Indonesi
an
Journa
l
of El
ect
ri
cal Engineer
ing
an
d
Comp
ut
er
Scie
nce
Vo
l.
1
3
,
No.
1
,
Jan
uar
y
201
9
,
pp.
339
~
346
IS
S
N: 25
02
-
4752,
DOI: 10
.11
591/ijeecs
.v1
3
.i
1
.pp
339
-
346
339
Journ
al h
om
e
page
:
http:
//
ia
es
core.c
om/j
ourn
als/i
ndex.
ph
p/ij
eecs
Optimal
placeme
nt of gri
d
-
co
nn
ec
ted phot
ovoltaic
gen
erato
rs
in
a power
system for
volt
age s
t
abilit
y enhanc
ement
Z
ett
y Adib
ah
Ka
m
aruz
z
ama
n,
A
z
ah M
ohamed
, Ramiz
i Moha
med
Cent
re
for
In
te
gr
at
ed
S
y
st
ems
Engi
nee
r
ing
and
A
dvanc
ed
T
ec
hno
logi
es
(In
te
gr
a)
,
Facul
t
y
of Engin
ee
ring
and
Buil
t Envi
ronm
ent
,
U
nive
rsiti Keba
ng
saa
n
Mal
a
y
s
ia, Bangi, 43600, Mal
a
y
si
a
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Ma
y
28
, 201
8
Re
vised
Oct
1
0,
2018
Accepte
d
Oct
25
, 201
8
The
h
igh
p
ene
t
rat
ion
of
photo
volt
aic
(PV
)
g
e
ner
ation
c
an
caus
e
m
a
n
y
te
chn
ic
a
l
issues
such
as
power
qual
i
t
y
and
impact
on
the
po
wer
s
y
stem
volt
ag
e
stabi
lit
y
.
To
improve
volt
age
stabilit
y
an
d
red
uci
ng
power
loss
in
a
power
s
y
stem
with
PV
gene
rat
or
s,
appr
opriate
pl
anni
ng
of
PV
gene
rat
ors
is
conside
red
b
y
opti
m
al
pl
acem
ent
of
PV
.
Th
us,
an
eff
e
ct
iv
e
heur
isti
c
opti
m
iz
ation
technique
such
a
s
the
W
ind
Dr
ive
n
Optimizati
on
(W
D
O)
te
chn
ique
is
app
li
ed
for
de
te
rm
i
ning
opti
m
al
lo
c
at
ion
of
PV
gen
er
at
ors
in
a
power
s
y
stem.
For
det
ermining
the
opt
imal
lo
ca
t
ion
of
PV
,
t
he
objecti
v
e
func
ti
on
considers
m
axi
m
iz
ing
t
he
Im
prove
d
Volta
g
e
Stabilit
y
Inde
x.
The
proposed
m
et
hod
for
opti
m
al
loc
ation
of
PV
ge
ner
at
ors
is
implemente
d
o
n
the
IEE
E
118
an
d
30
bus
tra
ns
m
i
ss
ion
sy
stems
and
the
69
-
rad
ia
l
d
istri
buti
o
n
s
y
stem.
Th
e
op
ti
m
iz
ation
resul
t
s
show
tha
t
integra
t
ing
PV
into
the
t
est
s
y
stems
improves
voltage
st
abi
lit
y
in
th
e
s
y
stem.
Com
par
ing
the
per
form
anc
e
o
f
t
he
W
DO
with
t
he
par
t
icle
sw
ar
m
opti
m
iz
at
ion
t
ec
hniqu
e,
it
is shown t
hat t
h
e
W
DO
te
chni
qu
e
give
s f
aste
r
con
ver
gence.
Ke
yw
or
d
s
:
Op
ti
m
al
locatio
n
Photo
vo
lt
ai
c
ge
ner
at
io
n
W
i
nd
dr
i
ven optim
iz
at
ion
Copyright
©
201
9
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
All right
s
reserve
d
.
Corres
pond
in
g
Aut
h
or
:
Ram
iz
i
Moh
am
ed
,
Ce
ntre fo
r In
te
gr
at
e
d
Syst
em
s
Enginee
rin
g
a
nd Ad
van
ce
d Tec
hnol
og
ie
s
(In
te
gr
a
)
,
Faculty
of E
ngineerin
g
a
nd B
uilt
Environm
ent
,
Un
i
ver
sit
i Ke
ba
ngsaan
Mal
ay
sia
,
Ba
ng
i,
43
600,
Ma
la
ysi
a
.
Em
a
il
:
ra
m
izi
@ukm
.ed
u.
m
y
1.
INTROD
U
CTION
The
i
ncr
easi
ng
gro
wth
of
l
oad
de
m
and
,
la
ck
of
act
ive
powe
r
ge
nerat
ion
as
well
as
te
ch
nical
const
raints
in
powe
r
syst
em
are
the
c
halle
ng
e
s
face
d
by
power
util
it
i
es
in
pro
vi
ding
reli
able
power
t
o
consum
ers.
I
nc
rease
in
loa
d
dem
and
s
fro
m
the
industri
al
,
com
m
ercial
an
d
resi
de
ntial
sec
tors
has
cause
d
congesti
on
i
n
powe
r
net
wor
k
a
nd
this
sit
ua
ti
on
le
ad
s
to
vo
lt
age
insta
bili
ty
in
power
syst
e
m
[1
]
.
U
ns
ta
bl
e
powe
r
syst
e
m
du
e
to
vo
lt
age
decli
ne
m
a
y
lead
to
volt
age
colla
ps
e.
T
her
e
f
or
e
,
m
a
intai
nin
g
volt
age
sta
bi
li
ty
is
a
m
ajo
r
co
nce
r
n
in
powe
r
syst
e
m
op
erati
on.
An
a
ppr
oach
t
o
i
m
pr
ove
volt
age
sta
bili
ty
in
power
syst
em
is
by
integrati
ng
renewable
e
nergy
base
d
distri
bute
d
ge
ner
at
io
ns
(
D
Gs)
w
hich
are
usual
ly
instal
le
d
in
powe
r
distrib
ution
syst
e
m
[2
-
5].
H
ow
e
ve
r,
the
in
te
gr
at
io
n
of
re
new
a
ble
e
nerg
y
based
D
Gs
i
nto
a
powe
r
s
yst
e
m
causes
c
ertai
n
te
chn
ic
al
ef
fec
ts
beca
us
e
t
he
netw
ork
desi
gn
init
ia
ll
y
do
e
s
not
c
on
si
der
the
inte
gr
at
io
n
of
D
G.
The
rap
i
d
gro
wth
of
D
G
in
s
ta
ll
at
ion
s
an
d
high
pen
et
r
at
io
n
of
D
G
can
c
ause
m
any
te
chn
ic
al
iss
ues
s
uch
as
powe
r
qual
i
ty.
In
te
gr
at
in
g
D
G
at
ran
dom
locat
ion
s
m
ay
cause
increase
d
po
wer
lo
sses
a
nd
v
oltage
i
ns
ta
bi
li
ty]
.
To
m
ai
ntain
vo
lt
age
sta
bili
ty
and
re
duce
powe
r
loss
,
it
i
s
i
m
po
rtant
to
place
DG
s
optim
al
ly
in
a
powe
r
syst
e
m
.
