Indonesi
an
Journa
l
of El
ect
ri
cal Engineer
ing
an
d
Comp
ut
er
Scie
nce
Vo
l.
13
,
No.
2
,
Febr
uar
y
201
9
, pp.
4
37
~
446
IS
S
N: 25
02
-
4752, DO
I: 10
.11
591/ijeecs
.v1
3
.i
2
.pp
437
-
446
437
Journ
al h
om
e
page
:
http:
//
ia
es
core.c
om/j
ourn
als/i
ndex.
ph
p/ij
eecs
Allocati
on
of
dist
ributed
ge
n
eration and
c
apacitor
bank
s in
distribu
tion syst
em
Oladep
o Ol
at
unde
1
, H
as
im
ah
Abdul
Rah
man
2
1,2
Cent
re
of Elec
tri
c
al
Ene
rg
y
S
ystems
(CEE
S),
I
nstit
ute of
Futur
e
En
erg
y
(IFE)
,
Univer
siti
Te
kno
logi
M
a
lay
sia
(U
TM)
,
Malay
s
ia
1
El
e
c
trica
l
and
E
le
c
troni
c
Eng
ineeri
ng
Dep
art
m
en
t,
Os
un
Sta
te Unive
rsit
y
,
Nig
eri
a
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
J
ul
8
,
2018
Re
vised
Sep
9
,
2018
Accepte
d
Se
p 23
, 201
8
Volta
ge
profile
and
power
losses
on
the
distri
buti
on
s
y
stem
is
a
func
ti
on
of
rea
l
and
imaginar
y
pow
er
lo
adi
n
g
condition.
Thi
s
ca
n
be
eff
e
ctive
l
y
m
ana
ge
d
through
th
e
con
t
roll
ed
real
and
r
ea
c
ti
ve
power
f
l
ow
b
y
opti
m
al
p
la
c
ement
of
ca
pa
ci
tor
banks
(CB)
and
d
istri
b
ute
d
g
ene
r
at
ors
(DG
).
Thi
s
p
aper
pre
sent
s
Adapti
ve
Part
ic
le
Sw
arm
Optimiza
ti
on
(AP
SO
)
to
eff
ic
i
ent
l
y
t
ac
kl
e
th
e
proble
m
of
sim
ult
an
eous
al
lo
cation
of
DG
and
CB
in
r
adi
a
l
distri
buti
o
n
s
y
stem
to
re
vamp
volt
age
m
agni
tude
a
nd
red
uce
po
wer
losses.
The
m
odifi
catio
n
to
the
conve
n
ti
onal
Part
ic
l
e
Sw
arm
Optimiza
ti
on
(PS
O
)
was
ac
hie
ved
b
y
rep
l
acing
the
ine
rtial
weigh
t
e
quat
ion
(W
)
in
the
vel
oc
i
t
y
updat
e
equation
,
base
d
on
th
e
par
ticl
e
best
e
xper
ie
n
ce
in
th
e
pre
v
ious
it
er
at
ion
.
Th
e
in
ert
i
al
we
ight
equ
at
ion
is
design
ed
to
v
ar
y
with
res
pec
t
to
the
it
er
at
ion
val
u
e
i
n
the
al
gor
it
hm
.
The
propos
ed
m
et
hod
was
inv
esti
gated
on
IEE
E
30
-
bus,
33
-
bus
and
69
-
bus
te
st
distri
bu
ti
on
s
y
stems
.
The
res
ult
s
show
a
signifi
c
ant
impr
ovement
in
th
e
rat
e
of
conve
r
genc
e
of
AP
SO
,
improved
volt
ag
e
profi
le and loss re
du
ct
io
n.
Ke
yw
or
ds:
Ad
a
ptive
pa
rtic
le
swar
m
op
ti
m
iz
ation
(APS
O)
Ca
pacit
or
ba
nk
s
Distrib
ution ge
ner
at
io
n
Distrib
ution sy
stem
In
e
rtia
l weig
ht
Vo
lt
age
pr
of
il
e
Copyright
©
201
9
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
:
Hasim
ah
Abd
ul
Rahm
an
,
C
entre
of Elec
tric
al
En
e
rg
y
S
yst
e
m
s (
CEES)
,
In
sti
tute
of F
utu
re
En
e
r
gy (IF
E),
U
niv
e
rsiti
Tekno
l
og
i M
al
ay
sia
(
UTM
),
81310 Jo
hor
B
ahru,
Mal
ay
sia
.
Em
a
il
:
hash
im
ahar@utm
.
m
y
1.
INTROD
U
CTION
The
global
in
crease
in
el
ect
rical
po
we
r
dem
and
and
t
he
natu
ral
de
plet
ion
in
f
ossi
l
fu
el
had
necessit
at
ed
th
e
al
te
rn
at
ive
s
ources
of
pow
er
ge
ne
rati
on
[
1]
.
Th
e
co
nve
ntion
al
m
et
ho
d
of
t
ran
sm
issi
on
li
ne
upgradi
ng
is
hig
hly
ca
pital
de
m
and
in
g
with
lim
it
ed
ex
pan
s
ion
i
n
powe
r
c
a
pacit
y.
I
n
m
od
er
nized
distri
bu
t
i
on
powe
r
syst
e
m
plan
ning
an
d
operati
on,
(DG
)
with
m
ic
ro
gr
i
d
(M
D)
had
be
ing
a
viable
al
te
rn
at
ive
a
nd
sol
ution
to
power
c
halle
ng
e
s
[2]
.
D
G,
eq
ually
know
n
as
dis
pe
rsed
ge
ner
at
io
n
an
nex
e
s
sm
al
l
-
scal
e
te
chnolo
gi
es
that
can
be
powe
re
d
by
re
ne
wab
l
e
ene
rg
y
sour
c
es
f
or
the
pro
duct
ion
of
el
ect
rici
ty
at
us
ers
’
vicinit
ie
s.
T
ec
hn
ic
al
adv
a
ntage
s
of
re
new
a
bl
e
ge
ne
rati
on
incl
ud
e r
edu
ct
io
n
of
pol
luti
on
, p
owe
r
loss
m
ini
m
iz
at
i
on,
volt
age
sta
bili
t
y
enh
a
ncem
ent
and
reli
abili
ty
i
m
pr
ovem
ent
in
distrib
utio
n
netw
ork.
It
can
be
m
od
el
le
d
as
in
du
ct
io
n
a
nd
synch
ron
ous
ge
ner
at
or
s
de
pe
nd
i
ng
on
it
s
e
xpect
ed
operati
on
m
od
e,
ei
the
r
to
ge
ner
at
e
r
eact
ive
powe
r
on
ly
or
gen
e
rati
on
of
both
real
a
nd
re
act
ive
pow
er
[
3]
.
D
G
ar
e
m
od
el
ed
at
unit
y
powe
r
fact
or
to
deliver
act
iv
e
pow
e
r
so
urces
.
The
i
nvolv
em
ent
of
non
-
pro
gr
am
m
able
ren
e
wa
ble
res
ources
raises
m
any
tech
nical
issue
s
in
the
op
e
rati
on
of
m
od
ern
powe
r
distri
bu
ti
on
netw
orks.
