Int
ern
at
i
onal
Journ
al of Ele
ctrical
an
d
Co
mput
er
En
gin
eeri
ng
(IJ
E
C
E)
Vo
l.
8
, No
.
6
,
Decem
ber
201
8
, p
p.
5399
~
5
408
IS
S
N:
20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v
8
i
6
.
pp
5399
-
5
408
5399
Journ
al h
om
e
page
:
http:
//
ia
es
core
.c
om/
journa
ls
/i
ndex.
ph
p/IJECE
A Diagn
ostic An
alyt
i
cs of H
ar
m
on
ic Source Si
gnatu
re
Recognit
ion by U
sing P
er
i
odogram
M.H
.
Jopri
1
, A.
R
.
Abdul
la
h
2
, T
.
Sut
ik
no
3
, M.
Manap
4
,
M
.
R
.
A
b G
hani
5
, A
.
S
.
Hussi
n
6
1
,4
Cent
er
for
Ro
boti
cs
and
Indus
tri
al Autom
at
ion
(CeRIA), Fac
ul
t
y
of
Eng
ineeri
ng
Technol
og
y
,
Univer
siti
Te
kn
i
kal
Ma
lay
sia
,
M
al
a
y
si
a
2
,5,6
CeRIA,
Fa
cu
lty
of
Elec
tri
c
al
Engi
ne
eri
ng,
Un
ive
rsiti T
ekni
ka
l
Malay
si
a
,
Mal
a
y
sia
3
Depa
rtment of
El
e
ct
ri
ca
l
Eng
in
ee
ring
,
Univ
ersitas
Ahm
ad
Dahlan (UAD
),
Indon
esia
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
J
un
9
, 201
8
Re
vised
Ju
l
14
,
201
8
Accepte
d
J
ul
30
, 2
01
8
Thi
s
pape
r
pre
s
ent
s
a
dia
gnosti
c
ana
l
y
t
ic
s
of
har
m
onic
source
signat
ur
e
rec
ogni
ti
on
of
r
ec
t
ifi
er
and
inv
ert
er
-
b
ase
d
lo
ad
in
the
distr
ibution
s
y
stem
with
single
-
poin
t
m
ea
surem
ent
a
t
th
e
poin
t
of
co
m
m
on
coupl
ing
b
y
u
ti
l
iz
ing
Periodogra
m
.
Signat
ur
e
rec
ogn
it
io
n
patter
n
is
used
to
disti
nguish
the
har
m
onic
source
s
ac
cur
a
te
l
y
b
y
obta
ini
ng
the
di
stribut
ion
of
har
m
onic
an
d
int
erh
armonic
c
om
ponent
s
and
the
har
m
oni
c
co
ntri
buti
on
cha
ng
es.
Thi
s
is
ac
hi
eve
d
b
y
u
sing
the
signifi
ca
n
t
signat
ure
rec
ognition
of
har
m
oni
c
produ
ci
ng
loa
d
obta
in
ed
from
ana
l
y
s
ing
the
har
m
onic
cont
ributi
on
cha
nges.
Based
on
vol
ta
g
e
and
cur
ren
t
sig
nat
ure
anal
y
s
is,
t
he
distri
bu
ti
on
o
f
har
m
oni
c
components
ca
n
be
divi
d
ed
in
t
o
thre
e
zone
s
.
To
disti
nguish
bet
wee
n
the
har
m
onic
produ
ci
ng
lo
ads,
the
har
m
onic
comp
onent
s
ar
e
obser
ved
a
t
th
ese
zone
s
to
g
et
the
signat
ure
re
cog
nit
ion
p
at
t
ern
.
T
he
result
demon
strat
e
that
per
iodogra
m
technique
a
cc
ur
atel
y
d
ia
gnose
a
nd
disti
nguish
the
t
y
pe
of
har
m
onic
sourc
e
s in
th
e
d
istri
but
i
on
s
y
st
em
Ke
yw
or
d:
Diag
nosti
c ana
ly
ti
c
Har
m
on
ic
s
our
ce
Sign
at
ur
e
an
al
ysi
s
Sign
at
ur
e
r
ec
ogniti
on
Perio
dogram
Copyright
©
201
8
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
Al
l
rights re
serv
ed
.
Corres
pond
in
g
Aut
h
or
:
M.
H
.
J
opri
,
Adva
nced Di
gital
Sign
al
P
r
oc
essing G
rou
p,
Faculty
of
Ele
c
tric
al
Engineer
ing
,
Un
i
ver
sit
i Te
knikal M
al
ay
sia
Mel
aka, Ha
ng Tu
a
h
Jay
a,
76100 D
ur
ia
n Tu
nggal, Mel
a
ka,
Mal
ay
sia
Em
a
il
: hatt
a@u
te
m
.ed
u.
m
y
1.
INTROD
U
CTION
The
inc
rease
d
and
wi
desprea
d
use
of
the
di
stribu
te
d
har
m
on
ic
s
ources
i
n
powe
r
syst
em
is
le
ading
to
qu
a
ntifyi
ng
th
e
ind
ivid
ual
ha
rm
on
ic
i
m
pa
ct
s
of
the
m
ul
ti
ple
har
m
on
ic
producin
g
lo
ads.
T
hese
ha
rm
on
ic
so
urces
a
re
ca
us
e
d
by
nonlin
ear
loa
ds
t
hat
are
at
ta
che
d
to
the
power
sys
tem
and
a
re
ge
ner
al
ly
un
des
irable.
The
a
bili
ty
to
m
ai
ntain
vo
lt
age
a
nd
c
urre
nt
sig
nals
with
c
onsta
nt
am
plit
ud
e
an
d
c
on
sta
nt
fun
da
m
ental
fr
e
qu
e
ncy
pr
es
ents
the
qu
al
it
y
of
el
ect
rical
powe
r
s
upplied
to
t
he
cu
stom
ers.
The
li
m
it
a
ti
on
of
ha
rm
on
ic
s
po
ll
utio
n
pr
ese
nt
at
the
po
i
nt
of
c
omm
on
cou
plin
g
(
PCC
)
had
bee
n
co
ntr
ol
by
IEE
E
Std.
519
to
im
pr
ov
e
t
he
powe
r
qual
it
y wh
ic
h
is
widel
y ac
cepted i
n
i
ndus
try
[1]
-
[
3]
. Th
e
interact
io
n betwee
n har
m
on
ic
so
urces
and the
ac
powe
r
s
up
ply
can
res
ult
in
a
bove
al
lo
wab
le
le
vels
of
ha
rm
on
ic
di
stortio
n,
over
he
at
ing
of
m
o
t
or
an
d
trans
form
ers,
ov
e
rloa
ding
of
phase
a
nd
ne
utral
c
onduct
ors
in
branc
h
di
stribu
ti
on
ci
rc
uits
or
inte
rf
e
r
ing
t
he
op
e
rati
on
of
se
ns
it
ive
el
ect
r
onic
eq
ui
pm
ent
[4]
,
[
5]
.
T
his
poses
a
po
te
ntial
threat
t
o
the
regular
pe
rform
ance
of
powe
r
syst
e
m
.
Id
e
ntifyi
ng
the
source
of
t
he
har
m
onic
distor
ti
on
is
not
sim
ply
by
j
us
t
l
ooking
at
the
wav
e
f
or
m
s.
D
evelo
ping
a
n
a
ccur
at
e
m
et
hod
a
nd
te
ch
niqu
e
f
or
i
den
ti
fica
ti
on
of
ha
rm
onic
source
s
is
re
qu
i
red
to
rem
ov
e
this
pro
blem
eff
ic
ie
ntly
,
especial
ly
wh
en
the
re
is
har
m
on
ic
pro
blem
in
a
syst
e
m
,
and
i
m
pr
ov
es
the
powe
r
qual
it
y
m
anag
em
ent.
Sever
al
m
eth
od
ologies
w
ere
pro
posed
ov
e
r
the
ye
ar
s
for
detect
in
g
an
d
identify
in
g
the
har
m
on
ic
source
locati
o
n
a
nd co
ntri
bu
ti
on
s
[
6]
. How
e
ve
r,
the con
ven
ti
ona
l t
echn
iq
ues
th
at
are
currently
used
for powe
r qu
al
it
y
m
on
it
or
in
g are
based o
n vi
su
al
of
volt
age
and cu
rr
e
nt
wa
vefor
m
s
[7]
,
[8]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
8
, N
o.
