TELKOM
NIKA
, Vol.13, No
.3, Septembe
r 2015, pp. 7
59~766
ISSN: 1693-6
930,
accredited
A
by DIKTI, De
cree No: 58/DIK
T
I/Kep/2013
DOI
:
10.12928/TELKOMNIKA.v13i3.1790
759
Re
cei
v
ed Ma
rch 2
6
, 2015;
Re
vised Ma
y 26, 2015; Accepted June 1
1
, 2015
Reconfiguration of Distribution Network with
Distributed Energy Resources Integration Using PSO
Algorithm
Ramado
ni Sy
ahputra*
1
, I
m
am Robandi
2
, Mochamad Ash
a
ri
2
1
Departme
n
t of Electrical En
gi
neer
ing,
Un
iver
sitas Muhamm
adi
ya
h Yo
g
y
ak
arta,
Jl. Ringr
oad B
a
rat T
a
mantirto, Kasihan, Yo
g
y
ak
ar
ta, Indo
n
e
sia 5
5
1
83, Ph
one: +
62-2
74-
3
876
56
2
Departme
n
t of Electrical En
gi
neer
ing,
Institu
t
T
e
knolog
i Se
pul
uh No
pem
b
e
r,
Su
ko
li
l
o
, Su
rabay
a
,
In
d
o
n
e
s
ia
6
0
1
1
1
,
Ph
on
e
:
+6
2
-
3
1
-
5
9
4
7
302
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: ramado
ni@
u
m
y
.ac.id
1
, rob
a
ndi@
ee.its.ac.i
d
2
, ashari@
ee.
its.ac.id
3
A
b
st
r
a
ct
T
h
is
pa
per pr
esents opti
m
a
l
reconf
i
gur
atio
n of ra
di
al
dis
t
ributio
n n
e
tw
ork w
i
th integr
a
t
ion
of
distrib
u
ted e
n
e
r
gy reso
urces
(DER) usi
ng
mo
difi
ed
p
a
rtic
le sw
arm
opti
m
i
z
at
io
n (PSO) alg
o
rith
m. T
h
e
advantages of
integr
ation of
DER in
distribution syst
em
are
m
i
nim
i
z
i
ng
power
losses, improving
v
o
lt
age
profil
es a
nd l
o
ad factors, e
l
i
m
i
nati
ng syste
m
upgr
ad
es, a
nd re
duc
ing
en
viron
m
e
n
tal
i
m
pacts. How
e
ve
r, the
prese
n
ce of
D
E
R cou
l
d a
l
so
cause tec
h
n
i
ca
l pro
b
le
ms
in v
o
ltag
e q
ual
ity and
syste
m
pr
otection.
Opti
ma
l
reconfi
gurati
o
n
of distri
but
io
n
netw
o
rk is su
bj
ected to
mini
mi
z
e
pow
er
loss
and t
o
i
m
pr
ove
voltag
e pr
ofil
e
in
order to en
ha
n
c
e the efficie
n
cy the distribut
ion syste
m
. In this study, reconfig
uratio
n
method is b
a
se
d
on
an i
m
pr
ove
d
PSO. T
he meth
od has
be
en tested in
a 60-
bus Ba
ntul dis
t
ributio
n
netw
o
rk of Yogyakar
t
a
Speci
a
l
Re
gio
n
pr
ovinc
e
, In
don
esia.
T
he
simulati
on
res
u
lts sh
ow
that
opti
m
al
r
e
co
nfigur
ation
of
th
e
network with integration of DE
R has successf
ully
enhancing the
efficiency of
the distribution system
.
Ke
y
w
ords
:
distrib
u
tion
n
e
tw
ork, recon
f
igurati
on, effi
ciency,
mo
difi
ed p
a
rticle s
w
arm o
p
timi
zation,
distrib
u
ted e
n
e
r
gy resourc
e
s.
Copy
right
©
2015 Un
ive
r
sita
s Ah
mad
Dah
l
an
. All rig
h
t
s r
ese
rved
.
1. Introduc
tion
Powe
r dist
rib
u
tion networks provide the
end lin
ks
be
tween tran
smissi
on
syst
ems a
nd
cu
stome
r
s. The networks
a
r
e
ge
neral
ly
ope
rated
i
n
radial
structure
amon
g t
he fee
d
e
r
s.
The
feeders
are
fitted with a
numbe
r of
switch
es th
at are normally clo
s
ed,
n
a
me
ly
sectio
nali
z
i
ng
swit
che
s
, or
norm
a
lly ope
ned, nam
ely tie swit
che
s
.
The obj
ective
of the re
co
n
f
iguration i
s
t
o
minimize a
c
tive power lo
sse
s
an
d to i
m
prove vo
lta
ge profile in
orde
r to imp
r
ove distri
buti
on
s
y
s
t
em
performanc
e
[1-2]. In [3], the efforts
of
recon
f
iguration
of the di
stributio
n network ha
ve
become th
e
first p
ublication. Th
e effort has be
en
made to
obt
ain the
mini
mal a
c
tive p
o
we
r
losse
s
usi
n
g
the tradition
al techni
que.
Other
tra
d
itional techniq
u
e
has
been
prop
osed in
[4].