Seve
ral
m
et
ho
ds
ha
ve
bee
n
de
velo
ped
f
or
de
te
rm
i
nin
g
opt
i
m
al
DG
placem
ent
us
in
g
ei
ther
analy
ti
c
or
he
ur
ist
ic
optim
iz
ation
a
ppr
oach
e
s.
I
n
[
6],
powe
r
sta
bili
ty
inde
x
is
us
ed
to
de
te
rm
ine
the
crit
ic
al
buses
in
a
syst
e
m
wh
ere
these
buses
a
r
e
con
si
der
e
d
a
s
op
ti
m
al
DG
locat
ion
s
.
A
m
od
i
fied
vo
lt
age
ind
e
x
is
us
e
d
as
an
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
1
3
, N
o.
1
,
Ja
nu
a
ry
201
9
:
339
–
346
340
obj
ect
ive
funct
ion
t
o
im
pr
ov
e
vo
lt
a
ge
sta
bili
ty
m
arg
in
an
d
m
ixed
-
inte
ger
nonlinea
r
pr
ogram
m
ing
is
ap
plie
d
to
so
lve
t
he
optim
iz
at
ion
prob
le
m
[7
]
.
In
add
it
io
n,
the
volt
age
sta
bili
ty
ind
e
x
(
VSI)
ha
s
been
de
velo
ped
t
o
determ
ine
the
op
ti
m
al
place
m
ent
a
nd
siz
in
g
of
DG
i
n
a
rad
ia
l
distrib
ution
syst
em
con
siderin
g
loa
d
grow
t
h
[8
]
.
Am
on
g
t
he
app
li
e
d
he
ur
ist
ic
op
tim
iz
at
i
on
te
c
hniq
ues,
the
pa
rtic
le
swar
m
op
ti
m
iz
a
ti
on
(
PS
O)
has
been
us
e
d
f
or
deter
m
ining
the
siz
e
an
d
locat
io
n
of
m
ulti
DG
s
consi
der
i
ng
m
ulti
-
obj
ec
ti
ve
ind
e
x
[9
]
.
I
n
a
dd
it
io
n,
consi
der
i
ng
he
ur
ist
ic
op
ti
m
izati
on
,
the
cuc
koo
sea
rch
al
go
rithm
[1
0]
and
a
con
str
uctive
heurist
ic
al
go
rithm
[11]
ha
ve
bee
n
us
e
d
for
s
olv
i
ng
the
optim
al
D
G
placem
ent
prob
le
m
.
In
a
ll
the
ab
ov
e
-
m
entione
d
m
et
ho
ds,
a
gen
e
ral
DG m
od
e
l
with
real
powe
r
in
j
ect
io
n has
bee
n
c
onsidere
d wit
hout
co
nsi
der
i
ng a
ny r
e
ne
wab
le
e
nergy.
On
e
ty
pe
of
D
G
that
is
widely
us
ed
in
m
od
ern
power
syst
e
m
is
the
phot
ovoltai
c
based
distribu
t
e
d
gen
e
rati
on
(
P
VDG).
T
he
P
VDG
syst
em
pro
vid
es
ad
di
ti
on
al
act
ive
powe
r
i
nt
o
the
net
wor
k,
there
by
facil
it
at
ing
reverse
powe
r
flo
w
an
d
re
du
ci
ng
ene
rg
y
co
nsum
ption
.
The
r
eact
ive
power
is
so
le
ly
pr
ovided
by
the
conve
ntio
na
l
gen
erat
or
s
a
nd
the
refo
re,
r
eact
ive
power
lim
it
at
ion
is
neg
le
ct
ed
w
he
n
find
i
ng
the
optim
a
l
PVDG
l
ocati
on
a
nd
siz
ing.
T
o
im
pr
ov
e
volt
age
sta
bili
ty
m
arg
i
n
an
d
redu
ci
ng
powe
r
los
s
in
a
powe
r
s
yst
e
m
with
PVD
Gs,
appr
opriat
e
plann
i
ng
of
P
VDGs
is
con
si
dered
by
optim
al
placem
ent
and
siz
ing
of
P
V.
So
m
e
researc
h
works
hav
e
bee
n
do
ne
f
or
determ
i
ning
opti
m
a
l
place
m
ent
and
s
i
zi
ng
of
P
VDG
co
ns
id
erin
g
vo
lt
ag
e
sta
bili
ty
i
m
pr
ov
em
ent
[1
2
-
14
]
.
In
[
12
]
,
t
he
PSO
te
ch
niqu
e
was
us
e
d
to
determ
ine
op
tim
u
m
place
m
e
nt
of
PVDG
an
d
w
ind
t
urbine
by
inco
rpor
at
i
ng
vo
lt
age
sta
bili
ty
factor
.
In
[
13
]
,
the
arti
fici
al
bee
-
colo
ny
op
ti
m
iz
ation
t
echn
i
qu
e
is
de
velo
ped
f
or
determ
ining
optim
al
siz
ing
of
hybri
d
PV
/
wind
tu
rb
i
ne/f
uel
cel
l
syst
e
m
by
m
a
xim
iz
ing
the
volt
age
c
ollapse
po
i
nt
us
i
ng
t
he
VSI.
I
n
[14],
the
sel
f
-
c
orre
ct
ion
al
gorith
m
was
us
e
d
in
deter
m
ining
opti
m
a
l
siz
ing
of
m
ulti
ple
PV
an
d
batte
ry
stora
ge
un
it
s
co
ns
ide
rin
g
volt
age
st
abili
ty
m
arg
in.
In
this
pa
pe
r,
a
relat
ively
new
he
ur
ist
ic
optim
iz
ation
te
ch
nique
know
n
a
s
wind
dr
i
ve
n
op
ti
m
iz
ation
(
WD
O
)
[
15]
is
app
li
ed
for
so
lvi
ng
the
op
tim
a
l
place
m
e
nt
of
P
V
DG
un
it
s.
In
the
pro
posed
optim
iz
at
ion
pro
blem
fo
rm
ulati
on
,
the
ob
j
e
ct
ive
functi
on
i
s
to
m
axi
m
iz
e
the
vo
lt
a
ge
sta
bili
ty
m
arg
in.
Vo
lt
age
sta
bili
ty
of
a
powe
r
syst
e
m
is
assessed
by
us
in
g
a
relat
ively
new
vo
lt
ag
e
sta
bili
ty
ind
ex
know
n
as
the
i
m
pr
ov
e
d
vo
lt
age
sta
bili
ty
ind
ex
(IVS
I)
[
16
]
.
Th
us
,
for
dete
rm
ining
the
optim
al
locat
ion
of
PVDG
,
t
he
obj
ect
ive
f
un
ct
io
n
consi
ders m
axim
iz
ing
the
IVS
I.
2.
THE
MATE
R
IAL
AND ME
THOD
This
sect
io
n
de
scribes
I
VSI
for
vo
lt
a
ge
sta
bili
ty
analy
sis,
PVD
G
m
od
el
li
ng
,
an
d
a
ppli
cat
ion
of
WDO
f
or
s
olv
i
ng
the o
ptim
izati
on
pro
blem
.
The
I
VSI
is
an
ind
ic
at
or
f
or
assessi
ng
w
heth
er
a
power
syst
e
m
is
sta
ble
f
ro
m
the
volt
age
sta
bili
ty
viewpoint
.
