Internati
o
nal
Journal of Ele
c
trical
and Computer
Engineering
(IJE
CE)
Vol.
3, No. 6, Decem
ber
2013, pp. 814~
822
I
S
SN
: 208
8-8
7
0
8
8
14
Jo
urn
a
l
h
o
me
pa
ge
: h
ttp
://iaesjo
u
r
na
l.com/
o
n
lin
e/ind
e
x.ph
p
/
IJECE
Siting and Sizing of DG for
Loss Reduction and Voltage Sag
Mitigati
on in RDS Using ABC Al
gorithm
K. Si
va Ram
u
du
1
, M. Padm
a
L
a
litha
2
, P.
Suresh B
a
b
u
3
Department o
f
E
l
ectrical and
Electro
ni
cs E
ngi
neeri
n
g
, A
I
TS
,
Rajam
p
et
, Ind
i
a
Em
ail:
sivaram.1810@gmail.co
m
1
, padm
ala
lith
a
_
m
a
redd
y
@
ya
h
oo.co.
i
n
2
, suresh
ram48@gmail.com
3
Article Info
A
B
STRAC
T
Article histo
r
y:
Received Aug 10, 2013
Rev
i
sed
O
c
t 18
, 20
13
Accepte
d Nov 4, 2013
In order to redu
ce th
e power lo
ss and to improve the voltag
e
p
r
ofile in
the
distribution
s
y
stem, distribu
ted
generato
rs
(DGs
) are
conn
ect
ed
to load
bus
.
To reduc
e the
to
tal power
loss in the s
y
st
em
, the
m
o
st im
portant process is to
identif
y
th
e prop
er lo
cation for
fixing and
sizing
of DGs. This paper pr
esents
a new methodolog
y
using a new populati
on based meta heuristic approach
nam
e
l
y
Artif
ici
a
l Be
e Colon
y
algori
t
hm
(ABC) for the
p
l
ac
em
ent o
f
Distributed G
e
n
e
rators (DG) in
the r
a
dial distr
i
b
u
tion s
y
s
t
ems to
reduce th
e
real
power loss,
to
im
prove th
e volt
a
ge
profil
e &
volt
a
ge sag
m
itigat
io
n
.
While these po
wer loss reduction, volta
ge prof
ile improvement and voltag
e
sag m
itigation
has significant
role in
lessonin
g
im
posed expenditures to
ut
i
l
ity
c
o
mpa
n
i
e
s.
T
h
e
powe
r
l
o
ss re
duc
t
i
on i
s
import
a
nt
fa
ct
or for ut
i
lity
companies because it
is directly propor
tion
a
l
to
the
compan
y
b
e
nefits in
a
competitiv
e electricity
market, while
reachin
g the better p
o
wer quality
standards is too
im
portant
as it
has vita
l eff
e
ct
on customer orientation. In
this paper an A
BC algorithm is develop
e
d
to ga
i
n
these goals a
ll
togeth
er. In
order to evaluat
e
sag m
itigation capabi
lit
y
of th
e proposed algorit
hm
, voltag
e
in voltag
e
sensit
i
v
e buses is inves
tig
ated. An ex
isting 20KV network (32-bus
s
y
s
t
em
) has
bee
n
chos
en as
tes
t
network and res
u
lts
are com
p
ared with th
e
proposed metho
d
in
the ra
dial distribution s
y
stem.
Keyword:
Artificial b
e
e co
lon
y
algo
rithm
Di
st
ri
b
u
t
e
d ge
nerat
i
o
n
Optim
al DG pl
acem
e
nt
Po
wer l
o
ss
re
d
u
ction
Vo
ltag
e
sag
m
i
tig
atio
n
Copyright ©
201
3 Institut
e
o
f
Ad
vanced
Engin
eer
ing and S
c
i
e
nce.
All rights re
se
rve
d
.
Co
rresp
ond
i
ng
Autho
r
:
K. Siva
Ram
u
d
u
Depa
rtem
ent of Electrical a
n
d
El
ect
ro
ni
cs E
n
gi
nee
r
i
n
g,
Ann
a
m
ach
arya In
stitu
te of Tech
no
log
y
an
d
Scien
ces
(AITS)
Tallap
a
k
a
, Raja
m
p
et, Bo
yan
a
p
a
lli, An
dhra Prad
esh
51
612
6, In
d
i
a, Ph
on
e:
+9
1 85
65
2
489
90
Em
a
il: siv
a
ra
m.18
10@g
m
ail.
co
m
1.
INTRODUCTION
Distribu
ted
g
e
n
e
ration
un
lik
e
cen
tra
lized electrical gene
ra
tion aim
s
to
ge
nerate electrica
l energy on
sm
a
ll scale
as
near as possi
ble to lo
ad
cen
ters, wh
ich
p
r
ov
id
e an
in
crem
en
tal cap
acity
to
p
o
wer syste
m
.
In
the
dere
g
u
l
a
t
e
d p
o
w
er m
a
rket
, conce
r
ns ab
out
t
h
e envi
ro
nm
ent
as wel
l
as econ
o
m
i
c i
ssues have l
e
d i
n
c
r
ease
d
i
n
t
e
rest
i
n
di
st
ri
b
u
t
e
d
gene
rat
i
ons
. Th
e em
erge
nce
of
ne
w
t
echn
o
l
o
gi
cal
al
t
e
rnat
i
v
es (
p
hot
ov
ol
t
a
i
c
sy
st
em
s,
wi
n
d
po
we
r, c
oge
ne
rat
i
o
n
,
et
c.) al
l
o
ws
ge
n
e
rat
i
n
g
pa
rt
of
the require
d
energy
clos
er
to
th
e po
in
t
of u
s
e,
i
m
p
r
ov
ing
q
u
ality lev
e
ls an
d m
i
n
i
mizin
g
th
e inv
e
stmen
t
s co
sts asso
ciated
with
o
f
transm
issio
n
and
d
i
stribu
tio
n sy
ste
m
s.
W
ith
el
ectricity
m
a
rk
et u
n
d
e
rgo
i
ng
tre
m
en
d
o
u
s
transform
a
tio
n
,
m
o
re
p
r
ice instab
i
lity in
the m
a
rket, ageing i
n
frastruc
ture a
n
d c
h
anging
regula
t
o
r
y
envi
ro
nm
ent
s
are
dem
a
ndi
ng
us
ers a
n
d
el
ect
ri
c
u
tilities to
explo
it b
e
n
e
fits
o
f
DG
[1,
2
]
.
DG app
lica
tio
n
s
are growing du
e t
o
en
v
i
ronmen
tal an
d eco
n
o
m
ic
i
ssues, t
e
c
h
n
o
l
ogi
cal
i
m
pro
v
e
m
e
nt
s, an
d
p
r
i
v
at
i
zat
i
on
of
p
o
we
r sy
st
e
m
s. DG ap
pl
i
cat
i
on,
h
o
we
v
e
r,
has
p
o
s
itiv
e and neg
a
tiv
e si
d
e
effects for
p
u
b
lic in
du
st
ries
and
co
nsu
m
ers [3
].
DG can be an altern
ative for
i
n
d
u
st
ri
al
, com
m
erci
al
and re
si
dent
i
a
l
appl
i
cat
i
ons.
DG m
a
kes use
of t
h
e
l
a
t
e
st
m
odern t
echn
o
l
o
gy
wh
i
c
h i
s
efficien
t, reliab
l
e, and
sim
p
l
e
en
oug
h
so
th
at it can
com
p
ete with
tr
ad
itio
n
a
l larg
e g
e
n
e
rators in so
m
e
areas [4].