W
it
h
co
nventi
on
al
rad
ia
ll
y
structu
re
d
dist
rib
ution
netw
orks,
s
ome
te
chn
ic
al
chall
eng
es
lim
it
the
integrati
on
of
D
G
un
it
s:
increase
fau
lt
cur
re
nt
m
agn
it
ud
e,
po
s
sible
reserv
e pow
e
r flo
w,
a
m
pacit
y violat
ion
s a
nd
vo
lt
a
ge variat
io
n from
the o
pe
rati
on lim
it
[4]
.
Conve
ntion
al
l
y,
capaci
tor
banks
(CB)
a
re
place
d
on
distrib
ution
syst
e
m
fo
r
re
act
ive
powe
r
com
pen
sat
ion.
It
is
a
reacti
ve
loa
d
gen
e
r
at
or
.
Its
util
izati
on
ben
e
fits
inclu
des:
m
i
nim
iz
at
ion
of
powe
r
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,
Vol
.
13
, N
o.
2
,
Fe
bru
ary 2
019
:
4
3
7
–
4
4
6
438
reducti
on
due
t
o
losse
s,
powe
r
facto
r
re
gula
ti
on
a
nd
m
ai
ntenan
ce
of
volt
age
re
gula
ti
on
s
on
l
oad
buses
of
t
he
netw
ork
[
5]
.
P
ow
e
r
los
s
re
duct
ion
an
d
volt
age
de
viati
on
im
pr
ov
em
ent
are
sig
nificant
a
nd
i
nf
l
uen
ti
al
on
t
he
op
e
rati
on
of
distribu
ti
on
netw
ork,
i
n
sup
plyi
ng
ec
onom
ic
al
and
high
-
qual
it
y
po
we
r
sup
pl
y
to
con
s
um
er
load
[5,
6]
.
Sizi
ng
and
al
locat
io
n
of
D
G
unit
s
with
capaci
to
r
banks
dem
and
s
th
oroug
h
inv
e
sti
gation
to
avoi
d
abno
rm
al
rise
in
vo
lt
age
on
distrib
ution
fee
der
s
,
li
ne
overl
oad
i
ng
and
unec
onom
ic
al
inv
estme
nt
co
st
.
Ther
e
f
or
e,
eval
uating t
he o
ptim
al
p
enetrati
on level,
DG loc
at
ion
alo
ng
with ca
pacit
or
ba
nks
for power
quali
ty
i
m
pr
ovem
ent, w
hile m
ai
ntaining
re
qu
ire
d o
pe
rati
ng con
diti
on is im
per
at
ive
[
8]
.
Abo
ut
te
n
ye
ar
s
ago,
di
ff
e
ren
t
researc
h
pa
pe
r
s
had
bee
n
pu
b
li
sh
ed
on
op
ti
m
al
locat
ion
and
siz
i
ng
of
DG
s
[
9]
.
A
ppreciable
resea
r
cher
s
ta
rg
et
e
d
integrati
on
of
capaci
to
r
ba
nk
s
i
n
distri
buti
on
syst
e
m
[10]
.
Howe
ver,
so
m
e
recent
pap
e
rs
fo
cu
s
on
op
ti
m
al
s
iz
ing
and
place
m
ent
of
DG
s
al
on
g
with
capaci
to
r
ba
nk
s
on
powe
r
distrib
ut
ion
syst
em
.
In
add
it
io
n,
PS
O
had
been
c
onvi
ncin
gly
i
m
ple
m
ented
in
pow
er
syst
e
m
research
es
and
oth
er
e
ng
i
neer
i
ng
fiel
ds
[11]
.
It
was
cl
early
sho
wn
that
P
SO
pe
rfo
rm
ed
bette
r,
f
ast
er
a
nd
acc
urat
e
in
com
par
ison
to
oth
e
r
m
et
ho
ds
[12]
.
The
vel
oc
it
y
equ
at
ion
of
PSO
wa
s
m
od
ifie
d
in
[
13
]
by
add
it
io
n
of
new
te
rm
s
and
eval
uation
of
di
ff
e
ren
ce
i
n
the
gl
ob
al
best
an
d
l
ocal
best
of
pa
rtic
le
s.
It
outp
erfor
m
ed
oth
e
r
PSO
m
od
ific
at
ion
in
te
rm
of
c
onverge
nce
s
pee
d,
qual
it
y
of
accuracy
wh
e
n
te
ste
d
on
va
rio
us
syst
em
m
od
el
s.
An
a
da
ptive
P
SO
w
as
prese
nted
in
[14]
,
the
pa
rtic
le
velocit
y
is
adj
ust
ed
u
si
ng
t
he
inerti
al
weig
ht
strat
eg
y
su
c
h
that
t
he
best
fitt
ed
pa
r
ti
cl
e
nav
igate
d
in
c
om
par
iso
n
to
s
peed
ins
pire
d
lo
w
fitt
ed
par
ti
cl
es.
I
n
[
15
]
,
the
op
ti
m
u
m
si
m
ul
ta
neo
us
D
G
an
d
capaci
to
r
placem
ent
was
pr
ese
nted
,
con
si
der
i
ng
power
los
s
m
inim
iz
at
ion
as
a
fitness
fun
ct
ion
.
T
he
m
eth
od
was
te
ste
d
on
sta
ndar
d
I
EEE
bus
syst
em
.
The
authors
in
[16]
presen
te
d
an
al
gorithm
fo
r
op
ti
m
al
al
loca
ti
on
of
D
G
usi
ng
im
pr
ove
d
Gen
et
ic
for
I
EEE
-
33
b
us
di
stribu
te
d
syst
e
m
to
com
pu
te
the
be
st
perform
ing
reacti
ve
po
we
r
co
ntr
ol
va
riables
f
or
act
ive
powe
r
loss
reducti
on
a
nd
m
a
xim
iz
e
vo
lt
age
pro
file
.
The
m
et
ho
d
re
duced
the
m
ulti
-
obj
ect
ive
f
unct
ion
into
si
ng
le
ob
j
ect
ive
to
a
vo
id
the
com
plexity
in
vo
l
ved
i
n
m
u
l
ti
-
obj
ect
ive
prob
le
m
.
In
[
17
]
,
the
load
var
i
at
ion
s
is
co
ns
i
der
e
d
to
ca
pture
the
syst
e
m
dyna
m
ic
natur
e.
The
loss
sensiti
vity
factor
s
was
us
ed
t
o
eval
uate
an
d
deter
m
ine
the
bu
s
wh
e
r
e
com
pen
sat
ion
i
s
required.
The
reafter
PS
O
w
as
e
m
plo
ye
d
to
evaluate
the
capaci
ti
es
of
the
capaci
to
r
ba
nks
to
be
place
d.
For
the
va
ryi
ng
s
ensiti
vity
facto
rs:
re
du
ce
d
po
wer
l
os
ses
with
be
tt
er
ene
rgy
con
se
r
vation
was
recorde
d.
C
om
par
at
ive
analy
sis
of
PSO
and
oth
er
te
chn
i
qu
e
s
li
ke
gen
et
ic
al
gor
it
h
m
,
ta
bu
search
a
nd
hybri
dizat
ion
m
et
ho
d sh
ows
that PS
O pro
duces
bette
r
sa
vi
ng
,
r
e
duced
lo
sses and
am
eli
or
at
e
d vo
lt
age
prof
il
e.