6
,
Dece
m
ber
201
8
:
5399
-
5
408
5400
Sign
al
proce
ss
ing
te
c
hniq
ues
ha
ve
bee
n
w
idely
us
e
d
for
autom
at
ic
power
qual
it
y
disturban
ce
s
detect
ion
a
nd
cl
assifi
cat
ion
[
9]
.
The
m
os
t
widely
adopte
d
ap
proac
h
in
sign
al
proce
ssing
is
sp
ect
ral
analy
sis
us
in
g
Four
ie
r
analy
sis
w
hich
is
F
ourier
t
ra
ns
f
or
m
[10]
.
F
ro
m
the
ap
plica
ti
on
of
Disc
re
te
Four
ie
r
T
ra
ns
f
or
m
(D
F
T), i
t i
s propo
s
ed
i
n
[11]
a
m
e
tho
d t
o
e
xtr
act
the am
plit
ud
e a
nd phase a
ng
le
of cu
rrent
s but co
ns
i
der
i
ng on
the
com
pone
nts
of
odd
num
ber
on
ly
(
from
t
he
fun
dam
enta
l
up
to
15
th
harm
on
ic
order).
I
t
us
e
d
m
od
el
s
base
d
on
A
rtific
ia
l
Neural
Net
work
(AN
N)
t
o
dev
el
op
an
a
ppli
cat
ion
f
or
identific
at
io
n
of
loa
ds
pr
ese
nt
in
the
el
ect
rical
syst
em
us
ing
bo
t
h
the
am
plit
ud
e
and
ph
as
e
a
ngle
of
the
m
eas
ur
e
d
c
urre
nts.
Unfortu
natel
y,
th
e
par
am
et
er
of
th
e
phase
a
ngle
of
t
he
c
urre
nts
can
c
om
pr
om
i
se
the
har
m
on
i
c
source
ide
ntific
at
ion
as
s
ho
wn
by
[12]
.
As
a
res
ult,
it
is
pro
posed
in
[
13
]
a
researc
h
on
th
e
m
e
tho
d
to
i
den
ti
fy
har
m
on
ic
sou
rces
usi
ng
t
he
a
m
plit
ud
e
of
the
m
easur
ed
current
only
.
A
n
at
tribu
te
sel
ect
or
nam
ed
Wrapp
e
r
is
com
bi
ned
with
the
D
FT
of
current
f
or
t
he data
pre
-
proces
sing
ste
p
t
o
obt
ai
n
the m
os
t re
le
van
t am
plit
ud
es
of
ha
rm
on
ic
cu
r
re
nt ord
e
r
s.
S
o,
these
sel
ect
ed
a
m
plit
ud
es
of
har
m
on
ic
cu
rrents
are
giv
e
n
as
inputs
to
the
intel
li
gen
t
syst
e
m
.
Usu
al
ly
the
sel
ect
ion
of
at
trib
utes
by
W
r
app
e
r
pro
vi
des
bette
r
resu
lt
s,
howev
e
r,
this
hig
hly
pe
rform
ance
com
es
with
a
high c
om
pu
ta
ti
on
al
c
os
t
[14]
due
to t
he
e
xha
us
ti
ve
sea
rc
h m
et
ho
d.
In
[15]
,
it
intro
duce
d
a
us
e
r
fr
ie
ndly
an
d
easy
-
to
-
us
e
to
ol,
nam
ed
“H
arm
o
Si
m
,”
to
pr
e
dict
a
nd
si
m
ulate
the
beh
avi
or
of
nonl
inear
l
oa
ds
.
U
sers
are
a
ble
to
us
e
this
to
ol
to
do
a
c
om
par
at
ive
stud
y
of
sever
a
l
loads
on
their
own
data.
It
us
ed
grap
hic
al
visu
al
iz
at
ion
of
di
ff
e
ren
t
par
am
et
ers
su
ch
a
s
total
ha
rm
on
ic
distor
ti
on
(T
H
D)
,
li
ne
cu
rr
e
nt,
powe
r
factor
an
d
crest
cof
act
or
.
Howe
ver,
if
the
power
syst
em
c
on
ta
i
ns
si
m
il
ar
ty
pe
of
loa
ds
,
their
cu
rr
e
nt
har
m
on
ic
sp
ect
ra
will
be
quit
e
sim
i
la
r.
The
n
it
will
be
dif
ficult
t
o
predict
the
co
rr
ect
harm
on
ic
loads
in
this
cases
w
he
n
co
ns
i
der
in
g
current
har
m
on
ic
s
only
[16]
.
Ther
e
f
or
e
it
is
not
al
ways
true
th
at
cur
re
nt
sp
ec
tra
it
sel
f
can
a
ble
to
identify
the
load
sp
ec
ia
ll
y
the
si
m
ilar
ty
pe
of
nonl
inear
loads
.
Fo
ll
owi
ng
this
co
nte
xt,
to
overc
om
e
this
diff
ic
ulty
,
the
propose
d
m
et
ho
d
util
iz
es
the
vo
lt
age
m
easur
em
ents
in
place
of
c
urren
t
m
easurem
ent
to
ident
ify
the
ty
pe
of
nonline
ar
lo
ad
[
17
]
.
T
he
vo
lt
age
ma
gn
it
ud
e
m
eth
od
ca
n
be
us
e
d
to
detect
the
har
m
on
ic
sou
r
ce,
howe
ve
r
th
e
proce
dure
is
no
t
pr
act
ic
al
be
cause
need
to
do
swit
chin
g
te
st
on
t
he
sit
e
[
18
]
.
I
n
real
tim
e,
accuracy,
quic
k
e
s
tim
a
ti
on
an
d
l
ow
c
os
t
are
im
po
rt
a
nt
and
m
os
t
of
th
e
pr
e
vious
st
udie
s
do
not
ta
ke
into
acc
ount
[
19]
,
[
20
]
.
Ba
se
d
on
a
bo
ve
disc
ussi
on,
the
li
m
i
t
at
io
n
of previ
ous tec
hn
i
qu
e
s ca
n be
overc
om
e u
sin
g peri
odogram
.
This
pa
per
pro
po
s
es
hi
gh
acc
ur
acy
,
fast
est
im
at
ion
and
c
ost
s
eff
ic
ie
nt
to
identify
an
d
di
agnose
ty
pe
of
M
HS
in
the
distrib
utio
n
syst
e
m
with
singl
e
-
point
m
easur
em
ent
at
the
PCC
by
util
iz
ing
pe
rio
dogra
m
.
The
per
i
odogram
i
s
a
m
et
ho
d
to
represe
nt
sign
al
s
in
tim
e
-
fr
e
qu
e
ncy
re
presentat
io
n
(T
FR).
The
diagnostic
analy
ti
cs
of
MHS
is
achie
ved
by
analy
z
e
the
volt
age
an
d
c
urre
nt
sign
al
s
by
us
i
ng
the
pe
rio
dogram
.
Fu
rt
her
m
or
e,
ha
rm
on
ic
ind
ic
ie
s
su
c
h
as
in
sta
ntane
ous
total
har
m
on
ic
dist
or
ti
on
a
nd
i
ns
t
antane
ous
total
non
-
har
m
on
ic
disto
rtion
are
us
e
d
to
dif
fer
e
ntiat
e
the
har
m
on
ic
and
inte
rh
a
rm
on
ic
com
po
ne
nts
in
the
sign
al
s
.
The
vo
lt
age
an
d
c
urre
nt
sig
natu
re
of
eac
h
har
m
on
ic
source
s
a
re
a
naly
zed
by
stud
yi
ng
t
he
ha
rm
on
ic
co
ntri
bu
ti
on
changes
.
Finall
y,
the
disti
ngui
sh
m
ent
of
MH
S
can
be
obta
i
ned
acc
urat
el
y
by
us
i
ng
t
he
s
ign
at
ure
r
eco
gnit
ion
patte
rn
t
hat
be
en
obta
ine
d
f
rom
the
d
ist
ribu
ti
on
o
f
ha
rm
on
ic
and
i
nterh
a
rm
on
ic
c
om
po
ne
nts
an
d
the h
ar
m
on
ic
con
t
rib
ution ch
ang
e
s.
2.
RESEA
R
CH MET
HO
D
A
com
m
on
typ
e
of
ha
rm
on
ic
producin
g
l
oads
are
sel
ec
te
d
an
d
m
od
el
le
d
in
Si
m
ulink
Ma
tl
ab
.