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
S
itin
g
an
d S
i
zi
n
g
o
f
DG for Lo
ss Redu
ction
a
n
d
Vo
ltag
e
Sag
Mitiga
tio
n in RDS
Using
…
(K. S
i
va
Ramud
u
)
81
5
In t
h
e literature, Single DG
Place
m
e
nt algor
ithm
has be
en applied t
o
DG
placem
ent and ABC
alg
o
r
ith
m
f
o
r th
e Sizi
n
g
of
DG
in th
e r
a
d
i
al
d
i
str
i
bu
tio
n sy
ste
m
s [
5
].
Th
e
A
BC algo
r
ithm
is a n
e
w popu
latio
n
b
a
sed
Meta h
e
u
r
istic app
r
o
a
ch
in
sp
ired
b
y
in
tellig
en
t
fo
rag
i
ng
b
e
h
a
v
i
o
r
o
f
h
o
n
e
yb
ee swarm
.
Th
e adv
a
n
t
age
o
f
ABC algo
rit
h
m
is th
at it d
o
e
s
n
o
t
requ
ire ex
tern
al p
a
ra
m
e
ters such
as
cro
ss
ove
r rat
e
an
d m
u
tation rate as
in
case of g
e
n
e
tic alg
o
r
ith
m
a
n
d
d
i
fferen
tial ev
o
l
u
tio
n. Th
e
o
t
h
e
r adv
a
n
t
age is th
at th
e g
l
o
b
a
l search
ab
i
lity
is
i
m
p
l
e
m
en
ted
b
y
in
tr
od
ucing n
e
ig
hbo
rh
ood so
ur
ce p
r
od
uctio
n
m
ech
an
ism w
h
ich
is si
m
ilar
to
m
u
tatio
n
process
.
In thi
s
pape
r, locations
of distributed
gene
rators are
identif
ie
d by single DG placem
ent
method [6]
an
d
ABC alg
o
rith
m wh
ich
tak
e
s th
e nu
m
b
er an
d
lo
cation
of D
G
s as i
n
p
u
t
has bee
n
de
vel
o
ped t
o
det
e
rm
i
n
e
th
e o
p
tim
al
siz
e
(s) of DG to
min
i
mi
ze real
powe
r losses in distribution s
y
ste
m
s. The advanta
g
es of rel
i
eving
AB
C
m
e
t
hod
fr
om
det
e
rm
i
n
at
i
on
of
l
o
cat
i
ons
of
D
G
s a
r
e im
pro
v
e
d
c
o
nve
r
g
ence
cha
r
act
eri
s
t
i
c
s an
d l
e
ss
com
put
at
i
on t
i
m
e
. Vol
t
a
ge a
n
d t
h
erm
a
l
cons
t
r
ai
nt
s are
c
ons
i
d
ere
d
[7]
.
The
co
nv
ent
i
o
nal
di
st
ri
b
u
t
i
o
n
gri
d
s a
r
e
co
mm
o
n
l
y fed
u
n
i
d
i
rection
a
l, a fau
lt
o
r
startin
g
a larg
e size
m
o
to
r in on
e b
u
s of th
e
grid
can cau
se voltag
e
sag i
n
b
u
ses i
n
vi
ci
ni
t
y
[8]
.
DG s
u
p
p
o
rt
s
vol
t
a
ge i
n
c
o
nnect
i
o
n
poi
nt
[9]
an
d t
h
i
s
effi
ci
ency
i
s
h
i
ghl
y
depe
n
d
ent
on t
h
e si
ze an
d l
o
c
a
t
i
on o
f
D
G
u
n
i
t
,
so o
p
t
i
m
a
l
si
zi
ng an
d l
o
cat
i
on
of
DG ca
n
gi
ve t
h
e
o
p
p
o
rt
uni
t
y
to
b
e
n
e
fit th
is p
o
t
en
tial capacity
to
m
i
tig
ate v
o
ltag
e
sa
gs, p
a
rticu
l
arly in
b
u
s
es with
sen
s
itiv
e lo
ads. Th
e
pr
o
pose
d
(AB
C
) base
d a
p
p
r
oach i
s
t
e
st
ed
on a
p
r
act
i
cal
32
-
bus
ra
di
al
di
st
ri
b
u
t
i
on
sy
st
em
and t
h
e s
cenari
o
s
y
i
el
ds effi
ci
en
cy
i
n
i
m
prove
m
e
nt
of
v
o
l
t
a
g
e
p
r
o
f
i
l
e
an
d
r
e
duct
i
o
n
o
f
vo
l
t
a
ge sags
an
d
p
o
we
r l
o
sses,
i
t
al
s
o
perm
its an increase in
powe
r t
r
ans
f
er capacit
y
and m
a
xim
u
m
loading.
2.
LOSSES I
N
A DIST
RIB
U
TI
ON
SYSTEM
Th
e t
o
tal
I
2
R l
o
ss
(P
lt
)
i
n
a
di
st
ri
but
i
o
n
sy
st
em
havi
ng
n
n
u
m
ber of
b
r
anc
h
es i
s
gi
ven
by
:
n
i
i
ti
lt
R
I
P
1
2
(1
)
Here I
ti
i
s
t
h
e
m
a
gni
t
ude
of t
h
e b
r
anc
h
c
u
r
r
e
nt
an
d R
i
is the resistance of
the
i
th
branc
h
respectively.
The
bra
n
c
h
cu
rre
nt
can
be
o
b
t
a
i
n
ed
f
r
om
the l
o
a
d
fl
o
w
s
o
l
u
t
i
o
n. T
h
e
b
r
anc
h
c
u
r
r
ent
has t
w
o c
o
m
pone
nt
s
,
active com
ponent (
Ia
i
) a
nd
react
i
v
e com
pone
nt
(
Iri
). T
h
e loss ass
o
ci
ated w
ith
th
e activ
e an
d
reactiv
e
com
pone
nts
of branc
h
c
u
rre
n
ts can be
written as
n
i
i
ai
la
R
I
P
1
2
(2
)
i
n
i
ri
lr
R
I
P
2
1
(3
)
Not
e
t
h
at
f
o
r
a
gi
ve
n c
o
n
f
i
g
ur
at
i
on
of
a si
n
g
l
e
-so
u
r
ce ra
di
al
net
w
or
k,
t
h
e l
o
ss
Pla ass
o
ciated with the
active com
ponent of branc
h
c
u
rrents ca
nnot
be m
i
nimized because
all active powe
r m
u
st be supplie
d by the
sou
r
ce at
t
h
e
r
oot
b
u
s.
Ho
we
ver
by
pl
aci
ng
DGs
, t
h
e act
i
v
e com
pone
nt
s
of
bra
n
c
h
c
u
r
r
e
nt
s are c
o
m
p
ensat
e
d
and l
o
sses d
u
e
t
o
act
i
v
e com
ponent
s o
f
b
r
anc
h
cu
rre
nt
s
are red
u
ce
d.
Thi
s
pa
per
pr
esent
s
a
m
e
t
hod t
h
at
minimizes the loss due to t
h
e activ
e com
p
one
n
t of the
bra
n
c
h
curre
nt
by
o
p
t
i
m
a
l
l
y
pl
aci
ng t
h
e D
G
s an
d
th
ereb
y red
u
ces th
e t
o
tal lo
ss
in
th
e
d
i
stri
b
u
t
i
o
n system
[1
0
]
.
3.
SINGLE DG
PLACE
MEN
T
ALGO
RIT
H
M
This algorithm
determ
ines
the optim
al size
and location
of
DG
units that should
be
placed in the
syste
m
to
min
i
mize lo
ss. First o
p
tim
u
m
sizes
o
f
DG
un
its fo
r all no
d
e
s are d
e
term
in
ed
fo
r
b
a
se case and
b
e
st
one i
s
c
h
ose
n
base
d o
n
t
h
e
m
a
xim
u
m
l
o
ss savi
n
g
.