This
w
ork
pr
esents
the
sim
ul
ta
neo
us
D
G
an
d
capa
ci
tor
ba
nk
al
loc
at
ion
in
distri
bu
ti
on
net
wor
k,
wh
il
e
co
ns
ide
rin
g
m
ini
m
iz
a
ti
on
of
po
wer
losses
an
d
volt
age
pro
file
m
axi
m
iz
a
ti
on
as
fitness
f
unct
io
n.
The
vo
lt
age
boun
dar
y
li
m
it
s
are
co
ns
ide
re
d
as
syst
em
c
on
st
raints.
T
he
m
od
ific
at
ion
was
achie
ve
d
by
rep
la
ci
ng
the
inerti
al
weig
ht
(w
)
in
the
ve
locit
y
up
date
base
on
the
su
ccess
achie
ve
in
exp
lo
rati
on
a
nd
exp
l
oitat
ion
of
the
search
sp
a
ce.
The
te
chn
i
cal
achievem
e
nt
of
a
par
ti
cl
e
in
it
s
cur
ren
t
locat
ion
is
dete
rm
ined
to
decide
ch
oic
e
of
inerti
al
we
igh
t
for
su
c
h
pa
rtic
le
fo
r
the
nex
t
it
erati
on.
In
ad
diti
on,
th
e
par
ti
cl
es
trai
l
ed
the
sensiti
vity
o
f
the
buses to
enh
ance
bette
r
c
onverge
nce.
2.
RESEA
R
CH MET
HO
D
As
sho
wn
is
1
give
s
the
pr
opos
e
d
m
at
he
m
at
ic
al
m
od
el
fo
r
t
he
sim
ul
ta
neous
D
G
a
nd
ca
pacit
or
placem
ent co
nsi
der
in
g
t
he fit
ness funct
io
n
a
s m
ini
m
iz
ation of total
acti
ve powe
r
loss
es
NL
i
I
R
K
f
M
i
n
i
i
NL
i
l
i
2
(1)
Su
c
h
that
NL
i
I
I
K
l
i
i
m
a
x
(2)
NL
i
V
V
V
i
i
i
m
a
x
m
i
n
(3)
NL
i
P
P
P
i
i
i
m
a
x
(4)
m
a
x
c
to
ta
l
QQ
(5)
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
All
oca
ti
on
of di
stri
bu
te
d ge
ne
ra
ti
on
and
c
ap
acitor
banks
in
d
ist
rib
ution sy
ste
m
(
Olade
po
Olatun
de
)
439
Wh
e
re
I
i
is l
ine c
urre
nt
in
bra
nch i
R
i
is resist
ance
of branc
h
i
V
i
is
node v
oltage at
node
i
P
i
is real
power i
n br
a
nc
h
i
Qi is reacti
ve
powe
r
i
n br
a
nch i
Q
cm
ax
is
m
axi
mu
m
size of
CB
Q
total
is t
otal re
act
ive pow
e
r d
e
m
and
K
i
re
pr
ese
nts t
he
to
polo
gical
sta
tus of t
he br
anch
e
s
N
is s
um
of
branch
e
s
NL
is t
he
set
of
br
a
nch
e
s
The
vo
lt
a
ge
st
abili
ty
pr
op
ose
s
pote
ntial
of
distrib
utio
n
ne
twork
within
volt
age
li
m
it
against
t
he
var
ia
ti
on in
loa
d dem
and
. T
he
f
it
ness
fun
ct
io
n
is e
xpresse
d as 6
.
2
42
2
2
,
.
.
.
,
m
i
n
4
4
s
s
i
L
i
i
L
i
i
L
i
i
L
i
iN
f
V
V
R
P
X
Q
X
P
R
Q
(6)
Mod
el
in
g of D
G un
it
p
rese
nte
d
in
P
Q node
m
od
e adopte
d are:
a)
DG
a
s
‘
Ne
gative
P
Q
loa
d’
m
od
el
of
P
Q
m
od
e.
T
he
D
G
is
si
m
ply
m
od
el
ed
as
a
c
onsta
nt
act
ive
(P)
a
nd
reacti
ve
(
Q)
powe
r
ge
ne
rati
ng
sou
rce.
T
he
giv
e
n
val
ues
of
the
D
G
m
odel
are
real
(P
DG
)
an
d
reacti
ve
(Q
DG
)
. T
he
m
od
ifie
d l
oa
d
at
bus i
with
D
G u
nit i
s expr
e
sse
d
as:
(7)
i
DG
i
l
o
a
d
i
l
o
a
d
Q
Q
Q
,
,
,
(8)
b)
DG
as
‘Consta
nt
Power
Fact
or’
m
od
el
of
PQ
m
od
e.
Sync
hr
onous
ge
ne
rato
r
an
d
po
wer
el
ect
ronic
base
d
DG
unit
s
was
m
od
el
ed
as
co
ns
ta
nt
power
f
act
or
m
od
el
.
T
he
po
wer
outp
ut
was
re
gu
l
at
e
d
by
c
ontrolli
ng
the
excit
in
g
c
urren
t
a
nd
tri
gg
e
r
a
ng
le
s
f
or
sync
hrono
us
ge
ner
at
or
a
nd
power
el
ect
ronic
ba
se
D
Gs
resp
ect
ively
.
A
s
sh
ow
n
in
9
a
nd
10
giv
e
the
reacti
ve
powe
r
an
d
eq
uiv
al
e
nt
current
in
j
e
ct
ion
of
the
D
G
resp
ect
ively
.
))
(
t
a
n
(
c
o
s
1
i
D
G
i
D
G
i
D
G
PF
P
Q
(9)
*
)
(
)
(
i
D
G
i
D
G
i
D
G
i
D
G
i
i
D
G
i
D
G
r
i
D
G
i
D
G
V
jQ
P
V
jI
V
I
I
(10)
Wh
e
re,
r
i
D
G
I
is real
com
po
ne
nt of
DG cu
rr
e
nt in
je
ct
ion
,
i
i
D
G
I
is Im
aginar
y c
om
po
ne
nt of
D
G
c
urren
t i
nj
e
c
ti
on
.
c)
The
DG
as
PV
node
is
c
ommon
ly
c
onsta
nt
vo
lt
age
m
od
el
.
The
gi
ven
val
ues
of
the
D
G
m
od
el
are
the
real
powe
r
a
nd
bus
vo
lt
age
m
agn
it
ude.
I
n
order
t
o
m
ai
ntain
in
g
c
on
sta
nt
vo
lt
age
,
vo
lt
ag
e
var
ia
ti
on
with
tim
e
is
m
a
intain
ed
a
ppr
ox
im
a
te
ly
zero
by
inj
ect
ing
require
d
reacti
ve
pow
er.
I
n
this
w
ork,
the
DG
w
a
s
config
ur
e
d
t
o
s
upply b
oth t
he real
an
d react
ive
powe
r wh
il
e the CB s
uppl
ie
d
the
reacti
ve
pow
e
r.