Fu
rt
her
m
or
e,
the
dia
gnos
ti
c
analy
ti
cs
of
HP
L
sig
nat
ur
e
rec
ogniti
on
is
perform
ed
by
util
iz
ing
the
per
i
odogram
.
2.1 Harm
on
ic
Prod
ucing
Lo
ad
A
c
omm
on
source
of
ha
rm
on
ic
c
urren
ts
in
powe
r
syst
em
s
su
c
h
as
powe
r
el
ect
ronic
c
onver
te
r
s.
Th
e
har
m
on
ic
-
pro
duci
ng
load
ca
n
hav
e
fairly
co
ns
ist
ent
ha
rm
on
ic
current
em
issi
on
cha
racte
risti
cs
ov
er
ti
m
e
or
each
m
ay
pr
esent
a
widely
-
va
ryi
ng
c
har
act
erist
ic
dep
e
nd
i
ng
on
the
c
on
t
ro
l
of
the
de
vice,
the
cha
ract
erist
ic
s
of
t
he
syst
em
,
and
oth
e
r
var
i
ables
(
IEEE
519
-
2014).
The
powe
r
el
ect
ronics
eq
uip
m
ent,
su
c
h
as
th
re
e
-
phase
recti
fiers
an
d
three
-
phase
in
ve
rter
with
pu
ls
e
width
m
od
ul
at
ion
(PWM
)
s
witc
hing
are
w
idely
us
ed
a
nd lead
to
the
distu
rb
a
nc
es
of
the
util
ity
and
the
e
nd
-
us
e
r’
s
e
quipm
ent
[21]
-
[2
3]
.
Th
us
,
in
order
to
identify
th
e
root
cause
a
nd
to
m
itigate
the
ha
rm
on
ic
distu
r
ban
ce
,
diag
nosti
c
analy
ti
cs
of
ha
rm
on
ic
so
urces
is
nece
ssary
to do so
.
2.2 Peri
odog
r
am
A
har
m
on
ic
s
pectr
um
of
non
-
li
nea
r
load
s
power
analy
s
er
wh
ic
h
cha
r
act
erizes
the
beh
a
vior
of
har
m
on
ic
source
s to
prov
i
de preli
m
inary d
at
a o
n
the se
ver
it
y of
the d
ist
ort
ion
pr
ob
le
m
is
widely
e
m
plo
ye
d
by
the
util
it
ie
s.
The
per
i
odogra
m
sp
ect
ral
represents
t
he
dis
tribu
ti
on
of
po
wer
sig
nal
ove
r
f
reque
ncy
an
d
it
is
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
A D
i
agnosti
c A
na
ly
ti
cs
of Har
monic
Sour
ce
Signa
t
ur
e Re
co
gn
it
io
n by
Usi
ng Perio
dogra
m
(
M.
H
J
op
ri
)
5401
base
d
on
a
F
ourier
se
ries
m
od
el
of
the
data,
th
at
is,
sp
eci
fical
ly
desig
ne
d
f
or
determ
ining
ha
rm
on
ic
com
po
ne
nts in
powe
r
sig
nals
[
24
]
,
[
25
]
.
I
t ca
n be
form
ulated
as
equati
on
1.
(
)
=
|
1
∫
(
)
−
2
2
−
2
|
2
(1)
Wh
e
re,
(
)
in freq
uen
cy
dom
ai
n;
(
)
is v
oltage
wa
ve
form
; VD
C i
s d
irect
c
urre
nt
vo
lt
age
and
2
is
syst
e
m
f
requenc
y.
2.3 I
n
st
antane
ou
s T
otal
Ha
r
mon
ic
Dist
ort
i
on
Total
har
m
on
i
c
distor
ti
on,
T
HD,
is
us
e
d
to
m
easur
e
of
how
m
uch
ha
rm
on
ic
co
ntent
in
a
wav
e
f
or
m
[26]
,
[
27]
,
[
8]
. T
he
total
har
m
on
ic
d
ist
or
ti
on
of a
wa
vefor
m
is d
e
fine
d
m
at
he
m
at
ic
ally as equ
at
io
n 2.
(
)
=
√
∑
ℎ
,
(
)
2
ℎ
=
2
1
(
)
(2)
Wh
e
re
V
h,rms
(
t
)
is R
MS h
arm
on
ic
volt
age
and
V
h,rms
(
t
)
is R
MS f
undam
ental
v
oltage
.
2.4
In
st
antane
ou
s T
otal
Non
-
H
armoni
c D
i
sto
r
tio
n
Be
sides
ha
rm
on
ic
,
a
si
gn
al
al
so
c
on
ta
in
s
int
erh
a
rm
on
ic
co
m
po
nen
ts
that
are
not
m
ulti
pl
e
integer
of
the
power
syst
e
m
fr
e
qu
e
ncy
[28]
,
[
29]
.
Re
fe
rr
in
g
to
t
his
a
s
the
t
otal
no
nharm
on
ic
disto
rtion,
(TnH
D),
the
expressi
on in
e
qu
at
io
n 3
[30]
:
(
)
=
√
(
)
2
−
∑
ℎ
,
(
)
2
ℎ
=
0
1
(
)
(3)
2.5 Pr
oposed
Te
chnique I
m
plem
ent
at
i
on
The
e
xec
ution
of
pro
posed
te
chn
i
qu
e
can
be
reali
sed
as
de
picts
in
Fi
gure
1
a
nd
Fig
ur
e
2.
Tw
o
ty
pe
s
of
com
m
on
an
d
widely
use
d
har
m
on
ic
s
our
ces
w
hich
are
the
th
ree
-
phas
e
recti
fier
a
nd
three
-
phase
i
nverter
with
pu
ls
e
w
i
dt
h
m
od
ulati
on
(PW
M
)
s
witc
hi
ng
[
31
]
,
[32]
ar
e u
se
d
i
n
this
r
esearch
.
Me
anwhil
e,
as
sh
w
own
i
n
Fi
gure
3,
the
m
e
asur
em
ent
data
are
analy
sed
by
us
in
g
per
i
odog
ram
and
fr
e
qu
e
ncy
s
pe
ct
ru
m
.
Fu
rt
herm
or
e,
the
harm
on
ic
ind
ic
es
wh
ic
h
a
re
inst
antane
ous
tota
l
har
m
on
ic
distor
ti
on
,
THD(
t
)
an
d
in
sta
ntane
ou
s
t
ot
al
non
-
har
m
onic
distor
ti
on,
T
nHD(t
)
are
cal
culat
ed.
T
he
c
on
t
rib
ution
c
ha
ng
e
s
and
distrib
utio
n
of
har
m
on
i
c
an
d
inter
ha
r
m
on
ic
com
ponen
ts
are
obta
ined
f
ro
m
the
dia
gnos
ti
c
a
na
ly
tic
process
.
Finall
y
,
the
disti
ng
uish
m
ent
of
ha
rm
on
ic
sourc
es
can
be
obta
ined
by
ob
se
rv
i
ng
t
he
si
gnat
ure
recog
niti
on
pat
te
rn
.
Figure
1
.
A
r
ec
ti
fier
-
base
d
l
oa
d.
Figure
2.
A
n
i
nverter
-
based lo
ad.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
8
, N
o.
6
,
Dece
m
ber
201
8
:
5399
-
5
408
5402
Figure
3. A
n
i
m
ple
m
entat
ion
of
har
m
on
ic
s
ource
diag
nosti
c analy
ti
cs.
3.
RESU
LT
S
A
ND AN
ALYSIS
The
res
ults
of
an
analy
sis
of
har
m
on
ic
ind
ic
es
,
ha
rm
on
ic
co
ntributi
on
c
ha
ng
es
a
nd
si
gn
at
ure
recog
niti
on
pat
te
rn
of
ha
rm
onic
so
urces
w
hich
a
re
recti
fier
-
base
d
l
oad
an
d
in
ve
rter
-
base
d
loa
d
pr
e
sent
ed
i
n
this sect
ion.
3.1 Rectifie
r
-
b
as
ed L
oad
In
t
his
case
st
ud
y,
t
he
sig
nat
ur
e
recog
niti
on
of
r
ect
ifie
r
-
ba
sed
loa
d
is
obser
ve
d
due
t
o
perform
th
e
har
m
on
ic
sou
r
ce
diag
nosti
c.
A
six
-
pulse
s
three
-
phase
r
ect
ifie
r
with
Pu
lse
W
i
dth
Modula
ti
on
(
P
W
M)
switc
hing
te
c
hniq
ue
is
sel
ect
ed
to
be
the
re
ct
ifie
r
-
bas
ed
l
oa
d.
