l
um
i
nan
c
e
hi
s
t
og
r
am
i
s
t
h
at
a
c
ol
o
ur
hi
s
t
og
r
am
c
an
b
e
al
s
o
ex
pr
e
s
s
ed a
s
“
T
hr
ee
-
C
ol
o
ur
H
i
s
t
og
r
am
s
,
"
eac
h of
w
hi
c
h s
ho
w
s
t
he br
i
ght
n
e
s
s
di
s
t
r
i
but
i
o
n of
eac
h
i
ndi
v
i
du
al
R
G
B
c
ol
ou
r
c
ha
nnel
I
n obs
e
r
v
i
ng
t
he s
ur
f
a
c
e
def
ec
t
,
t
he i
m
age qu
al
i
t
y
pl
ay
s
an i
m
por
t
ant
r
ol
e.
S
o i
t
i
s
i
m
por
t
a
nt
t
o
m
ai
nt
ai
n
t
h
e
i
m
age
i
n
a
hi
gh
qu
al
i
t
y
bef
or
e
i
t
i
s
pr
o
c
es
s
ed
t
o
get
a
goo
d
a
n
i
m
age
hi
s
t
og
r
am
.
T
he
t
er
m
of
i
m
age
q
ual
i
t
y
,
t
ec
h
ni
c
al
l
y
r
ef
er
t
o
i
m
age
a
t
t
r
i
but
es
l
i
k
e
i
m
age
r
e
s
ol
ut
i
on,
c
ont
r
as
t
,
b
r
i
g
ht
nes
s
,
noi
s
e and
c
o
nt
r
a
s
t
-
t
o
-
noi
s
e r
a
t
i
o (
C
N
R
)
.
T
hi
s
i
m
a
g
e q
ual
i
t
y
at
t
r
i
but
es
c
ha
nge
du
r
i
n
g
t
he
edi
t
i
ng,
c
on
v
e
r
t
i
ng
an
d
upl
oadi
ng
p
r
oc
es
s
.
T
o
m
ai
nt
ai
n
t
he
i
m
age
qual
i
t
y
,
t
he
t
r
adi
t
i
onal
m
et
hod
by
us
i
ng t
he
er
r
o
r
s
um
m
at
i
on
m
et
hod c
a
n
be
us
e
d.
B
ut
i
n r
ec
ent
dev
el
opm
ent
t
he di
f
f
er
ent
m
et
hod l
i
k
e
t
he hi
s
t
ogr
a
m
-
b
as
ed
i
m
a
ge qual
i
t
y
i
ndex
w
a
s
us
e
d t
o
m
ai
nt
ai
n t
h
e
i
m
age
qual
i
t
y
[
13]
.
O
t
he
r
m
et
hod
t
o m
ai
nt
ai
n t
h
e i
m
age
qual
i
t
y
i
s
by
u
s
i
n
g
t
he
B
ay
es
i
an
i
m
a
ge
e
s
t
i
m
at
i
on
t
o
i
m
p
r
ov
e
t
he
C
N
R
[
14
]
.
I
n
or
de
r
t
o
i
m
pr
ov
e
t
he
c
ont
r
a
s
t
,
i
t
c
an
be
done
by
en
ha
nc
i
n
g
t
he
i
m
a
ge
hi
s
t
o
gr
a
m
w
r
a
ppi
n
g
[
15]
.
2.
R
ese
ar
ch
M
et
h
o
d
T
her
e
a
r
e
t
h
r
ee
s
t
ep
s
s
ho
ud
b
e
d
one
t
o
a
c
hi
ev
e
t
he
pur
po
s
e
of
t
hi
s
r
e
s
ea
r
c
h;
s
am
pl
es
pr
ep
ar
at
i
on,
P
D
T
e
s
t
i
ng a
nd pi
c
t
ur
i
n
g t
he s
am
pl
e
s
ur
f
ac
e bef
o
r
e
and af
t
er
t
h
e P
D
t
es
t
.
T
w
o
di
f
f
er
ent
pol
y
m
er
s
,
nam
el
y
or
gani
c
pol
y
m
er
s
(
N
R
)
an
d
i
nor
g
ani
c
pol
y
m
er
s
(
L
D
P
E
)
,
w
i
t
h
di
f
f
er
ent
c
om
p
os
i
t
i
o
ns
ar
e
m
i
x
ed
dur
i
ng
s
am
pl
e
p
r
epar
at
i
on.
T
o
c
h
ec
k
t
h
e
ef
f
ec
t
of
P
D
,
P
D
m
ea
s
u
r
em
e
n
t
i
s
don
e.
B
ef
or
e
and
af
t
e
r
t
he
P
D
t
es
t
,
t
he
s
am
pl
e
s
u
r
f
ac
e
i
m
age
w
a
s
t
ak
e
n
us
i
ng
an
el
e
c
t
r
o
ni
c
mi
c
r
os
c
o
pe
w
i
t
h
20x
opt
i
c
al
z
oom
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
K
O
M
NI
K
A
I
SSN
:
1
6
9
3
-
6
930
T
he
C
o
r
r
el
at
i
on
of
S
t
at
i
s
t
i
c
al
I
m
age
and
P
ar
t
i
al
D
i
s
c
h
ar
ge
P
ul
s
e
C
ount
of
L
D
PE
-
N
R
…
(
A
u
l
i
a
)
979
2
.