The
bac
kwar
d/
forw
a
r
d
swee
p
(BFS)
te
ch
ni
qu
e
for
loa
d
f
low
analy
sis
was
ad
opte
d
f
or
the
wor
k.
The
rati
o
of
r
esi
sta
nce
to
re
act
ance
(R/X
)
in
distri
bu
ti
on
syst
e
m
is
high
a
nd
the
c
on
ven
ti
onal
pow
er
fl
ow
m
et
ho
ds
li
ke
the
Ne
wton
-
R
aphso
n
ca
nnot
be
perfect
ly
fitt
ed,
ta
king
into
c
ognizan
ce
their
m
et
ho
d
of
der
i
vation.
T
he
refor
e
,
it
m
ay
ex
per
ie
nce
c
onve
r
gen
ce
dif
f
ic
ulty
in
so
l
vin
g
po
wer
flo
w
in
rad
ia
l
distribu
ti
on
netw
ork.
I
n
th
e
BFS
te
ch
nique,
t
he
netw
ork
syst
em
is
as
su
m
ed
to
be
ba
la
nced
a
nd
is
represe
nted
by
a
n
equ
i
valent
si
ng
le
li
ne
dia
gr
am
.
T
he
a
naly
sis sta
rts
f
ro
m
one
branc
h
to
a
nother
in
a
se
quen
ti
al
way
unti
l
al
l
the
br
a
nc
hes
in
the
fee
der
s
ha
ve
be
en
trace
d.
T
he
bus
volt
ages,
exce
pt
the
sou
rce
bu
s
,
a
r
e
ini
ti
al
l
y
assum
ed
to
be
on
e
(
1)
pu
at
a
ng
le
z
er
o
(0).
Ba
sed
on
these
volt
ages
a
nd
giv
e
n
real
an
d
reacti
ve
powe
r
,
the
branc
h
cu
rr
e
nts,
procee
ding
f
rom
the
end
bus
es
to
the
s
our
ce,
are
e
valuat
ed
an
d
sa
ve
d
(Back
ward
S
w
eep).
It
is
ho
wev
e
r,
requires
a
se
quent
ia
l
pr
oc
ed
ure
to
ens
ur
e
the
netw
ork
branc
hes
are
rig
htly
traced.
T
he
br
a
nch
i
ncide
nce
ta
bl
e
was
use
d
to
a
void
s
kippin
g
of
bu
s
ses
an
d
br
anch
e
s.
T
hen,
br
a
nc
h
cu
rr
e
nt
s,
are
com
pu
te
d
to
cal
culat
e
the
real
and reacti
ve
po
wer l
os
s
on t
he
n
et
w
ork
. T
he c
urren
t at
t
he
s
ource e
nd is
now ob
ta
ine
d usi
ng 11:
i
DG
i
l
o
a
d
i
l
o
a
d
P
P
P
,
,
,
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,
Vol
.
13
, N
o.
2
,
Fe
bru
ary 2
019
:
4
3
7
–
4
4
6
440
=
∑
=
1
≠
+
∑
.
=
1
=
1
≠
+
(
∑
+
=
1
≠
∑
.
=
1
=
1
≠
)
)
/
∗
(11)
=
+
(12)
.
=
−
=
(13)
.
=
2
(14)
.
=
2
(15)
Wh
e
re:
∑
=
1
≠
is sum
m
ation
of act
ive loa
d powe
r
c
onnect
ed
to
en
ti
re
consu
m
er’
e
nd
buses;
∑
=
1
≠
is sum
m
ation
of r
eact
ive
loa
d powe
r
c
onne
ct
ed
to t
he
e
ntire c
on
s
um
er’
e
nd buses;
∑
.
=
1
=
1
≠
is sum
m
ation
of branc
h (ij
) r
eal
p
owe
r
los
s
acro
s
s the
all
net
work branc
he
s;
∑
.
=
1
=
1
≠
is sum
m
ation
of branc
h (ij
) r
eact
ive powe
r
l
os
s ac
r
os
s t
he a
ll
n
et
wor
k br
a
nch
e
s;
∗
is t
he
c
onju
gate of the
sou
rce
vo
lt
age;
I
is c
urren
t at
t
he
s
ource e
nd;
is i
m
ped
ance
of branc
h
i
j
;
is resist
ance
of
br
a
nch ij;
is rea
ct
an
ce
of branc
h
i
j
;
.
is v
oltage
dro
p
ac
ro
s
s
br
a
nc
h
ij
are
volt
age at
bu
s
j
a
nd i res
pe
ct
ively
;
is cu
rr
e
nt th
rough b
us
i t
o j.
The
forw
a
r
d
s
weep
is
the
n
be
gin
s
at
t
he
s
ource
t
o
the
en
d
of
t
he
fee
de
r
s
to
c
o
m
pu
te
vo
lt
age
dr
op
us
in
g
13,
c
urr
ent
(
),
real
a
nd
reacti
ve
po
wer
los
ses
us
i
ng
14
a
nd
15
resp
ect
ively
.
T
he
e
valu
at
ed
total
powe
r
los
ses
are
c
om
par
ed
to
init
ia
l
valu
es
com
pu
te
d
by
ass
um
ing
on
e
pe
r
unit
vo
lt
age
f
or
al
l
buse
s
.
If
the
di
ff
e
rence
exce
eds
the
tolerance
lim
i
ts,
the
so
urce
current
is
re
-
c
om
pu
te
d
us
i
ng
11
,
in
te
rm
s
of
t
he
new
ly
ob
ta
ine
d
los
ses,
a
nd
t
he
path
retract
i
ng
ope
rati
on
is
repea
te
d.
The
com
pu
ta
ti
on
proce
dure
is
re
peate
d
un
ti
l
the
var
ia
t
ion
in
losses
be
tween
su
c
cess
ive
values
of
the
source
c
urr
ent
is
within
the
sp
eci
fied
tol
eran
ce
lim
it
. Th
e P
SO pa
rtic
le
’s
vel
oc
it
y and
posit
ion i
s m
at
he
m
atical
ly
m
od
el
ed
as:
(16)
Wh
e
re
k
i
v
is
a
ge
nt
i
velocit
y
at
it
erati
on
k,
w
is
weig
htin
g
functi
on,
12
c
a
n
d
c
are
acc
el
erati
on
c
oe
ffi
ci
ents
,
rand
is
ra
ndom
nu
m
ber
between
0
a
nd
1,
k
i
x
and
k
j
x
are
a
ge
nt
i
a
nd
j
cu
rr
e
nt
posit
io
n
at
it
erati
on
k
resp
ect
ively
,
i
p
b
e
s
t
is
the
pb
est
of
age
nt
i,
an
d
gb
e
st
is
the
gbest
of
th
e
gr
oup.
The
pa
rtic
le
po
sit
io
n
is
updated
usi
ng
17.