T
he
volt
ag
e
an
d
c
urren
t
s
ign
al
in
tim
e
do
m
ai
n
wh
e
n
this
recti
fier
-
base
d
load
util
iz
ed
as
the
har
m
on
ic
pro
du
ci
ng
loa
d
show
n
in
Fig
ur
e
4(
a
).
Cl
early
sh
ow
n
that
the
vo
lt
a
ge
an
d
c
urren
t
sign
al
s
a
re
not
a
pu
re
sin
usoi
dal
f
or
m
.
As
s
how
n
in
Fig
ur
e
4(b
),
t
he
harm
on
ic
ind
ic
es
wh
ic
h
are
T
HD
(
t)
and
T
nHD(t
)
in
the
ra
nge
of
1.0
93%
to
1.098
%
an
d
1.2
95%
to
1.3%
,
res
pe
ct
ively
.
These
ha
rm
on
ic
in
dices
sho
w
that
the
vo
l
ta
ge
sign
al
co
ns
ist
of
ha
rm
on
ic
and
inte
rharm
on
ic
com
p
on
e
nts
.
Nev
e
rtheless
,
the
ha
rm
on
ic
distor
ti
on
le
vel
sti
ll
in
the
per
m
it
te
d
ran
ge
of
I
EEE
Stan
dard 5
19
-
20
14,
w
he
re
the
total
h
arm
on
ic
distor
ti
on c
an
not m
or
e tha
n 8
%.
The
cu
rr
e
nt
s
pe
ct
ru
m
of
signa
l
sh
own
in
Fi
gure
4(c)
.
Wh
e
reas
the
ha
rm
on
ic
current
exi
st
at
25
0Hz,
900Hz
a
nd
10
00Hz,
res
pecti
vely
.
I
n
the
m
eantim
e,
the
inter
har
m
on
ic
com
po
ne
nts
disco
ver
e
d
at
375Hz
,
425Hz, 57
5H
z
,
an
d
67
5H
z
, r
e
sp
ect
ively
. A
s
this 6
-
pu
lse
s re
ct
ifie
r,
the h
ar
m
on
ic
o
rd
er s
t
art at
5
th
wh
il
e the 3
rd
har
m
on
ic
an
d
belo
w
are
el
im
inate
d.
As
sho
wn
in
Fig
ur
e
4(d
),
the
vo
lt
ag
e
sign
al
com
pri
sing
f
our
ha
r
m
on
i
c
and
t
wo
inter
ha
rm
on
ic
com
po
ne
nts,
i
nd
i
viduall
y.
Ne
xt,
to
hav
e
a
go
od
unde
rstan
ding
of
c
onve
rter
-
ba
sed
load, a
d
ia
gn
ost
ic
an
al
yt
ic
o
f c
onve
rter
-
base
d
loa
d
is
p
e
rform
ed
an
d
e
xp
la
i
n
in
n
e
xt secti
on.
(a)
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
A D
i
agnosti
c A
na
ly
ti
cs
of Har
monic
Sour
ce
Signa
t
ur
e Re
co
gn
it
io
n by
Usi
ng Perio
dogra
m
(
M.
H
J
op
ri
)
5403
(b)
(c)
(d)
Figure
4. A
Frequ
e
ncy S
pectr
um
. a)
A insta
ntane
ous
vo
lt
a
ge
a
nd curre
nt
sign
al
,
b)
T
H
D
(
t
)
and
T
nHD(t
)
of
vo
lt
age
sig
nal,
c) Freq
ue
ncy s
pectr
um
o
f
ha
r
m
on
ic
an
d
i
nte
rh
a
rm
on
ic
of c
urren
t
sig
nal,
d) F
reque
ncy
sp
ect
r
um
o
f ha
rm
on
ic
and int
erh
a
rm
on
ic
of
vo
lt
age
sig
nal.
3.2 Co
nv
er
ter
-
based L
oad
This
case
stu
dy
seeks
to
a
ddress
the
si
gn
at
ur
e
recog
niti
on
of
in
ver
te
r
-
ba
sed
loa
d
a
nd
the
fin
ding
sh
oul
d
m
ake
an
i
m
po
rta
nt
con
t
rib
ution
to
the
har
m
on
ic
so
urce
diag
nosti
c.
The
th
r
ee
-
phase
AC
-
DC
-
AC
conve
rter
with
Pu
lse
W
idt
h
Modula
ti
on
(
P
WM)
switc
hing
te
chn
i
qu
e
w
as
sel
ect
ed
to
be
the
inv
e
rter
-
base
d
load.
Fi
gure
5(a)
pr
ese
nts
t
he
volt
age
a
nd
c
urren
t
sig
nal
i
n
ti
m
e
do
m
ai
n
w
hen
the
th
re
e
-
phase
in
ver
t
er
was
util
iz
ed
as
the
har
m
on
ic
pro
du
ci
ng
loa
d.
F
ur
t
her
m
or
e,
Fi
gure
5(b)
sho
w
the
ha
rm
on
ic
ind
ic
es
w
hich
ar
e
THD(
t
)
and
T
nHD(t
)
in
the
range
of
1.4
65%
to
1.475%
and
6.0
1%
t
o
6.05%,
resp
ect
ively
.
This
harm
on
ic
ind
ic
es
s
how
t
hat
the
har
m
onic
and
i
nter
harm
on
ic
com
po
ne
nts
are
e
xist
in
the
si
gn
al
.
H
ow
e
ve
r,
the
ha
rm
on
ic
distor
ti
on
le
vel
sti
ll
in
the
perm
itted
ra
ng
e
of
I
EEE
Stan
da
rd
519
-
2014,
wh
e
re
t
he
total
ha
rm
on
ic
distor
ti
on
cannot m
or
e th
an 8%.
The
sp
ect
ru
m
of
cu
rr
e
nt
dra
wn
by
the
inv
e
rter
show
n
in
Figure
5(c).
W
her
e
by
the
harm
on
ic
cur
ren
t
exist
at
25
0Hz,
350H
z
,
600H
z,
900Hz
and
1000Hz
,
res
pe
ct
ively
.
Me
anw
hile,
the
inter
har
m
on
ic
com
pone
nts
disco
ver
e
d
at
575Hz,
675Hz
and
875
Hz,
r
especti
vely
.
Fi
gure
5(
d)
s
hows
the
f
re
qu
e
nc
y
sp
ect
ru
m
of
vo
lt
age
and
it
was
disti
nguis
hed
f
our
har
m
on
ic
an
d
six
interha
rm
on
ic
com
po
ne
nt
s
in
the
vo
lt
ag
e
sign
al
,
res
pe
ct
ively
.
Pr
e
vious
analy
sis
of
h
a
rm
on
ic
and
inter
ha
r
m
on
ic
con
trib
ution
of
recti
fi
er
-
base
d
load
is
done
an
d
f
ur
t
her
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
8
, N
o.
6
,
Dece
m
ber
201
8
:
5399
-
5
408
5404
analy
sis
on
c
onve
rter
-
base
d
load
is
necessa
ry
due
t
o
c
on
cl
ud
e
the
hypothesis
of
si
gnat
ur
e
rec
ogniti
on
of
HP
L.
In
a
ddit
ion
,
no
pr
e
vi
ou
s
resear
ch
disti
nguish
e
d
the
HP
L
base
d
o
n
ha
rm
on
ic
and
inte
rh
a
r
m
on
ic
con
t
rib
ution o
f
volt
age a
nd c
urre
nt.
(a)
(b)
(c)
(d)
Figure
5
.
AC
-
DC
-
AC P
WM
conve
rter a
) A
instanta
ne
ous
vo
lt
age
and
cu
rr
e
nt sig
nal,
b)
THD(
t
)
and
TnHD(t
)
of volt
age si
gnal
, c)
Fr
e
qu
e
nc
y spectr
um
o
f harm
on
ic
a
nd inter
har
m
on
ic
of c
urren
t si
gn
al
, d) Fre
quency
sp
ect
r
um
o
f ha
rm
on
ic
and int
erh
a
rm
on
ic
of
vo
lt
age
sig
nal.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
A D
i
agnosti
c A
na
ly
ti
cs
of Har
monic
Sour
ce
Signa
t
ur
e Re
co
gn
it
io
n by
Usi
ng Perio
dogra
m
(
M.