1
.
S
a
m
pl
e
P
re
pa
ra
t
i
on
T
he c
om
po
s
i
t
i
ons
of
t
he
pr
ep
ar
ed s
a
m
pl
es
ar
e t
a
bul
at
e
d i
n T
abl
e 1.
T
he
LD
P
E
/
N
R
bl
end
s
w
e
r
e
m
ade
up
of
N
R
w
i
t
h
di
f
f
e
r
ent
c
om
po
s
i
t
i
ons
,
r
a
ngi
ng
0
w
t
%
-
10
w
t
%
,
20
w
t
%
and
3
0
w
t
%
.
T
he
t
em
per
at
ur
e
f
or
bot
h
r
ol
l
m
i
l
l
s
w
a
s
s
et
a
t
150°C
du
r
i
n
g
t
he
bl
endi
n
g
pr
o
c
e
s
s
.
T
he
t
hr
ee
c
at
ego
r
i
es
of
s
am
pl
e
s
w
e
r
e
pr
e
pa
r
ed
by
f
i
r
s
t
m
el
t
i
ng
t
he
L
D
P
E
,
and
t
he
n
a
s
m
al
l
am
o
unt
of
N
R
o
r
na
n
os
i
l
i
c
a w
as
g
r
adu
al
l
y
add
e
d unt
i
l
t
he N
R
di
s
pe
r
s
ed
w
el
l
.
T
he
s
a
m
e pr
oc
e
s
s
w
a
s
r
epe
at
ed f
o
r
c
at
eg
or
i
e
s
B
and
C
.
T
he
nex
t
pr
oc
es
s
i
s
t
o
p
r
ep
ar
e t
he
s
a
m
pl
e
as
r
equi
r
ed
by
C
I
G
R
E
M
et
h
od t
o
obt
ai
n
t
he
r
e
q
ui
r
e
d t
hi
c
k
n
es
s
us
i
n
g t
he
m
ol
di
ng
m
a
c
hi
ne.
T
h
e
s
a
m
e
t
em
per
at
ur
e
w
a
s
al
s
o s
et
at
t
he m
ol
di
n
g m
ac
hi
ne.
I
n or
d
er
t
o g
e
t
a good m
el
t
of
s
am
pl
e,
t
he
s
am
pl
e w
a
s
heat
ed
bet
w
een
5
-
8
m
i
n
ut
es
bef
or
e
pr
e
s
s
ur
e w
a
s
a
ppl
i
ed
t
o
t
he m
o
ul
d.
T
he
s
am
pl
e
w
as
t
hen
c
ool
e
d
d
ow
n
us
i
n
g
t
h
e
w
at
e
r
c
ool
i
ng
m
a
c
hi
n
e
f
or
at
l
ea
s
t
8
m
i
nut
es
or
u
n
t
i
l
t
he
r
oom
t
em
p
er
at
ur
e
w
as
a
c
hi
ev
ed.
T
abl
e
1.
S
am
pl
es
p
r
ep
ar
ed
f
or
t
hr
ee
di
f
f
er
ent
c
at
ego
r
i
es
S
am
pl
e Nam
e
S
i
l
i
c
a (
gr
am
)
LDP
E
(g
ra
m
)
NR
(g
ra
m
)
A1
-
100
-
A2
-
90
10
A3
-
80
20
A4
-
70
30
2
.
2
.
E
xp
er
i
m
en
t
al
S
et
u
p
A
50
H
z
A
C
v
ol
t
age s
ou
r
c
e w
i
t
h
a pe
a
k
v
al
ue
of
6.
5 k
V
r
m
s
w
a
s
a
ppl
i
ed
t
o t
he t
e
s
t
el
ec
t
r
ode
w
hi
l
e
t
he
pl
ane
el
ec
t
r
o
de
w
as
ear
t
he
d.
T
w
o
(
2)
di
f
f
e
r
ent
p
r
ot
ec
t
i
o
n
uni
t
s
w
e
r
e
i
n
s
t
al
l
ed
t
o
pr
ot
ec
t
t
he
s
am
pl
e
a
nd
m
eas
ur
i
ng
un
i
t
f
r
om
o
v
er
c
ur
r
ent
and
ov
er
v
ol
t
age.
T
he
hi
gh
v
ol
t
a
ge
r
es
i
s
t
o
r
w
a
s
i
ns
t
al
l
e
d i
n s
e
r
i
es
w
i
t
h t
he
s
am
pl
e t
o l
i
m
i
t
t
he r
et
ur
ni
n
g hi
gh
c
ur
r
en
t
i
n c
as
e of
a
br
ea
k
d
ow
n.
A
ddi
t
i
onal
pr
ot
ec
t
i
on
s
y
s
t
em
w
a
s
a
c
hi
ev
ed
by
t
h
e
s
pher
e
gap,
w
hi
c
h
f
un
c
t
i
on
s
a
s
a
h
i
g
h
v
o
l
t
a
g
e
li
mi
t
er
.
