TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol. 12, No. 8, August 201
4, pp. 5976 ~ 5984
DOI: 10.115
9
1
/telkomni
ka.
v
12i8.612
6
5976
Re
cei
v
ed Ap
ril 23, 2014; Revi
sed
Ju
n
e
3, 2014; Acce
pted Ju
ne 15,
2014
SVC Placement for Voltage Profile Enhancement Using
Self-Adaptive Firefly Algorithm
Sel
v
arasu. R*, Sur
y
a Kalav
a
thi.
M
Department of EEE, JN
T
UH,
Hy
der
abad, India
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: selvaras
unav
een
@gma
il.co
m
A
b
st
r
a
ct
Static VAR Co
mp
ens
ator (SVC) is one of th
e s
hunt type F
A
CT
S devices
for providi
ng r
eactive
pow
er sup
port
in pow
er sys
tems n
e
tw
ork and its p
l
ac
e
m
e
n
t repres
en
ting the l
o
cati
on a
nd si
z
e
h
a
s
signific
ant
infl
u
ence
o
n
e
n
h
a
n
ce
me
nt of v
o
ltage
pr
ofile.
T
h
is pap
er pres
ents
a
firefly
alg
o
rith
m bas
e
d
opti
m
i
z
at
ion
strategy for
pl
ace
m
e
n
t of SV
C i
n
pow
er syst
e
m
s w
i
th a vi
ew
of
mini
mi
z
i
n
g
vo
l
t
age
devi
a
tio
n
at
the lo
ad
bus
es
to en
hanc
e th
e lo
ad
bus v
o
lt
ages. T
h
is
met
hod
uses
a se
l
f
-adaptiv
e sch
e
m
e f
o
r tuni
ng t
h
e
firefly par
am
et
ers. The proposed strategy
is
tested
on t
h
ree IEEE test system
s.
The
obt
ained res
u
lts are
pro
m
isi
ng a
nd
show
the effectiv
en
ess of the prop
osed
appr
oach.
Ke
y
w
ords
: fire
fly algor
ith
m
, SVC, voltage pr
ofile
Copy
right
©
2014 In
stitu
t
e o
f
Ad
van
ced
En
g
i
n
eerin
g and
Scien
ce. All
rig
h
t
s reser
ve
d
.
1. Introduc
tion
In re
cent yea
r
s, the
po
wer system i
s
fa
cing
ne
w cha
llenge
s. The
voltage devia
tion du
e
to co
ntinuo
u
s
lo
ad va
riat
ion a
nd
electric p
o
wer transfe
r limitat
ions were
o
b
se
rved
due
to
rea
c
tive power un
balan
ce
s. Also it ca
use
s
a hig
h
impact on
power
syste
m
se
curity a
n
d
reliability. He
nce
this conti
nuou
sly in
cre
a
sin
g
lo
ad
de
mand
nee
d t
o
be
mo
nitored o
r
ob
serv
ed to
avoid the tra
n
smi
ssi
on lin
es overl
oad
e
d
and poo
r lo
ad bu
s voltage profile. Co
n
s
tru
c
tion of n
e
w
gene
ration f
a
cilitie
s and
transmissio
n
netwo
rk
will
not be an
efficient way to solve th
ese
chall
enge
s. Since it is complicated they in
volve huge in
stallat
i
on co
st, environm
ent imp
a
ct,
political, la
rg
e di
spla
cem
e
nt of po
pulati
on a
nd la
nd acq
u
isitio
n. One of
the si
mplest way
s
for
minimizi
ng the voltage de
viation rather than co
n
s
tru
c
ting ne
w ge
neratio
n syst
ems is throu
g
h
providin
g opti
m
al quantity of reactive
po
wer sup
p
o
r
t
at
approp
riate buses.
The po
we
r electro
n
ics ba
sed FACTS d
e
vi
ce
s, devel
oped by Hin
gora
n
i NG [1
], have
been effe
ctively used fo
r flexible ope
rat
i
on and
co
nt
rol of the power sy
stem through
cont
rolli
ng
their parameters. They have the capability to
control the vari
ous electri
c
al parameters in
transmissio
n netwo
rk i
n
order to a
c
hiev
e better sy
stem perfo
rma
n
ce. FACTS d
e
v
i
c
e
s
can
b
e
divided in
t
o
sh
u
n
t
c
o
n
n
ec
t
e
d
,
se
r
i
e
s
c
o
n
n
e
c
t
ed
an
d
a co
mbina
t
ion
of b
o
th
[
2
]
.