H
J
op
ri
)
5405
3.3 Vol
tage
and C
urrent
Si
gnature
Analys
is
Althou
gh
e
xtensiv
e
resea
rc
h
has
bee
n
car
r
ie
d
ou
t
on
harm
on
ic
so
urce
diag
nosti
c,
no
sing
le
st
udy
exists
w
hich
a
dequatel
y
stu
dy
the
relat
io
nship
of
co
ntri
buti
on
im
pact
of
har
m
on
ic
a
nd
inter
har
m
on
ic
a
nd
the
HP
L
t
hat
in
t
he
powe
r
distrib
ution
syst
em
.
The
si
gn
a
t
ur
e
analy
sis
of
vol
ta
ge
an
d
c
urre
nt
is
pe
rfo
rm
e
d
du
e
to
stud
y
the
im
pact
of
the
co
ntri
bu
ti
on
t
o
the
s
yst
e
m
.
Fu
rthe
r
m
or
e,
this
anal
ysi
s
enh
a
nce
our
unde
rstan
di
ng
on
vo
lt
age
an
d
c
urre
nt
sig
natu
re
patte
r
n
to
wards
th
e
dia
gnost
ic
analy
sis
of
HP
L
s.
As
s
how
n
i
n
Fi
gure
6,
th
e
har
m
on
ic
co
ntr
ibu
ti
on
cha
nge
s
(H
CC
)
cl
earl
y
sh
ows
that
the
vo
lt
age
c
on
tr
ibu
ti
on
cha
nge
s
of
co
nverter
-
base
d
load
are
dom
i
nan
t
at
m
os
t
of
f
re
qu
e
ncies
com
par
ing
to
the
recti
fier
-
ba
sed
loa
d.
I
nt
eresti
ng
ly
,
at
275Hz,
300Hz,
350Hz
,
675Hz
a
nd
1025Hz
,
only
har
m
on
ic
an
d
interha
rm
on
ic
vo
lt
age
c
om
ponen
ts
of
c
onve
rter
-
base
d
loa
d
we
re
exist
with
100%
rate
of
H
CC
.
Tu
rn
i
ng
now
to
t
he
vo
lt
age
si
gn
at
ur
e
of
recti
fier
-
bas
ed
lo
a
d
that
on
ly
has
a
do
m
inant
harm
on
ic
co
m
po
ne
nt
at
700H
z
a
nd
900Hz,
re
s
pecti
vely
.
Me
anwh
il
e
at
700Hz,
the
HCC
rate
was
100%
.
This
par
a
gr
a
ph
ha
s
analy
sed
the
vo
lt
age
sig
natur
e
,
w
hile
the
nex
t
par
a
gr
a
ph
will
exp
la
in
the
an
a
ly
si
s o
f
c
urre
nt sig
natur
e
of
H
PLs.
Figure
6
.
V
oltage
Si
gn
at
ur
e
Analy
sis.
Wh
at
fo
ll
ows
i
s
an
ex
planati
on
of
cu
rr
e
n
t
sig
natu
re
of
recti
f
ie
r
-
ba
sed
l
oad
and
c
onve
rter
-
base
d
loa
d.
On
e
of
the
sig
natu
re
com
pone
nt
that
co
ntribute
in
the
de
ve
lop
m
ent
of
HPL’s
sig
natu
re
r
ecognit
ion
is
curren
t
sign
at
ur
e.
As
sh
ow
n
in
Fig
ure
7,
th
e
m
os
t
strikin
g
res
ult
to
e
m
erg
e
fro
m
the
resu
lt
i
s
that
the
HCC
s
of
conve
rter
-
base
d
loa
d
hav
e
be
com
e
so
do
m
inant.
Th
e
HCC
rate
f
or
conve
rter
-
base
d
loa
d
wer
e
100%
at
350Hz,
600Hz
and
87
5H
z
.
T
hese
ar
e:
250Hz,
350Hz,
60
0H
z
,
87
5H
z
,
900Hz
a
nd
10
00Hz
.
Com
par
i
ng
t
he
current
sig
natu
re
of
c
onver
te
r
-
base
d
loa
d,
i
t
can
be
seen
th
at
the
CC
OTs
for
recti
fier
-
ba
sed
load
with
r
at
e
of
100%
at 3
75H
z an
d 425
Hz.
I
n
the
m
eantim
e, it was
dom
inant to
o
at
575Hz a
nd yi
el
ded a 58%
of
HC
C.
A
stron
g
relat
ion
s
hip
of
vo
l
ta
ge
an
d
cu
rrent
sign
at
ur
e
ver
s
us
HCC
is
interest
ing
be
cause
th
is
sign
ific
a
nt
rela
ti
on
s
hip
ca
n
be
util
iz
ed
in
di
agnostic
the
ab
ov
e
sai
d
H
PLs
.
I
n
t
his
sect
io
n,
it
is
e
xpla
in
ed
tha
t
the H
CC
for
volt
age and
c
urr
ent sign
at
ur
e
of co
nv
e
rter
-
ba
s
ed
loa
d
are s
o do
m
inant co
m
par
e to
recti
fie
r
-
base
d
load.
Figure
7
.
Cu
rr
e
nt Signat
ure
A
naly
sis.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
8
, N
o.
6
,
Dece
m
ber
201
8
:
5399
-
5
408
5406
3.4 S
i
gnat
u
re
Recogni
tion o
f
H
arm
on
ic
P
rodu
ci
n
g
L
oa
d
The
res
ults
obta
ined
f
r
om
the
pr
el
im
inary
analy
sis
wer
e
a
naly
se
and
t
he
sign
ific
a
nt
rec
ogniti
on
of
har
m
on
ic
pro
duci
ng
load
was
ob
ta
ine
d.
W
it
h
ref
e
r
to
the
Figure
7,
the
di
st
ribu
ti
on
of
ha
rm
on
ic
com
p
on
e
nts
can
be
div
i
ded
into
t
hr
ee
zo
ne
s;
zo
ne
1
(15
0H
z
to
450Hz
)
,
zo
ne
2
(
450H
z
to
750Hz)
an
d
z
on
e
3
(
750Hz
t
o
1050Hz
).
F
urt
her
m
or
e,
by
ta
bu
la
ti
ng
the
ha
rm
on
ic
and
i
nterh
a
rm
on
ic
co
m
po
nen
ts
as
de
picts
in
Fig
ure
8,
th
e
m
os
t
striking
r
esult
to
em
erge
from
the
dat
a
is
in
t
he
z
one
1
a
nd
z
one
3,
t
he
ta
bu
la
ti
on
of
har
m
on
i
c
an
d
interha
rm
on
ic
com
po
ne
nts
ar
e
higher
i
n
the
se
zo
ne.
Wh
e
r
eby
f
or
in
ve
rter
-
base
d
loa
d,
i
n
zo
ne
1,
five
vo
lt
ag
e
har
m
on
ic
an
d
2
cu
rr
e
nts
harm
on
ic
com
po
ne
nts
we
re
disti
ngui
s
hed.
I
n
z
on
e
2,
t
hr
ee
c
urren
ts
har
m
on
ic
an
d
two
vo
lt
a
ge
ha
rm
on
ic
com
po
ne
nts
wer
e
di
sti
nguish
e
d
in
the
sig
nal.
F
urt
her
m
or
e,
in
z
on
e
3,
thre
e
volt
ag
e
har
m
on
ic
an
d
three
c
urre
nt
ha
rm
on
ic
com
po
nen
ts
wer
e
det
ect
ed
in
the
si
gn
al
,
res
pecti
ve
ly
.
In
the
m
ea
nti
m
e,
as
show
n
in
F
igure
9,
the
si
gn
at
ur
e
rec
ogniti
on
of
recti
fier
-
base
d
loa
d,
eac
h
z
on
e
consi
st
le
ss
ha
rm
on
ic
com
po
ne
nts
com
par
e
to
the
inv
erte
r
-
base
d
loa
d
sig
natu
re
rec
ogniti
on.
It
can
th
us
be
su
ggest
e
d
th
at
the
conve
rter
-
base
d
loa
d
pro
du
ce
d
m
or
e
ha
rm
on
ic
an
d
i
nterha
rm
on
ic
com
ponen
ts
at
z
one
1
an
d
z
on
e
3
c
om
par
e
to r
ect
ifie
r
-
ba
s
ed
loa
d. Th
us,
the d
ia
gnos
ti
c
of
HPLs s
ucce
ssfu
ll
y d
one
by observin
g
t
he si
gn
at
ur
e
rec
ogniti
on
patte
rn
at
z
one
1,
z
one
2
an
d
zon
e
3.