Most of tradit
i
onal techniq
ues d
o
not n
e
ce
ssarily
se
cure in glob
a
l
minima. In rece
nt decade
s,
the use
of artificial intell
igen
ce (AI) i
n
vari
ou
s fie
l
ds ha
s attracted the i
n
terest of ma
ny
resea
r
chers [
5
-7]. In terms of optimization of di
strib
u
t
ion network
config
uratio
n, the techniq
u
e
s
based on AI
have also be
come
so
meth
ing of interest
for many re
sea
r
che
r
s, a
s
can b
e
seen
in
[8-18]. The
u
s
e of g
eneti
c
algorith
m
(G
A) for net
wo
rk re
co
nfigu
r
a
t
ion method t
o
minimi
ze t
h
e
active po
wer
loss ha
s bee
n prop
osed in
[8]. In
[9] and [10], the method
s of sim
u
lated an
neal
ing
in large
scale
distributio
n system for acti
ve powe
r
loss re
ductio
n
p
u
rpo
s
e h
a
ve been p
r
e
s
ent
ed.
A methodolo
g
y based o
n
GA with the fundam
ental
l
oop for network reconfigu
r
ation ha
s be
en
pre
s
ente
d
by
Mendo
za
et al. [11]. Another vers
io
n of the GA for netwo
rk re
configuration has
been
pro
p
o
s
ed in [12]. T
hey have d
e
v
eloped
a G
A
method b
a
s
ed
on the
matroid
and
grap
h
theorie
s. In [
13], the a
ppl
ication
of a
n
t col
ony
o
p
timization
(A
CO) m
e
thod
for pl
aceme
n
t of
se
ctionali
z
ing
switche
s
i
n
distrib
u
tion
n
e
twor
ks
h
a
s been propo
sed.
Netwo
r
k reconfigu
r
atio
n
based o
n
a
si
mple b
r
an
ch
excha
nge te
chniqu
e of
sin
g
le loo
p
ha
s
been
propo
sed in [1
4]. In this
approa
ch, lo
ops sele
ction
seque
nce affects the opt
i
m
al config
uration to minimize the po
wer
loss. In [15],
the approa
ch
of harmony sea
r
ch
algo
rithm (HSA)
wa
s use
d
to reconfigure larg
e-
scale di
stri
bu
tion network
in or
der to redu
ce p
o
wer losse
s
. Th
e
app
roa
c
h i
s
con
c
e
p
tuali
z
ed
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No. 3, September 20
15 : 759 – 766
760
usin
g the m
u
si
cal process of har
m
o
n
y
searchin
g in perfe
ct st
ate. The use
of fuzzy multi-
obje
c
tive techniqu
e for op
timal netwo
rk reconfig
u
r
ati
on ha
s bee
n
pre
s
ente
d
in
[1, 2], and [16].
The te
chni
qu
e of p
a
rti
c
le
swarm
optimi
z
ation
(PSO
) for
network
reconfigu
r
atio
n pu
rpo
s
e
h
a
s
been
present
ed in [1
7-18]. In their wo
rk, crite
r
ia fo
r
selectin
g a
m
e
mbe
r
ship fu
nction
for
ea
ch
obje
c
tive are
not provide
d
, so re
du
ction
of losses h
a
s
yet to reach o
p
timal re
sults.
In re
cent ye
ars, th
e a
ppl
ication
of
re
newable
ene
rgy sou
r
ces
has be
com
e
popul
ar
becau
se of depletin
g su
pplie
s of fossil ene
rg
y a
nd enviro
n
m
ental issue
s
.
Indone
sia has
committed to
utilizing
rene
wabl
e ene
rgy
sou
r
ces to
g
enerate ele
c
tricity. Many potential are
a
s f
o
r
developm
ent
of DE
R, on
e
of whi
c
h
Bant
ul, Yogyakart
a
Spe
c
ial
Re
gion
provin
ce
. Gene
ration
of
electri
c
ity de
rived from
ren
e
wa
ble e
nerg
y
sou
r
ce
s i
s
calle
d di
strib
u
ted en
ergy reso
urce
s (DE
R
)
[19-20]. Th
e
advantag
es
o
f
DER inte
gration in
di
stri
bution
syste
m
are
re
du
ci
ng po
we
r lo
sse
s
,
improvin
g vo
ltage p
r
ofile
s and
load
factors,
eli
m
i
nating
syste
m
upg
ra
de
s and
re
du
ci
ng
environ
menta
l
impacts [2
1-22]. Integration of
DER in distri
bu
tion system
has b
e
co
me
an
intere
sting
ch
alleng
e
fo
r re
sea
r
che
r
s
to find the most appro
p
riate
method in the plannin
g
and
operation
of the di
stri
but
ion
system
[23-2
4
]. In th
is p
ape
r, an
modified
P
S
O algo
rith
m is
pre
s
ente
d
to solve di
stribu
tion netwo
rk reconfigu
r
atio
n pro
b
lem in
the pre
s
en
ce of distrib
u
te
d
energy resou
r
ce
s for re
du
cing
po
we
r l
o
ss a
n
d
im
proving voltag
e
profile
while
ra
dially of t
h
e
netwo
rk i
s
m
a
intaine
d
. In this stu
d
y, all obje
c
tive fun
c
tion
s are si
multaneo
usly
weig
hted, which
is a new i
s
sue in
a m
u
lti-objective
opti
m
ization
[1-2]. The PSO
a
ppro
a
ch i
s
te
sted i
n
a
60
-bus
Bantul distri
b
u
tion network of Yogyak
art
a
Special
Re
gion province
, Indonesi
a
.