I
t
is
us
e
d
i
n
the
obj
ect
i
ve
f
unc
ti
on
in
fi
nd
i
ng
the
opti
m
al
loc
at
io
n
of
P
V
DG
for
i
m
pr
ov
in
g
vo
lt
age
sta
bili
ty
in
a
power
syst
em
.
This
sect
ion
al
so
discusse
s
on
t
he
ap
plic
at
ion
of
WDO
te
ch
niqu
e
in
determ
ining
PVD
G
opti
m
al
place
m
ent
.
I
n
this
stu
dy,
the
entire
m
od
el
ing
a
nd
sim
ulati
on
is carrie
d ou
t
usi
ng MATL
AB
softwa
re.
2.1
.
I
VS
I
for
vo
l
tage st
ab
il
ity
a
sse
ssmen
t
The
de
velo
pme
nt
of
IVSI
f
or
volt
age
sta
bili
ty
assessm
ent
i
s
base
d
on
th
e
current
a
nd
po
wer
flo
w
at
a
bus
in
a
po
w
er
net
wor
k.
T
he
m
a
the
m
at
ic
a
l
form
ula
fo
r
I
VS
I
f
or
n
-
bus
syst
e
m
is
der
ived
in
[16]
an
d
it
is
giv
e
n by,
0
2
1
4
1
.0
c
o
s
s
in
c
o
s
s
in
n
ij
ij
i
i
j
n
j
ij
ij
ij
ij
ij
ij
j
G
B
P
Q
I
V
S
I
V
G
B
(1)
wh
e
re
P
i
+
Q
i
is
the
loa
d
act
i
ve
a
nd
reacti
ve
powe
rs
at
t
he
receivin
g
e
nd
bu
s
,
V
j
is
t
he
volt
age
at
the
se
nd
i
ng
end
bus,
δ
ij
is
the
phase
a
ng
le
diff
e
ren
ce
bet
ween
se
ndin
g
end
a
nd
receiv
ing
e
nd
bu
se
s
and
G
ij
an
d
B
ij
are
th
e
el
e
m
ents in
the
bus
a
dm
ittance m
a
trix.
The
I
VS
I
is
an
ind
ic
at
or
to
as
sess
vo
lt
age
st
abili
ty
a
t
the
b
us
es
in
a
powe
r
syst
e
m
.
The
IV
S
I
has
a
value
betwee
n
0
an
d
1.
A
bus
is
sai
d
to
be
sta
ble
if
the
IVS
I
value
is
le
ss
than
1
(
IVSI
<1
).
H
oweve
r,
a
bu
s
i
s
sai
d
to
ex
per
ie
nce
a
volt
age
colla
ps
e
if
the
IV
S
I
val
ue
e
qu
al
s
to
1.
W
hen
run
ning
a
si
m
ulati
on
to
assess
vo
lt
age
insta
bi
li
ty
us
ing
I
VS
I
,
the
loa
ding
m
arg
i
n
at
a
bus
is
incr
eased
gr
a
du
al
ly
un
ti
l
the
I
VS
I
value
re
aches
appr
ox
im
at
ely 1
.
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Op
ti
m
al
pla
ce
men
t
of
gr
id
-
c
onnected
ph
otov
oltaic g
ener
at
or
s i
n a power
…
(
Zett
y Ad
ib
ah K
amar
uz
z
aman
)
341
2.2
.
I
VS
I
for
vo
l
tage st
ab
il
ity
a
sse
ssmen
t
In
t
his
stu
dy,
t
he
sta
ti
c
m
od
el
of
PVD
G
is
ba
sed
on
a
sin
gl
e
-
di
od
e
e
quiva
le
nt
ci
rcu
it
[
12
]
wh
e
re
the
act
ive pow
e
r
is
expres
sed
as,
PV
PV
O
C
SC
P
N
V
I
(2)
in which
,
,
1000
S
C
S
C
S
T
C
i
C
r
e
f
G
I
I
K
T
T
(3)
,
O
C
O
C
S
T
C
v
C
r
e
f
V
V
K
T
T
(4)
20
800
CA
N
C
O
T
T
T
G
(5)
wh
e
re
G
is
th
e
so
la
r
irra
dian
ce
set
at
co
ns
t
ant
10
00
W
/m
2
,
T
C
is
the
s
ol
ar
ope
rati
ng
te
m
per
at
ur
e,
V
OC
is
the
op
e
n
ci
rc
uit
vo
lt
age,
I
SC
is
the
sh
ort
ci
rcu
it
c
urren
t,
N
PV
is
P
V
m
od
ule
num
ber,
TA
is
the
a
m
bient
tem
per
at
ur
e
wh
ic
h
sat
at
c
on
sta
nt
35°C,
NOCT
is
t
he
nom
inal
op
erati
ng
cel
l
te
m
per
at
ur
e,
K
i
is
t
he
short
ci
rc
uit
current
coeffic
ie
nt,
K
v
is
the
open
ci
r
cuit
volt
age
co
eff
ic
ie
nt,
T
ref
is
the
ref
e
re
nce
op
e
rati
ng
te
m
per
at
ur
e
at
25°
C.
All
the
el
ect
rical
c
har
act
erist
ic
s
a
t
sta
nd
ar
d
te
st
conditi
on
(
STC
)
are
ta
ke
n
fro
m
the
PV
datasheet
pro
vid
e
d
by
the
m
anu
fact
ur
e
r.
2.3
.
F
ormul
ati
on o
f the
Op
timi
z
at
ion P
ro
blem
The
op
ti
m
iz
ati
on
pro
blem
fo
rm
ulati
on
cons
iders
m
axi
m
iz
i
ng
volt
age
sta
bili
ty
m
arg
in
of
a
powe
r
syst
e
m
.
The
search
for
opti
m
al
so
luti
on
i
s
subj
ect
e
d
to
the
fo
ll
owin
g
obj
ect
ive
fun
ct
ion
an
d
te
ch
nical
const
raints.
T
o
determ
ine
the
op
ti
m
al
PV
DG
locat
ion
,
t
he
weak
e
st
bus
in
a
te
st
syst
e
m
has
to
be
dete
r
m
ined
by
cal
culat
ing
IV
S
I
at
each
lo
ad
bus.
F
ro
m
t
he
m
a
the
m
at
ical
fo
rm
ula
in
(7
)
,
the
IVSI
va
lue
is
between
0
an
d
1,
su
c
h
t
hat
if
a
bus
has
an
I
VS
I
value
of
1,
it
m
eans
that
the
bu
s
is
the
weak
e
st
bus
f
r
om
vo
lt
age
i
nst
abili
t
y
view
po
i
nt.
Th
us, the I
V
SI
at e
ach lo
a
d
bus as
an
obj
ect
ive
f
un
ct
io
n need
t
o
be
m
axi
m
iz
e
d
f
or
determ
ini
ng
the
weak
e
st b
us
.
T
he
m
at
he
m
at
ical form
ula o
f o
bj
ect
iv
e f
unct
ion f
or opti
m
al
PVDG
locati
on is
giv
e
n by,
1
m
a
x
i
f
I
V
S
I
(6)
The
te
ch
nical
const
raints
c
onsidere
d
in
the
op
ti
m
iz
ation
pro
blem
are
the
PVDG
,
volt
ag
e
and
pow
e
r
flo
w
const
rain
ts.