Th
e
pr
esent
res
ults
ar
e
sign
i
ficant
in
diag
nosin
g
th
e
recti
fier
-
base
d
loa
d
and co
nverter
-
base
d
loa
d
t
hat conn
ect
e
d
t
o
t
he
syst
em
.
Figure
8. Sig
na
ture
Re
c
ogniti
on of C
onve
rter
-
base
d
Loa
d.
Figure
9
.
Sig
na
ture
Re
c
ogniti
on of Rect
ifie
r
-
Ba
sed
Loa
d
4.
CONCL
US
I
O
N
The
m
ai
n
con
cern
of
this
r
esearch
is
to
diag
nose
ty
pe
of
har
m
on
ic
pro
du
ci
ng
loa
d
w
hich
ar
e
connecte
d
i
n
the
distri
bu
ti
on
syst
e
m
and
act
ed
as
ha
rm
on
ic
so
urces
by
ut
il
iz
ing
per
i
odogram
te
chn
iq
ue
.
Th
e
m
ai
n
con
trib
ut
ion
of
this
res
e
arch
is
the
sig
na
ture
rec
ogn
it
ion
patte
rn
of
ha
rm
on
ic
so
urc
es
that
can
be
us
e
d
to
disti
nguish
ty
pe
of
har
m
on
ic
so
urces
.
Tw
o
c
omm
on
and
widely
us
e
d
of
ha
rm
on
ic
produ
ci
ng
loa
d
wh
ic
h
a
r
e
three
-
phase
re
ct
ifie
r
an
d
th
re
e
-
phase
i
nv
e
rt
er
are
m
od
el
le
d
a
nd
ch
os
e
n
i
n
this
resea
rch.
Ba
sed
on
si
gnat
ure
recoginit
on
pa
tt
ern
,
both
ty
pe
s
of
har
m
on
i
c
producin
g
lo
ad
can
be
dist
inguishe
d
plai
nly
by
ob
se
r
vin
g
t
he
zon
e
1,
z
one
2
an
d
z
one
3.
The
i
nv
e
rter
-
ba
sed
l
oad
al
wa
ys
has
m
or
e
ha
rm
on
ic
com
po
ne
nts
for
eac
h
z
on
e
com
par
e to t
he
r
ect
ifie
r
-
base
d l
oad ty
pe.
Con
se
quently
,
con
sid
erin
g
the
ex
planati
on
,
the
per
i
odog
ram
te
chn
iq
ue
accuratel
y
diagnose
d
an
d
disti
nguish
e
d
t
ype of
har
m
on
i
c sour
ce
s in
th
e pow
e
r dist
rib
ution sy
ste
m
.
ACKN
OWLE
DGE
MENTS
This
resear
ch
is
su
pp
or
te
d
by
Advan
ce
d
Di
gi
ta
l
Sign
al
Pr
oc
essing
La
bo
ra
tory
(ADS
P
Lab).
Sp
eci
al
than
ks
al
so
to
the
Faculty
of
Ele
ct
rical
Eng
inee
rin
g
an
d
Faculty
of
E
nginee
rin
g
Tec
hnology
of
U
ni
ver
sit
i
Tek
nik
al
Ma
la
ysi
a
Me
la
ka
(
UTeM)
,
Ce
nte
r
f
or
Ro
boti
cs
an
d
I
ndust
ria
l
Au
t
om
ation
(CeR
IA)
of
U
TeM
,
Mi
nistry
of
H
igh
e
r
Ed
ucati
on
Ma
la
ysi
a
(MOH
E
)
an
d
Mi
nistry
of
S
ci
ence,
Tech
nolo
gy
an
d
Inn
ov
at
io
n
(MOST
I)
f
or
giv
in
g
t
heir
c
oope
rati
on
a
nd
fun
ding
f
or
thi
s
resea
rch
wit
h
gr
a
nt
num
ber
06
-
01
-
14
-
SF
00119
L0
0025. T
heir
su
pp
or
t i
s
grate
fu
ll
y ac
kn
ow
l
edg
e
d.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
A D
i
agnosti
c A
na
ly
ti
cs
of Har
monic
Sour
ce
Signa
t
ur
e Re
co
gn
it
io
n by
Usi
ng Perio
dogra
m
(
M.
H
J
op
ri
)
5407
REFERE
NCE
S
[1]
A.
R.
Abdulla
h
,
M.
Mana
p
,
M.
R
.
Yus
off,
and T.
Sutikno,
“
A fa
st
loc
a
li
z
at
ion
of
m
ult
iple
h
armonic
source
s for
rec
t
ifi
er
lo
ads
b
y
uti
l
izing
pe
riod
ogra
m
,
”
Te
lk
om
nik
a
(
Tele
comm
unic
ati
on
Comp
ut.
El
e
ct
ron.
Co
ntro
l.
,
vol
.
15
,
no.
1
,
2017
.
[2]
M.
H.
Jopri
,
A
.
R.
Abdull
ah, M.
Mana
p,
M
.
R
.
Y
usoff,
T
.
Sutikno
,
and
M.
F.
Habb
an,
“
An i
m
prove
d
detec
t
ion
and
c
la
ss
ifi
c
atio
n
technique of
h
armonic
signa
ls
in
power
d
istri
b
uti
on
b
y
utilizin
g
spec
trogr
am,”
Int.
J. Ele
c
tr.
Comput
.
Eng
.
,
v
ol.
7
,
no
.
1
,
2017
.
[3]
IEE
E
,
I
EE
E
R
ecom
mende
d
Practice
s and
R
equi
re
ments
for Harm
onic
Con
trol
in
El
e
ct
rica
l
Pow
er
Syste
ms
.
1993.
[4]
S.
M.
Nawi
,
J.
J
ohar
i, a
nd
A.
F
.
Abidin,
“
A fuzz
y
logic
appl
i
ca
t
i
on
for
id
entificat
ion
of
h
armonics
disturba
n
ce
s
source
s,”
Natl
.
P
ower
Eng. Conf. PECon
2003
-
P
roc.
,
pp.
27
–
31,
2003.
[5]
M.
H.
Abdul
la
h
,
A.R.
;
Norddin
,
N.;
Abidin
,
N
.
Q.
Z.
,
“
Leaka
ge
cur
ren
t ana
l
y
sis
on
pol
y
m
eric and
n
on
-
pol
y
m
er
ic
insula
ti
ng
m
at
er
i
al
s using tim
e
-
fr
eque
nc
y
distri
bu
ti
on,
”
Pow
er
En
ergy
(
PE
C
on)
,
2
012
IEEE
In
t. C
onf.
,
no.
Dec
ember, pp. 2
–
5,
2012
.
[6]
A.
Abdulla
h
,
M
.
Jopri,
and
M
.
M
ana
p,
“
An Im
prove
d
Spec
trogra
m
to
Id
ent
i
f
y
Mult
ipl
e
Harm
onic
S
ourc
es
in
Distribut
ion
S
y
st
em wit
h
Inv
ert
er
Loa
ds,
”
Proc.
,
vol.
7
,
no
.
4
,
pp
.
1355
–
1365,
201
7.
[7]
N.
H
.
T. Huda
,
A.
R.
Abdulla
h
,
and
M.
H.
Jopri
,
“
Pow
er
qual
ity
s
igna
ls d
et
e
ct
ion
using S
-
tra
nsfor
m
,
”
in
Proce
ed
ings o
f
t
he
2013
I
EEE
7t
h
Inte
rnat
ional
Powe
r E
ng
ine
eri
ng
and
Optimiza
ti
on
Conf
ere
nc
e, P
EOCO
2013
,
2013
.
[8]
A.
R.
Abdulla
h
,
N.
S.
Ahm
ad, N. Ba
har
i,
M.
Man
ap,
A
.
Jidin
,
“
Short
-
ci
r
cui
t
sw
it
c
hes
fau
l
t ana
l
y
si
s of
voltage
source
inv
erter
u
sing spe
ct
rogr
a
m
,
”
Elec
tr.
Ma
c
h.
Syst
.
(
ICEMS)
,
2013
Int. Conf.
,
pp.
1808
–
1813,
2013.
[9]
A.
Rahi
m
Abdul
la
h,
N.
H
.
T
.
H. A
hm
ad,
N. a. Abidull
ah
,
N
.
H.