2. Rese
arch
Metho
d
2.1. Problem Formulation
The pu
rpo
s
e
of optimizatio
n of distributi
on net
work is to minimize
power lo
sses and to
improve volta
ge p
r
ofile
whi
l
e ra
diality of
the network i
s
mai
n
taine
d
. The o
p
timiza
tion co
nst
r
ain
t
s
are loa
d
flow equation
s
, u
pper a
nd lo
wer limits
of b
u
s voltage
s, and up
per a
n
d
lowe
r limits o
f
line cu
rrents.
Formul
ation o
f
power lo
ss
minimization
can b
e
expre
s
sed a
s
follo
ws:
n
k
k
k
k
k
MIN
loss
v
Q
P
r
P
1
2
2
2
,
)
(
(1)
Subject to:
f(x
)
=
0
(2)
max
,
min
,
k
k
k
v
v
v
(3)
max
,
min
,
k
k
k
i
i
i
(4)
Whe
r
e
P
loss,MIN
is a cost fu
nction
of acti
ve power lo
ss;
n
is th
e nu
mber
of bra
n
c
h;
r
k
i
s
resi
st
ance
at bus
k-t
h
;
P
k
and
Q
k
are active an
d
rea
c
tive po
wers,
re
spe
c
tively;
v
k
is voltage at bu
s
k-
th
;
v
k,m
i
n
and
v
k,
max
are lower a
nd uppe
r voltage limits at
bus
k-
th
, res
p
ec
tively;
i
k
is
curre
n
t at bu
s
i-
th
;
and
i
k,min
a
nd
i
k,m
a
x
are lower a
nd up
pe
r cu
rre
nt limits at bus
k-t
h
, respec
tively
.
2.2. Modified
Particle S
w
arm Optimization
Eberh
a
rt and
Kennedy [25] has pu
bli
s
he
d
the pa
rticle swa
r
m
optimization
(PSO)
algorith
m
in 1
995. The
algo
rithm wa
s in
spired
by a
flock
of bird
s m
o
vement
in searchin
g of food.
The movem
e
nt model can
be used a
s
a
powerful opt
i
m
ize
r
. In one
n-dim
e
n
s
iona
l sea
r
ch sp
ace,
let us a
s
sum
e
that the po
sition of the i
-
th individu
al is X
i
= (x
i1
, ...
, x
id
, ...
,
x
in
) a
nd the spee
d
of
the i-th individual is
S
i
= (
s
i1
, ...
, s
id
, ...
,
s
in
). The particle best exp
e
rien
ce i-th i
s
re
co
rded a
n
d
r
e
pr
es
e
n
t
ed
b
y
Pbest
i
=
(pbe
sti
1
, ..., p
bes
t
id
, ..., pbest
in
). Th
e b
e
st glo
bal
po
sition fo
r swarm
sea
r
c
h
is
Gbe
s
t
i
= (gbe
st
1
, ..., gbest
d
, ..., gbest
n
). The modifie
d
velocity of each parti
cle
is
cal
c
ulate
d
ba
sed
on th
e p
e
rsonal
initial
velocity,
the
distan
ce
fro
m
the p
e
rson
al be
st po
siti
on,
and the di
sta
n
ce fro
m
the global b
e
st p
o
sition (Figu
r
e 1), as
sho
w
n in the following equ
ation:
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Re
config
urati
on of Dist
ribu
tion Network with
Distribute
d
Energy
… (Ram
adoni S
y
ahputra)
761
)
(
)
(
)
(
)
(
)
(
2
2
)
(
1
1
)
(
)
1
(
t
i
i
t
i
i
t
i
t
i
X
Gbest
rand
c
X
Pbest
rand
c
S
S
(5)
Equation
(5
) dete
r
mine
s
the velo
city vector of
the
i-th
parti
cle.
The
r
efo
r
e, t
he late
st
positio
n of the particl
e ca
n
be determi
ne
d by using th
e equatio
n:
)
1
(
)
(
)
1
(
t
i
t
i
t
i
S
X
X
(
6
)
Whe
r
e
i
=
1,
2, ..
.,
N
is the index of each p
a
rticl
e
;
t
is the numb
e
r of iteration
s
;
rand
1
(
) a
n
d
rand
2
(
) are a
rando
m num
ber bet
wee
n
0 and 1; and
N
is the num
ber of the swarm.
Inertia wei
ght
s
ω
can b
e
d
e
termin
ed by the equatio
n:
t
t
t
max
min
max
max
)
1
(
(
7
)
Whe
r
e
ω
ma
x
is the maxim
u
m inertia
weight;
ω
mi
n
is
the minimu
m inertia
wei
ght;
t
ma
x
is the
maximum nu
mber of itera
t
ions; and t is the act
ual
numbe
r of iteration
s
.
The
value of ine
r
tia
weig
ht decre
ase lin
early from 0.9 to 0.4.
Figure 1. The
optimization
con
c
e
p
t usin
g PSO
The modifie
d
PSO algorith
m
is de
scribe
d as follo
ws:
1.
Input the data
of dist
ributio
n netwo
rk a
n
d
initialize the
param
eters of PSO.
2.
Run the p
r
og
ram of power flow to measur
e the fitness (active po
wer lo
ss) of ea
ch
particl
e (p
be
st) and sto
r
e it with t
he be
st value of fitness
(gbe
st).
3.
Upd
a
te veloci
ty of particle usin
g (5
).
4.
Upd
a
te po
sition of particl
e usin
g (6
).
5.
De
cre
a
se the
inertia wei
g
h
t
(
ω
) linearly from 0.9 to 0.4.
6.