The
PVDG
is
l
i
m
it
ed
by
the
avail
able
powe
r
sources
at
any
giv
en
locat
ion
.
T
hus,
it
is
necessa
ry
to
c
on
st
raint
the
c
apacit
y
betwee
n
it
s
m
axi
m
um
and
m
ini
m
um
le
vel
of
a
ct
ive
powe
r,
wh
ic
h
is
giv
e
n by,
,
m
i
n
,
m
a
x
D
G
D
G
D
G
P
P
P
(7)
The
syst
em
o
pe
rati
on sho
uld
be kept
withi
n ±5
%
of
volt
ag
e d
e
viati
on and
the
vo
lt
ag
e c
onstrai
nt is
g
i
ve
n by,
0
.9
5
1
.0
5
i
p
u
V
p
u
(8)
The p
ow
e
r flo
w
e
qu
at
io
ns t
o be sati
sfie
d
ar
e
g
ive
n b
y,
,,
i
P
V
D
G
i
i
D
P
P
P
(9)
,,
i
P
V
D
G
i
i
D
Q
Q
Q
(10)
in which
,
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
1
3
, N
o.
1
,
Ja
nu
a
ry
201
9
:
339
–
346
342
1
c
o
s
(
)
N
i
i
i
j
j
i
j
i
j
j
P
V
Y
V
(11)
1
s
i
n
(
)
N
i
i
i
j
j
i
j
i
j
j
Q
V
Y
V
(12)
wh
e
re
N
is
the
nu
m
ber
of
bu
s,
P
i,PVDG
and
Q
i,PVDG
are
the
PV
ge
ne
rated
real
and
react
ive
powe
rs
at
bu
s
i,
resp
ect
ively
a
nd
P
i,D
a
nd
Q
i,D
are the
r
eal
a
nd
r
eact
ive
powe
r dem
and
s at b
us
i,
res
pecti
ve
ly
.
2.4
.
Ap
pli
ca
t
ion
of W
DO f
or
O
pt
im
al L
ocatio
n
of
PV
DG
WDO
is
base
d
on
the
Ne
wton’s
sec
ond
l
ow
of
m
otion
w
hich
is
us
e
d
to
descr
i
be
the
m
ot
ion
of
ai
r
par
cel
s
on
t
he
earth’s
atm
os
phere
[
15]
.
I
n
the
WD
O
t
opol
og
y,
the
posit
ion
of
sm
al
l
a
ir
par
cel
s
with
r
andom
velocit
ie
s
in
a
n
N
-
dim
ension
al
sp
ace
is
update
d
based
on
t
he
ph
ysi
c
al
equ
at
io
ns
t
hat
go
vern
la
r
ge
-
scal
e
atm
os
ph
e
ric
m
otion.
T
he
m
ov
em
ent
of
an
a
ir
par
cel
in
on
e
dim
ension
is
con
t
rib
uted
f
r
om
the
m
ov
e
m
ent
of
ai
r
par
cel
s
in
ano
t
her
dim
en
sion.
The
velo
ci
ty
of
ai
r
parce
ls
is
der
ive
d
fr
om
the
concept
of
a
n
ai
r
par
cel
m
ov
ing
with th
e w
in
d
i
n wh
ic
h
the
up
datin
g velocit
y i
s g
i
ve
n by,
d
im
m
a
x
m
a
x
m
in
11
o
th
e
r
o
ld
n
e
w
o
ld
o
ld
o
ld
o
ld
P
c
u
u
u
g
x
R
T
x
x
PP
(13)
wh
e
re
u
new
an
d
u
old
are
the
updated
vel
ocity
and
c
urren
t
ve
locit
y,
resp
ect
ively
,
x
max
is
the
distance
fro
m
the
highest
-
pr
es
sur
e
point,
P
max
i
s
the
m
axi
m
um
pr
essur
e
,
P
o
ld
is
the
press
ure
at
the
c
urre
nt
locat
io
n,
T
is
the
tem
per
at
ur
e
,
g
is t
he gra
vitat
ion
al
acc
el
erati
on, a
nd
R
,
α
,
a
nd
c
a
re c
on
sta
nts.
The
thir
d
te
r
m
of
eq
uation
(13)
is
a
f
orc
e
towards
t
he
locat
ion
of
m
axi
m
u
m
pr
essur
e
w
hich
is
represe
nted
by
the
global
be
s
t
locat
ion
of
th
e
optim
iz
at
ion
pro
blem
.
On
ce
the
new
vel
oc
it
y
is
cal
culat
e
d,
t
he
po
sit
io
n of ai
r parcel
will
b
e
updated
usi
ng t
he follo
wing e
qu
at
io
n.
n
e
w
o
l
d
n
e
w
x
x
u
t
(14)
wh
e
re
x
old
a
nd
x
new
are
the
cu
r
ren
t
posit
ion
a
nd
ne
w
posit
io
n
of
the
ai
r
parce
l,
res
pecti
vel
y
and
the
tim
e
ste
p
∆t
is assu
m
ed a
s 1.
The
WDO
is
app
li
ed
for
sol
vin
g
op
ti
m
al
locat
ion
of
P
VDGs
an
d
it
is
descr
ibe
d
in
te
rm
s
of
a
flo
wch
a
rt as shown
in
Fig
ur
e
1.
F
ro
m
the f
lo
wch
a
rt, th
e al
gorithm
starts with the in
it
ia
li
z
at
ion
stage
w
he
re all
input
pa
ram
et
e
rs
are
def
i
ned
t
o
est
ablish
t
he
pro
ble
m
bo
unda
ries.
A
fter
set
ti
ng
the
in
put,
the
first
ste
p
of
the
op
ti
m
iz
ation
pro
blem
is
to
find
th
e
optim
al
PVDG
locat
io
n.
I
n
this
al
gor
it
h
m
,
the
PV
D
G
is
represe
nted
as
a
n
ai
r
par
cel
. D
ur
i
ng
t
he
sim
ulatio
n, the
pr
ess
ur
e o
f
eac
h
ai
r pa
rcel (ob
j
ect
ive
functi
on)
is e
va
luate
d
at
it
s cu
r
ren
t
po
sit
io
n.
The
n,
the
popula
ti
on
is
ranke
d
ba
sed
on
it
s
pres
su
re
an
d
t
he
ve
locit
y
is
evaluated
us
i
ng
eq
uatio
n
(13).
The
po
sit
ion
of
the
ai
r
pa
rcel
for
the
ne
xt
it
erati
on
is
update
d
us
i
ng
(
14).
T
his
proce
dure
co
ntin
ues
un
ti
l
it
re
aches
the
m
axi
m
u
m
nu
m
ber
of
it
erati
on
.
Finall
y,
the
be
st
locat
ion
of
the
la
st
it
erati
on
is
rec
orde
d
a
s
the
op
ti
m
iz
ation
r
e
su
lt
.
2.5
.
Ap
pli
ca
t
ion
of PSO
for
Op
timal L
oc
at
i
on
of P
VDG
The
PS
O
is
an
evo
luti
ona
ry
com
pu
ta
ti
on
al
gorithm
based
on
a
so
ci
al
-
ps
yc
ho
l
og
ic
al
m
et
a
phor
[17].