T
h
e
Static V
a
r
Comp
en
sato
r (SVC) a
n
d
Static Synch
r
onou
s
Com
p
ensator
(S
T
A
T
C
OM) b
e
l
o
n
g
to th
e
sh
u
n
t
c
o
nn
ec
t
ed d
e
v
ice and a
r
e i
n
use fo
r a long time. C
o
n
s
equ
e
n
tly
,
they are v
a
ria
b
l
e
sh
u
n
t r
e
a
c
t
o
r
s
,
w
h
i
c
h
i
n
j
e
c
t
o
r
a
b
s
o
r
b
r
e
a
c
t
i
v
e
p
o
w
e
r
i
n
o
r
d
e
r
t
o
c
o
n
t
r
o
l
the
vo
ltage
a
t
a
giv
e
n
bus
[
3
]
.
T
h
y
r
i
s
t
o
r
Controlled S
e
rie
s
Compe
n
sato
r (T
CS
C) a
nd Static Synchro
nou
s Serie
s
Co
mpen
sato
r (S
SSC)
are
seri
es
c
o
n
n
e
c
t
e
d
d
e
v
ic
es for
c
o
n
t
r
o
lling th
e a
c
tiv
e
p
o
w
er i
n
a lin
e b
y
v
a
r
y
i
n
g
th
e lin
e
r
e
a
c
t
a
nce
.
T
h
e
y
a
r
e
in ope
r
at
io
n
at a
f
e
w p
l
aces
but are
st
ill in the st
age o
f
d
e
v
e
lop
m
ent
[
4
]
.
Unified Po
we
r Flo
w
Co
ntro
ller (UPF
C) b
e
long
s to
co
mbination of
shu
n
t
and se
ries devices an
d
is
a
b
le
to
c
o
nt
r
o
l a
c
ti
ve
po
w
e
r
,
r
e
activ
e
po
w
e
r
a
n
d
vo
lta
g
e
m
a
gnitu
d
e
sim
u
lta
n
eo
usly or
s
e
p
a
r
a
t
e
l
y
[5]. These devices can facilita
t
e the
control
of power flow, increase the
power transfer
capability, red
u
ce the generation cost, im
prove t
he security and enh
ance the stabi
lity o
f
the pow
er
sy
st
ems.
I
n
r
e
c
e
n
t
y
e
a
r
s
,
t
h
e
S
V
C
attract
s
the system
en
gine
ers
an
d re
se
arche
r
s
for p
r
ovidin
g
rea
c
tive po
wer
sup
port i
n
power
syste
m
s a
nd
its pl
acem
ent h
a
s signifi
cant i
n
fluence on
b
u
s
voltage
p
r
ofil
e.
The
in
s
t
a
l
l
a
t
i
o
n
o
f
S
V
C
s
c
a
n
b
e
d
e
s
c
r
i
bed as
an optimiza
t
i
on problem with
objectives o
f
simultaneously mini
mizing
the
voltage d
e
via
t
ions and
improving
th
e voltage
pro
f
ile
while satisfy
i
ng sy
st
em const
r
aint
s.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
SVC Placem
ent for Voltag
e Profile Enh
ancem
ent
Usi
ng Self-Adapt
ive Firefly…
(Selva
ra
su. R)
5977
Different n
a
ture in
spi
r
ed
meta-h
euri
s
ti
c alg
o
rithm
s
su
ch a
s
G
e
netic Algo
rith
m (GA),
Simulated a
n
nealin
g (SA),
Ant Colo
ny Optimizati
o
n
(ACO
), Bee
s
Algorithm
s (BA), Different
ial
Evolution (DE), and
Parti
c
le S
w
a
r
m
Optimizatio
n
(PSO)
and
B
a
cteri
a
l fo
rag
i
ng o
p
timizati
on
algorith
m
etc [6] have bee
n appli
ed in
solving
th
e F
A
CTS pla
c
e
m
ent proble
m
s. GA ha
s
bee
n
prop
osed to identify the optimal locatio
n
of mu
lti type FACTS d
e
vice
s in a p
o
we
r sy
stem
to
improve the l
oada
bility [8]. PSO has b
e
en appli
ed to
find the optim
al locatio
n
of FACTS devi
c
es
con
s
id
erin
g
cost of i
n
stalla
tion an
d sy
stem lo
a
dabilit
y [9]. PSO ha
s b
een
propo
sed
to
sele
ct
the
optimal location and pa
ra
meter setting of SVC and TCSC to mitigate small sig
nal oscillation
s i
n
multi ma
chin
e po
we
r
syst
em [10]. Bee
s
Algo
rithm
h
a
s
been
p
r
op
ose
d
to d
e
termine the
opti
m
al
allocation of
FACTS d
e
vice
s fo
r maxi
mizing
t
he
a
v
ailable tran
sfer capabilit
y [11]. Bacte
r
ial
Fora
ging al
g
o
rithm ha
s b
een p
r
op
ose
d
for loss
mi
nimizatio
n
an
d voltage sta
b
ility improve
m
ent
[12] Bacterial
Foragi
ng alg
o
rithm ha
s b
een u
s
ed to
find the optim
al locatio
n
of UPFC devi
c
es
with obje
c
tive
s of minimizi
n
g
the losse
s
and imp
r
ovin
g the voltage profile [13].