I
t
i
s
not
ed t
hat
by
us
i
ng t
he
C
I
G
R
E
M
et
hod I
I
,
and w
i
t
h t
he gi
v
en
ex
per
i
m
ent
al
c
ondi
t
i
o
ns
,
t
he P
D
a
c
t
i
v
i
t
i
es
onl
y
o
c
c
ur
i
n t
he v
oi
d ar
ea.
F
i
gur
e
1
s
how
s
t
he
l
abor
at
or
y
s
et
-
up
f
or
t
h
e
P
D
t
e
s
t
s
y
s
t
em
.
F
i
gur
e
1
(
a
)
i
s
t
he
s
c
h
em
a
t
i
c
di
a
gr
a
m
a
nd
F
i
gur
e
1
(b
)
i
s
t
he
pi
c
t
u
r
e
of
t
he
r
eal
P
D
m
eas
ur
em
ent
s
et
-
up
.
I
t
c
o
n
s
i
s
t
s
of
an
A
C
hi
gh
v
ol
t
a
ge
s
u
ppl
y
an
d
i
t
s
m
eas
ur
i
ng
s
y
s
t
em
,
a
t
es
t
c
el
l
c
on
s
t
r
u
c
t
ed
ac
c
o
r
di
n
g
t
o
t
he
C
I
G
R
E
M
et
hod
I
I
(
C
M
-
I
I
)
[
4
]
,
an
d
a
dat
a
ac
qui
s
i
t
i
on
s
y
s
t
e
m
(
m
ade
up
of
a
P
i
c
o
s
c
ope
5
20
3
dev
i
c
e
a
nd
a
P
C
)
.
HV
transfor
mer
HV resistor
Step down
transformer
Picoscope
From power
supply
High Pass
Filter
CIGRE
Method II
Test Cell
Netbook
and data
logger
Sphere
gap
(
a)
(
b)
F
i
gur
e
1.
T
es
t
s
et
-
up
f
or
t
h
e
par
t
i
al
di
s
c
har
g
e
t
es
t
,
t
he
phot
o
of
t
he
ex
per
i
m
e
nt
al
s
et
-
up
3.
R
esu
l
t
s
a
nd
A
na
l
y
s
i
s
3
.
1
.
P
D
P
ul
s
e
C
oun
t
C
h
a
ra
c
t
e
ri
s
t
i
c
s
of
LD
P
E
-
N
R
C
om
pos
i
t
e
s
Tw
o
di
f
f
er
ent
anal
y
s
es
w
e
r
e
c
ar
r
i
e
d out
on L
D
P
E
-
N
R
s
am
pl
es
,
P
D
c
ount
a
n
d s
u
r
f
a
c
e
i
m
age a
nal
y
s
i
s
.
P
D
c
ount
anal
y
s
i
s
w
as
done f
o
r
al
l
s
am
pl
e
s
an
d i
m
age a
nal
y
s
i
s
w
a
s
d
one f
or
s
am
pl
e
A
1
an
d
A
3.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN
:
1
6
9
3
-
6
930
T
E
L
K
O
M
NI
K
A
Vo
l
.
15
,
N
o
.
3,
S
ept
em
ber
20
17
:
977
–
9
83
980
3.
1.
1
C
um
ul
a
t
i
v
e
P
D
P
ul
s
e
C
ou
nt
F
i
gur
e 2
s
ho
w
s
t
he c
um
ul
at
i
v
e pos
i
t
i
v
e and negat
i
v
e P
D
pul
s
e
c
ount
s
obt
ai
n
ed f
r
om
di
f
f
er
ent
L
D
P
E
-
N
R
c
om
p
o
s
i
t
i
on
s
ov
e
r
a 60
m
i
nut
e
s
t
r
es
s
dur
at
i
on.
I
t
c
a
n b
e s
een
t
hat
t
h
e
negat
i
v
e
P
D
pul
s
e
c
ou
nt
i
s
hi
ghe
r
t
h
an
t
he
po
s
i
t
i
v
e
P
D
p
ul
s
e
c
ou
nt
ex
c
e
pt
f
or
s
am
pl
e
A
1.
T
he
i
nc
r
em
e
nt
of
t
he
m
ea
s
u
r
ed
P
D
pul
s
e
c
ou
nt
f
or
ev
er
y
t
en
m
i
nut
e
s
i
s
s
om
ew
hat
l
i
ne
ar
i
ndi
c
at
i
ng
a
c
on
s
t
ant
P
D
pul
s
e c
o
unt
f
or
t
he gi
v
e
n t
i
m
e i
nt
er
v
al
ex
c
ept
f
or
s
a
m
pl
e A
3.
T
h
e t
ot
al
P
D
p
ul
s
e
c
ou
nt
af
t
er
s
i
x
c
ons
ec
ut
i
v
e c
y
c
l
e
s
(
6
0 m
i
nut
es
)
of
hi
gh v
ol
t
age
s
t
r
es
s
w
a
s
ab
out
2000 f
or
al
l
s
am
pl
es
i
n
c
at
egor
y
A
ex
c
ept
f
o
r
s
am
p
l
e
A
3,
w
hi
c
h
w
a
s
8
000.