In PS
O
to
polo
gy, th
e
b
e
st i
nd
ividu
al
posit
io
n of pa
rtic
le
, i is r
e
pr
ese
nted
by
,
1
2
3
...
i
i
i
i
i
n
P
b
e
s
t
P
b
e
s
t
P
b
e
s
t
P
b
e
s
t
P
b
e
s
t
(15)
wh
e
reas t
he be
st p
os
it
ion
of t
he
e
ntire s
warm
in n
-
dim
ension
al
searc
h
s
pa
ce is re
pr
ese
nt
ed by
1
2
3
...
i
i
i
i
i
n
G
b
e
s
t
G
b
e
s
t
G
b
e
s
t
G
b
e
s
t
G
b
e
s
t
(16)
An
i
nd
i
vidual
in
a
swa
rm
app
r
oac
hes
the
optim
u
m
throu
gh
it
prese
nt
po
sit
io
n,
pr
e
se
nt
velocit
y,
pr
e
vious
e
xper
ie
nces
of
Pb
es
t
and
G
best
a
nd
the
ex
per
ie
nc
e
of
it
s
nei
ghbours.
T
he
pa
r
ti
cl
e
of
ne
w
posit
ion
and v
el
ocity
ar
e d
et
erm
ined u
sing t
he follo
w
ing
e
quat
io
ns
,
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Op
ti
m
al
pla
ce
men
t
of
gr
id
-
c
onnected
ph
otov
oltaic g
ener
at
or
s i
n a power
…
(
Zett
y Ad
ib
ah K
amar
uz
z
aman
)
343
(
1
)
1
k
k
k
i
i
i
X
X
t
v
(17)
(
1
)
(
)
(
)
(
)
(
)
(
)
1
1
2
2
k
k
k
k
k
k
i
i
i
i
i
i
V
V
c
r
P
b
e
s
t
X
c
r
G
b
e
s
t
X
(18)
W
he
re
i
=
1,
2,
3,
…
,
n
is
the
num
ber
of
dim
ensio
n
of
each
par
ti
c
le
,
c
1
and
c
2
are
weigh
t
factors,
r
1
an
d
r
2
are
rand
om
nu
m
b
er
bet
ween
(0,1)
an
d
k
is
the
num
ber
of
it
erati
on
tim
e
s.
The
i
ner
ti
a
weig
hting
fa
ct
or
i
s
represe
nted by
ω
of t
he
f
ollo
w
ing
e
quat
io
n,
,
m
a
x
m
a
x
m
in
m
a
x
k
k
ite
r
ite
r
(19)
S
t
a
r
t
I
npu
t
t
e
s
t
s
ys
t
e
m
da
t
a
C
r
a
t
e
r
a
ndo
m
i
ni
t
i
a
l
a
i
r
pa
r
c
e
l
r
e
pr
e
s
e
nt
i
ng P
V
D
G
pl
a
c
e
m
e
nt
R
U
N
L
oa
d F
l
ow
C
a
l
c
ul
a
t
e
1
st
obj
e
c
t
i
ve
f
unc
t
i
on
=
M
axI
V
SI
(
i
)
E
val
uat
e
pr
e
s
s
ur
e
f
or
e
a
c
h ai
r
par
c
e
l
R
e
c
or
d
G
l
oba
l
B
e
s
t
U
pda
t
e
pos
i
t
i
on a
nd vel
oc
i
t
y
C
he
c
k t
he
s
t
op c
r
i
t
e
r
i
on
F
i
nd a
nd s
a
ve
t
he
opt
i
m
um
P
V
D
G
l
oc
a
t
i
on
E
nd
I
npu
t
W
D
O
par
a
m
e
t
e
r
s
No
Y
e
s
Figure
1.
Im
ple
m
entat
ion
of
WDO al
gorith
m
f
or
so
lvi
ng opti
m
a
l l
ocati
on of PVD
G
S
t
a
r
t
I
npu
t
t
e
s
t
s
ys
t
e
m
da
t
a
I
npu
t
P
S
O
pa
r
a
m
e
t
e
r
C
r
e
a
t
e
r
a
ndo
m
i
ni
t
i
a
l
pa
r
t
i
c
l
e
r
e
pr
e
s
e
nt
i
ng P
V
D
G
pl
a
c
e
m
e
nt
R
un
L
oa
d
F
l
ow
C
al
c
ul
at
e
1
s
t
obj
e
c
t
i
v
e
f
unc
t
i
on
=
M
ax
I
V
SI
(
i
)
R
e
c
or
d
P
be
s
t
,
G
be
s
t
C
he
c
k t
he
s
t
op
c
r
i
t
e
r
i
on
U
pda
t
e
pos
i
t
i
on a
nd v
e
l
oc
i
t
y
E
nd
F
i
nd a
nd s
a
ve
t
he
opt
i
m
um
P
V
D
G
No
Y
e
s
Figure
2.
Im
ple
m
entat
ion
of
PSO al
gorithm
for
so
lvi
ng opti
m
a
l l
ocati
on of PVD
G
The
P
SO
te
c
hniqu
e
is
a
pp
li
ed
for
s
olv
in
g
optim
a
l
locat
ion
of
PVDG
s
an
d
it
is
descr
ibe
d
in
te
rm
s
of
a
flo
wc
har
t
as
show
n
i
n
Fi
gure
2.
T
he
fi
rs
t
ste
p
of
t
he
optim
iz
at
ion
prob
le
m
is
to
fin
d
the
opti
m
a
l
PVDG
locat
ion
. In
th
e
PS
O
al
gorith
m
,
the P
VDG
is represen
te
d
a
s a p
arti
cl
e in the
swa
rm
p
opul
at
ion
. Th
e al
gorith
m
sta
rts
by
init
ia
li
zi
ng
al
l
the
in
pu
t par
am
et
ers
as
show
n
in
t
he
flo
wch
a
rt.
D
ur
i
ng
t
he
sim
ul
at
ion
,
P
best
(
obje
ct
ive
functi
on)
is
e
va
luate
d
at
it
s
c
urren
t
posit
ion
us
i
ng
e
qu
at
io
n
(15
)
wh
e
rea
s
G
best
is
eval
ua
te
d
us
in
g
e
qu
at
ion
(16).
The
ne
w
po
sit
ion
a
nd
velocit
y
of
pa
r
ti
cl
e
are
up
dated
us
in
g
eq
uat
ion
(
17)
an
d
(
18).
T
his
proc
edure
con
ti
nues
un
ti
l
it
reaches
the
m
axi
m
u
m
nu
m
ber
of
it
eratio
n.
Finall
y,
G
best
of
the
la
st
i
t
erati
on
is
recor
ded
a
s
the opti
m
iz
a
ti
on
resu
lt
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
1
3
, N
o.
1
,
Ja
nu
a
ry
201
9
:
339
–
346
344
3.
RESU
LT
S
AND DI
SCUS
S
ION
The
sim
ulatio
n
resu
lt
s
for
optim
al
place
m
ent
of
PVDGs
in
the
IEEE
30
bus,
IE
EE
118
bu
s
a
nd
69
-
rad
ia
l
distrib
ution
syst
em
s
us
ing
WDO
a
nd
PSO
are
prese
nted
an
d
disc
usse
d.
Sim
ulati
on
s
wer
e
car
ried
out
to
determ
ine
the
weak
bu
ses
in
the
te
st
syst
e
m
s
us
in
g
IVSI.