Perform viola
t
ion of particl
e positio
n:
If particle po
sition pos(j)>m
p
, then pos(j)=mp
Else if particl
e positio
n po
s(j
)
<m
p, then
pos(j)=1.
7.
Perform viola
t
ion of particl
e velocity:
If particle velocity vel(j)>m
v, then vel(j)=mv
Else if particl
e velocity vel(j)<-mv, then p
o
s(j
)
= -mv.
8.
De
cre
a
se the
inertia wei
g
h
t
(
ω
) linearly from 0.9 to 0.4.
9.
Rep
eat step
s 2-8 until a cri
t
eria is obtai
n
ed.
Modificatio
n
s of PSO al
go
rithm in
this
study
a
r
e
in t
he
sixth an
d
seventh
ste
p
s
a
bove,
while i
n
the o
r
iginal PSO
a
l
gorithm
doe
s not exist
[25]
. The ste
p
s
a
r
e u
s
eful to
a
v
oid violation
s
of both spe
e
d
and po
sition
of particle
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No. 3, September 20
15 : 759 – 766
762
3. Results a
nd Discu
ssi
on
In the
se
ction
,
the imp
r
ove
d
PSO al
gorit
hm
is teste
d
on a
practi
cal
60-bu
s Bant
ul po
we
r
distrib
u
tion
system. Bantul
po
wer di
strib
u
tion
system
is lo
cate
d in
distri
ct of Ba
ntul, Yogyaka
r
ta
Special
Re
gi
on. Yogyaka
r
ta Spe
c
ial
Regi
on is
on
e of the pro
v
ince
s in Ind
one
sia
which
is
located in
Java islands. T
he results of
optimal
re
co
nfiguratio
n of
60-bu
s 20
-kV Bantul ra
di
al
distrib
u
tion n
e
twork
with
DER inte
gration u
s
ing
th
e
prop
osed m
e
thod to mini
mize a
c
tive p
o
we
r
losse
s
and to
improve the
voltage qualit
y of the sy
stem are p
r
esen
ted. The syst
em con
s
i
s
ts
of
13 feed
ers th
at are
po
we
red by two 6
0
MVA po
we
r
transfo
rme
r
s, but this stu
d
y
has fo
cu
se
d
only on feede
rs of 6, 7, an
d 11. Selecti
on of t
he thre
e feeders is
becau
se the
most comple
x in
terms
of network length, n
u
mbe
r
of loa
d
s, an
d
pote
n
tial integration of
DE
R. T
he sy
stem h
a
s 6
0
buses and 5
7
se
ction
s
, as sho
w
n
i
n
Figure
2.
T
he switch of
the syste
m
con
s
i
s
ts of
57
se
ctionali
z
ing
switch
es an
d 5 tie
switch
es. Se
ctio
n
a
lizing
switche
s
of the
sy
ste
m
are
clo
s
ed
in
norm
a
l condit
i
ons
while tie
swit
che
s
a
r
e
open i
n
no
rm
al co
ndition
s.
Load
and
branch data
of the
60-b
u
s di
strib
u
tion n
e
two
r
k ca
n
be fo
und
in [1
3]. The
five tie
swit
che
s
a
r
e
58,
59,
60, 61
an
d
6
2
.
The total lo
a
d
of the radi
al syste
m
is
2654
7 kW
a
nd the initial
power l
o
ss
of the sy
ste
m
is
656.20
kW. T
he ba
se of the system i
s
V=20
kV and S
=
60 MVA.
Figure 2. 60-bus Yogya
k
a
r
ta radial di
stri
bution net
work in initial co
n
f
iguration
T
abl
e 1. DER Locatio
n and
Capa
city of 60-Bu
s
Y
ogya
k
arta Ra
dial Di
stributio
n System
Bus
Number
DER Active Power
(k
W
)
DER Po
w
e
r
Fac
t
or
DER Reactive
Power (k
VAr
)
8 250
0.8
187.50
13 250
0.9
121.08
20 300
1
0
32 400
0.9
193.73
36 300
1
0
47 250
0.9
121.08
59 300
0.8
225
The Initial co
nfiguratio
n of
the netwo
rk wi
thout DE
R integ
r
ation
is sh
own in
Figure
2
while after reco
nfiguratio
n is sho
w
n
in Figure 3.
In order to
analyze the
impact of DER
integratio
n to distrib
u
tion n
e
twork, we h
a
ve insta
lle
d as ma
ny as seven DERs o
n
buses of 8,
13,
20, 32, 3
6
, 4
7
, and
59, re
spe
c
tively, as sho
w
n
in
Ta
ble 1. Th
e DER mod
e
ls t
hat have u
s
e
d
i
n
our
study co
nsi
s
t of both
sola
r ph
otovoltaics an
d wind farm
s. S
e
lectio
n of th
e DER type
s is
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Re
config
urati
on of Dist
ribu
tion Network with
Distribute
d
Energy
… (Ram
adoni S
y
ahputra)
763
becau
se mo
st potential in
the ar
ea of Y
ogyakarta.
We have a
s
su
m
ed that p
o
w
er facto
r
of
all
DER
sol
a
r
ph
otovoltaics a
r
e unity, whil
e
wind
farm
s
are
ran
g
ing
from 0.8 to
0.9
(lag
ging
). Th
e
PSO pa
rame
ters that h
a
ve be
en
used
to 6
0
-b
us Y
ogyakarta
ra
dial di
strib
u
tion
system
a
r
e
con
s
i
s
ts of p
opulatio
n si
ze of 25 an
d maximum iteration of 200.