Thi
s
pr
ocess i
s
rep
eat
ed i
f
m
u
l
t
i
pl
e D
G
l
o
cat
i
ons
are
requ
ired
b
y
m
o
d
i
fying
th
e b
a
se syste
m
b
y
in
sertin
g a
DG unit in
to
th
e system
o
n
e
-b
y-on
e
[11
]
.
3.
1. Me
th
od
ol
og
y
Ass
u
m
e
t
h
at
a
si
ngl
e-
so
urce
r
a
di
al
di
st
ri
b
u
t
i
on
sy
st
em
wi
t
h
n
b
r
anc
h
es a
nd a
D
G
i
s
t
o
be pl
ace
d a
t
bus
m
a
nd
r is a set of bra
n
ches connecte
d
between the
s
ource and bus m
.
The DG
produces active current I
dg
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
IJECE Vol. 3, No. 6, D
ecem
ber 2013
:
814 – 822
81
6
and
f
o
r a
radi
al
net
w
or
k i
t
cha
nge
s o
n
l
y
t
h
e a
c
t
i
v
e com
pone
nt
o
f
cu
rre
nt
of
bra
n
c
h
set
r.
T
h
e cu
rr
ent
s
of
ot
he
r
bra
n
c
h
es a
r
e
unaffected. T
h
us ne
w active
curre
nt I
ai
new
of
the
i
th
bra
n
c
h
i
s
g
i
ven
by
dg
i
ai
new
ai
I
D
I
I
(4
)
Whe
r
e D
i
=1;
if branc
h
i
r
=
0
;
other
w
ise
The l
o
ss
P
la
dg
associ
at
ed
w
i
t
h
t
h
e act
i
v
e
com
pone
nt
o
f
bra
n
ch
cu
rr
ent
s
i
n
ne
w s
y
st
em
(whe
n
DG
i
s
connected)
is given by
n
i
i
dg
i
ai
dg
la
R
I
D
I
P
1
2
)
(
(5
)
The sa
vi
n
g
S i
s
t
h
e
di
ffe
re
nce
bet
w
ee
n e
quat
i
on
(
2
)
an
d
(
5
)
and
i
s
gi
ve
n
by
n
i
i
dg
i
dg
ai
i
dg
la
la
R
I
D
I
I
D
P
P
S
1
2
)
)
(
2
(
(6
)
The
DG c
u
rrent I
dg
t
h
at
pr
ovi
d
e
s m
a
xim
u
m
savi
n
g
ca
n
be
o
b
t
a
i
n
ed
f
r
om
0
)
(
2
1
i
dg
n
i
i
ai
i
dg
R
I
D
I
D
I
S
(7
)
The
D
G
c
u
r
r
en
t
fo
r m
a
xim
u
m
savi
ng
i
s
r
i
i
r
i
i
ai
i
n
i
i
n
i
i
ai
dg
R
R
I
R
D
Ri
D
I
I
1
1
(8
)
The c
o
rres
ponding
DG size i
s
P
dg
=V
m
I
dg
(9
)
Whe
r
e V
m
is th
e
v
o
ltag
e
m
a
g
n
itud
e
of
b
u
s-
m
. T
h
e
optimum
size of
DG at eac
h
bus
is dete
rm
ined usi
n
g
equation (9).T
h
en savi
ng for
each
DG
is determ
ined
usi
ng e
q
uation (6).The
DG wi
th highest sa
ving is
candi
date loca
tion for single
DG placem
ent. Whe
n
the
candi
date bus is identified and DG is placed, the
pr
ocess
i
s
re
pe
at
ed t
o
i
d
e
n
t
i
f
y
su
bse
que
nt
bu
ses f
o
r
D
G
pl
a
c
em
ent
[1
2]
.
3.2. Algorithm
for
Single DG Placement
Step
1
:
Co
ndu
ct lo
ad
flo
w
an
alysis for th
e orig
in
al
syste
m
.
Step 2:
Calculate I
dg
an
d DG
si
ze usi
n
g
e
quat
i
o
ns
(
8
)
& (9
) fo
r bu
se
s
i
=2…
n
.
St
ep
3:
Det
e
rm
i
n
e savi
ng
usi
n
g e
quat
i
on
6
,
fo
r
bu
ses
i
=2…
n
.
St
ep
4:
Ide
n
t
i
f
y
t
h
e m
a
xi
m
u
m
savi
ng
and
t
h
e c
o
rres
p
on
di
n
g
D
G
si
z
e
.
Step
5:
The corres
p
onding bus is candidate bus whe
r
e
DG can be
placed. M
odi
fy th
e active load at this bus
and conduct t
h
e load fl
ow aga
i
n.
Step 6:
Ch
eck wh
eth
e
r th
e sav
i
ng
ob
tain
is m
o
re th
an
1
k
W. If yes,
g
o
to
step
2
.
Ot
h
e
rwise,
go
to
n
e
x
t
step
.
Step
7:
pri
n
t all the ca
ndi
date locations t
o
place DG source
s a
n
d the sizes.
Since the
DGs
are added t
o
the system
one by one, th
e si
z
e
s obt
ai
ned
by
si
ngl
e D
G
pl
acem
e
nt
al
gori
t
hm
are
lo
cal op
ti
m
a
n
o
t
g
l
ob
al op
t
i
m
u
m
so
lu
tio
n. Th
e g
l
o
b
a
l
o
p
tim
al so
lu
tio
n
is
ob
tain
ed if m
u
ltip
le DGs are
sim
u
ltaneously
placed in t
h
e s
y
ste
m
by using ABC algor
ithm
.
This
m
e
thod is e
x
pl
ained
in next section.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
S
itin
g
an
d S
i
zi
n
g
o
f
DG for Lo
ss Redu
ction
a
n
d
Vo
ltag
e
Sag
Mitiga
tio
n in RDS
Using
…
(K. S
i
va
Ramud
u
)
81
7
4.
ARTIFICIAL
BEE COLONY
ALGORITHM
(AB
C
)
Artificial Bee Colony (ABC) is one of the
m
o
st r
ecently
defi
ned algorithm
s
by Dervis
Kara
boga in
20
0
5
, m
o
t
i
v
at
ed by
t
h
e i
n
t
e
l
l
i
gent
be
ha
vi
o
r
of h
o
n
ey
bees
. AB
C
as an opt
im
i
zati
on t
ool
pr
o
v
i
d
es a po
p
u
l
a
t
i
o
n
b
a
sed
search
pro
c
edu
r
e in
wh
ich
ind
i
v
i
du
als called
fo
od
p
o
s
ition
s
are
m
o
d
i
fied
b
y
th
e artificial b
e
e
s
with
tim
e
and the
bee’s aim
is to
discover the
pl
aces of
f
ood s
o
urces
with hi
gh necta
r
am
ount a
n
d
finally the
one
with
th
e h
i
gh
est n
ectar. In
this alg
o
r
ith
m
[1
1
,
12
], th
e
co
lo
n
y
of artificial b
ees co
n
s
ist
s
o
f
th
ree grou
p
s
o
f
bees:
em
pl
oy
ed bees
, o
n
l
o
o
k
e
rs an
d sc
out
s
.
Fi
rst
hal
f
of t
h
e col
ony
c
o
n
s
i
s
t
s
of t
h
e em
pl
oy
ed a
r
t
i
f
i
c
i
a
l
bees
and the second half includes the onlo
okers.