Durin
g
the
si
m
ula
ti
on
,
eac
h
load
bus
is
ev
al
uate
d
ind
ivi
du
al
ly
by
m
axi
m
iz
ing
the
IVSI
as
th
e
obj
ect
ive
f
unct
ion.
The
l
oa
d
facto
r
is
in
creased
by
a
certai
n
per
ce
ntage
of
t
he
base
val
ue
un
ti
l
I
VSI
a
pp
ro
ac
hes
1,
wh
i
ch
is
t
h
e
volt
age
c
ollapse
poi
nt.
T
he
first
bu
s
that
exp
e
riences vo
lt
age
colla
ps
e
i
s
sai
d
to
be
t
he
weak
est
b
us
in
the
syst
em
a
nd
it
is
m
ark
ed
as
an
opti
m
a
l
PVDG
locat
ion
.
Th
us
,
the
PVDG
is
instal
le
d
at
this
bu
s
.
F
or
m
ulti
ple
nu
m
ber
s
of
P
V
DG,
the
first
PVD
G
wi
ll
b
e
integrate
d
us
in
g
the
bu
s
ra
nking
ap
proac
h.
T
he
first
P
V
DG
will
be
i
nteg
rated
at
t
he
weake
st
bus
(e.
g.
B
us
61
for
the
69
-
ra
dial
distrib
ution
syst
e
m
)
wh
e
re
as
the
sec
ond
PVDG
will
be
integrate
d
at
th
e
seco
nd
wea
ke
st
bus
(e.
g.
B
us
64
for
the
69
-
ra
dial
distri
buti
on
sy
stem
).
Sim
i
la
r
proce
dure
is
a
pp
li
ed
f
or
the
IEEE
30
bus
s
yst
e
m
and IE
EE
118 bu
s
syst
em
.
Table
1,
Table
2
an
d
Ta
ble
3
sho
w
the
I
V
SI
val
ues
with
var
ia
ble
load
factor
s
f
or
th
e
69
-
ra
dial
distrib
ution
sy
stem
,
IEEE
30
bu
s
,
an
d
IEE
E
118
bus
,
res
pe
ct
ively
.
The
load
facto
r
of
each
bus
is
incre
ased
i
n
per
ce
ntage
f
rom
the
base
lo
ad
value
.
Re
s
ul
ts
in
Ta
ble
1
sh
ow
that
the
op
ti
m
al
PV
D
G
locat
i
on
is
Bus
61
because
at
loa
d
facto
r
of 1
90
%,
the I
V
SI
v
a
lue
eq
uals
1.0
an
d
B
us
61
is
s
ai
d
to
e
xp
e
rien
ce
a
volt
age
c
ol
la
ps
e.
Howe
ver,
oth
e
r bu
ses
in
the s
yst
e
m
still op
e
rate wit
hi
n vo
lt
age sta
bili
ty
lim
it
w
it
h
IVSI values
less tha
n 1.
Re
su
lt
s
in
Ta
bl
e
2
sho
w
that
the
opti
m
a
l
PV
D
G
locat
io
n
of
t
he
IE
EE
30
bus
syst
em
is
Bus
30.
It
is
the
first
bus
t
o
reac
h
volt
ag
e
colla
ps
e
poi
n
t
(IVS
I=1)
at
a
load
fact
or
of
150%
.
The
nex
t
op
ti
m
al
PVDG
locat
ion
is
Bus
26 a
nd the
n fol
lowed b
y B
us
5.
Table
1
.
IVSI
with loa
d
i
ncr
e
m
ent f
or th
e
69
-
ra
dial dist
ri
buti
on
syst
em
Load
Factor
(%)
W
DO
PSO
Bu
s 6
5
Bu
s 6
4
Bu
s 6
1
Bu
s 6
5
Bu
s 6
4
Bu
s 6
1
100
0
.26
2
2
0
.40
9
2
0
.43
1
6
0
.26
2
2
0
.40
9
2
0
.43
1
6
120
0
.32
8
6
0
.51
2
7
0
.54
0
0
0
.32
8
6
0
.51
2
7
0
.54
0
0
140
0
.40
7
9
0
.63
3
0
.66
3
5
0
.40
1
8
0
.62
6
6
0
.65
9
1
160
0
.52
1
1
0
.77
1
4
0
.80
0
3
0
.48
3
3
0
.75
3
4
0
.79
1
0
180
0
.59
3
6
0
.91
0
0
0
.97
1
8
0
.57
5
0
0
.89
6
0
0
.93
8
9
190
0
.70
9
3
0
.98
4
2
1
.0
0
.62
5
6
0
.97
4
6
1
.0
Table
2
.
IVSI
with loa
d
i
ncr
e
m
ent f
or the
IE
EE 30
bus syst
e
m
Load
Factor
(%)
W
DO
PSO
Bu
s 5
Bu
s 2
6
Bu
s 3
0
Bu
s 5
Bu
s 2
6
Bu
s 3
0
100
0
.36
8
2
0
.16
0
0
0
.20
5
7
0
.35
0
0
0
.14
7
9
0
.19
5
9
120
0
.18
0
1
0
.30
5
7
0
.31
9
1
0
.44
1
7
0
.23
7
1
0
.31
4
0
140
0
.59
8
4
0
.44
7
3
0
.58
9
4
0
.55
8
9
0
.42
3
7
0
.57
7
2
150
0
.67
0
3
0
.70
1
9
1
.0
0
.65
8
3
0
.69
5
8
1
.0
Re
su
lt
s
of
t
he
IEEE
11
8
bus
syst
e
m
dep
ic
te
d
Ta
ble
3
sho
w
th
at
the
first
bus
t
o
e
xperie
nce
volt
age
colla
ps
e
is
Bus
117
an
d
then
fo
ll
owe
d
by
Bus
1
and
B
us
13.
Th
us,
Bus
117
is
co
ns
ide
red
as
th
e
op
ti
m
a
l
locat
ion
for
P
VDG.
From
the
resu
lt
s,
each
te
st
syst
e
m
gi
ves
sim
il
ar
op
tim
al
PV
DG
l
oc
at
ion
us
in
g
th
e
tw
o
diff
e
re
nt
optim
iz
at
ion
te
ch
niques;
WD
O
a
nd
PSO.
H
oweve
r,
the
IVSI
value
at
each
loa
d
fact
or
var
ie
s
wit
h
diff
e
re
nt
op
ti
m
iz
at
ion
te
chni
qu
es
.
Eval
uation
f
or
the
be
st
op
ti
m
iz
at
ion
te
ch
nique
w
il
l
no
t
be
perf
or
m
ed
consi
der
i
ng
t
ha
t
the
dif
fer
e
nces
in
t
he
I
VS
I
val
ues
a
re
not
sig
nifi
cant.
H
oweve
r,
the
t
wo
di
ff
e
ren
t
op
ti
m
iz
ation
te
chn
i
qu
e
s;
WDO
an
d
P
SO
a
r
e
com
par
ed
in
te
rm
s
of
conv
erg
e
nce
c
har
ac
te
risti
c.
Figure
s
3,
4
and
5
sho
w
the
co
nv
e
r
gen
ce
c
har
act
erist
ic
for
opti
m
u
m
PV
DG
locat
io
n
for
the
69
rad
ia
l
-
distrib
ution
sy
stem
,
30
bus
syst
em
and
118
bus
s
yst
e
m
,
resp
ect
ively
.