T
h
us
t
h
e
P
D
c
o
unt
f
or
A
3
i
s
f
our
t
i
m
es
hi
g
her
t
han t
ho
s
e f
o
r
s
a
m
pl
e
s
A
1
,
A
2,
and A
4.
A
hi
gh P
D
p
ul
s
e
c
ou
nt
i
n
di
c
at
e
s
a hi
g
h
det
er
i
o
r
at
i
o
n
l
ev
el
of
s
am
pl
e
s
ur
f
a
c
e
d
ue
t
o
t
he
el
ec
t
r
i
c
al
s
t
r
e
s
s
.
.
(
a)
(
b)
(
c)
(
d)
F
i
gur
e
2
C
um
ul
at
i
v
e
p
os
i
t
i
v
e
and
ne
gat
i
v
e
P
D
pul
s
e
c
ount
s
f
o
r
di
f
f
er
ent
LD
P
E
-
N
R
c
om
po
s
i
t
i
o
ns
ov
er
60
m
i
n
ut
es
s
t
r
e
s
s
d
ur
at
i
on;
(
a)
s
am
pl
e
A
1
,
(
b)
s
a
m
pl
e
A
2,
(
c
)
s
am
pl
e
A
3,
and
(
d
)
s
am
pl
e
A4
3.
1.
2
.
Tot
a
l
P
D
P
ul
s
e
C
ou
nt
F
i
gur
e
3
s
ho
w
s
t
h
e t
ot
al
P
D
pul
s
e
c
o
unt
s
b
ot
h f
or
pos
i
t
i
v
e an
d neg
at
i
v
e af
t
er
60
m
i
nut
es
el
ec
t
r
i
c
s
t
r
e
s
s
f
o
r
LD
P
E
-
N
R
c
om
po
s
i
t
es
.
T
he t
ot
al
neg
at
i
v
e P
D
pul
s
e c
ount
s
f
or
al
l
s
am
pl
es
a
r
e
hi
ghe
r
t
han
t
ha
t
f
or
t
he
po
s
i
t
i
v
e
P
D
,
ex
c
ept
f
or
s
am
pl
e
A
1.
T
he
hi
ghe
r
t
ot
al
c
ount
f
or
t
he negat
i
v
e P
D
i
ndi
c
at
es
t
hat
t
he
LD
P
E
-
N
R
i
s
m
or
e s
u
s
c
ept
i
bl
e t
o s
u
r
f
ac
e
at
def
e
c
t
s
c
om
p
ar
ed t
o
pur
e
L
D
P
E
.
A
goo
d i
n
s
ul
at
i
on m
at
er
i
al
s
h
oul
d
hav
e
an
equ
al
P
D
pul
s
e
c
o
unt
f
or
bot
h
po
s
i
t
i
v
e
and
ne
gat
i
v
e
w
he
n
s
u
bj
e
c
t
ed
t
o
a
s
y
m
m
et
r
i
c
al
v
ol
t
ag
e
s
t
r
e
s
s
.
F
i
gur
e
4
s
ho
w
s
t
h
at
t
he
t
o
t
al
P
D
pul
s
e
c
ou
nt
s
f
o
r
di
f
f
er
ent
LD
P
E
-
N
R
di
el
e
c
t
r
i
c
s
am
pl
e
s
(
A
1,
A
2
a
nd
A
4)
a
r
e
bel
o
w
60
00,
ex
c
e
pt
s
am
pl
e
A
3
,
w
hi
c
h
i
s
al
m
os
t
1
6,
000.
T
he
t
ot
al
P
D
p
ul
s
e
c
ou
nt
of
s
a
m
pl
e A
1
i
s
s
l
i
g
ht
l
y
hi
gher
t
h
an t
hat
f
or
s
a
m
pl
e A
4,
but
t
he t
ot
al
P
D
pul
s
e c
ount
f
or
s
am
pl
e A
2 i
s
m
uc
h l
o
w
e
r
t
h
an t
ho
s
e f
o
r
s
am
pl
es
A
1
,
A
3 and
A
4.
A
hi
gh
er
w
ei
ght
pe
r
c
ent
a
ge
(w
t
%
)
o
f
N
R
i
s
e
x
p
e
c
t
e
d
t
o
i
nc
r
ea
s
e
t
he
P
D
r
e
s
i
s
t
a
nc
e.
I
t
w
as
f
oun
d
t
hat
t
he
P
D
pul
s
e
c
o
un
t
f
o
r
s
am
pl
e
A
3
i
s
gr
eat
e
r
t
h
an
t
hos
e
f
o
r
s
am
pl
es
A
2
and
A
1.
T
hi
s
h
ap
pen
s
b
ec
au
s
e
of
m
any
f
a
c
t
or
s
,
i
nc
l
u
di
ng
t
he
pos
s
i
bl
e
ex
i
s
t
enc
e
of
v
oi
ds
dur
i
ng
s
am
pl
e
pr
ep
ar
at
i
o
n
w
hi
c
h
t
he
n
l
ead
s
t
o
a
l
ow
er
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
K
O
M
NI
K
A
I
SSN
:
1
6
9
3
-
6
930
T
he
C
o
r
r
el
at
i
on
of
S
t
at
i
s
t
i
c
al
I
m
age
and
P
ar
t
i
al
D
i
s
c
h
ar
ge
P
ul
s
e
C
ount
of
L
D
PE
-
N
R
…
(
A
u
l
i
a
)
981
P
D
r
e
s
i
s
t
anc
e t
han t
hat
f
o
r
t
he ba
s
e o
r
r
ef
er
en
c
e i
n
s
ul
at
i
o
n
m
at
e
r
i
al
.