For e
v
ery food source, the
r
e
i
s
onl
y
o
n
e em
pl
oy
ed
bee. I
n
ot
h
e
r
wo
rd
s, t
h
e
n
u
m
ber of em
pl
oy
ed bees i
s
e
q
ual
t
o
t
h
e
n
u
m
b
er
of
f
o
o
d
s
o
urces a
r
ou
n
d
t
h
e hi
ve. T
h
e e
m
pl
oy
ed
bee whose food source has
been aba
n
do
ne
d
becom
e
s a scout
[
1
3]
. Th
us,
A
BC syste
m
com
b
ines local searc
h
carri
ed
o
u
t
by
em
pl
oy
ed a
nd
onl
oo
ke
r bee
s
,
an
d gl
obal
sea
r
ch m
a
nage
d
b
y
onl
oo
ker
s
an
d sc
out
s
,
at
t
e
m
p
t
i
n
g
t
o
bal
a
nce e
x
pl
orat
i
o
n a
n
d e
x
pl
oi
t
a
t
i
o
n
p
r
oc
ess [
14]
.
Th
e
ABC algorith
m
creates a rando
m
l
y d
i
st
ri
bu
ted
i
n
itial p
opu
latio
n
of so
lu
tion
s
(
f
=
1,
2,…
.
.,
T
nf
),
whe
r
e ‘
f
’ sign
i
f
ies th
e size
o
f
p
opu
latio
n
and
‘
T
nf
’ is the
num
b
er of em
pl
oyed bees. Eac
h
s
o
lution
x
f
is a D-
di
m
e
nsi
onal
ve
ct
or, w
h
e
r
e D i
s
t
h
e num
ber of pa
ram
e
t
e
rs to be o
p
t
i
m
i
zed. The p
o
si
t
i
on
of a fo
o
d
-s
o
u
r
ce, i
n
t
h
e AB
C
al
go
r
i
t
h
m
,
represe
n
t
s
a possi
bl
e s
o
l
u
t
i
on t
o
t
h
e
o
p
t
i
m
i
zat
i
on pr
obl
em
, and t
h
e
nect
ar am
ount
of a
food
sou
r
ce co
rresp
ond
s to
th
e q
u
a
lity (fitn
ess v
a
l
u
e) of th
e asso
ciated
so
lu
tion
.
After in
itializa
tio
n
,
th
e
p
opu
latio
n
of th
e p
o
s
itio
n
s
(so
l
u
tion
s
) is su
bj
ected
to
repeate
d
cycles of the searc
h
processe
s for the
em
pl
oy
ed, o
n
l
o
o
k
e
r
, an
d sco
u
t
bees (cy
c
l
e
= 1, 2, …,
MC
N
), where
MC
N
i
s
t
h
e
m
a
xi
m
u
m
cy
cl
e num
ber of
t
h
e searc
h
pr
o
cess. The
n
, a
n
em
pl
oy
ed bee
m
odi
fi
es t
h
e posi
t
i
on (
s
ol
ut
i
o
n) i
n
her m
e
m
o
ry
de
pe
ndi
ng
on t
h
e
local inform
ation (vis
ual inform
at
ion) a
n
d tests the ne
ctar am
ount (fitness value) of the
ne
w
position
(m
odi
fi
ed sol
u
t
i
on)
. If t
h
e
n
ect
ar am
ount
of t
h
e n
e
w o
n
e
i
s
hi
gher t
h
a
n
t
h
at
of t
h
e
p
r
evi
ous
one
, t
h
e bee
me
m
o
rizes th
e n
e
w po
sition
an
d
forg
ets th
e o
l
d
on
e. Ot
h
e
rwise, sh
e
k
eeps th
e p
o
s
ition
o
f
th
e
p
r
ev
iou
s
o
n
e
i
n
her
m
e
m
o
ry
. A
f
t
e
r al
l
em
pl
oy
ed
bees
ha
ve c
o
m
p
l
e
t
e
d t
h
e
s
earch process t
h
ey s
h
are
the nectar
inform
ation
of
th
e food
sou
r
ces an
d
th
ei
r positio
n
in
fo
rm
at
io
n
with
t
h
e
o
n
l
oo
ker
bees wa
i
t
i
ng i
n
t
h
e
da
nce area. An onlooke
r
bee eval
uates the nectar inform
ation taken from
al
l em
pl
oy
ed bees and c
h
o
o
ses
a fo
od s
o
u
r
ce
wi
t
h
a
probability related to its
nect
ar am
ount. The sam
e
proc
e
d
ure of position
m
odification and
selection
c
r
iterion
use
d
by
t
h
e
e
m
pl
oy
ed bees
i
s
appl
i
e
d t
o
onl
oo
ke
r be
e
s
. The
g
r
eedy
-
sel
ect
i
on
pr
o
cess i
s
sui
t
a
bl
e fo
r
u
n
c
on
strain
ed
o
p
tim
izat
io
n
prob
lem
s
. Th
e
p
r
ob
ab
ility o
f
selectin
g
a
fo
od
so
urce
P
f
b
y
on
look
er
bees is
calcu
lated
as
fo
llo
ws:
nf
T
f
f
f
fitness
fitness
P
1
(1
0)
Whe
r
e
fitn
ess
f
is th
e fitn
ess valu
e o
f
a so
lu
ti
o
n
f ,
a
nd
T
nf
is th
e to
tal n
u
m
b
e
r of fo
od-sou
r
ce p
o
s
itions
(sol
ut
i
o
n
s
)
or,
i
n
ot
he
r w
o
r
d
s
,
hal
f
o
f
t
h
e col
ony
si
ze. C
l
ear
l
y
, resul
t
i
ng
fr
om
usi
ng (
1
0
)
,
a go
od
fo
o
d
s
o
u
r
ce
(so
l
u
tio
n)
will attract
m
o
re on
loo
k
e
r
b
ees t
h
an a
b
a
d on
e. Sub
s
eq
u
e
n
t
to on
loo
k
e
rs sel
ectin
g
th
ei
r
p
r
eferred
fo
o
d
-s
ou
rce,
t
h
ey
pr
od
uce a
n
e
i
g
h
b
o
r
fo
o
d
-s
ou
rce
po
si
t
i
on
f+1
to the sele
cted one
f,
a
nd com
p
are the
nectar
am
ount
(fi
t
n
es
s val
u
e
)
o
f
t
h
at
nei
g
hb
o
r
f+1
p
o
s
ition
with
th
e o
l
d
po
sition
.
The sam
e
sel
ect
i
on cri
t
e
ri
on
use
d
b
y
th
e e
m
p
l
o
y
ed
b
ees is app
lied
to
o
n
l
o
o
k
e
r b
ees as well. Th
is seq
u
e
n
c
e is rep
eated
until a
ll o
n
l
o
o
k
e
rs are
di
st
ri
b
u
t
e
d.
F
u
rt
herm
ore, i
f
a sol
u
t
i
o
n
f
does
not im
prove for a speci
fied
num
ber
o
f
tim
e
s (lim
it), the
e
m
ployed
bee associated w
ith this s
o
lution abandons it,
and s
h
e
becomes a scout a
nd sea
r
c
h
es for a ne
w
rando
m
fo
od
-so
u
rce
po
sitio
n. On
ce t
h
e
n
e
w
p
o
sitio
n
is
d
e
term
in
ed
, ano
t
h
e
r
ABC algo
rith
m
(
MCN
) cycle
starts. Th
e same p
r
o
cedu
r
es are
rep
eated un
til th
e st
op
p
i
n
g
criteria are m
e
t. In
o
r
d
e
r to
d
e
termin
e a
nei
g
hb
o
r
i
n
g f
o
o
d
-
so
ur
ce
po
si
t
i
on (
s
ol
ut
i
o
n)
t
o
t
h
e
ol
d
o
n
e i
n
m
e
m
o
ry
,
t
h
e
AB
C
al
go
ri
t
h
m
al
ters
one
ran
d
o
m
l
y
chos
en pa
ram
e
t
e
r and
kee
p
s t
h
e re
m
a
i
n
i
ng pa
ram
e
t
e
rs u
n
cha
n
ge
d. I
n
ot
her
w
o
r
d
s,
by
ad
di
n
g
t
o
t
h
e
cur
r
ent
c
h
osen
param
e
t
e
r val
u
e t
h
e
p
r
o
d
u
ct
of
t
h
e
uni
fo
r
m
vari
ant
[-
1,
1]
an
d t
h
e
di
ffe
rence
bet
w
e
e
n t
h
e
cho
s
en pa
ram
e
t
e
r val
u
e an
d ot
he
r “ran
d
o
m
”
sol
u
t
i
on
para
m
e
t
e
r val
u
e, t
h
e nei
g
h
b
o
r
f
o
od
-s
ou
rce p
o
si
t
i
on i
s
created. T
h
e following e
x
pres
sion ve
rifies that
)
(
mg
old
fg
old
fg
new
fg
x
x
u
x
x
(1
1)
Whe
r
e
f
m
and b
o
t
h are
nf
T
,.....