The
res
ults
in
the
fig
ure
s
show
t
hat
th
e
W
D
O
giv
es
faster
conve
rg
e
nce
w
i
th m
ini
m
u
m
o
bj
ect
iv
e f
unct
ion val
ues
c
ompare
d
t
o
the
PSO tech
nique.
Table
3
.
IVSI
with loa
d
i
ncr
e
m
ent f
or the
IE
EE 11
8 bus sys
tem
Load
Factor
(%)
W
DO
PSO
Bu
s 1
3
Bu
s 1
Bu
s 1
1
7
Bu
s 1
3
Bu
s 1
Bu
s 1
1
7
100
0
.03
2
4
0
.01
0
2
0
.08
9
5
0
.03
2
4
0
.01
0
2
0
.08
9
5
120
0
.11
0
6
0
.09
0
5
0
.14
0
7
0
.10
1
0
0
.08
3
7
0
.13
4
7
140
0
.12
6
5
0
.11
7
7
0
.16
3
2
0
.11
9
6
0
.10
1
4
0
.15
3
1
160
0
.31
4
3
0
.29
0
4
0
.38
7
9
0
.30
2
1
0
.28
2
0
.37
3
3
165
0
.83
7
9
0
.91
1
5
1
.0
0
.82
6
1
0
.89
5
9
1
.0
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Op
ti
m
al
pla
ce
men
t
of
gr
id
-
c
onnected
ph
otov
oltaic g
ener
at
or
s i
n a power
…
(
Zett
y Ad
ib
ah K
amar
uz
z
aman
)
345
Figure
3
.
Co
nverg
e
nce c
ha
rac
te
risti
c fo
r
opti
m
u
m
PVDG
locati
on
us
in
g WDO
and PS
O for t
he
69 bus
rad
ia
l
distrib
ution sy
ste
m
Figure
4
.
Co
nverg
e
nce c
ha
rac
te
risti
c fo
r
opti
m
u
m
PV
DG
locat
ion
Usi
ng
WDO a
nd P
S
O
f
or the
I
E
EE
30 bus sy
ste
m
Figure
5
.
Co
nverg
e
nce c
ha
rac
te
risti
c fo
r
opti
m
u
m
PV
DG
l
oc
at
ion
us
in
g W
DO an
d PS
O
f
or the
IEEE
11
8
bu
s
syst
em
Table
4
s
hows
the
MA
TLAB
tim
e
si
m
ulatio
n
f
or
the
im
plem
entat
ion
of
WDO
a
nd
PS
O
te
ch
nique
s
in
determ
ining
optim
al
loc
at
ion
of
FV
T
G
ge
ner
at
or
.
The
WDO
te
chn
i
qu
e
has
f
ast
er
tim
e
s
i
m
ula
ti
on
com
par
ed
t
o
t
he
P
SO
te
ch
ni
qu
e
w
he
re
dif
fer
e
nces
ca
n
be
seen
f
or
the
IE
E
E
30
bus
an
d
69
bu
s
s
yst
e
m
si
m
ulati
on
s.
T
her
e
fore,
it
ca
n
be
c
oncl
ude
d
that
the
WD
O
te
ch
nique
is
m
os
t
su
it
able
to
be
us
e
d
in
PVDG
op
ti
m
al
locati
on
determ
inati
on
.
Table
4
.
M
AT
LAB sim
ulati
on
ti
m
e fo
r op
ti
m
al
locati
on
of PVD
G
Op
ti
m
izatio
n
T
echn
iq
u
e
MA
TL
AB
si
m
u
lat
io
n
ti
m
e (
s
)
IE
E
E
1
1
8
bu
s
IE
E
E
3
0
bus
6
9
bu
s
syste
m
W
DO
8
1
6
.703
8
0
5
.418
8
0
4
.531
PSO
8
1
6
.831
8
1
0
.924
8
0
7
.999
4.
CONCL
US
I
O
N
S
This
pap
e
r
ha
s
presente
d
t
he
ap
plica
ti
on
of
WDO
f
or
determ
ining
optim
al
PV
D
G
locat
ion
by
m
axi
m
iz
ing
volt
age
sta
bili
ty
m
arg
in,
IVSI
as
a
n
obj
e
ct
ive
functi
on.
The
WDO
is
app
li
e
d
to
s
ol
ve
the
op
ti
m
iz
ation
pro
blem
and
it
i
s
com
par
ed
wi
th
the
PSO
te
c
hn
i
qu
e
.
The
W
DO
a
nd
P
SO
t
echn
i
qu
e
s
are
te
ste
d
on
the
IEEE
30
bu
s
syst
em
,
IEEE
118
bus
syst
e
m
and
69
-
rad
ia
l
distrib
ution
syst
em
.
Com
par
ison
on
t
he
perform
ance
of
the
WDO
a
nd
PS
O
te
ch
ni
ques
s
howe
d
th
at
the
WD
O
t
echn
i
qu
e
gi
ves
faster
c
onve
r
gen
ce
com
par
ed
to
the
PS
O
te
ch
ni
qu
e
.
A
naly
sis
of
res
ults
pro
ve
d
that
integra
ti
on
of
PVDG
on
the
wea
ke
st
bu
s
cou
l
d
im
pr
ov
e
powe
r
syst
em
vo
lt
age
sta
bili
ty
by
increasin
g
the
value
of
IV
S
I.
In
c
rease
d
P
VDG
pen
et
rati
on
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
1
3
, N
o.
1
,
Ja
nu
a
ry
201
9
:
339
–
346
346
le
vel
al
so
incr
eases
volt
age
sta
bili
ty
and
im
pr
ov
es
bu
s
vo
lt
age
pro
file
s.
The
m
axi
m
um
le
vel
of
PVD
G
pen
et
rati
on
is
determ
ined
by
ens
ur
i
ng
t
hat
the
bus
volt
ag
e
does
no
t
e
xc
eed
the
up
per
vo
lt
age
lim
it
of
the
no
m
inal vo
lt
a
ge
.
ACKN
OWLE
DGE
MENT
We
would l
ike
to tha
nk Unive
rsiti
K
eba
ngsa
an
Ma
la
ysi
a f
or
prov
i
ding the
f
ina
ncial
sup
por
t
on this
pro
j
ect
un
der
P
roject Co
de G
GP
-
2017
-
01
1.
REFERE
NCE
S
[1]
S
y
ed
Mus
ta
ffa
S
A,
Mus
iri
n
I,
O
thman
M
M,
Ma
nsor
M
H.
Chaot
ic
Muta
ti
on
Im
m
une
Evol
uti
on
ar
y
Program
m
ing
for
Volta
ge
Sec
urity
wi
th
the
Presenc
e
of
DG
PV
.
Indone
sian
Journal
of
El
e
ct
ri
ca
l
Eng
ineer
ing
and
Com
pu
te
r
Scie
nc
e, Vol. 6,
No.
3,
pp.
721
~
729,
2017
.
[2]
Zi
chun
Y
et
a
l.
Anal
y
tical
m
et
h
od
of
the
impact
of
distri
bute
d
g
ene
ra
ti
on
on
static
voltage
stabili
t
y
of
distr
ibut
io
n
net
work
and
it
s
developm
ent
.