Firefly Algorit
hm (FA), which is a n
a
ture-in
s
pi
red m
e
ta-he
u
ri
stic
algorith
m
, ha
s bee
n
sug
g
e
s
ted fo
r solving
opti
m
ization
p
r
ob
lems [6
-7]. It
has be
en
wid
e
ly applie
d in
solving
seve
ral
optimizatio
n probl
em
s, to
name a few: eco
nomi
c
dispatch [14
-
16],
and unit
com
m
itment [17]
etc.
Ho
wever, the
imprope
r ch
oice of FA pa
ramete
rs
affe
cts the conve
r
gen
ce a
nd
may lead to sub
-
optimal soluti
ons. T
here is thus a
nee
d
for devel
o
p
i
ng better
stra
tegies fo
r opt
imally sele
cti
n
g
the FA para
m
eters with
a view of obtaining t
he gl
obal be
st sol
u
tion be
side
s achievin
g b
e
tter
conve
r
ge
nce. Self Adaptiv
e FA (SAFA
)
ba
sed
strate
gies have
be
en p
r
op
osed
to minimi
ze t
he
transmissio
n loss throu
gh p
l
acin
g SVCs [
18] and UPF
C
s [19].
In this
paper, a s
e
lf adaptive firefly al
gorithm
ba
sed st
rategy i
s
p
r
opo
se
d for SVC
placement
wi
th a view of
minimizi
ng v
o
ltage d
e
viations
be
side
s
enha
nci
ng lo
ad bu
s volta
ges.
The
strate
gy identifie
s t
he o
p
timal l
o
catio
n
s an
d
the SVC p
a
ram
e
ters. S
i
mulation
s a
r
e
performed on three IEEE t
e
s
t
s
y
s
t
ems
us
ing
MATLAB s
o
ftware
pack
a
ge and the res
u
lt
s
are
pre
s
ente
d
to demon
strate the effect
iven
ess of the pro
posed ap
pro
a
c
h.
2. Po
w
e
r Flow
Model Of
SVC
The SVC eit
her i
n
je
cts
o
r
ab
so
rb
s
re
active p
o
wer in o
r
de
r to
regul
ate the
voltage
magnitud
e
at
the poi
nt of
con
n
e
c
tion to
the AC
net
work and
its equivalent ci
rcuit
of
va
riab
le
su
sceptan
ce
model is
sho
w
n in Figu
re
1.
The lin
eari
z
e
d
equ
ation
repre
s
e
n
ting t
he total
su
sceptan
ce
s
vc
B
as state va
riabl
e i
s
given by the followin
g
equ
a
t
ion:
00
0
k
k
k
i
i
i
svc
i
svc
P
Q
B
Q
B
(
1
)
At each iterati
on (
k
), the variable shunt su
sceptan
ce,
s
vc
B
is updated.
1
kk
k
s
vc
svc
s
vc
B
BB
(2)
Based o
n
the
equivalent ci
rcuit of
SVC, the curre
n
t drawn by SVC i
s
:
s
vc
svc
i
Ij
B
V
(3)
Rea
c
tive po
wer drawn by SVC, which is al
so rea
c
tive power inje
cted at bus i, Q
SVC
is
:
2
s
vc
i
i
svc
QQ
V
B
(
4
)
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 8, August 2014: 597
6 –
5984
5978
I
sv
c
Bu
s
i
B
sv
c
Figure 1. Vari
able Susce
p
tance Model o
f
SVC
3. Firefly
Algorithm
FA is a recen
t
nature in
spi
r
ed m
e
ta-h
eu
ristic
algo
rith
ms which ha
s be
en d
e
veloped
by
Xin She Yan
g
at Cam
b
rid
ge University in 2007 [6].
The alg
o
rith
m mimics the
flashing
beh
avior
of fireflies. It is si
milar to
other
optimization
alg
o
rith
ms em
ployin
g swa
r
m inte
lligen
ce
su
ch
a
s
PSO. But FA
is found to ha
ve supe
rio
r
p
e
rform
a
n
c
e in
many ca
ses
[7].
3.1. Classica
l Firefly
Algorithm
The n
u
mbe
r
of fireflies i
n
the swa
r
m i
s
kno
w
n
as th
e
popul
ation
si
ze,
N
. The selection
of popul
ation
size de
pen
ds on the
spe
c
if
ic optim
i
z
atio
n problem. T
houg
h, typica
lly a popul
ation
size of 20 to
50 is u
s
e
d
for PSO and FA
for most a
ppl
ication
s
[9, 1
5
]. Each
th
m
firefly is denoted
by a vector
m
x
as
:
12
,,
nd
mm
m
m
x
xx
x
(
5
)
The se
arch
space is limite
d
by
the following in
equalit
y const
r
aints.