Sham
sudin,
“
Perfor
m
anc
e Eva
lu
ation
of
R
eal
Pow
er
Quali
t
y
D
isturba
nc
es
Anal
y
sis Us
ing
S
-
Tr
a
nsform
,
”
Appl. Mec
h.
Mat
er.
,
v
ol.
752
–
753
,
pp
.
1343
–
1348,
2015.
[10]
N.
H.
H
.
and
Ab
idul
la
h
,
N
.
A,
Abdulla
h
,
A.
R.
,
Z
uri_Sha’a
m
eri, A
.
,
Sham
sudin,
N.H.,
Ahm
ad, “Rea
l
-
Ti
m
e
Pow
er
Quali
t
y
D
isturba
nc
es
Det
e
ct
ion
and
Cla
ss
if
ic
a
ti
on
S
y
stem,
”
World
App
l. Sc
i
.
J
.
,
vol
.
32
,
no
.
8,
pp
.
1637
–
1651,
2014
.
[11]
D.
Sriniva
san
,
W
.
S.
Ng,
and
A.
C.
L
ie
w,
“
Neura
l
-
Network
-
Bas
ed
Signa
ture Recognit
ion
for
Ha
rm
onic
Source
Ide
nti
f
ic
a
ti
on
,
”
I
EE
E
Tr
ans.
Pow
er
D
el
i
v.
,
vo
l. 21
,
no
.
1
,
pp
.
398
–
405,
Jan
.
2006
.
[12]
S.
Vara
d
an and E.
B
.
Makr
am,
“
Prac
tical c
onsid
e
rat
ions
in
th
e ap
pli
c
at
ion
of
n
eur
al
n
et
works
to
th
e
ide
nti
f
icati
on
of harm
onic
lo
ads,
”
E
le
c
tr. Powe
r
Syst.
Re
s.
,
vo
l. 3
0,
no
.
2
,
pp
.
103
–
106,
1994
.
[13]
R.
A.
S.
Fern
and
es,
S.
M.
R
anz
in
i,
I
.
N
.
Silv
a, a
n
d
M.
Ol
eskovic
z
,
“
W
rap
per
at
tr
ib
ute
se
le
c
tor and i
nte
lligent
s
y
stems
applied to
th
e
id
ent
i
ficat
ion
of
r
eside
n
ti
a
l
har
m
onic sourc
es,
”
IEEE Power
Ene
rgy
Soc
.
Ge
n.
Me
et.
,
2012.
[14]
M.
A.
Ha
ll a
nd
G.
Holm
es,
“
Be
nchmarki
ng
Attr
ibut
e
Selecti
on
Te
chn
ique
s for
Discre
te Cl
ass D
at
a
Mining,”
IEE
E
Tr
ans.
Knowl.
Data
Eng
.
,
vol.
15
,
no
.
6
,
pp
.
1437
–
1447
,
20
03.
[15]
J.
Baptista
,
R
.
M
ora
is,
A
.
Va
le
nt
e
,
S.
Soare
s,
M.
C
ande
i
as,
and
M.
J.
C.
S.
R
ei
s
,
“
Harm
oSim
:
A t
ool
for
har
m
onic
distor
t
ion
sim
ula
t
ion
a
nd
assess
m
ent
of
nonli
n
ea
r
loa
ds,
”
Comput.
App
l.
Eng.
Educ.
,
vol
.
22,
no
.
2
,
pp
.
340
–
348,
2014
.
[16]
D.
Saxen
a, K.
V
erma,
and
S.
Sin
gh,
“
Pow
er
quali
t
y
ev
ent c
l
assificat
ion: a
n
over
v
iew
and
k
e
y
issues
,
”
In
t. J.
Eng.
S
ci.
Te
chno
l.
,
vol
.
2
,
no
.
3
,
p
p.
186
–
199
,
201
0.
[17]
a. Ba
nshw
ar
an
d
a
.
K
.
Cha
ndel,
“
Ide
nti
fi
cation of
har
m
onic source
s using fuzz
y
logi
c
,
”
2010
J
t. I
nt.
Con
f.
Powe
r E
le
c
tron. Dr
iv
es
Ene
rgy
S
yst.
2010
Pow
er
India
,
pp
.
1
–
7
,
2
010.
[18]
M.
H.
Jopri
,
N
.
A.
Abidullah,
G.
Z. Peng, a
nd
A.
R.
Abdull
ah, “A
new
two
poin
ts
m
et
hod
for identif
y
dom
ina
nt
har
m
onic
disturb
anc
e
using f
r
equ
ency
and
ph
ase
s
pec
t
rogr
am,”
Int
.
R
ev. El
e
ct
r.
En
g.
,
vol. 9, no. 2,
pp.
453
–
459
,
2014.
[19]
A.
Abdulla
h
,
M
.
Jopri,
and
M
.
M
ana
p,
“
An Im
prove
d
Spec
trogra
m
to
Id
ent
i
f
y
Mult
ipl
e
Harm
onic
S
ourc
es
in
Distribut
ion
S
y
st
em wit
h
Inv
ert
er
Loa
ds,
”
Proc.
,
vol.
7
,
no
.
4
,
p
.
1
1591,
2017
.
[
20]
P.
Supri
y
a and T
.
N.
Padm
ana
bh
an
Nam
bia
r
,
“
Re
vie
w of
h
armoni
c
source
id
ent
if
i
ca
t
ion
t
ec
hn
iques
,
”
Int
.
Rev.
El
e
ct
r.
Eng.
,
vo
l
.
7
,
no
.
3
,
pp
.
45
25
–
4531,
2012
.
[21]
J.
Afs
har
ia
n
,
D.
D.
Xu,
B
.
W
u,
B
.
Gong,
and
Z
.
Y
ang,
“
Th
e
Opt
imal
PW
M Modula
ti
on
and
Com
m
u
ta
ti
on
Scheme
for Thre
e
-
Phase
Isola
te
d
Buck
Matr
ix T
ype
Re
ct
if
ie
r
,
”
IE
EE
Tr
ans.
Pow
e
r E
lectron.
,
vol
.
8993,
no
.
c
,
pp.
1
–
1
,
2017
.
[22]
S.
Som
kun,
“
Fast
DC Bus Voltag
e
Contro
l
of
Sin
gle
-
Phase
PW
M
Rec
tifiers
using
A Ri
pple Volt
ag
e
Est
imator,”
pp.
0
–
5
,
2016
.
[23]
T.
F.
P.
W
.
M.
R
.
Under
,
“
Model
ing,
Modul
at
ion
,
and
Con
trol of
t
he
Thr
ee
-
Phase
Four
-
Sw
it
ch
PW
M Re
ctifie
r
Under
Balanc
ed Volta
ge
,
”
IEEE T
rans
.
Powe
r E
l
ec
tron.
,
vo
l. 31,
no.
7
,
pp
.
4892
–
4905,
2016
.
[24]
M.
H.
Jopri
,
A
.
R.
Abdull
ah, N.
M.
Kass
im,
M. M
ana
p,
N
.
A.
N
gat
iman,
and
M. R.
Yus
off,
“
Lo
c
al
i
za
t
ion
of
Multi
ple Harmonic
Sourc
es
for I
nver
te
r
Lo
ads
Ut
il
izing
Per
iodogr
am,”
J. Tek
nol
.
,
vol.
8
,
no
.
2
,
pp
.
87
–
91,
2016
.
[25]
M.
Z
.
R.
Zub
er Ahm
adi
,
A.
Jid
i
n,
K.
B.
Jaffa
r
,
M.
N.
Othm
an
,
R.
N.
P.
Nag
araj
an,
“
Mi
nimiz
at
i
on
of
torqu
e
rippl
e
utilizing b
y
3
-
L
CHM
I in
DTC,
”
Proc. 20
13
IEEE
7
th
Int
.
Powe
r E
ng
.
Opt
i
m.
Conf. PEOCO
2013
,
no
.
June,
pp
.
636
–
6
40,
2013
.
[26]
A.
R.
Abdulla
h
,
M.
Mana
p
,
M.
F.
Habb
an, a
nd
T. Suti
kno,
“
An a
c
cur
ate cla
ss
ifica
t
ion
m
et
hod
of
ha
rm
onic
signal
s in
power distribut
ion
s
y
st
em b
y
utilising
S
-
tra
nsform
,
”
Te
l
komnik
a
(
Teleco
mm
unic
ati
on
Co
mput.
El
e
ct
ron.
Control.
,
vol
.
15
,
no
.
1
,
2017.