E
v
en t
ho
ugh
s
am
pl
e A
4
has
a
hi
ghe
r
P
D
r
e
s
i
s
t
a
nc
e
t
han
s
a
m
pl
e
A
3,
t
he
v
al
ue
i
s
onl
y
s
l
i
g
ht
l
y
bet
t
er
t
han
t
hat
f
or
s
a
m
pl
e
A1
.
F
i
gur
e
3
.
P
os
i
t
i
v
e
and
neg
at
i
v
e
P
D
pul
s
e
c
ou
nt
s
of
di
f
f
er
ent
LD
P
E
-
N
R
c
om
p
os
i
t
i
ons
ov
e
r
60
m
i
nut
es
s
t
r
e
s
s
dur
at
i
on;
(
a
)
s
am
pl
e
A
1
,
(
b)
s
am
pl
e
A
2,
(
c
)
s
am
pl
e
A
3,
and
(
d)
s
am
pl
e
A
4
F
i
gur
e
4
.
T
ot
al
P
D
pul
s
e
c
ount
s
of
di
f
f
er
ent
LD
P
E
-
N
R
c
om
po
s
i
t
i
on
ov
er
60
m
i
n
ut
es
s
t
r
e
s
s
dur
at
i
o
n;
(
a)
s
am
pl
e
A
1
,
(
b)
s
a
m
pl
e
A
2
,
(
c
)
s
am
pl
e
A
3,
and
(
d)
s
am
pl
e
A
4
T
he
abov
e
a
nal
y
s
i
s
of
P
D
pul
s
e
c
o
unt
s
ho
w
s
t
hat
s
am
pl
e
A
2
i
s
t
he
b
es
t
and
s
am
pl
e
A
3
i
s
t
h
e
w
o
r
s
t
.
T
he
addi
t
i
o
n
of
N
R
t
o
LD
P
E
w
i
t
h
s
m
al
l
e
r
t
han
10
w
t
%
of
N
R
s
ee
m
s
t
o
i
m
pr
ov
e
t
he
P
D
r
e
s
i
s
t
a
nc
e
of
t
he
LD
E
P
-
N
R
c
om
po
s
i
t
e.
3.
2.
S
u
r
f
ace
I
m
ag
e
an
al
y
si
s
o
f
L
D
PE
-
N
R
Co
m
p
o
s
i
te
s
F
i
gur
e
5
s
ho
w
s
t
he
i
m
ag
e
of
s
am
pl
e
A
1
an
d
A
3
bef
or
e
an
d
af
t
er
t
he
hi
g
h
v
ol
t
a
ge
s
t
r
e
s
s
had
bee
n
a
p
pl
i
ed
f
o
r
60
m
i
nut
es
.
T
he
di
f
f
er
e
nc
es
i
n
t
he
s
ur
f
a
c
e
c
o
ndi
t
i
on
s
b
ef
or
e
an
d
af
t
er
t
h
e
t
es
t
w
er
e
c
l
e
ar
l
y
i
d
e
n
t
i
f
i
e
d
.
S
ev
er
al
r
a
n
dom
l
y
l
o
c
at
e
d
da
r
k
s
pot
s
w
i
t
h
v
a
r
i
abl
e
s
i
z
es
a
r
e
s
een
i
n
t
he
af
t
er
s
t
r
e
s
s
i
m
age.
T
h
e
dar
k
s
pot
s
ar
e
c
r
e
at
ed
a
r
ea
d
ue
t
o
t
h
e
c
a
r
bo
ni
z
at
i
on
phe
nom
e
non
as
a
r
e
s
ul
t
of
P
D
ac
t
i
v
i
t
i
es
.
I
t
i
s
not
ed t
hat
t
he da
r
k
s
pot
s
ar
e onl
y
v
v
i
abl
e us
i
n
g a m
a
gni
f
y
i
n
g
c
am
e
r
a.
U
s
i
n
g t
he
w
av
el
et
anal
y
s
i
s
ap
p
r
oa
c
h,
t
h
e
s
u
r
f
ac
e
r
oug
hn
es
s
of
t
he t
w
o i
m
a
ge
s
c
ou
l
d
be
di
s
t
i
n
gui
s
h
ed
ea
s
i
l
y
.
T
abl
e
2
s
ho
w
s
t
he
hi
s
t
o
g
r
a
m
o
f
L
D
PE
-
N
R
c
om
po
s
i
t
es
(
s
am
pl
e
s
A
1
and
A
3)
.
A
s
c
a
n
be
s
ee
n,
t
he
s
ha
pe
of
t
he
hi
s
t
ogr
a
m
c
han
g
es
f
r
om
bef
o
r
e
s
t
r
es
s
t
o
af
t
er
s
t
r
e
s
s
.