()
(
)
vv
v
x
mi
n
x
x
m
a
x
1,
2
,
,
vn
d
(
6
)
Initially, the p
o
sition
s of the fi
reflies are gene
rated fro
m
a uniform distrib
u
tion u
s
ing the follo
win
g
equatio
n:
()
(
)
()
vv
v
v
m
x
x
m
i
n
xm
a
x
xm
i
n
r
a
n
d
(7)
Her
e
,
rand
is a
random
num
b
e
r b
e
twe
en 0
and
1, take
n from
a unif
o
rm di
stri
buti
on.
The initial distribution do
es not significa
ntly affe
ct the
perform
an
ce
of the algorithm. Every time
the algo
rithm
is executed a
nd the optimi
z
ation p
r
o
c
e
s
s sta
r
ts
with a different set of initial poin
t
s.
Ho
wever, in
each ca
se, the algorit
hm
searche
s
for the optimum solu
tion. In the case of multiple
possibl
e set
s
of solutio
n
s,
the pro
p
o
s
e
d
algo
rithm
may conve
r
g
e
on differen
t
solution
s e
a
ch
time. Although each of those solution
s
will be valid
a
s
they all will satisfy the re
quire
ment.
The light inte
nsity of the
th
m
fi
refly,
m
I
is given by:
()
mm
I
Fi
t
n
e
s
s
x
(8)
The attra
c
tiveness bet
wee
n
the
th
m
and
th
n
firefly,
,
mn
is given by:
2
,
m
a
x
,
,
mi
n
,
,
,
mi
n
,
,
()
e
x
p
(
)
mn
m
n
mn
m
m
n
m
n
r
(9)
Whe
r
e
j
i
r
,
is Ca
rtesia
n dista
n
c
e bet
wee
n
i
-th and
j
-th firefly
.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
SVC Placem
ent for Voltag
e Profile Enh
ancem
ent
Usi
ng Self-Adapt
ive Firefly…
(Selva
ra
su. R)
5979
2
,n
1
nd
kk
mm
n
m
n
v
rx
x
x
x
(10
)
The value
of
mi
n
is taken a
s
0.2 and the v
a
lue of
ma
x
is taken
as 1.
is anothe
r
con
s
tant who
s
e value is
related to the
dynamic
ran
ge of the sol
u
tion sp
ace. The po
sition
o
f
firefly is
upd
ated in
ea
ch
iterative
ste
p
. If the lig
h
t
intensity of
th
n
firefly is larger than the
intensity of th
e
th
m
firefly, then the
th
m
firefly moves
towards the
th
n
firefly and its
motion
at the
th
k
iteration is d
enoted by the
following e
q
u
a
tion:
,n
(
k
)
(
k1
)
(
k1
)
(
k1
)
0
.
5
mm
m
n
m
xx
x
x
r
a
n
d
(11)
The inte
nsity
of each firefly
is
comp
ared
with
all
othe
r fireflie
s a
n
d
the po
sition
s of the
fireflies a
r
e u
pdated u
s
ing
(9). After an adeq
uate nu
m
ber of iterat
ions, ea
ch firefly converg
e
s
to
the same p
o
sition in the se
arch sp
ace a
nd the glob
al optimum is a
c
hieve
d
.
3.2. Self Adaptiv
e
Firefl
y
Algorithm
In the ab
ove
narrated
FA, ea
ch firefly of
the swa
r
m
travel a
r
ou
n
d
the p
r
obl
e
m
sp
ace
taking
into a
c
cou
n
t the
re
sults o
b
tained
by
others an
d
still applyin
g
it
s own
ran
domized mov
e
s
as
well. P
e
rf
orma
nce of
the FA
can
b
e
imp
r
ov
ed
by tuning
three p
a
ramete
rs.
The
ra
nd
om
movement fa
ctor (
) i
s
ve
ry effective o
n
the
pe
rform
ance
of FA
wh
ose valu
e is
comm
on
ly
cho
s
e
n
in the
ran
ge 0
and
1. A larg
e value of
makes the
move
ment to explo
r
e the
sol
u
ti
o
n
throug
h the di
stan
ce search spa
c
e a
nd
smalle
r value
of
tends to facilitate lo
cal
sea
r
ch.