2
,
1
. Th
e
m
u
l
tip
lier
u
i
s
a rand
om
num
ber bet
w
een
[-1
, 1]
and
D
g
,....
2
,
1
.
In ot
he
r wo
r
d
s,
x
fg
is th
e
g
th
p
a
ram
e
ter o
f
a so
lu
tion
x
f
that was select
ed to
be m
odified.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
IJECE Vol. 3, No. 6, D
ecem
ber 2013
:
814 – 822
81
8
Wh
en
th
e
food-sou
r
ce po
sition
h
a
s
b
e
en
aban
don
ed, th
e em
p
l
o
y
ed
b
ee asso
ciated
w
ith
it b
eco
m
e
s a s
c
o
u
t
.
Th
e scou
t pro
d
u
ces a co
m
p
letely n
e
w
fo
od
so
urce
po
sitio
n
as fo
llo
ws:
)
min(
)
max(
)
min(
)
(
fg
fg
fg
new
fg
x
x
u
x
x
(1
2)
Wh
ere
(12
)
app
lies to
all
g
param
e
ters an
d
u
i
s
a ra
n
d
o
m
nu
m
b
er bet
w
een
[
-
1
,
1]
.
If a
pa
ram
e
t
e
r
val
u
e
produce
d
using (11) and/or
(12) e
x
cee
ds its
pre
d
eterm
i
ned lim
i
t,
the pa
ra
meter can
be s
e
t to an accept
a
ble
val
u
e.
I
n
t
h
i
s
p
a
per
,
t
h
e
val
u
e
of t
h
e
param
e
ter excee
di
n
g
it
s li
m
i
t is fo
rced
to
t
h
e
n
earest
(d
iscrete)
b
oun
d
a
ry
li
mit v
a
lu
e asso
ciated
with
it. Fu
rth
e
rm
o
r
e, th
e rando
m
m
u
l
tip
lier n
u
m
b
e
r
u
is set to
b
e
b
e
tween
[0
, 1
]
in
stead
o
f
[-1, 1
]
. Thu
s
, th
e ABC
alg
o
rith
m
h
a
s
th
e fo
llowing
co
n
t
ro
l p
a
rameters: 1
)
th
e co
lon
y
size (CS), th
at
consists of em
ployed be
es
T
nf
pl
us
o
n
l
o
o
k
er
bees
T
nf
;
2
)
th
e lim
i
t
v
a
lu
e, wh
ich
is th
e
nu
m
b
er of trials fo
r
a
fo
o
d
-s
ou
rce
p
o
s
i
t
i
on (
s
ol
ut
i
o
n
)
t
o
be
aba
n
do
ned;
a
n
d
3)
t
h
e
m
a
xim
u
m
cy
cl
e num
ber
MC
N
.
The
pr
op
ose
d
AB
C
al
go
ri
t
h
m
for
fi
n
d
i
n
g s
i
ze of
DG at
se
lected location
to minim
i
ze th
e real powe
r loss is a
s
fo
llows:
Step
1
:
In
itialize th
e fo
od-sou
r
ce
p
o
sitio
n
s
x
f
(
sol
u
t
i
ons
p
o
pul
at
i
o
n)
,
whe
r
e
f
=1, 2,….., T
nf
. The
x
f
so
lu
ti
on
fo
rm
is as follo
ws.
St
ep
2:
C
a
l
c
ul
at
e t
h
e n
ect
ar am
ount
o
f
t
h
e
p
o
pul
at
i
o
n
by
m
eans of
t
h
ei
r fi
t
n
ess
va
l
u
es
usi
n
g
powerloss
Fitness
1
1
(1
3)
St
ep
3:
Pro
d
u
ce
nei
g
h
b
o
r
s
o
l
u
t
i
o
ns f
o
r t
h
e em
pl
oy
ed bee
s
by
u
s
i
n
g (
1
1
)
a
nd e
v
a
l
uat
e
t
h
em
as indi
cat
ed
by
Step 2.
Step
4:
Apply the
gree
dy selection process.
St
ep
5:
If al
l
onl
oo
ke
r
bees a
r
e
di
st
ri
b
u
t
e
d,
g
o
t
o
St
e
p
9.
Ot
he
r
w
i
s
e,
g
o
t
o
t
h
e
ne
xt
st
ep.
Step
6
:
Calcu
l
ate th
e prob
ab
ility v
a
lu
es
P
f
fo
r th
e so
l
u
tio
ns
x
f
u
s
i
n
g (1
0)
Step
7
:
Pr
odu
ce
n
e
ighb
or
so
lu
tion
s
fo
r
t
h
e selected on
loo
k
e
r
b
ee, d
e
p
e
nd
ing
on th
e
v
a
lu
e, usin
g (8
) and
evaluate t
h
em
as Step 2 indicates.
Step 8:
Follow
Ste
p
4.
Step
9:
Determ
ine the abandone
d s
o
l
u
tion
for the s
c
out
bees, if it
exists, and re
place it with a com
p
letely
n
e
w so
l
u
tion
usin
g (1
2) and
ev
alu
a
te t
h
em
a
s
ind
i
cated
in Step
2.
Step
1
0
:
Me
m
o
rize th
e
b
e
st so
lu
tion
at
tain
ed
so
far.
Step
1
1
:
If cycle = MC
N, st
o
p
an
d prin
t resu
lt.
Oth
e
rwise fo
llow St
ep
3
.
5.
VOLTAGE S
A
G
MITI
GATION
The
Propose
d
approach d
eals with
voltage
sag
propa
gation mitig
ation. This takes i
n
to
account the
nu
m
b
er of
b
u
s
e
s
t
h
at
expe
ri
ence v
o
l
t
a
ge sa
g. (
1
4) S
h
o
w
s
t
h
e pr
o
pose
d
f
unct
i
o
n t
o
m
i
nim
i
ze
t
h
e vol
t
a
ge sa
g
an
d th
ereb
y i
n
creasing
th
e voltag
e
am
p
litu
d
e
s in
rad
i
al
d
i
stribu
tio
n system
[1
5
]
.
h
m
l
sp
V
V
V
V
3
2
1
(1
4)
In w
h
ich
V
l
i
s
t
h
e
n
u
m
b
er of bus
es wi
t
h
vol
t
a
ge
am
pl
i
t
ude
dr
o
p
bel
o
w 0.
1 p.
u,
V
m
is the nu
m
b
er of
bus
es wi
t
h
v
o
l
t
a
ge am
pl
i
t
ude bet
w
ee
n 0.
4 p
.
u an
d 0.
7 p
.
u a
nd fi
nal
l
y
V
h
i
s
t
h
e num
ber of
buses
wi
t
h
v
o
l
t
a
ge
am
pl
i
t
ude bet
w
een
0.
7 p.
u a
nd
0.