The minimu
m and maxi
mum
voltages a
r
e
set at 0.90 a
nd 1.00 p.u., respe
c
tively. The re
sult
s o
f
the case stu
d
y are sho
w
n
in
Figure 3, Fi
gure
4, Figu
re 5, a
nd T
able 2.
Net
w
ork re
confi
guratio
n
u
s
in
g
improved PSO
algorith
m
ha
s re
sulted that
there a
r
e fo
u
r
tie swit
ch
es that must b
e
clo
s
ed, i.e., swit
che
s
of 5
7
,
58, 59,
and
6
0
, whil
e the
section
a
lizin
g
swit
che
s
to
b
e
op
ene
d a
r
e
switch
es of
8, 9, 27,
and
43,
as sho
w
n in
Table 2.
Figure 3. 60-bus Bantul
ra
dial distri
butio
n netwo
rk aft
e
r re
co
nfiguration
Figure 4. Power lo
ss disp
ersi
on of 60
-b
us
Bantul radi
al distrib
u
tion
test system
Figure 4 sh
o
w
s p
o
wer lo
ss dispersio
n
before
re
conf
iguratio
n, after installi
ng
DER, and
after reconfig
uration
for 60
-bu
s
Ba
ntul
radial
dist
ri
buti
on te
st
syste
m
. It can
b
e
o
b
se
rved
that t
he
magnitud
e
of the powe
r
loss of each bu
s depend
s on the length of line betwe
en the bus a
nd the
size of each load bu
s. It is sho
w
n that the long
er
the
line, the gre
a
ter the po
wer lo
ss. Simil
a
rly,
from Fig
u
re
4
,
it is also
sh
own th
at the
greate
r
the
lo
ad that i
s
served by a b
u
s,
the greate
r
the
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
9
30
TELKOM
NIKA
Vol. 13, No. 3, September 20
15 : 759 – 766
764
power lo
ss. It can b
e
se
en
that the pre
s
ence of DE
R
as ma
ny as
seven units
on
buse
s
of 8, 1
3
,
20, 32, 36, 4
7
, and 59 ha
s the effect
s on the po
we
r loss
red
u
ctio
n of the syste
m
, espe
cially
on
buses clo
s
e
s
t
to
the DER. Before re
conf
iguratio
n the
netwo
rk a
s
a
base case, to
tal active p
o
wer
loss und
er st
udy is 656.2
0
kW. Total a
c
tive power
lo
ss afte
r instal
ling as ma
ny as five DERs is
474.86
kW,
while total
activ
e
po
we
r l
o
ss
after rec
onfig
uration
of
net
work with
DE
R inte
gratio
n
is
294.71
kW, as shown in T
able 2. From
the Table
can
also b
e
seen
that integrati
on of five DERs
has re
sulte
d
i
n
re
du
ction o
f
powe
r
lo
ss.
Percent
ag
e o
f
powe
r
lo
ss
redu
ction
afte
r in
stalling th
e
DERs i
s
27.
63%, whil
e
percenta
ge
o
f
powe
r
lo
ss
after reconfi
guratio
n of n
e
twork
with
DG
integratio
n is 55.09%. The
s
e re
sult
s ha
ve proved
tha
t
the reconfig
uration of the
network hav
e a
con
s
id
era
b
le
influen
ce on t
he re
du
ction
of active
po
wer lo
ss in
dist
ribution
syste
m
. Redu
ction
of
power loss is
certai
nly imp
r
oving the
efficien
cy of
th
e
distrib
u
tion
n
e
twork. T
able
2 al
so
repo
rt
ed
that the effici
ency
of the
distrib
u
tion
n
e
twork of
6
0
-
bu
s Ba
ntul radial
system
in the
o
r
igin
al
con
d
ition i
s
97.53%. Th
e
efficien
cy h
a
s
in
crea
sed
to 98.2
1% after inte
gration
of a
s
ma
ny
as
seven
DERs in the sy
ste
m
. Afte
r integration
of the DERs, opti
mization i
s
carri
ed o
u
t on
the
netwo
rk configuratio
n. The
re
su
lt sho
w
ed that an i
n
cre
a
si
ng in
e
fficiency b
e
a 98.89%
after
reconfigu
r
atio
n is achieved.
For voltag
e p
r
ofile of the
n
e
twork, it is i
n
tere
sting to
find that with
integratio
n of
DER i
n
60-b
u
s Bantu
l
radi
al di
strib
u
tion net
wo
rk, voltage
qu
al
ity of each
b
u
s i
s
imp
r
ove
d
, as
sh
own i
n
Figure 5. The voltage qu
ality is to be
improve
d
further by doin
g
reco
nfig
uration of distrib
u
t
ion
netwo
rk tha
n
ever befo
r
e. It should b
e
n
o
ted in
the re
sults that onl
y
a voltage magnitud
e
alon
g
the main
fee
der
of bu
s i
s
pre
s
e
n
ted.
Before
re
conf
iguratio
n the
netwo
r
k a
s
a
ba
se
ca
se, i
t
is
resulted th
at the hig
hest
voltage ma
g
n
itude i
s
1.