The influe
nce
of other
solu
tions i
s
controll
ed by the v
a
lue of attra
c
tiveness of e
quation
(9), which ca
n be adju
s
te
d by modifying two pa
ra
meters
ma
x
and
. In gene
ral th
e value of
ma
x
is cho
s
en
in
the
ran
ge
of 0 to
1
and
t
w
o li
miting
case
s
ca
n b
e
defined:
The
algorith
m
perfo
rms coo
perative l
o
cal
sea
r
ch
with
the bri
ght
e
s
t firefly stron
g
l
y
determini
n
g
othe
r firefli
e
s
positio
ns,
especi
a
lly in its neig
hbo
rho
o
d
, wh
en
ma
x
= 1 and only
no
n
-
co
ope
rative distrib
u
ted
rand
om search with
ma
x
= 0. On the othe
r hand, the value of
dete
r
mine
s the variation o
f
attractivene
ss with i
n
crea
sing
dista
n
ce
from
comm
unicated firef
l
y. Generall
y
the absorp
t
ion
coeffici
ent
is ch
osen in
th
e in th
e rang
e of 0
to
1
0
.
Indeed, th
e
choice of th
ese pa
ram
e
ters
affects th
e fin
a
l solution
an
d the
co
nverg
ence of
th
e al
gorithm. In
thi
s
p
ape
r, the
para
m
eters
,
mi
n
and
are tune
d throug
h a self-ada
ptive mech
ani
sm.
Each fi
refly fo
r a
problem
with
nd
control variables
will be
defined to encompass
nd
+3
deci
s
io
n vari
a
b
les in th
e p
r
opo
sed
form
u
l
ation involvin
g self-ad
aptive techniq
ue.
The a
ddition
a
l
three de
ci
sio
n
variable
s
re
pre
s
ent
m
,
min,
m
and
m
. A firefly is
repres
ented as
:
12
mi
n
,
m
,,
,
,
,
nd
mm
m
m
m
m
xx
x
x
(13)
Each firefly posse
ssing t
he solution v
e
ctor,
m
,
min,
m
and
m
unde
rg
o the
whol
e
sea
r
ch
pro
c
e
ss.
Duri
ng iteration
s
,
the FA prod
uce
s
be
tte
r o
ff-spri
ng
s through Equ
a
tio
n
s. (9
) and
(11)
usin
g the
p
a
ram
e
ters
a
v
ailable i
n
t
he firefly of Equatio
n. (13), th
ere
b
y enh
an
cing
the
conve
r
ge
nce of the algorith
m
.
4. Proposed
Strateg
y
The SVCs a
r
e to b
e
in
stalled at a
ppropriate l
o
cations
with o
p
timal pa
ramet
e
rs tha
t
minimize the
voltage deviations to en
h
ance the loa
d
bus voltag
e profile. Thi
s
pap
er aim
s
to
develop a me
thodolo
g
y that perform
s SVC pla
c
eme
n
t to enhan
ce th
e load bu
s vo
ltage profile.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 8, August 2014: 597
6 –
5984
5980
4.1. Objectiv
e Functio
n
The l
oad
bu
s
voltage
can
b
e
b
r
ou
ght to
the n
o
rm
al val
ue of
1.0
pe
r
unit throug
h t
a
ilorin
g
the obje
c
tive function for
minimizi
ng th
e sum of
de
viations of all
load bu
s voltages from the
nominal volta
ge of 1.0 per
unit. The obje
c
tive function
is formul
ated
as:
Minimize
nload
i
i
V
u
x
1
1
)
,
(
(14)
Whe
r
e,
nload i
s
the n
u
mbe
r
of load
buse
s
.
V
i
is the Voltage magnitu
de
at bus i.
4.2. Problem Cons
traints
4.2.1. Equality
Constraints
The equ
ality con
s
trai
nts a
r
e the load flo
w
equ
ation gi
ven by:
(,
)
Gi
D
i
i
PP
P
V
for PV and
PQ buses
(15
)
(,
)
Gi
Di
i
QQ
Q
V
for PQ buses
(16)
Whe
r
e
Gi
P
and
Gi
Q
repre
s
e
n
t the real an
d re
active power i
n
j
e
cted by the
gene
rato
r at
bus
i
, res
p
ec
tively.
D
i
P
and
D
i
Q
re
p
r
esent the re
al and rea
c
ti
ve powe
r
dra
w
n by the load at
bus
i
, res
p
ec
tively.
4.2.2. Inequa
lit
y
Constrai
nts
Voltage Co
nstraints
mi
n
m
a
x
ii
i
VV
V
for PQ bus
es
(17)
Rea
c
tive Power gen
eration
limit
mi
n
m
a
x
Gi
Gi
Gi
QQ
Q
for PV bus
es
(18
)
Whe
r
e
mi
n
Gi
Q
and
ma
x
Gi
Q
are the uppe
r a
nd lower limit of reactive po
wer
sou
r
ce i.