[27]
A.
R.
Abdulla
h
,
N.
A.
Abidu
ll
ah
,
N.
H.
Sham
sudin,
N.
H.
H
.
Ahm
ad,
"
Powe
r qual
i
ty
signa
ls c
lassif
ic
ati
on
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
8
, N
o.
6
,
Dece
m
ber
201
8
:
5399
-
5
408
5408
sy
stem
using time
-
freque
n
cy di
str
ibut
ion
"
,
vol
.
49
4
–
495.
2014
.
[28]
M.
Jopri,
A.
Ab
dull
ah
,
T
.
Sut
ikn
o,
and
M.
Mana
p,
“
A Uti
l
isat
ion
of
Im
prove
d
G
a
bor
Tra
nsform
f
or
Harm
onic
Signal
s Det
ec
t
io
n
and
C
la
ss
ifica
t
ion
Anal
y
sis,
”
In
t.
J.
,
vo
l. 7, no. 1
,
pp
.
21
–
28
,
20
1
7.
[29]
N.
A.
Abidu
ll
ah
,
A.
R
.
Abdull
ah, N.
H.
Sham
sudin,
N.
H. T.
H.
A
hm
ad,
“
Rea
l
-
ti
m
e
power
qualit
y
signal
s
m
onit
oring
s
y
st
e
m
,
”
Proc
ee
ding
-
2013
IE
EE St
u
dent
Con
f. R
es.
Dev
.
SCOReD
2
013
,
no
.
De
ce
m
ber
,
pp
.
433
–
438,
2015
.
[30]
A.
R.
Abdulla
h
,
G.
Z
.
Peng
,
S.
A.
Ghani
,
and
M
.
H.
Jopri,
“
A ne
w ve
c
tor
dr
aft m
ethod for
h
armonic
source
det
e
ct
ion
a
t
poin
t
of com
m
on
coupl
ing,”
Pro
c. 20
14
IEEE
8
th
Int
.
Powe
r E
ng
.
Opt
i
m.
Conf. PEOCO
2014
,
no
.
Marc
h,
pp.
110
–
114,
2014
.
[31]
K.
B.
Naga
sa
i
an
d
T. R. J
y
o
thsna, “H
armoni
c
Ana
l
y
sis
and
Appl
icati
on
of
PW
M T
ec
hniqu
es
for Th
ree
Phase
Inve
rte
r
,
”
Int
.
Res
.
J
.
Eng
.
Te
chn
ol.
,
pp.
1228
–
12
33,
2016
.
[32]
L.
R
.
L. V.
R
aj, A
.
Jidin, K.
Abdul
Kari
m
,
T
.
Suti
kno,
R.
Sundram
,
“
Im
prove
d
torq
ue
con
trol pe
rfor
m
anc
e
of
dire
c
t
torqu
e con
trol
f
or
5
-
phase
i
nduct
ion
m
ac
hin
e,
”
Int
.
J. P
ower
El
e
ct
ron.
Dr
iv
e
Syst.
,
vol
.
3
,
no
.
4,
2013
.
BIOGR
AP
HI
ES
OF
A
UTH
ORS
Mohd
Hatt
a
Jopri
was
born
in
J
ohor,
Mal
a
y
s
ia
on
1978.
H
e
re
c
ei
ved
his
B.
Sc
f
rom
Univer
siti
Te
knologi
Mal
a
y
sia
in
2000
a
nd
Ms
c
.
in
Elec
tr
ic
a
l
Pow
er
Engi
ne
eri
ng
fro
m
Rhei
nisch
-
W
estfä
li
sche
Techni
sche
Hochs
chul
e
Aa
che
n
(
RW
TH),
Germ
an
y
in
2011.
Sin
ce
2005
,
he
h
as
bee
n
an
acade
m
ia
staff
in
th
e
U
nive
rsiti
Te
kn
ik
al
Mal
a
y
s
ia
Me
l
aka
(UT
eM)
an
d
cur
ren
t
l
y
h
e
pursuing
his Ph
D i
n
th
e area
of
power
qualit
y
.
Abdul
Rahi
m
A
bdull
ah
,
PhD
w
as
born
in
Keda
h,
Malay
sia
on
1979.
He
rec
ei
v
ed
his
B.
Eng.
,
Master
Eng.,
PhD
Degre
e
fro
m
Univer
sit
i
T
e
knologi
Malay
si
a
in
2001,
200
4
and
2011
in
El
e
ct
ri
ca
l
Engi
n
ee
ring
and
Digi
t
al
Signa
l
Proce
s
sing
,
respe
ct
iv
ely
.
He
is
cur
r
entl
y
an
As
socia
t
e
Profess
or
with
the
Depa
rtment
of
El
e
ct
ri
cal
En
gine
er
ing,
Chi
ef
of
Advanc
ed
Digit
al
Sign
al
Proce
ss
ing
(AD
SP
)
La
b
and
Ce
nte
r
of
Ex
ce
l
le
nt
(COE)
Chie
f
fo
r
Univer
siti
T
ek
nika
l
Mal
a
y
s
ia
Mela
ka
(UTe
M)
.
Tol
e
Sutikno
is an Ass
oci
at
ed
Pro
fessor i
n
Dep
artm
ent
of Electric
al
Engi
ne
eri
ng
a
t
Univer
si
ta
s
Ahm
ad
Dahla
n
(UA
D),
Indone
sia.
He
has
comp
le
t
ed
his
studie
s
for
B.
Eng.,
M.E
ng.
,
Ph.D.
in
El
e
ct
ri
ca
l
Engi
n
ee
ring
from
Di
ponegor
o
Unive
rsit
y
,
Gadj
ah
Mada
Univer
si
t
y
and
Univer
siti
Te
knologi
Malay
sia
,
r
espe
ctively
.
His
rese
a
rch
i
nte
rests
in
cl
ud
e
t
he
field
of
industri
al
elec
troni
c
s
and
informatics,
power
elec
tro
nic
s,
FP
GA
appl
icati
ons,
embedde
d
s
y
st
ems
,
dat
a
m
ini
ng
,
informati
on
tech
nolog
y
and
digit
al
li
bra
r
y
.
Mus
ta
fa
Mana
p
was
born
in
K
ual
a
Lumpur,
Malay
s
ia
on
19
78.
He
rec
e
ive
d
his
B.
Sc
fro
m
Univer
siti
Tech
nologi
Malay
sia
in
2000
and
Ms
c.
in
El
ectrical
Engi
ne
eri
ng
from
Univer
siti
Te
knik
al
Malays
ia
Mela
ka
(UT
eM)
2016.
Since
2006,
he
has
bee
n
an
acade
m
ia
staff
in
the
Uni
ver
siti
T
eknikal
Mal
a
y
si
a
Mela
k
a
(UTe
M).
His
rese
arc
h
in
t
ere
sts
are
power
el
e
ct
ron
i
cs
and
drive
,
i
nstrum
en
ta
ti
on
,
and
DS
P
appl
i
ca
t
ion.
Datuk
Dr.
Mohd.
Ruddin
Ab.
Ghani
cur
ren
t
l
y
a
Profess
or
of
El
e
ct
ri
ca
l
E
ngin
ee
ring
with
th
e
Facul
t
y
of
Elec
t
ric
a
l
Engi
ne
eri
n
g,
Univer
siti
T
e
knika
l
Mal
a
y
si
a
Mela
ka
(UTe
M
).
His
rese
arc
h
int
er
est
ar
e
s
y
ste
m
engi
nee
r
ing, energ
y
con
trol a
n
d
distri
but
ion au
tomati
on.
Adibah
S
y
ah
irah
Hus
sin
was
b
orn
in
Johor,
Malay
s
ia
on
19
90
.
She
rec
e
ived
her
B.
Eng
in
El
e
ct
ri
ca
l
Engi
n
ee
ri
ng
from
Dal
housie
Univer
sit
y
Can
ada
in
20
14
.
She
has
wo
rke
d
in
Te
xas
Instrum
ent
s
El
ectroni
cs
Malay
si
a
as
Te
st
Engi
n
eer
since
2014.
Cu
rre
ntly
,
she
is
a
m
aste
r
degr
ee
student
rese
arc
h
er
in
power
qua
li
t
y
in
Elec
tri
c
al
E
ngineeri
ng
at
Univer
siti
Te
k
nika
l
Ma
lay
s
ia
Mela
ka
(UTe
M)
.
Evaluation Warning : The document was created with Spire.PDF for Python.