9 p.
u
du
ri
ng
vol
t
a
g
e
sag
occu
rre
nce.
Wei
g
ht
i
ng c
o
e
ffi
ci
ent
s
are
de
fi
ne
d as
,
and
to
m
a
g
n
i
fy vo
l
t
ag
e su
ppo
rt in m
o
re ill-co
nd
i
tio
n
e
d bu
ses.
6.
ILLUSTRAT
I
VE E
X
AMPLE
An
exi
s
t
i
n
g
2
0
K
V
net
w
o
r
k
w
h
i
c
h i
s
m
odel
e
d
wi
t
h
3
2
buse
s
i
s
st
u
d
i
e
d
.
T
h
i
s
m
e
di
um
vol
t
a
ge fee
d
e
r
whi
c
h i
s
l
o
cat
ed i
n
s
out
h K
h
o
r
asa
n
p
r
o
v
i
n
ce i
n
Ira
n
ha
s
severe
voltage problem
s especially in peak loa
d
ho
u
r
s a
n
d
i
n
e
n
d
feede
r
a
r
eas.
Fi
gu
re
1 s
h
ows
si
n
g
l
e
l
i
n
e
di
agram
of t
h
i
s
ne
t
w
o
r
k
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
S
itin
g
an
d S
i
zi
n
g
o
f
DG for Lo
ss Redu
ction
a
n
d
Vo
ltag
e
Sag
Mitiga
tio
n in RDS
Using
…
(K. S
i
va
Ramud
u
)
81
9
Fi
gu
re
1.
Si
n
g
l
e
Li
ne
Di
ag
ra
m
of t
h
e Test
Net
w
or
k
Tabl
e 1.
T
h
e
L
o
ad
Dat
a
o
f
32
-B
us Sy
st
em
Bus No
P
L
in KW
Q
L
in
KVA
r
Load K
V
A
Sending
end node
Reciving
end node
Resistance in
ohm
s
Reactance in
ohm
s
1
-
-
-
2 360
63
425
1
2
1.
32
0.
48
3 0
0
0
2
3
1.
56
0.
56
4 170
30
200
3
4
0.
78
0.
28
5 445
78
525
3
5
1.
4
0.
51
6 467
82
550
5
6
3.
3
1.
2
7 85
15
100
6
7
7.
8
2.
8
8 -
-
-
7
8
3.
4
1.
2
9 573
101
675
8
9
6.
8
2.
4
10
85
15
100
9
10
12
4.
6
11
42
7
50
9
11
7
2.
8
12
85
15
100
8
12
4.
6
1.
7
13
85
15
100
12
13
2.
9
1
14
85
15
100
13
14
12.
5
4.
5
15
63
11
75
14
15
11.
7
4.
2
16
488
86
575
15
16
3.
1
1.
1
17
148
26
175
13
17
5.
4
1.
9
18
191
33
225
17
18
4.
6
1.
7
19
127
22
150
18
19
6.
8
2.
4
20
-
-
-
1
20
4.
4
1.
6
21
190
33
225
20
21
4.
4
1.
6
22
127
22
150
21
22
11.
2
4
23
297
52
350
22
23
7.
8
2.
8
24
-
-
-
23
24
7.
8
2.
8
25
467
82
550
24
25
3.
9
1.
4
26
276
48
325
24
26
5.
1
1.
8
27
85
15
100
20
27
5.
8
2.
8
28
430
75
505
20
28
11.
7
4.
2
29
191
33
225
28
29
5.
08
1.
8
30
-
-
-
29
30
4.
4
1.
6
31
85
15
100
30
31
9.
8
1.
3
32
63
11
75
30
32
5.
6
2
7.
R
E
SU
LTS AN
D ANA
LY
SIS
Fi
rst
l
o
ad fl
o
w
i
s
con
duct
e
d fo
r 3
2
-
b
us t
e
st
sy
st
em
. The po
wer l
o
ss
due t
o
act
i
v
e
com
pone
nt
o
f
current
is 1015.9 kW
and pow
er loss due to
reac
tive component
of
the current
is
42.
5752
kW. A program
is
written i
n
“M
ATLAB” to i
m
ple
m
ent single DG
placem
ent algorithm
.
For the
fi
rst iteration the
m
a
xim
u
m
savi
n
g
i
s
occu
r
r
i
n
g at
bus 8.
The can
di
dat
e
l
o
cat
i
on f
o
r
D
G
i
s
bus 8
wi
t
h
a l
o
ss savi
n
g
of 4
7
5
.
8
8
59
k
W
. Th
e
opt
i
m
u
m
si
ze
of
DG at
b
u
s 8 i
s
2.
66
5
6
M
W
. B
y
assum
i
ng
2.
66
5
6
M
W
D
G
i
s
con
n
ect
ed at
bus 8
of ba
s
e
syste
m
an
d
is co
nsid
er
ed
as
b
a
se case.
N
o
w
th
e cand
i
d
a
t
e
lo
catio
n
is bu
s w
ith
0
.
4
908 MW
size and
th
e lo
ss
sav
i
ng
is 120
.61
4
2
KW. Th
is
p
r
o
cess is rep
e
ated
till th
e
lo
ss sav
i
ng
is in
si
g
n
i
fican
t. Th
e
resu
lts are shown
i
n
Tabl
e 2.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
IJECE Vol. 3, No. 6, D
ecem
ber 2013
:
814 – 822
82
0
Table 2. Single
DG placem
ent
res
u
lts
Iteration No
Bus No
DG S
i
z
e
(MW)
Saving (KW
)
1 8
2.
6656
502.
09
92
2 30
0.
4908
120.
61
42
3 29
0.
4383
87.
067
0
4 27
0.
4405
76.
277
3
The candi
date locations
for DG placem
ent are taken
from
s
i
ngle DG place
ment algorithm i.e. 8, 30,
29
, 2
7
.
W
i
t
h
t
h
ese l
o
cat
i
o
ns
,
si
zes of
D
G
s
cor
r
es
po
n
d
i
n
g
t
o
gl
o
b
al
s
o
l
u
t
i
on a
r
e det
e
rm
i
n
ed
by
usi
ng
AB
C
Al
g
o
ri
t
h
m
descri
be
d i
n
sect
i
on
3. The si
zes
of D
G
s are de
pen
d
e
n
t
on t
h
e
num
ber of
D
G
l
o
cat
i
o
n
s
. G
e
neral
l
y
i
t
i
s
not
po
ssi
bl
e t
o
i
n
st
al
l
m
a
ny
DGs
i
n
a
gi
v
e
n r
a
di
al
sy
st
e
m
. Here 4 ca
se
s are c
o
nsi
d
e
r
e
d
.
In
case
1
o
n
l
y
one
DG in
stallatio
n
is assu
m
e
d
.
In
case
2
two DGs, i
n
case
3 t
h
ree
DGS and in the la
st case four
DGs a
r
e
assu
m
e
d
to
b
e
in
stalled
.
DG sizes in
th
e fo
ur op
tim
a
l
lo
cat
io
n
s
, to
tal real p
o
wer l
o
sses
b
e
fo
re and
after
DG
i
n
st
al
l
a
t
i
on f
o
r
fo
ur
cases a
r
e
gi
ve
n i
n
Ta
bl
e
3.
Tabl
e
3. R
e
s
u
l
t
s
o
f
32
-B
us
sy
s
t
em
Case Bus
Locations
DG S
i
z
e
(KW
)
Total Size
(MW)
Losses Before D
G
installation (KW)
Losses After
D
G
Installation (KW)
Savings
(KW)
1 8
2.
6656
2.
6656
1058.
4
462.
64
22
595.
75
78
2
8 0.
2188
0.
7096
378.
65
82
679.
74
18
30
0.