SVC Con
s
trai
nts
10
0
1
0
0
SV
C
M
VA
R
Q
M
V
A
R
(19)
Whe
r
e,
SV
C
Q
= Rea
c
tive p
o
we
r inje
cted
at SVC place
d
bus in p.u
The firefly of the pro
p
o
s
ed
SVC placeme
n
t proble
m
is
defined a
s
:
1S
V
C
1
m
i
n
,
S
V
C
m
i
n
,
S
V
C
,
N
,
,
m
i
n
,
{
(
L
,
Q
,
,
,
)
.
.
..(
L
,
Q
,
,
,
)
.
.
...
...
..
.(
L
,
Q
,
,
,
)
}
mm
m
m
M
M
m
m
m
N
T
C
S
C
N
N
N
N
x
(20
)
The Self Ad
aptive FA (S
AFA) search
es fo
r
o
p
tim
a
l solution
b
y
maximizin
g
the lig
ht
intensity
m
I
, like
the fitne
s
s f
unctio
n
in
an
y other st
o
c
h
a
stic optimi
z
ation te
chni
q
ues.
The
ligh
t
intensity fun
c
tion can
be o
b
t
ained by tran
sformi
ng the
voltage devia
tion functio
n
and the
volta
ge
c
o
ns
traint into
m
I
function a
s
:
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
SVC Placem
ent for Voltag
e Profile Enh
ancem
ent
Usi
ng Self-Adapt
ive Firefly…
(Selva
ra
su. R)
5981
1
1
I
Max
m
(21)
A populatio
n
of fireflies i
s
ra
ndo
mly gene
rated
a
nd the inten
s
ity of each
firefly is
cal
c
ulate
d
usi
ng (1
8). Based on the lig
ht intensity,
each firefly is moved to the
optimal sol
u
tio
n
throug
h (1
1)
and the iterati
v
e process c
ontinue
s till the algorith
m
converg
e
s.
5. Simulation Resul
t
s an
d Discus
s
io
ns
The effective
ness of the prop
osed SAFA fo
r optimally placing t
he SVC devi
c
e
s
to
minimize the
voltage deviation in
the power
system has been
tested on IEEE-14, IEEE-30 and
IEEE-57 bus tes
t
s
y
s
t
ems
us
ing
MATLAB 7.5. The
line data and bus
data
for the three
tes
t
system
s
are
take
n fro
m
[2
0]. The limit
s
for the
contro
l and
de
pen
d
ant varia
b
le
s
and th
e
cho
s
en
rang
e fo
r
self
ada
ptive pa
rameters
are
given in
Tabl
e 1. Th
e p
o
p
u
lation
si
ze,
N
for all
the te
s
t
system
s is ta
ken a
s
30 a
n
d
the numbe
r of iterations,
K
ma
x
, is consi
dere
d
as 2
00.
IEEE 14 bus
s
y
stem:
The
system
co
mp
rise
s
20 tran
smissi
on li
ne
s, five gene
rato
r bu
se
s
(Bus
No. 1,2,
3,6 and 8
)
an
d nine lo
ad b
u
se
s. Simu
lat
i
ons a
r
e
ca
rri
ed out with di
fferent numb
e
rs
of SVCs an
d
it is found that three SV
Cs a
r
e
suffi
ci
ent to reali
z
e
the satisfa
c
t
o
ry perfo
rma
n
ce.
The results i
n
term
s of the locatio
n
s
a
nd the
SVC
para
m
eters a
r
e given in
T
able 2. Th
e bus
voltages befo
r
e and after
placi
ng thre
e SVCs are pre
s
ente
d
in Table3. It is clear from this tabl
e
that SAFA algorithm ide
n
tifies the optim
al plac
e
m
ent of SVC to enhan
ce the bu
s voltage prof
ile.
The Co
mpa
r
i
s
on of loa
d
b
u
s voltage
s with and wi
tho
u
t SVC place
m
ent is shown in Figure 2.