T
EL
KO
MN
IKA
I
ndonesia
n
Journ
al
of
Elec
tr
ical
Engi
ne
eri
ng,
Vo
l.
11
,
No.
9,
pp.
5018~5029,
201
3.
[3]
Vovos
P
N,
Ki
pra
kis
A
E,
W
a
ll
a
ce
A
R,
Harr
ison
G
P.
Cent
r
al
i
ze
d
a
nd
d
istributed
voltage
c
ontrol
:
Im
pact
o
n
distri
bute
d
gen
er
at
ion
penetration
.
IE
EE T
r
ansa
cti
on
on
Pow
er
S
y
s
te
m
s,
22(1), pp.
476
-
83,
2007
.
[4]
Cai
sheng
W
,
Ne
hrir
MH
.
Anal
y
t
ic
a
l
appr
o
ac
hes
for
opti
m
al
p
la
c
ement
of
distr
ib
ute
d
gen
erati
on
source
s
in
powe
r
s
y
stems
.
IE
EE T
ran
sac
t
ion
on
Pow
er
S
y
st
ems
,
19
(4),
pp
.
2068
-
20
76,
2004
.
[5]
Acha
r
y
a
N,
Mah
at
P,
Mithu
la
n
an
tha
n
N.
An
an
aly
t
ical
appr
o
ac
h
f
or
DG
al
locati
on
in
primar
y
distr
ibut
ion
n
et
work
.
Inte
rna
ti
ona
l
Jou
rna
l El
e
ct
r
ic
Po
wer,
28(10)
,
pp
.
669
-
78,
2006
.
[6]
Am
an
M
M,
Jas
m
on
G
B,
Mokh
li
s
H,
Baka
r
A
H
A.
Optimal
pla
ce
m
ent
and
sizing
of
DG
base
d
on
a
new
power
stabi
lit
y
inde
x
a
nd
li
n
e
losses.
Elec
tr
ic
a
l
Pow
er a
nd
Ene
rg
y
S
y
st
e
m
s,43,
pp.
1296
-
1304,
2012
.
[7]
Al
Abri
R
S,
El
-
Saada
n
y
Eha
b
F,
Atwa
Yass
er
M.
Optimal
pla
c
ement
and
sizi
n
g
m
et
hod
to
im
prove
the
voltag
e
stabi
lit
y
m
arg
in
in
a
distri
buti
on
sy
st
em
using
di
stribut
ed
gen
era
t
ion.
IEEE
Tra
ns
ac
t
ion
on
Pow
er
Sy
st
ems
,
28(1),
pp.
326
-
34
,
201
3.
[8]
Murt
y
V
V,
Ashwa
ni
K.
Optimal
place
m
ent
of
DG
in
rad
ia
l
distri
buti
on
s
y
s
te
m
s
base
d
on
new
volt
age
stab
ili
t
y
inde
x
und
er loa
d
growth. Elec
t
rical
Pow
er and En
erg
y
,
69
,
pp
.
246
-
56,
2015
.
[9]
Kavia
ni
-
Aran
i
A.
Optimal
pla
c
e
m
ent
and
sizi
ng
of
distri
bute
d
ge
ner
ation
unit
s
using
co
-
evol
ut
ion
ar
y
pa
rticle
sw
arm
opti
m
iz
ation
a
lg
orit
hm
s.
TE
LK
OM
NIK
A
Indone
sian
Jour
nal
of
El
e
ct
ri
cal
Eng
i
nee
ring
,
Vol.
13
,
No.
2
,
pp.
247
~
256,
2015
.
[10]
Morave
j
Z,
Ak
hla
ghi
A.
A
novel
appr
oa
ch
ba
sed
on
cuc
koo
sea
rch
for
DG
al
lo
ca
t
ion
in
distri
buti
on
ne
twork.
El
e
ct
ri
c
Pow
er E
ner
g
y
S
y
s
te
m
s, 44,
pp
.
672
-
679,
2013.
[11]
Ross
et
i
Gus
ta
vo
J
S,
de
Oliv
ei
r
a
Edi
m
ar
J,
de
Olivi
er
a
Le
on
ar
do
W
,
CS
il
va
I,
Pere
s
W
.
Optimal
al
locati
on
o
f
distri
bute
d
gene
r
at
ion
wi
th
re
con
figura
t
ion
in
e
lectr
i
c
distri
bu
ti
on
s
y
stems
.
Elec
tric
Pow
er
S
y
stem Re
sea
rch
,
103,
p
p.
178
-
83,
2013
.
[12]
Ka
y
al
P,
Chand
a
C
K.
Placem
e
nt
of
wind
and
s
ola
r
b
ase
d
DG
s
in
distri
bu
ti
on
s
y
stem
for
power
loss
m
ini
m
iz
at
i
on
and
vol
ta
ge
stab
i
li
t
y
improvem
ent
.
Journa
l
on
Ele
ct
ri
ca
l
Pow
er
an
d
Ene
rg
y
S
y
st
e
m
s 53,
pp.
795
-
8
09,
2013
.
[13]
Nasira
ghdam
H,
and
Jadid
S.
Optimal
h
y
brid
PV
/WT/FC
sizi
ng
and
distri
buti
on
s
y
stem
re
conf
igur
at
ion
using
m
ult
i
-
obje
c
ti
ve
ar
ti
fi
cial
b
ee c
olon
y
a
lg
orit
hm
.
Journal
on
Solar Ene
rg
y
,
86,
pp.
3057
-
71
,
2012
.
[14]
Hung
D
Q,
Mit
hula
nan
tha
n
N
,
Bansal
R
C.
I
nt
egr
ation
of
PV
and
BES
un
it
s
i
n
comm
erc
ia
l
di
stribut
ion
s
y
ste
m
s
conside
ring
en
er
g
y
loss a
nd
vol
t
a
ge
stab
il
i
t
y
.
Journal
on
Appli
ed
E
ner
g
y
,
113
,
pp.
1
162
-
70,
2014
.
[15]
Ba
y
r
aktar
Z
,
Ko
m
urc
u
M,
W
ern
er
D
H.
W
ind
d
rive
n
opt
imizati
on
(W
DO
):
A
novel
n
at
ure
-
inspire
d
opt
imizatio
n
al
gorit
hm
and it
s
application
to el
ec
tromagne
ti
cs
2
010.
[16]
Yang
C
F,
L
ai
G
G,
L
ee
C
H,
Su
C
T,
Chang
G
W
.
Optimal
s
et
ti
ng
of
r
eact
iv
e
compensat
ion
devi
c
es
with
an
improved
volt
ag
e
stabi
l
ity
ind
ex
for
volt
ag
e
stab
il
ity
enha
n
ce
m
e
nt.
E
le
c
trica
l
Po
wer
and
Ene
rg
y
S
y
stems
,
37,
p
p.
50
-
57,
2012
.
[17]
He
J,
Guo
H.
A
m
odifi
ed
par
t
ic
l
e
sw
arm
opti
m
iz
at
ion
a
lgorithm
.
TE
LKOM
NIK
A
Indone
sian
Jour
nal
of
E
lectr
i
cal
Engi
ne
eri
ng,
Vo
l.
11
,
No
.
10
,
62
09
~ 6215, 2013.
Evaluation Warning : The document was created with Spire.PDF for Python.