0
0
p.u. on
bu
s 1,
while th
e lowest volt
age
magnitud
e
is 0.910 p.u. on bus 60, a
s
sh
own in Figu
re
5 and Table
2. In Figure 5
,
it can be se
en
that on the
o
r
iginal
conditi
on of the
net
work, the
fart
her
away fro
m
the
sub
s
ta
tion lo
cation,
the
lowe
r the
a
m
plitude
of t
he b
u
s’
s volt
age. Integ
r
ati
on of
DE
R
has resulted
in in
crea
sin
g
of
voltage mag
n
i
tude. After in
tegration
of
DER i
n
60
-b
us Ba
ntul di
stribution
network, th
e hig
h
e
st
voltage magn
itude is 1.00 p.u. on bus 1,
while the
lowest voltage magnitud
e is 0.934 p.u. on bu
s
60, as sho
w
n
in Figure 5 a
nd Table 2. It can be o
b
se
rved from Fig
u
re 5 that integratio
n of DER
as m
any a
s
seven unit
s
on
buses
of 8,
13, 20,
32, 3
6
, 47, and
59
has th
e st
ro
ng effect
s on
the
voltage profil
e improvem
e
n
t, espe
cially
on buses
cl
o
s
e
s
t to the DER. The voltage imp
r
ove
m
ent
is o
c
curred
al
most the
enti
r
e b
u
s, ex
ce
pt for
bu
s
1,
becau
se the
magnitud
e
of
the voltage
has
reached its m
a
ximum limit.
Figure 5. Voltage profile of 60-b
u
s Ba
nt
u
l
radial di
strib
u
tion test syst
em
Furthe
rmo
r
e,
optimization
of network
conf
ig
uratio
n
usi
ng im
pro
v
ed PSO alg
o
rithm
on
60-b
u
s Bant
ul net
work
with
DER i
n
tegration
ha
s bee
n de
mo
nstrate
d
. Th
e re
sult
s of
the
optimizatio
n can al
so be
see
n
in Figu
re 5 and
Ta
ble 2. Here, it can be se
en that network
reconfigu
r
atio
n usin
g improved PSO has the st
rong
impact of b
u
s’
s voltage
magnitud
e
. After
reconfigu
r
atio
n, the hi
ghe
st voltage
m
agnitud
e
i
s
kept 1.00
p.u.
on
bu
s 1,
while
the l
o
we
st
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Re
config
urati
on of Dist
ribu
tion Network with
Distribute
d
Energy
… (Ram
adoni S
y
ahputra)
765
voltage ma
gn
itude is 0.95
4
p.u. on b
u
s
60. Thi
s
volta
ge ma
gnitud
e
is b
e
tter tha
n
the mag
n
itude
of the voltage before recon
f
iguring
the n
e
twork. The
s
e re
sults p
r
ov
e that the distribution n
e
twork
reconfigu
r
atio
n with
DER
integratio
n u
s
ing im
prove
d
PSO meth
od ha
s b
een
su
ccessful i
n
improvin
g the
performan
ce
of 60-bu
s
Ba
ntul radial di
stribution sy
stem.
T
abl
e 2.
The
Simulation Result
s of 60-B
u
s Bantul Ra
dial Di
stributi
on Net
w
ork
T
e
st Case of
Distribution Net
w
ork
Parameters of A
nal
y
s
is
Active Pow
e
r
Loss (
k
W)
Percentage
of Loss
Reduction
(%)
Efficiency
of
Distribution
N
e
tw
o
r
k (%
)
Minimum
Voltage
(p.u.)
Max
i
mum
Voltage
(p.u.)
Tie Sw
itches
to be Closed
Sectionalizing
Sw
itches to be
Open
Without DER
integration befo
r
e
reconfiguration
656.20
-
97.53
0.910
(V
60
)
1.00
(V
1
)
NA NA
With DER
integration befo
r
e
reconfiguration
474.86
27.63
98.21
0.934
(V
60
)
1.00
(V
1
)
NA NA
With DER
integration after
reconfiguration
294.71
55.09
98.89
0.954
(V
60
)
1.00
(V
1
)
57
58
59
60
8
9
27
43
4. Conclusio
n
A study of optimal re
confi
guratio
n of radial
dist
ribut
ion netwo
rk with the integration of
DER
usi
ng
modified PS
O algo
rithm
is p
r
e
s
ented
in this
pap
er. The
stu
d
y
was ba
se
d on
minimizi
ng a
c
tive power l
o
sse
s
and i
m
provin
g voltage qu
ality in ord
e
r to en
han
ce di
strib
u
tion
system p
e
rfo
r
man
c
e. Th
e
methodol
og
y was te
st
ed
on a p
r
a
c
tical 60-bu
s Ba
ntul distri
buti
on
system of Yo
gyaka
r
ta Spe
c
ial
Regio
n
p
r
ovince
, Indo
nesi
a
. Base
d
on the n
u
m
e
rical results,
it
wa
s sho
w
n t
hat the alg
o
rithm is effect
ive in
enha
n
c
ing
efficien
cy of the two
test dist
ributi
o
n
system
s. Fo
r
a 60
-bu
s
Ban
t
ul dist
ri
butio
n syste
m
, the
efficien
cie
s
i
n
the o
r
igin
al
con
d
ition, aft
e
r
integratio
n of
seven
DE
Rs,
and
after
net
work
re
co
nfig
uration
a
r
e
9
7
.53%, 98.21
%, and 9
8
.89
%
,
respe
c
tively.
Also, integrati
on of DER h
a
s re
sulte
d
in improved vol
t
age profile i
n
the test rad
i
al
netwo
rks. After optimal re
config
uratio
n
of the network, the volta
ge profile i
s
to be improv
ed
further.