S
a
m
pl
e
A
3
s
ho
w
s
a
l
ar
ge
r
c
han
ge
t
ha
n
s
am
p
l
e
A
1.
T
h
e
p
e
ak
v
al
ue
i
s
r
e
duc
ed
bel
o
w
120,
00
0
pi
x
el
s
a
nd
i
nt
en
s
i
t
y
w
it
h
a
l
o
w
pi
x
el
app
ear
s
.
A
s
c
an
be
s
ee
n
t
hat
ad
di
t
i
on
of
N
R
t
o
L
D
P
E
c
h
ang
es
t
he
s
hape
of
t
he
hi
s
t
og
r
am
t
o
bec
om
e
s
l
i
m
m
er
.
T
h
e c
hang
e of
hi
s
t
ogr
a
m
s
hape
i
ndi
c
at
e
s
t
h
e c
h
ange i
n
t
he
c
or
r
e
s
p
ondi
n
g
s
t
at
i
s
t
i
c
al
p
ar
am
et
er
s
.
T
he
c
han
ge
of
hi
s
t
og
r
am
s
hape i
ndi
c
at
e
s
t
hat
t
he
c
ol
our
o
f
L
D
PE
-
N
R
s
u
r
f
ac
e
c
h
ange
d du
r
i
n
g t
he hi
gh
v
ol
t
age s
t
r
e
s
s
du
e t
o P
D
a
c
t
i
v
i
t
i
es
.
I
t
i
s
unde
r
s
t
o
od t
hat
dur
i
ng a
P
D
i
nc
i
dent
,
t
he def
e
c
t
i
v
e s
ur
f
a
c
e
w
i
l
l
be c
r
eat
e
d a
s
a r
es
ul
t
of
t
he
c
ar
boni
z
at
i
o
n
pr
o
c
e
s
s
.
C
ar
boni
z
at
i
on
oc
c
ur
s
be
c
au
s
e
t
he
P
D
ge
ne
r
a
t
es
he
at
an
d
c
an
b
ur
n
a
nd
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN
:
1
6
9
3
-
6
930
T
E
L
K
O
M
NI
K
A
Vo
l
.
15
,
N
o
.
3,
S
ept
em
ber
20
17
:
977
–
9
83
982
dam
ag
e t
he
s
ur
f
a
c
e
of
t
he i
ns
ul
at
or
.
T
he dam
aged
s
ur
f
a
c
e
ar
e
a
c
an b
e i
dent
i
f
i
ed by
t
he dar
k
c
ol
o
r
as
s
o
c
i
at
ed
w
i
t
h
t
he
d
ef
ec
t
i
v
e
ar
e
a.
F
i
gur
e
5.
T
he
s
ur
f
a
c
e
i
m
a
g
es
of
LD
P
E
-
N
R
c
om
po
s
i
t
e
s
am
pl
e
A
1
(
1
00:
0)
bef
o
r
e
and
af
t
er
60
m
i
nut
es
hi
gh
v
ol
t
age
s
t
r
e
s
s
T
abl
e
2.
L
D
PE
-
N
R
c
om
po
s
i
t
es
i
nt
e
ns
i
t
y
hi
s
t
ogr
am
of
s
am
pl
e
A
1
an
d
s
am
pl
e
A
3
bef
or
e
a
nd
af
t
er
60
m
i
nut
es
hi
g
h
v
ol
t
age
s
t
r
e
s
s
S
am
pl
e
Hi
s
togr
am
B
efor
e
A
fter
A1
A2
4
.
C
o
n
c
l
u
s
i
o
n
R
e
s
ea
r
c
h
ha
s
s
ho
w
n
a
s
t
r
ong
c
or
r
el
at
i
o
n
bet
w
een
t
he
PD
pul
s
e
c
ount
w
i
t
h
t
h
e
ex
t
ent
of
dam
ag
e on t
he s
u
r
f
a
c
e of
t
he s
am
pl
e
s
ur
f
a
c
e
LD
P
E
-
N
R
.
T
he r
e
s
ul
t
s
i
n
di
c
at
e
t
hat
t
he s
ur
f
ac
e
i
m
age
s
t
at
i
s
t
i
c
al
anal
y
s
i
s
c
an
be
u
s
ed
a
s
a
t
o
ol
t
o
j
us
t
i
f
y
t
he
t
o
t
a
l
of
t
he
PD
pul
s
e
c
ount
on
t
he
s
ur
f
a
c
e
f
o
r
di
f
f
er
ent
s
am
pl
e
s
.
A
ckn
o
w
l
ed
g
e
me
n
t
s
T
he a
ut
hor
s
w
oul
d l
i
k
e
t
hank
U
ni
v
er
s
i
t
a
s
A
n
da
l
as
,
P
ada
ng,
I
ndone
s
i
a
(
C
ont
r
ac
t
N
um
b
er
012/
U
N
.
1
6.
09.
D
/
P
L/
2016
a
nd
U
ni
v
er
s
i
t
i
T
e
k
nol
ogi
M
al
ay
s
i
a,
J
oho
r
B
a
hr
u,
M
al
ay
s
i
a
f
or
s
up
por
t
i
n
g t
h
i
s
r
es
ea
r
c
h
w
i
t
h t
he
go
o
d qu
al
i
t
y
equ
i
pm
ent
.