Table 1. Co
ntrol Varia
b
le
s
Minimum
Max
i
mum
Power
sy
stem
variables
VM (per
unit)
0.95
1.1
SV
C
Q
(MVAR)
-100
100
Self Adaptive
Parameters
0
0.5
0.2
1
0
1
Table 2. Opti
mal Location,
Paramete
r of SVC
for IEEE 14- Bus
Sys
t
em
Number
of SVC
Locations
(Line
N
u
mb
e
r)
Q
(MVAR)
3
17
19
15
-17.947
-32.052
-50.00
Table 3. Bus
Voltages of I
EEE 14- Bus
Sys
t
em
Bus No
Bus Voltages
Without SVC
With SVC
1 1.060
1.060
2 1.040
1.040
3 1.005
1.005
4 0.984
0.999
5 1.000
1.001
6 1.065
1.065
7 0.998
1.007
8 1.085
1.085
9 1.002
0.998
10 1.005
1.000
11 1.031
1.018
12 1.004
1.000
13 1.025
1.013
14 0.998
0.999
IEEE 30 bus s
y
stem:
The system ha
s 41 tran
smi
ssi
on line
s
a
nd six gene
rator bu
se
s
(Bus
No. 1, 2
,
5, 8, 11 an
d 13). T
he si
mulation
stud
y is perfo
rme
d
with six SV
Cs, a
s
they can
prod
uce ad
eq
uate pe
rform
ance for 3
0
b
u
s te
st syst
e
m
. The resul
t
s in term
s of
the location
s
and
the SVC pa
ra
meters are gi
ven in Table
4. The bu
s voltage
s befo
r
e and afte
r pl
acin
g thre
e SVCs
are presente
d
in Table 5. It is seen fro
m
this
table that the identified
placeme
n
t
of SVC enh
ance
the bus volta
ge profile. Th
e Com
pari
s
o
n
of load
bu
s
voltages
with and with
out SVC pla
c
eme
n
t
is
s
h
ow
n
in
F
i
gu
r
e
3
.
IEEE 57 bus
s
y
stem:
The
system
ha
s 8
0
tran
smi
ssi
o
n
line
s
an
d
seven ge
nerat
or b
u
ses
(Bus
No. 1, 2
,
3, 6, 8, 9 a
nd 12
). The
simulati
o
n
re
sults in te
rm
s of the locati
ons a
nd the
SVC
para
m
eters a
r
e p
r
e
s
ente
d
in Table
-
6.Th
e bu
s vo
ltage
s
befo
r
e and after
pla
c
ing
three
SV
Cs a
r
e
pre
s
ente
d
in
Table 7. It is
see
n
from thi
s
tabl
e th
at the identified
placem
ent of S
V
C enh
an
ce t
h
e
bus voltag
e profile. The
Comp
ari
s
o
n
of load bu
s voltage
s with
and with
out SVC placem
ent is
s
h
ow
n
in
F
i
gu
r
e
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 8, August 2014: 597
6 –
5984
5982
Figure 2. Co
mpari
s
o
n
of Load Bus Volt
age
Magnitud
e
s o
f
IEEE 14 Bus System
Table 4. Opti
mal Location,
Paramete
r of
SVC for IEEE
30- Bus
Sys
t
em
Number of SV
C
Locations
(Line
N
u
mb
e
r)
Q
(MVAR)
6
26
33
24
14
18
19
11.619
7.931
7.442
-40.237
-37.021
-11.632
Table 5. Bus
Voltages of I
EEE 30- Bus
System
Bus
No
Bus Voltages
Bus
No
Bus Voltages
Without
SVC
wi
t
h
SVC
Without
SVC
wi
t
h
SVC
1 1.060
1.060
16
1.033
1.003
2 1.043
1.043
17
1.023
1.001
3 1.021
1.013
18
1.016
0.997
4 1.012
1.003
19
1.011
0.999
5 1.010
1.010
20
1.014
1.001
6 1.012
1.004
21
1.014
0.998
7 1.013
1.008
22
1.015
1.000
8 1.010
1.010
23
1.017
0.998
9 1.042
1.008
24
1.009
0.999
10 1.026
1.008
25 1.010
1.002
11 1.082
1.082
26 0.993
0.996
12 1.052
1.013
27 1.020
1.012
13 1.073
1.073
28 1.010
1.002
14 1.036
1.001
29 1.000
0.998
15 1.030
1.000
30 0.989
0.997
Figure 3. Co
mpari
s
o
n
of Load Bus Volt
age
Magnitud
e
s o
f
IEEE 30 Bus System
Table 6.Opti
mal Location,
Paramete
r of SVC for IEEE 57- Bus System
Number of SV
C
Locations
(Line Numbe
r
)
Q
(MVAR)
7
26
33
24
14
18
19
11.619
7.931
7.442
-40.237
-37.021
-11.632
Figure 4. Compari
s
on of Load Bus Volt
age Magnitudes of IEEE 57 Bus System
0.
96
0.
97
0.
98
0.
99
1
1.
01
1.
02
1.
03
1.
04
1.
05
4
5
7
9
10
11
12
13
14
Bus Voltage
Magnitudes
Load Buses
Wi
t
h
out
SVC
Wi
t
h
SVC
0.96
0.97
0.98
0.99
1
1.01
1.02
1.03
1.04
1.05
1.06
3
6
9
1
21
51
7
1
92
1
2
32
52
7
2
9
Bus Voltage
Magnitudes
Load Buses
Without SVC
With SVC
0.
92
0.
94
0.
96
0.
98
1
1.
02
1.