Ackn
o
w
l
e
dg
ements
The autho
rs gratefully ackno
w
led
ge the co
nt
ributi
ons of the Dire
cto
r
ate
Gene
ral of
High
er Edu
c
a
t
ion (DIKTI),
Minist
ry of Rese
arch, Te
chnolo
g
y and
High
er Edu
c
a
t
ion, Repu
blic of
Indone
sia, for funding this rese
arch.
Referen
ces
[1]
Sy
ahputra R, Robandi I, Ashari M
.
Optimization
of Distrib
ution N
e
t
w
or
k Confi
gurati
on w
i
t
h
Integr
atio
n
of
Distrib
uted Energ
y
Res
our
ces
Usi
ng
E
x
te
nde
d F
u
zz
y M
u
lti-o
b
jectiv
e M
e
thod,
Inter
nati
ona
l Rev
i
ew
of Electrical En
gin
eeri
n
g
. 20
1
4
; 9(3): 629-6
3
9
.
[2]
Sy
ahputra R,
Robandi I, Ashari M.
Optimal
Distributi
on
N
e
tw
ork Rec
onfi
gurati
on
w
i
th Penetrati
on of
Distribut
ed En
ergy Res
ource
s.
1st Internationa
l Conf
eren
ce on Informat
i
on T
e
chno
log
y
, Com
put
e
r
and El
ectrical
Engi
neer
in
g (ICIT
A
CEE 2014). Semaran
g
. 201
4: 386-
391.
[3]
Merlin A, Back
H.
Search for
a mini
mal-l
o
ss oper
ating s
p
a
n
n
in
g tree co
nfi
gurati
on i
n
an
urba
n pow
er
distrib
u
tion sys
tem.
5th PSCC
Confere
n
ce. C
a
mbrid
ge UK.
197
5: 1-18.
[4]
Lin
W
M
, Chi
n
HC. A
Ne
w
Ap
proac
h for
Dist
r
ibuti
on F
e
ed
e
r
Rec
onfig
urati
on for
L
o
ss R
e
ductio
n
an
d
Service R
e
stor
ation
.
IEEE Trans on PWRD
. 199
8; 13(3): 87
0-87
5.
[5]
Afrisal H, F
a
ris M, Utomo P,
Soesanti I, Andri F.
Porta
b
le
smart s
o
rtin
g a
nd
grad
in
g
ma
chin
e for fru
i
ts
usin
g co
mpute
r
vision.
Proce
edi
ng - 201
3 Internatio
nal C
o
n
f
er
ence o
n
Co
mputer, Contro
l, Informatic
s
and Its Applic
a
t
ions: "Rec
ent Chal
le
nges i
n
Comp
uter, Con
t
rol and Inform
atics", IC3INA. 201
3.
[6]
Nugr
oho HA, F
a
isal N, So
esa
n
ti I, Chorida
h
L.
Analysis of d
i
gital
ma
mmog
r
ams for det
ection of bre
a
s
t
cancer
. Proc
e
edi
ng - 20
14
Internatio
nal
Confer
enc
e
o
n
Comp
uter, Contro
l, Informatics and It
s
Appl
icatio
ns: "Ne
w
Ch
al
len
g
e
s
and Opp
o
rtu
n
ities i
n
Big Da
ta", IC3INA. 2014.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No. 3, September 20
15 : 759 – 766
766
[7]
Rohm
ah RN,
Susanto A, So
esanti I.
Lun
g tubercu
losis i
d
entificati
on
b
a
s
ed on statisti
cal feature of
thoracic X-ray
.
201
3 Intern
atio
nal C
onfer
enc
e on Qu
alit
y i
n
Researc
h
, QiR 201
3 - In Co
nju
n
ction
w
i
t
h
ICCS 201
3: T
h
e 2nd Inter
nati
ona
l
Conf
erenc
e on Civ
ic Spa
c
e. 2013.
[8]
Nara K, Sh
ios
e
A, Kitaga
w
a
M. Implement
ati
on
of Genet
ic Algor
ithm fo
r
Distributi
on S
y
stem
L
o
ss
Minimum R
e
co
nfigur
ation
.
IE
EE T
r
ans on PW
RD
. 1992; 7(
3): 1044-
10
51.
[9]
Augu
gli
a
ro A,
Duso
nchet L,
Ippol
ito M. Minimum L
o
sses
Reco
nf
igur
ati
on of Distri
but
ion N
e
t
w
or
k
throug
h Co
ntro
l of
T
i
e-S
w
itc
h
es.
IEEE Trans on PWRD
. 20
03; 18(3): 7
62-
771.
[10]
Jeon YJ, Kim JC, Kim JO, Shin JR. An Efficient
Simu
late
d Anne
ali
ng fo
r Net
w
o
r
k Rec
onfig
uratio
n i
n
Larg
e
-Scal
e
Di
stributio
n
.
IEEE Trans on PWRD.
200
2; 17(4
)
: 1070-1
0
7
8
.
[11]
Mend
oza J, Lopez R, Moral
e
s
D, Lopez
E. Minimal los
s
reconfig
urati
on usi
ng GA w
i
t
h
restricte
d
pop
ulati
on a
n
d
address
ed o
p
e
rators.
IEEE Trans on PS
. 2006; 21(
2): 948
-954.
[12]
Enach
e
a
nu B,
Raiso
n
B, Ca
ir
e R, Dev
a
u
x
O
.
Radi
al N
e
t
w
o
r
k Reco
nfigur
ation Us
in
g Gen
e
tic Alg
o
rithm
Based o
n
the
Matroid T
heor
y.