11
k
k
k
ii
x
x
v
(17)
The wei
ghti
ng
factor i
s e
xpres
sed
as
m
a
x
m
in
m
a
x
m
a
x
ww
w
w
I
tre
I
tre
(18)
k
i
k
i
i
k
i
k
i
x
g
b
e
s
t
r
a
n
d
c
x
p
b
e
s
t
r
a
n
d
c
wv
v
2
2
1
1
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
All
oca
ti
on
of di
stri
bu
te
d ge
ne
ra
ti
on
and
c
ap
acitor
banks
in
d
ist
rib
ution sy
ste
m
(
Olade
po
Olatun
de
)
441
Wh
e
re
w
is
the
weig
hting
f
unc
ti
on
,
m
a
x
w
and
m
i
n
w
are
m
axi
m
u
m
and
m
ini
m
u
m
weigh
ts
,
m
a
x
I
t
r
e
and
I
t
r
e
are
m
axim
u
m
and
c
urr
ent
it
erati
on
re
sp
ect
ively
.
T
he
cho
ic
e
of
i
ne
rtia
weig
ht
pl
ay
s
a
sign
ific
a
nt
ro
le
in
t
he
pe
r
f
or
m
ance
of
the
par
ti
cl
e
swa
r
m
op
ti
m
iz
at
ion
.
The
m
od
ific
a
ti
on
on
t
he
ine
r
ti
al
weigh
t
(
w)
in
the
velocit
y
updat
e
was
base
d
on
the
s
ucces
s
achieve
in
e
xp
l
or
at
io
n
an
d
exp
loit
at
io
n
of
the
sea
rch
sp
ace.
The
te
ch
nical
achievem
ent
of
a
par
ti
cl
e
in
it
s
cur
ren
t
lo
cat
ion
is
deter
m
ined
to
deci
de
ch
oice
of
i
ner
ti
al
weig
ht
f
or
s
uc
h
pa
rtic
le
for
the
ne
xt
it
erati
on.
I
n
a
dd
it
io
n,
the
pa
rtic
le
s
trai
le
d
the
sensit
ivit
y
of
the
buses
to
enh
a
nce
bette
r c
onve
rg
e
nce.
The
Fi
gure
1
s
hows
t
he
propose
d
al
gorithm
for
t
he
ada
pti
ve
base
P
SO
f
or
sim
ultaneou
s
locat
io
n
of
DG
an
d
CB
.
If
the
P
D
is
the
t
ot
al
power
dem
and
on
distrib
ut
ion
netw
ork.
Q
D
is
the
total
reacti
ve
power
on
th
e
netw
ork. The
re
fore:
D
D
D
S
P
Q
(19)
S
D
de
fines
t
he t
otal netw
ork
l
oad
i
ng. Re
pr
es
enting a
par
ti
cl
e b
y:
1
2
3
,
4
,
,
,
....
....
np
P
P
P
P
P
P
(20)
If
Nb
re
prese
nt
the n
um
ber
o
f
b
use
s o
n
the net
work
su
c
h
t
ha
t:
(b
=
2,3
,4…N
b)
r
e
pr
ese
nt
the possible locat
io
n
of DG
or CB
on th
e
n
et
wor
k.
Th
us
:
22
D
G
C
B
L
N
b
a
n
d
L
N
b
Figure
1.
Flo
w
char
t
of the
pro
po
s
ed
m
et
ho
d
The
st
ru
ct
ur
e
o
f
eac
h partic
le
is sh
own
i
n
Fi
gure
2
.
Inp
u
t netwo
rk d
ata and
r
u
n
po
wer
f
lo
w
Calcu
late po
wer
lo
ss
es an
d
vo
ltag
e stab
ility
Gen
erate
initial
po
p
u
latio
n
Evalu
ate t
h
e f
itn
ess
f
u
n
ctio
n
f
o
r
each ag
en
t
Co
m
p
u
te Pbes
t
an
d
Gbes
t f
o
r
th
e
p
o
p
u
latio
n
p
Calcu
late and
up
d
ate velo
city
an
d
po
s
itio
n
Usin
g
(
1
6
)
an
d
(
1
7
)
b
ase o
n
Pbes
t an
d
Gbes
t
Meeting
end
criter
io
n
?
Retu
rn b
est so
lu
tio
n
Start
Yes
N0
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,
Vol
.
13
, N
o.
2
,
Fe
bru
ary 2
019
:
4
3
7
–
4
4
6
442
np
L
DG
P
DG
Q
DG
L
CB
Q
CB
Figure
2.
Str
uc
ture of
p
a
rtic
le
1
2
3
4
5
n
p
n
p
n
p
n
p
n
p
n
p
(21)
np
represe
nts the
t
otal n
um
ber
of
par
am
et
ers
in a
g
ive
n pa
rtic
le
o
r
d
im
ension
of
par
ti
cl
e.
Wh
e
re:
L
DG
is l
ocati
on
of DG
P
DG
is real p
ow
er
of
D
G
Q
DG
is react
ive
pow
e
r of D
G
L
CB
is l
ocati
on
of CB
Q
CB
is react
ive
powe
r of
CB
3.
RESU
LT
S
A
ND AN
ALYSIS
The
pr
opos
e
d
al
gorithm
is
si
m
ula
te
d
us
in
g
the
IEEE
30,
33
a
nd
69
-
bus
rad
ia
l
net
w
ork
with
base
vo
lt
age
of
12.
66kV
an
d
the
base
ap
par
e
nt
powe
r
is
10
M
VA.
Fig
ur
e
3
and
Fi
gure
4
s
how
the
sin
gle
li
ne
diag
ram
o
f
33
-
bu
s
syst
em
an
d 6
9
-
bus syst
em
[
18
]
.
Figure
3.
Sin
gl
e li
ne
dia
gr
am
o
f
33
-
bus
distri
bu
ti
on test
syst
e
m
Figure
4.
Sin
gl
e li
ne
dia
gr
am
o
f
69
-
bus
distri
bu
ti
on test
syst
e
m
3.1.
B
as
e
Ca
s
e Sy
s
tem
A
n
al
ys
is
In
it
ia
ll
y,
load
flo
w
analy
sis
was
car
ried
out
on
30
-
bus
syst
e
m
fo
r
the
determ
inati
on
of
volt
ag
e
prof
il
e
an
d
po
wer
los
s.
Total
power
los
s
is
0.008
7
+
0.002
6i
(0.00
91)
in p
er u
nit.
F
or
33
-
bus
s
yst
e
m
,
t
he
total
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
All
oca
ti
on
of di
stri
bu
te
d ge
ne
ra
ti
on
and
c
ap
acitor
banks
in
d
ist
rib
ution sy
ste
m
(
Olade
po
Olatun
de
)
443
powe
r
loss
of
t
he
syst
e
m
is
0.
0027
+
0.
00
11
i
(0
.
0029)
in
pe
r
un
it
.
Th
e
bas
e
case
vo
lt
age
of
the
33
-
bus
s
yst
e
m
befor
e
the
inte
gr
at
io
n
of
DG
and
CB
into
t
he
netw
ork
is
pr
ese
nted
in
T
able
1.
The
vo
lt
age
m
agn
it
udes
in
Table
1
is
plo
tt
ed
a
gains
t t
heir
r
es
pecti
ve bus
num
ber
s is s
how
n
i
n
Fi
gure
5
.
Ta
bl
e
1
.