W
e
al
s
o
l
i
k
e t
o t
han
k
E
ngi
ne
er
i
ng
F
ac
ul
t
y
of
U
ni
v
er
s
i
t
a
s
A
nda
l
as
f
or
s
upp
or
t
i
ng
t
he
publ
i
c
at
i
on
p
r
o
c
e
s
s
of
t
hi
s
pap
e
r
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
K
O
M
NI
K
A
I
SSN
:
1
6
9
3
-
6
930
T
he
C
o
r
r
el
at
i
on
of
S
t
at
i
s
t
i
c
al
I
m
age
and
P
ar
t
i
al
D
i
s
c
h
ar
ge
P
ul
s
e
C
ount
of
L
D
PE
-
N
R
…
(
A
u
l
i
a
)
983
R
ef
er
en
ces
[1
]
M H
i
ki
t
a
,
K Ya
ma
d
a
,
A N
a
k
a
mu
ra
,
T
Mi
zu
t
a
n
i
,
A
O
ohas
i
,
M
I
eda.
M
eas
u
r
em
ent
s
of
par
t
i
al
d
i
s
c
har
ges
b
y
c
om
put
er
a
nd
an
al
y
s
i
s
of
par
t
i
al
d
i
s
c
ha
r
ge
di
s
t
r
i
bu
t
i
o
n
b
y
t
he
M
o
nt
e
C
ar
l
o
m
et
h
od
.
El
e
ct
ri
ca
l
I
ns
ul
at
i
on,
I
E
E
E
T
r
ans
ac
t
i
ons
on.
199
0;
25
:
453
-
468.
[2
]
W
Z
i
om
ek
,
M
R
ef
or
m
at
,
E
K
uf
f
el
.
A
pp
l
i
c
at
i
o
n
of
ge
net
i
c
al
gor
i
t
hm
s
t
o
p
at
t
er
n
r
ec
ogn
i
t
i
o
n
of
def
ec
t
s
i
n
GIS
.
D
i
el
ec
t
r
i
c
s
and E
l
ec
t
r
i
c
al
I
ns
ul
at
i
o
n,
I
E
E
E
T
r
ans
ac
t
i
ons
on.
200
0
;
7
:
161
-
1
68.
[3
]
B Ya
zi
c
i
.
S
t
at
i
s
t
i
c
al
pat
t
er
n
ana
l
y
s
i
s
of
p
ar
t
i
al
di
s
c
h
ar
g
e m
eas
ur
em
ent
s
f
or
q
ual
i
t
y
as
s
es
s
m
ent
of
i
ns
ul
at
i
o
n
s
y
s
t
em
s
i
n
h
i
g
h
-
v
o
l
t
age el
ec
t
r
i
c
al
m
ac
hi
ner
y
.
I
n
d
u
st
ry
Ap
p
l
i
ca
t
i
o
n
s,
I
EEE
T
ra
n
sa
ct
i
o
n
s
o
n
.
200
4;
40
:
157
9
-
159
4.
[4
]
A
Lap
p
,
H
G
Kra
n
z.
T
he us
e o
f
t
he C
I
G
R
E
dat
a f
or
m
at
f
or
P
D
di
a
gn
os
i
s
a
ppl
i
c
at
i
o
ns
.
D
i
e
l
e
ct
ri
cs a
n
d
El
e
ct
ri
ca
l
I
ns
ul
at
i
on,
I
E
E
E
T
r
ans
ac
t
i
o
ns
on
2
000;
7
:
10
2
-
11
2.
[5
]
S
u
w
ar
no,
Y
S
uz
uok
i
,
F
Ko
mo
ri
,
T
M
i
z
ut
ani
,
P
ar
t
i
al
d
i
s
c
h
ar
ges
d
ue
t
o
e
l
e
c
t
r
i
c
al
t
r
eei
ng
i
n
po
l
y
m
er
s
:
P
has
e
-
r
es
ol
v
e
d an
d t
i
m
e
-
s
eq
uenc
e o
bs
er
v
a
t
i
on a
nd
an
al
y
s
i
s
.
J
our
n
a
l
o
f
Ph
ysi
cs D
:
Ap
p
l
i
e
d
Ph
ys
i
cs.
199
6;
29
:
292
2
-
293
1.
[6
]
MH
A Au
l
i
a
,
Z
u
l
k
ur
n
ai
n
A
b
dul
-
M
al
ek
,
Y
anu
a
r
Z
.
Ari
ef
,
K
w
an Y
i
e
w
La
u,
N
ov
i
z
o
n.
Pa
rt
i
a
l
D
i
sch
a
rg
e
C
har
ac
t
er
i
s
t
i
c
s
i
n
L
LD
P
E
-
N
a
t
ur
al
R
ubb
er
-
B
l
en
ds
:
C
or
r
e
l
at
i
ng
E
l
ec
t
r
i
c
al
Q
uant
i
t
i
es
w
i
t
h
S
ur
f
ac
e
D
egr
adat
i
on.
J
our
n
al
of
E
l
ec
t
r
i
c
al
E
n
gi
n
eer
i
n
g &
T
ec
hnol
og
y
.