IEEE
Trans on PS
. 2008; 23
(1): 186-1
95.
[13]
F
a
lag
h
i H,
Hag
h
ifam MR. AC
O-Based Met
h
od for
P
l
acem
e
n
t of Sectio
nal
i
z
ing S
w
i
t
ches
i
n
Distri
butio
n
Net
w
orks Usi
n
g a F
u
zz
y
Apr
o
ach
.
IEEE Trans on PWRD
. 2009; 24(
1): 268
-276.
[14]
F
a
raha
ni V, Vahi
di B,
Ab
ya
neh HA. R
e
co
nfigur
ation
an
d Cap
a
citor Pl
aceme
n
t Simu
ltane
ousl
y
f
o
r
Energ
y
L
o
ss R
educti
on.
IEEE Trans on PS
. 201
2: 27(2
2
): 587-5
95.
[15]
Rao
RS, Nara
simham SV
L, Raju
MR. Opti
mal
N
e
t
w
ork Reco
nfigur
atio
n
of
L
a
rge-Sc
ale
Distrib
utio
n
S
y
stem Usi
ng
Harmon
y
S
ear
ch Algor
ithm
. IEEE Trans on PS
. 2011; 26(
3
)
.
[16]
S
y
a
h
p
u
tra R,
Rob
and
i I, Ash
a
ri M.
Rec
onfi
g
uratio
n of D
i
stri
butio
n N
e
tw
ork w
i
th DG Usin
g
Fu
zz
y
M
u
lti-
obj
ective M
e
thod.
201
2 Int
e
rnati
ona
l C
o
nferenc
e o
n
I
nnov
atio
n, Ma
nag
ement
an
d
T
e
chnolo
g
y
Rese
arch (ICIMT
R 2012). Melacc
a, Mala
ys
ia. 201
2: 316-
3
21.
[17]
S
y
a
h
p
u
tra
R,
Rob
and
i I, As
hari
M. PSO
Based
Multi-
ob
jective
Optimiz
a
tion
for
Reco
nfigur
ation
o
f
Radi
al D
i
strib
u
t
ion Net
w
o
r
k.
Internatio
na
l Journ
a
l of Ap
pl
ied E
ngi
ne
erin
g Rese
arch.
2
015; 1
0
(6)
:
145
73-1
4
5
86.
[18]
Niknam
T
,
Me
ymand
HZ
, Moj
a
rrad
HD. A
pr
actical
multi-
ob
jective
PSO al
gorithm
for o
p
ti
mal o
per
atio
n
mana
geme
n
t o
f
distributi
on n
e
t
w
o
r
k
w
i
t
h
re
gard to fu
el c
e
ll po
w
e
r
pla
n
ts.
Renew
ab
le E
nergy.
20
11
;
36: 152
9-1
544.
[19]
S
y
a
h
p
u
tra R,
Rob
and
i I, Ashari
M. Perform
ance A
nal
ys
is
of W
i
nd T
u
rbin
e as a D
i
strib
u
ted Gen
e
rati
on
Unit in
Distrib
ution S
y
stem
.
Internatio
na
l Journ
a
l of C
o
mputer
Sci
ence
& Informati
on
T
e
chno
logy
.
201
4; 6(3): 39-
56.
[20]
Ashari M, An
a
m
S, Surojo. A
W
i
de R
a
n
ge F
u
zz
y
Based M
a
ximum P
o
w
e
r
Poin
t T
r
acker
for Improving
the Efficienc
y a
nd Sizi
ng of P
V
S
y
stems.
Jur
nal of Electric
al
Engin
eeri
ng.
2
011; 11(
2): 52-
56.
[21]
Ashari M, Na
yar CV. Optimum Operation
Stra
teg
y
an
d Econom
ic Ana
l
y
s
is of a PV-Batter
y
-ma
in
s
H
y
brid U
n
i
n
terr
uptib
le Po
w
e
r Supp
l
y
.
Jo
urn
a
l
of Renew
a
b
le
Energy
. 20
01; 22(1): 24
7-2
5
4
.
[22]
Ashari M,
Na
yar CV. A Gri
d
-interactiv
e
Ph
otov
olta
ic U
n
in
terruptib
le P
o
w
e
r Su
pp
l
y
S
y
stem
Usin
g
Batter
y
Stor
ag
e an
d a B
a
ck
up Di
ese
l
Gen
e
rator.
IEEE Transactions on Energy
Conv
ersion
. 200
0;
15(3): 34
8-3
5
3
.
[23]
Rob
and
i I. Desain sistem te
nag
a moder
n: optimis
asi, log
i
ka fuzz
y
,
alg
o
r
i
tma gen
etika. Yog
y
ak
arta:
Andi Pu
blis
her.
2006.
[24]
Jamal A, S
y
a
hputra
R. UP
F
C
Base
d o
n
Ad
a
p
tive
Ne
uro-F
u
zz
y for
Po
w
e
r
F
l
o
w
Contro
l
of
Multimachine Po
w
e
r
S
y
stem
s.
Internatio
nal
Journ
a
l of E
n
gin
eeri
ng Sc
ie
nce Inve
ntio
n.
201
3; 2(1
0
):
05-1
4
.
[25]
Kenn
ed
y J, Eberhart RC.
Par
t
icle sw
arm opt
imi
z
a
t
io
n.
Proc IEEE Int Conf Neural N
e
t
w
or
ks. 1995; 4:
194
2-19
48.
Evaluation Warning : The document was created with Spire.PDF for Python.