4908
3
8 0.
1825
1.
0432
336.
21
74
722.
18
26
30
0.
4224
29
0.
4383
4
8 0.
1518
1.
3353
311.
55
74
746.
84
26
30
0.
3594
29
0.
3836
27
0.
4405
Du
e t
o
th
e installatio
n
o
f
t
h
e th
ree
DG’s at th
e d
e
termin
ed
lo
cation
s
with
t
h
e co
rresp
ond
ing
det
e
rm
i
n
ed at
si
zes, t
h
e
t
o
t
a
l
real
p
o
w
er
l
o
ss
o
f
t
h
e
sy
st
em
i
s
red
u
ce
d
fr
o
m
1015
.9
K
W
t
o
28
3.
6
5
8
K
W
wi
t
h
a
m
a
xim
u
m
saving
of 7
3
2
.
2
4
2
K
W. The res
u
l
t
s are sho
w
n i
n
t
a
bl
e 4. Sim
i
l
a
rl
y
due t
o
t
h
e i
n
t
r
od
uct
i
o
n of
DG i
n
to
th
e
syste
m
th
e
v
o
ltag
e
pro
f
ile as b
e
en
im
p
r
ov
ed
wh
ich
is
represen
ted
i
n
th
e b
e
l
o
w Tab
l
e 5
.
Tabl
e 4.
R
e
s
u
l
t
s
f
o
r real
po
we
r
l
o
ss
befo
re and
after
DG in
st
allatio
n
Losses Before D
G
Installation (K
W
)
Losses After
D
G
i
n
stallation (K
W)
Real Power L
o
ss
1015.
9
283.
65
8
Tab
l
e
5
.
Resu
lts fo
r
vo
ltag
e
p
r
o
f
ile
b
e
fo
re and
after
DG in
st
allatio
n
,
Vo
ltag
e
sag
m
itig
ati
o
n.
Bus No
Volt
ages Bef
o
re
DG
Ins
t
allat
i
on
Volt
ages A
f
t
e
r DG
Inst
allat
i
on
Volt
ages Af
t
e
r Vo
lt
age sag
m
i
t
i
gat
ion
1 1.
0000
1.
0000
1.
0000
2 0.
9768
0.
9875
0.
9924
3 0.
9601
0.
9833
0.
9882
4 0.
9601
0.
9833
0.
9882
5 0.
9464
0.
9808
0.
9857
6 0.
9159
0.
9765
0.
9815
7 0.
8539
0.
9759
0.
9809
8 0.
8319
0.
9801
0.
9851
9 0.
8169
1.
0128
1.
0176
10
0.
7938
0.
9943
0.
9992
11
0.
8150
1.
0112
1.
0161
12
0.
8134
0.
9646
0.
9697
13
0.
8022
0.
9553
0.
9604
14
0.
7923
0.
9471
0.
9522
15
0.
7865
0.
9422
0.
9474
16
0.
7858
0.
9416
0.
9468
17
0.
7870
0.
9427
0.
9479
18
0.
7817
0.
9382
0.
9435
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
S
itin
g
an
d S
i
zi
n
g
o
f
DG for Lo
ss Redu
ction
a
n
d
Vo
ltag
e
Sag
Mitiga
tio
n in RDS
Using
…
(K. S
i
va
Ramud
u
)
82
1
19
0.
7773
0.
9345
0.
9398
20
0.
9469
0.
9580
0.
9791
21
0.
9321
0.
9434
0.
9650
22
0.
8947
0.
9065
0.
9290
23
0.
8730
0.
8852
0.
9082
24
0.
8544
0.
8668
0.
8903
25
0.
8544
0.
8668
0.
8903
26
0.
8469
0.
8594
0.
8832
27
0.
9423
0.
9534
0.
9747
28
0.
9198
0.
9312
0.
9943
29
0.
9093
0.
9209
0.
9964
30
0.
9058
0.
9174
1.
0033
31
0.
9058
0.
9174
1.
0151
32
0.
9044
0.
9160
1.
0020
8.
CO
NCL
USI
O
N
In
t
h
i
s
pa
per,
a si
n
g
l
e
D
G
pl
acem
e
nt
m
e
t
hod
i
s
pr
op
ose
d
t
o
fi
n
d
t
h
e
op
t
i
m
a
l
l
o
cat
i
ons
o
f
DG
an
d
Artificial Bee
Co
lon
y
(ABC) alg
o
rith
m
is p
r
op
o
s
ed
t
o
fin
d
th
e op
tim
a
l
sizes of
DGs fo
r m
a
x
i
m
u
m lo
ss
redu
ction
of
rad
i
al d
i
stri
b
u
tio
n
system
s
is p
r
esen
ted.
Du
e t
o
h
i
gh
e
m
p
l
o
y
m
en
t of vo
ltag
e
sensitiv
e
eq
u
i
p
m
en
ts, vo
ltag
e
sup
port in
sen
s
itiv
e lo
ads are a
great con
cern
for u
tility co
m
p
an
ies. Besi
d
e
network
ope
rat
i
o
n c
o
st
i
s
di
rect
l
y
affe
ct
ed by
p
o
we
r
l
o
ss.
In
t
h
i
s
pa
per a
l
o
ng
an
d
hi
g
h
l
y
l
o
ade
d
20
K
V
fee
d
e
r
h
a
s bee
n
u
n
d
e
r in
v
e
stigatio
n
to
fi
n
d
op
ti
m
a
l lo
catio
n
an
d
sizing
fo
r DG
u
n
its to
su
ppo
rt 30
%
o
f
th
e feed
er lo
ad
. Th
is
DG
pe
net
r
at
i
o
n l
e
vel
i
s
reas
ona
bl
e
d
u
e t
o
eco
n
o
m
i
c consi
d
e
r
at
i
o
n
s
.
T
h
e
pre
s
ence
o
f
t
w
o
hi
ghl
y
vol
t
a
g
e
sen
s
itiv
e lo
ad
s in
th
is feed
er h
a
v
e
m
a
d
e
network
exp
a
n
s
io
n
p
l
ann
e
rs th
ink
of
v
o
ltage sag
m
i
t
i
g
a
ti
o
n
and
v
o
ltag
e
su
ppo
rt in
th
ese
bu
ses. Besid
e
, b
ecause of
ill-con
d
iti
o
n
e
d
n
e
twork, Lo
ss red
u
c
tion
and
v
o
ltag
e
pro
f
ile
im
pro
v
em
ent
have
bee
n
o
f
great
c
o
ncer
n as
wel
l
.
O
b
jec
t
i
v
e
f
unct
i
o
ns wi
t
h
di
ffe
re
nt
app
r
oaches
ha
ve bee
n
defi
ned a
nd
d
i
ffere
nt
scena
r
i
o
s wer
e
i
nve
st
i
g
at
ed. R
e
su
l
t
s
show si
gni
fi
cant
re
duct
i
o
n i
n
p
o
w
er l
o
ss i
n
ad
d
ition
to
voltag
e
p
r
ofile im
p
r
o
v
e
m
e
n
t
. Cap
a
b
ility o
f
p
r
op
o
s
ed
m
e
t
h
od
for
v
o
ltag
e
sag
m
i
t
i
g
a
tio
n
i
n
sen
s
itiv
e
b
u
ses h
a
s b
e
en
also
i
n
v
e
stig
ated
and
was
p
r
ov
ed to
b
e
with
in
accep
t
ab
le lim
i
t
s.
REFERE
NC
ES
[1]
CIGRE WG 37
–23. “Impact of increasing contribution of is
persed generation on power sy
st
em”. Final Repor
t.
September 1998
.
[2]
Dugan RC, Price SK. “
Issues fo
r distributed generations in the US
”. Proc. I
E
EE PES, Winter
Meeting
.