Base
C
ase
Volt
age for 33 Bus
Network
Bu
s n
o
Vo
ltag
e witho
u
t DG&
CB
1
0
.98
6
0
+
0
.0228
i
0
.98
6
3
2
0
.97
5
2
+
0
.0301
i
0
.97
5
7
3
0
.96
9
6
+
0
.0338
i
0
.97
0
2
4
0
.96
4
2
+
0
.0375
i
0
.96
4
9
5
0
.94
8
9
+
0
.0431
i
0
.94
9
9
6
0
.94
4
0
+
0
.0420
i
0
.94
4
9
7
0
.94
0
4
+
0
.0452
i
0
.94
1
5
8
0
.93
4
8
+
0
.0479
i
0
.93
6
0
9
0
.92
9
5
+
0
.0503
i
0
.93
0
9
10
0
.92
8
8
+
0
.0509
i
0
.93
0
2
11
0
.92
7
7
+
0
.0519
i
0
.92
9
2
12
0
.92
2
0
+
0
.0543
i
0
.92
3
6
13
0
.91
9
5
+
0
.0547
i
0
.92
1
1
14
0
.91
8
2
+
0
.0554
i
0
.91
9
9
15
0
.91
6
9
+
0
.0561
i
0
.91
8
6
16
0
.91
4
6
+
0
.0565
i
0
.91
6
3
17
0
.91
4
0
+
0
.0568
i
0
.91
5
8
18
0
.98
5
5
+
0
.0230
i
0
.98
5
8
19
0
.98
2
0
+
0
.0244
i
0
.98
2
3
20
0
.98
1
2
+
0
.0246
i
0
.98
1
5
21
0
.98
0
5
+
0
.0248
i
0
.98
0
8
22
0
.97
1
9
+
0
.0318
i
0
.97
2
4
23
0
.96
5
5
+
0
.0345
i
0
.96
6
1
24
0
.96
2
4
+
0
.0359
i
0
.96
3
1
25
0
.94
7
6
+
0
.0440
i
0
.94
8
6
26
0
.94
6
0
+
0
.0452
i
0
.94
7
1
27
0
.93
8
2
+
0
.0482
i
0
.93
9
4
28
0
.93
2
8
+
0
.0503
i
0
.93
4
2
29
0
.93
0
7
+
0
.0518
i
0
.93
2
1
30
0
.92
6
5
+
0
.0529
i
0
.92
8
0
31
0
.92
5
5
+
0
.0531
i
0
.92
7
0
32
0
.92
5
2
+
0
.0531
i
0
.92
6
7
F
igure
5
.
V
oltage
prof
il
e
for 3
3
-
bus
netw
ork
The
br
a
nc
h
pr
of
il
e
f
or
33
-
bus
syst
em
by
plo
tt
ing
the
c
urren
t
m
agn
it
ude
a
gainst
the
ir
res
pecti
ve
br
a
nc
h
is
sh
own
in
Fig
ure
6.
For
69
-
bus
syst
e
m
,
initial
load
flo
w
was
si
m
ulate
d
to
get
the
base
volt
age
at
each
bu
s
a
nd
t
he
base
t
otal
po
we
r
loss
of
th
e
syst
e
m
witho
ut
placem
ent
of
de
vices.
T
he
base
case
total
pow
e
r
loss
is
0.
02
38
+
0.
0107i
(0.
0261)
in
pe
r
un
it
.
T
he
base
case
vo
lt
age
m
agn
it
ud
e
s
is
plo
tt
ed
again
st
their
resp
ect
ive
bus
and that
of the
br
a
nc
h
c
urren
t
pro
file
is s
how
n
in
Fig
ure
7
a
nd Fig
ure
8.
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,
Vol
.
13
, N
o.
2
,
Fe
bru
ary 2
019
:
4
3
7
–
4
4
6
444
Figure
6
.
Cu
rr
e
nt m
agn
it
ud
e
for 3
3
-
bus
netw
ork
Figure
7.
V
oltage
prof
il
e
for 6
9
-
bus
netw
ork
Figure
8.
Cu
rr
e
nt m
agn
it
ud
e
for 6
9
-
bus
netw
ork
3.2.
Ef
fects o
f
D
G
and
CB
P
lacement
Applyi
ng
the
pro
po
se
d
m
et
ho
d
f
or
the
optim
u
m
DG
a
nd
CB
locat
ions
with
t
heir
c
orres
pondi
ng
siz
es,
the
c
ompu
te
d
r
es
ults
ar
e
pr
ese
nte
d
in
Table
2
.
T
he
e
ff
ect
of
DG
a
nd
CB
placem
e
nt
is
al
so
e
valu
at
ed
by
cal
culat
ing
the
powe
r
syst
em
lo
sses
re
duct
io
n
an
d
syst
e
m
vo
lt
age
im
pr
ov
e
m
ent.
Table
3
sh
ows
e
ff
ect
of
D
G
placem
ent
on
netw
ork
pe
rform
ance.
The
ba
se
case
of
33
-
bus
syst
e
m
and
the
i
m
pr
oved
vo
lt
age
m
agn
it
ud
e
is
sh
ow
n
in
Fi
gur
e 9
.
Ther
e
is
an
im
pro
vem
ent
in
v
oltage
pro
file
after
the
placem
ent
of
D
G
an
d
CB
on
bus
26
and
bu
s
21
resp
ect
ively
.
T
he
lo
west
volt
age
t
hat
is
bus
17
with
it
volt
age
i
ncr
eas
es
f
ro
m
0.915
8
to
0.948
1
pe
r
un
it
vo
lt
age
.
T
his
is
an
in
dicat
ion
that
the
DG
and
CB
instal
le
d
has
im
pr
ov
ed
the
vo
lt
age
s
at
each
bus
of
th
e
netw
ork.
T
he
blu
e
li
ne
point
ou
t
the
base
c
ase
vo
lt
age
tre
nd
wh
il
e
the
r
ed
li
ne
po
i
nt
out
the
vo
lt
age
prof
il
e
after
DG
an
d
CB
placem
ent.
The
69
bu
s
base
case
volt
age
m
agn
it
ude
,
an
d
t
he
c
orr
esp
onding
im
pr
ove
d
vo
lt
age
w
e
re
pl
otted aga
inst t
heir res
pecti
ve bu
s
es.
The
volt
age
prof
il
e
plo
t
is sho
wn in Fi
gure
10
.
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
All
oca
ti
on
of di
stri
bu
te
d ge
ne
ra
ti
on
and
c
ap
acitor
banks
in
d
ist
rib
ution sy
ste
m
(
Olade
po
Olatun
de
)
445
Table
2.
A
pp
li
cat
ion
of
Pro
pose
d
Al
gorith
m
for
Sim
ultaneous
D
G
a
nd C
B
Plac
em
ent
a
nd
Sizi
ng
Test
Sy
ste
m
Prop
o
sed
M
eth
o
d
DG
Po
sitio
n
DG Size
CB
Po
sitio
n
CB
Size
(kVar)
P(kW)
Q(kVar
)
30
-
Bu
s
12
3
7
.2
2
2
8
5
3
.9
17
3
8
0
8
.9
33
-
Bu
s
26
1
.04
7
3
5
6
.6
21
2
9
0
8
.3
69
-
Bu
s
36
8
9
1
.0
2
7
5
3
4
0
.8
42
2
0
9
2
2
9
.2
Figure
9.