2016.
[7
]
M
N
edj
ar
,
A
B
er
ou
al
.
A
gi
ng
und
er
A
C
v
o
l
t
a
ge
of
p
ol
y
ur
et
h
ane
by
us
i
ng
W
ei
bul
l
s
t
at
i
s
t
i
c
.
2007
I
ee
e
I
nt
er
nat
i
o
na
l
C
onf
er
e
nc
e o
n S
ol
i
d D
i
e
l
ec
t
r
i
c
s
,
V
ol
s
1 and
2
,
200
7:
654
-
657.
[8
]
H
G
Kra
n
z.
P
a
r
t
i
al
D
i
s
c
har
ge
E
v
al
u
at
i
o
n
of
P
ol
y
et
h
y
l
e
ne
C
abl
e
-
M
at
er
i
al
b
y
P
h
as
e
A
n
gl
e
A
nal
y
s
i
s
.
I
eee T
r
ans
a
c
t
i
ons
on E
l
ec
t
r
i
c
al
I
n
s
ul
at
i
o
n.
1981;
EI
-
17
:
15
1
-
15
5.
[9
]
HN
N
a
gam
a
ni
,
S
N
M
oor
c
h
i
ng
.
A
s
t
udy
on t
he i
nf
l
u
enc
e of
par
t
i
a
l
di
s
c
har
ges
o
n t
he p
er
f
or
m
a
nc
e
o
f
s
o
lid
d
ie
l
e
c
t
r
ic
s
.
I
n
C
o
nd
uc
t
i
on
an
d B
r
e
ak
d
o
w
n
i
n
S
o
l
i
d
D
i
el
ec
t
r
i
c
s
,
19
98
.
I
C
S
D
'
98.
P
r
oc
ee
di
n
gs
of
t
he 199
8 I
E
E
E
6t
h
I
nt
er
nat
i
on
al
C
o
nf
er
enc
e
on
.
199
8
:
127
-
130.
[
10]
G
P
aol
et
t
i
,
A
G
ol
ub
ev
.
P
ar
t
i
al
di
s
c
har
ge
t
heor
y
a
nd t
ec
hn
ol
o
gi
es
r
el
at
e
d t
o t
r
adi
t
i
ona
l
t
es
t
i
n
g
m
et
h
ods
of
l
ar
ge r
ot
at
i
n
g ap
p
ar
at
us
.
P
ho
eni
x
,
A
Z
,
U
S
A
.
1999
:
967
-
98
1.
[
11]
PH
F
Mo
rsh
u
i
s.
D
egr
a
dat
i
on o
f
s
ol
i
d d
i
el
ec
t
r
i
c
s
due t
o i
nt
er
nal
p
ar
t
i
a
l
di
s
c
har
g
e:
s
om
e t
hou
ght
s
o
n
pr
o
gr
es
s
m
ad
e and
w
h
er
e t
o go no
w
.
D
i
e
l
e
ct
ri
cs a
n
d
El
e
ct
ri
ca
l
I
ns
u
l
at
i
on,
I
E
E
E
T
r
ans
ac
t
i
ons
o
n.
200
5;
12
:
905
-
913.
[
12]
F
G
uas
t
av
i
no,
A
D
ar
dan
o,
A
R
at
t
o,
E
T
o
re
l
l
o
,
M H
o
y
o
s,
JM
G
om
ez
-
El
vi
ra
,
et
al
.
R
e
si
st
a
n
ce
to
s
ur
f
ac
e
par
t
i
al
di
s
c
har
ges
of
LD
P
E
nan
oc
o
m
p
os
i
t
es
.
I
n
E
l
ec
t
r
i
c
al
I
ns
ul
at
i
on
a
nd
D
i
el
ec
t
r
i
c
P
hen
om
e
na,
2
007.
C
E
I
D
P
2
0
07.
A
nnu
al
R
e
por
t
-
C
onf
er
e
n
c
e on
.
20
07
:
2
44
-
2
47.
[
13]
Y
Y
al
m
an.
A H
i
st
ogr
am
bas
ed
I
m
age Q
ual
i
t
y
I
nde
x
.
P
r
z
e
gl
a
d E
l
ek
t
r
ot
ec
h
ni
c
z
ny
.
2
012;
88
:
126
-
1
29
.
[
14]
AH
Ba
y
d
u
sh
,
C
E F
l
o
y
d
Jr.
I
m
pr
ov
e
d i
m
a
g
e qu
al
i
t
y
i
n d
i
g
i
t
al
m
am
m
o
gr
a
ph
y
w
i
t
h i
m
a
g
e pr
oc
es
s
i
n
g
.
Me
d
Ph
ys.
20
0
0;
27
:
150
3
-
15
0
8.
[
15]
M
G
r
undl
a
nd
,
N
A
D
odgs
on
.
A
ut
om
at
i
c
c
ont
r
as
t
en
h
an
c
em
ent
b
y
hi
s
t
ogr
am
w
ar
pi
n
g.
C
om
put
e
r
V
i
s
i
o
n and G
r
a
phi
c
s
.
20
06;
32
:
293
-
30
0.
Evaluation Warning : The document was created with Spire.PDF for Python.