2002;
1
:
121–126.
[3]
M Fotuhi-Firuzabad, A R
a
jab
i
-Ghahnavi
e. “An
Analy
t
ical Meth
od to Consid
er
DG Impacts on Distribution S
y
s
t
em
Reliab
ili
t
y
”. I
E
E
E
/PES Displacem
e
nt and
Distrib
u
tion C
onf
eren
ce &
Exhib
ition
,
Asia and Pacef
ic. 2005;
9:
1-6.
[4]
W
El-Khattam
,
MMA
Salam
a
. “
D
istributed
gen
e
rat
i
on tech- no
logies, d
e
finitio
ns and benefits”.
El
ectr
i
c Powe
r
Sy
ste
m
s Re
se
ar
ch
. 2004; 71: 119
-128.
[5]
E Diaz-Dorado,
J Cidras,
E Mig
u
ez. “A
pplicatio
n of evo
l
ution
a
r
y
algor
ithms for the plann
i
ng of
urban distr
i
butio
n
networks of med
i
um voltag
e
”.
I
E
EE Trans. Pow
e
r Systems
. 2002
; 17(3): 879-
884.
[6]
M Mardaneh, GB Gharehpetian
.
“Siting and sizing
of DG units
using GA and
OPF based tech
nique”.
TENCON.
IEEE Reg
i
on 10
Conference
. 200
4; 3: 331-334.
[7]
Silvestri A Berizzi, S Buonanno.
“Distributed generation plan
ni
ng using genetic algorithms”.
Ele
c
tric
Powe
r
Engineering, Po
wer Tech
Budap
est 99, Inter. Co
nference.
1999:
257.
[8]
R Gnativ, JV Milanovic. “Voltag
e
Sag Propagatio
n In Sy
stems With Embedded Genera
tion And Induction Motors”.
Power Engineering Societ
y
Summer Meeting
,
20
01.
IEEE.
2001
;
1: 474 –
479.
[9]
Angelo Baggini
. Handbook of
Power Qualit
y
,
Fi
rst Edition,
th
e Atrium
, Southern Gate, Chich
e
ster, West Susse
x
PO19 8SQ, England John Wiley
& Sons Ltd
,
20
.
[10]
M
P
a
dm
a Lalit
ha, VC Veera
Redd
y, N Us
ha. “
O
ptim
al
DG placem
ent for
m
a
xim
u
m
los
s
reduct
i
on in rad
i
al
distribution
s
y
stem
”. Internatio
nal journal of
em
erging technologies
&
applications
in
engin
e
ering, technolog
y
&
sc
ie
nce
s
. 2009
;
2(2): 719-723
.
[11]
M
P
a
dm
a Lali
th
a, N S
i
v
a
ram
i
R
e
dd
y, VC Ve
era
Redd
y. “
O
ptim
al DG P
l
ac
em
en
t for M
a
xim
u
m
Los
s
Reduction
in
Radial Distribution S
y
stem
Usin
g ABC algorith
m
”.
Internationa
l Journal o
f
Reviews in Computing
, ISSN : 2076
-
3328. 2010; 3: 4
4
-52.
[12]
M Padma Lalitha, VC Veer
a
Redd
y
,
N Sivar
a
mi Redd
y
.
“Application of fu
zzy
and ABC
algorithm for D
G
placement for
minimum loss
in Radial Distr
i
bution S
y
st
em
”. Ir
ani
a
n J
our
nal of
El
ec
tric
al
& E
l
ec
troni
cs
Engineering. 20
10; 6(4): 248-25
6.
[13]
Dervis Karabog
a and B
a
hriy
e Bastur
k. “
A
rti
f
ici
a
l Be
e Colo
n
y
(ABC) Opti
m
i
zation Algor
i
t
hm
for Solvin
g
Constrained
Optimization
Problems”. Springer-V
erlag
,
IFSA 200
7, LNAI 4529
. 2
007: 789–798.
[14]
Karaboga D and
Basturk B. “On the performance
of artif
icial bee colon
y
(ABC)
algorithm”.
Els
evier Applied Soft
Computing
. 200
7; 8: 687–697.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
IJECE Vol. 3, No. 6, D
ecem
ber 2013
:
814 – 822
82
2
[15]
Hemamalini S and Sishaj P Si
m
on. “Economic load dispatch
with valve-po
int effect using artificial b
ee co
lo
n
y
algorithm”.
xxxii national system
s conferen
ce, NS
C 2008
. 2008
: 1
7
-19.
[16]
Fahad S Abu-Mouti and
ME El-Hawar
y
.
“Optimal Distributed
Ge
neration Allo
cation and Sizin
g
in Distribu
tio
n
S
y
s
t
em
s
via
art
i
f
i
ci
al b
e
e
Colon
y
algorithm
”
.
I
E
EE transactions
o
n
power delivery
. 2011; 26(4).
[17]
S
M
F
a
ras
hbas
h
i-As
taneh, A Da
s
t
fan. “
O
ptim
al
P
l
acem
en
t
and
S
i
zing of DG fo
r Los
s
Reduct
i
o
n
, Voltag
e
P
r
ofi
l
e
Im
provem
e
nt and Voltage Sag
Mitigat
ion”
.
International Conference on
Renew
able Energies and Power Quality
(
I
CREPQ’10)
Granada (
Spain)
. 2010.
BIOGRAP
HI
ES OF
AUTH
ORS
K.
Siva Ramudu
was born in 1989. He receiv
e
d B.Tech
Degree from JNTUA, Anantapur in
the
y
e
ar 2011. At pr
esent persuing
M.Tech
in Annam
achar
y
a
Institu
t
e of Technolog
y
and
Scien
ces,
Rajampet,
Andhra Pradesh
,
India. Area
of
intr
est
distribution
s
y
stems.
M
.
P
a
dma Lalitha
is graduated from JNTU, An
athapur in
El
ect
r
i
ca
l & Elec
troni
cs
Engineer
ing
in the
ye
ar 199
4. Obta
ined Pos
t
gradu
a
te
degr
e
e
in PSOC fro
m
S.V.U, Tirup
a
thi
in th
e
ye
ar
2002. Having 14
y
e
ars of experience in teach
ing in
graduate
and post graduate lev
e
l. She had 10
intern
ation
a
l jou
r
nal public
at
ions and 10 internati
onal and nation
a
l conferen
ces to her credit. She
pulished near
ly 54 papers in various natio
na
l & int
e
rnat
i
onal journa
ls & conferen
ces.
S.V.University
,
Tirupathi award
e
d doctor
a
te
in
the
y
ear 2011 Pr
esently
working as Professor and
HOD of
EEE department in
Ann
a
m
ach
arya In
stitu
te of
Techn
o
l
o
g
y
an
d Scien
ces
,
Rajampet
A
n
d
h
ra Pra
d
es
h, In
dia
. Ar
eas
o
f
int
e
res
t
inclu
d
e rad
i
al
dis
t
ri
bution s
y
s
t
em
s
,
artificial intelligence
in power s
y
stems,
ANN.
P
.
Sur
e
s
h Babu
was born in 1984. He is gradu
a
ted fro
m JNTU, H
y
der
a
bad in
2
006. Receiv
ed
PG degree from S.V.U Colleg
e
of Engineering
,
Tirupathi in
the
y
e
ar 2009
. He h
a
s 6
y
ears of
tea
c
hing exper
i
ence
. P
r
esentl
y working as
Assistant P
r
ofessor in EEE departm
e
nt in
Ann
a
m
ach
arya In
stitu
te of Tech
no
log
y
an
d
Scien
ces
, Raj
a
mpe
t
,
Andhra Prades
h,
India
.
Area
of i
n
trest
Power systems.
Evaluation Warning : The document was created with Spire.PDF for Python.