V
oltage
prof
il
e
for 3
3
-
bus a
fter
place
m
ent
Figure
10.
V
oltage
prof
il
e
for 69
-
bu
s
after
p
l
ace
m
ent
Table
3.
Im
pact o
f
Pro
posed
Algorithm
for
Si
m
ultaneo
us
DG an
d
CB
Plac
e
m
ent
an
d
Si
zi
ng
Test
S
y
ste
m
Prop
o
sed
M
eth
o
d
Po
wer
lo
ss
W
ith
o
u
t DG
&
CB
Po
wer
lo
ss
W
ith
DG
&
CB
% Total
lo
ss
redu
ctio
n
% Vo
ltag
e
i
m
p
rov
e
m
en
t
Si
m
u
latio
n
Ti
m
e
Plo
ss
(kW)
Qlo
ss
(kVar)
Plo
ss
(kW)
Qlo
ss
(kVar)
30
-
Bu
s
874
256
402
84
5
5
.1%
35%
1
.48
1
6
s
33
-
Bu
s
266
108
102
57
60.
2%
42%
3
.49
0
9
s
69
-
Bu
s
2383
1074
1191
482
5
0
.9%
45%
1
1
.18
8
5
s
A
sig
nifica
nt
im
pr
ov
em
ent
is
obser
ve
d
in
F
igure
10
a
fter
DG
an
d
CB
place
m
ent.
Bus
53
has
th
e
sm
a
ll
est
vo
lt
ag
e
of
0.9
322
be
fore
D
G
an
d
CB
place
m
ent
and
ha
ve
incr
eased
to
0.9
65
after
th
e
place
m
ent
.
The
blu
e
li
ne
po
i
nt
out
the
base
case
volt
age
tre
nd
w
hil
e
the
r
ed
li
ne
sh
ows
the
tre
nd
of
t
he
vo
lt
a
ge
a
fter
placem
ent.
3.
3
.
C
ompari
so
n
of the
Method
w
ith
E
xisting Me
tho
d
and
Net
w
or
k w
itho
u
t DG
an
d C
B
A
com
par
at
iv
e
exam
inati
on
of
the
pro
po
sed
m
et
ho
d
is
furthe
r
m
ade
with
ot
her
optim
iz
at
ion
te
chn
iq
ues
i
n
pr
e
vious
wor
ks
for
est
ablis
hi
ng
it
s
ef
fici
ency.
Mo
re
s
pe
ci
fical
ly
,
sal
i
ent
featu
res
s
uch
a
s
conve
rg
e
nce
a
nd
te
ch
nical
im
pr
ov
em
ent
pr
edicat
ed
the
a
ppr
oach
a
dopt
ed
in
this
w
ork.
T
his
MPSO
ver
sio
n
was
giv
e
n
e
na
bling
par
am
et
e
rs
to
ac
hieve
op
ti
m
al
so
luti
on
by
e
xten
ding
it
erati
on
nu
m
ber
as
hi
gh
as
10
00
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,
Vol
.
13
, N
o.
2
,
Fe
bru
ary 2
019
:
4
3
7
–
4
4
6
446
with
po
pu
la
ti
on
of
200
par
ti
cl
es,
an
d
eq
ua
ll
y
achieve
a
bette
r
an
d
fast
conve
rg
e
nce
tim
e
of
1.481
6s
ec
,
3.490
9s
ec
an
d
11
.
1885s
for
30
-
bus,
33
-
bus
an
d
69
-
bu
s
syst
e
m
resp
ect
ively
co
m
par
e
to
[
19
]
.
PSO,
m
ul
ti
o
bj
ect
i
ve
g
e
netic
algorit
hm
(
MOG
A)
a
nd b
i
og
e
ogra
phy base
d o
pti
m
iz
at
ion
(
BB
O)
te
chn
iq
ue we
re
u
se
d
in
[20]
.
T
he
pr
opos
e
d
D
G
a
nd
CB
al
locat
ion
m
et
ho
d
is
al
so
hel
pful
f
or
r
edu
ct
io
n
of
network
powe
r
lo
ss
an
d
vo
lt
age
im
pr
ove
m
ent w
it
h hig
hest
value of
60.2% a
nd
45
%
resp
ect
ively
.
4.
CONCL
US
I
O
N
Op
ti
m
al
DG
and
CB
place
m
ent
pro
blem
has
bee
n
s
ol
ved
by
us
i
ng
A
dap
ti
ve
pa
r
ti
cl
e
swar
m
op
ti
m
iz
ati
on
(
AP
S
O)
te
c
hniqu
e
.
T
he
m
od
ific
at
ion
was
a
chieve
d
by
repl
aci
ng
the
i
nert
ia
l
weigh
t
(
w)
in
th
e
velocit
y
updat
e.
The
pro
pos
ed
locat
io
n
a
nd
siz
e
of
act
iv
e
powe
r
a
nd
r
eact
ive
powe
r
of
D
G
wit
h
re
act
ive
powe
r
of
CB
ha
d
been
com
pu
te
d
by
s
umm
i
ng
up
t
he
loa
ds
an
d
lo
sses
of
al
l
the
bran
ch
es
on
the
distri
bu
ti
on
netw
ork.
Th
rough
t
he
opti
m
al
al
locat
ion
of
D
G
a
nd
CB
by
the
pro
po
se
d
te
ch
nique,
th
e
act
ive
an
d
re
act
ive
powe
r
loss
es a
re r
e
duced
and
the syst
em
v
oltage
prof
il
e im
pr
ove
d.
REFERE
NCE
S
[1]
Mogali
gunta
Sa
nkar
aiah1.
Suresh
Redd
y
S.
Vijay
a
Kum
ar
M.
G
W
O
base
d
optim
al
reac
ti
v
e
po
wer
coor
din
at
io
n
o
f
DF
IG
,
ULTC
and
ca
pa
ci
tors
.
I
ndonesian
Journal
of
Elec
tri
cal
Engi
ne
ering
and
Computer
Sci
e
nce
.
2018
;
(11)3
:
805
–
813.
[2]
Moh
an
V.
Alber
T.
Optimal
sizing
and
sitt
ing
of
distribut
ed
gen
e
ration
using
Parti
cle
Swarm
Optimizati
on
Guide
d
Gene
tic
Al
gori
th
m.
Inte
rn
at
ion
al
Confer
ence
on
I
nte
lligent
Com
puti
ng,
Instrum
en
ta
ti
on
and
Contr
ol
Technol
ogi
es.
2017;
(10)5:
709
–
720.
[3]
La
kshm
i
P
rasa
nn.
,
Jain
A.
R
anjith
Kum
ar.
Opti
mal
Distribute
d
Gene
ration
Pl
a
c
eme
nt
Us
ing
Hybrid
Techni
que
.
As
ia
Paci
f
ic Pow
er
and
Ene
rg
y
Engi
ne
eri
ng
Con
fer
ence. 2017.
[4]
Cal
don
R
Copp
a
M.
Sgarboss
a
R.
Turr
i
R
.
A
simplif
ie
d
algori
thm
for
OL
TC
cont
rol
in
ac
ti
v
e
distr
ibution
M
V
net
works.
AE
IT Annual
Confer
e
nce
.
2013
;
1
–
6.
[5]
Olat
unde
O.
T
ola
A.
Pow
er
Quali
t
y
Im
prov
ement
in
E
le
c
tr
ic
a
l
Distribution
Network.
Ame
rican
Journal
o
f
Engi
ne
ering
R
ese
arch
,
2016
;
(5)
12:
224
–
227.
[6]
Mashhadi
O.
Sa
degh
M.
The
in
fl
uen
ce
of
sm
art
c
ontroll
er
of
th
e
distribut
ed
ge
nerati
on
sour
ces
and
capac
i
tor
banks
on
loss
minimizati
on
an
d
profil
e
improve
ment
in
th
e
el
e
ct
ric
it
y
distri
buti
on
ne
tworks
vol
tag
e,
”
20th
El
e
ct
ri
ca
l
Pow
er
Distribut
i
on
Co
nfe
ren
c
e
(
EPDC).
2015
;
272
–
27
6.
[7]
Nabil
Bin
Muhtazaru
ddin.
Dist
ribut
ion
Pow
er
Loss
Minim
iz
at
ion
via
Distrib
ute
d
Gene
ra
ti
on
,
Capa
c
it
or
an
d
Network
Rec
on
f
igura
ti
on
.
Indon
esian
Journal
of
El
e
ct
rica
l
Eng
in
ee
ring a
nd
Computer
Sc
ie
nc
e.
2
017;
(5)3:
488.
[8]
Ka
y
al
P.
As
hish
T.
Chand
a
C
.
Simult
an
eous
p
lac
eme
n
t
and
si
zing
of
ren
ewab
le
DG
s
and
ca
paci
tor
ban
ks
i
n
distribut
ion
n
et
w
ork.
Int
ern
a
ti
ona
l
Confer
ence
on
Circ
uit
s
,
Pow
er and
Com
pute
r
T
ec
hnolog
y
.
2014
;
607
–
611.
[9]
Ta
n
W
.
Hass
an
M.
Maji
d
M
.
Abdul
Rahman
H.
Optimal
distr
ib
ute
d
r
ene
wabl
e
gene
ra
ti
on
p
la
nn
ing:
A
rev
ie
w
of
diffe
ren
t approache
s.
Re
n
ewabl
e
and
Sustainab
le E
nergy
R
evei
ws.
2013;
(
18):
626
–
645.
[10]
Jega
dee
san
M.
Simult
aneous
Place
ment
of
Mul
ti
-
Dg
Us
ing
Hy
brid
Optimization
Me
thod
.
I
E
EE
In
te
rn
at
ion
a
l
Confer
ence
on
I
nnovat
ions
in
Gr
ee
n
Ene
rg
y
and
Hea
lt
h
ca
r
e
T
ec
h
nologi
es.
2017;
1
–
7.
[11]
Raouf
A.
A
mul
t
i
-
objective
distri
bute
d
gen
erati
on
all
ocation
and
s
izing
using
swar
m
int
el
l
ige
n
ce
b
ased
algorit
hms
.
19th
IE
EE Medi
te
rra
n
ea
n
Eletrotec
hni
ca
l
Confer
e
nce
.
2018;
281
–
286.
[12]
Maha
ja
n
S.
Va
dher
a
S.
Opt
imal
Siz
ing
and
D
epl
o
y
ing
of
Distribut
ed
Gen
era
t
ion
Unit
us
ing
a
Modifi
ed
Multi
objecti
v
e
P
art
i
cl
e
Sw
arm Opti
m
iz
ation,
IE
E
E
Tr
ansacti
on
o
n
Industrial App
li
cation.
2016;
(
52)3.
[13]
Singh
S.
A
Rev
i
ew
on
Part
ic
l
e
S
warm
Optimiza
tion
Algorit
hm
,
In
te
rnational
Jour
nal
of
Scientif
ic and
Engi
n
ee
ring
Re
search
.
2014;
(5)4.
[14]
Ojha
R.
Das
M.
An
Adapti
v
e
Approac
h
for
Modif
y
ing
Ine
rti
a
W
ei
gh
t
usi
ng
Parti
c
le
Sw
arm
Optimis
at
io
n.
Inte
rnational
Jo
urnal
of
Comput
er
Scienc
e. 2012
;
(
9)5:
105
–
112.
[15]
Am
an
M.
Jasm
o
n
B.
Sol
angi
K
.
Baka
r
A.
Mok
hl
is
H.
Optimum
Sim
ult
ane
ous
D
G
and
Cap
acitor
Placem
ent
on
th
e
Basis
of
Minim
iz
a
ti
on
of
Pow
er
Losses.
2013;
Inte
rnational
J
ournal
of
Comp
ute
r
and
El
ectri
cal
Engi
n
ee
ring
.
2013
,
(5)5:516
–
522.
[16]
A
hm
adi
A.
Khade
m
i
H.
Mous
avi
H.
Haj
iz
ad
eh
A.
Optimal
Allo
ca
t
i
on
of
DG
b
y
using
Im
prove
d
Gene
tic
for
I
E
E
E
33
Bus S
y
st
ems
.
Technical
Jour
n
al
of
Engi
n
ee
rin
g
and
App
li
ed
Sc
ie
nc
e.
2014;
29
–
35.
[17]
Mansour
H.
Ab
del
sal
am
A.
Na
bil
M.
Sall
am
A
.
Optimal
capacitor
banks
all
ocat
ion
in
distribut
ion
systems
wit
h
distribut
ed
g
en
er
ators
using
antl
ion
op
ti
miz
er
.
19th
Int
ern
at
ion
al
Midd
le
-
Ea
st
Pow
er
S
ystem
Confer
ence.
2017
;19
–
21.
[18]
Sam
ir
M.
Dawoud.
Xiangni
ng
Li
n.
Fir
as
M.
Qasim
Kam
il
Mohs
in.
H
y
br
id
Method
for
Optim
al
Pla
ce
m
ent
of
Multi
ple
SP
V
Based
Multi
p
le
RDG
s
in
Microgr
id
S
y
st
em.
In
donesian
Journ
al
of
El
e
ct
ric
al
Engi
ne
ering
a
nd
Computer
Scien
ce
.
2016
;
4)2
:
29
8
–
304.
[19]
Gkaida
t
zi
s
P.
Bouhouras
A.
Doukas
D.
Sgouras
K.
La
bridi
s
D
.
Appl
ic
a
ti
on
and
ev
alua
ti
on
of
U
PSO
to
OD
GP
in
radial
Distributi
on
Net
works
.
Int
ern
ational Conference on European
En
erg
y
Mark
.
2016.
[20]
Haz
em
N.
E
lsha
hed
M.
A.
Z.
O
sm
an
H.
Optimal
Pl
a
ce
men
t
of
Dispatchabl
e
an
d
Non
-
Dispatch
able
Distribut
ed
Gene
ration
o
f
D
if
fe
ren
t
Te
chnologies.
In
te
rn
at
io
nal
Midd
le
Ea
st
Pow
er
S
y
stems
Confer
ence. 201
7;19
–
21.
Evaluation Warning : The document was created with Spire.PDF for Python.