04
4
1
01
41
7
2
02
32
6
2
93
23
5
3
84
14
44
7
5
05
35
6
Bus Voltage
Magnitudes
Load Buses
Wi
t
h
out
SVC
Wi
t
h
SVC
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
SVC Placem
ent for Voltag
e Profile Enh
ancem
ent
Usi
ng Self-Adapt
ive Firefly…
(Selva
ra
su. R)
5983
Table 7. Bus
Voltages of I
EEE 57- Bus
System
Bus
No
Bus Voltages
Bus
No
Bus Voltages
Without
SVC
wi
t
h
SVC
Without
SVC
wi
t
h
SVC
1 1.040
1.040
30
0.960
0.982
2 1.010
1.010
31
0.934
0.967
3 0.985
0.985
32
0.948
0.997
4 0.981
0.981
33
0.946
0.995
5 0.976
0.976
34
0.957
1.025
6 0.980
0.980
35
0.964
1.008
7 0.978
0.983
36
0.974
1.002
8 1.005
1.005
37
0.983
1.006
9 0.982
0.982
38
1.011
1.000
10 1.001
1.001
39
0.981
1.004
11 0.975
0.995
40
0.971
0.998
12 1.015
1.015
41
0.997
1.000
13 0.979
0.986
42
0.966
0.972
14 0.970
0.976
43
1.010
1.008
15 0.988
0.997
44
1.015
1.005
16 1.014
1.014
45
1.035
1.027
17 1.018
1.018
46
1.032
1.032
18 1.000
0.999
47
1.026
0.999
19 0.970
0.985
48
1.034
1.001
20 0.963
0.986
49
1.019
1.020
21 1.006
0.997
50
1.046
1.012
22 1.008
0.998
51
0.976
1.028
23 1.006
0.995
52
0.968
0.989
24 0.995
0.998
53
0.968
0.979
25 0.980
0.997
54
0.968
1.001
26 0.956
0.979
55
1.033
1.023
27 0.976
0.988
56
0.968
0.987
28 0.990
1.005
57
0.964
0.975
29 1.003
1.019
It is very cle
a
r
from th
e ab
ove discu
s
sio
n
s
that the
propo
sed SAF
A is able
to redu
ce to
the lo
ss to t
he lo
we
st p
o
ssi
ble
by opti
m
ally
pla
c
ing
and
dete
r
mi
ning th
e p
a
rameters
of S
V
C
whe
n
co
mpa
r
ed to othe
r
optimizatio
n algorith
m
s.
In addition th
e self ada
ptive nature of
the
algorith
m
av
oids
re
peate
d
ru
ns for fi
xing the o
p
timal FA p
a
ra
meters by a
trial a
nd e
r
ro
r
pro
c
ed
ure an
d provide
s
th
e best
po
ssibl
e
para
m
eters values.
6. Conclusio
n
In this p
ape
r
a ne
w SAFA
has bee
n p
r
o
posed to i
d
e
n
tify the optimal lo
cation
s of SVC
and th
eir
parameter with
a view
of mi
nimizin
g
the
voltage devia
tions to
enh
a
n
ce
the lo
ad
bu
s
voltage profile. Simulation
s re
sult
s in te
rms
of lo
catio
n
s, SVC p
a
ra
meters an
d the bu
s voltag
es
have been presented for t
h
ree I
EEE test systems. It has
been found
that the i
dentified location
and SVC pa
ramete
rs by
the SAFA
are
able to
enha
nce the
bus voltage
profile
and
the
develop
ed al
gorithm i
s
sui
t
able for pract
i
cal ap
plicatio
ns.
Referen
ces
[1]
Hing
o
ra
ni
NG, G
y
u
g
y
i I.
Und
e
rstand
in
g
FACT
S: Concepts
an
d T
e
chn
o
lo
g
y
of Fle
x
ib
le
AC
transmissio
n
s
y
stems. Ne
w
Y
o
rk: IEEE Press. 2000.
[2
]
Ma
th
ur R
M
, V
a
rm
a
R
K
.
T
h
y
r
i
s
tor
-
ba
sed F
A
C
T
S C
ont
ro
l
l
e
r
s f
o
r Electrica
l
T
r
a
n
sm
i
s
si
on
s
y
stem.
P
i
scata
w
a
y
:
IEEE P
r
ess. 2002.
[3]
Am
briz-
p
e
r
e
z
H,
A
c
h
a
and E,
F
u
e
r
te
-Esqu
i
ve
l C
R
. A
d
va
nce
d
SV
C mod
e
l
s
f
o
r Ne
w
t
o
n
Ra
p
h
so
n lo
a
d
flo
w
a
n
d Ne
w
t
o
n
opt
i
m
al
po
w
e
r flo
w
stu
d
ie
s.
IEEE T
r
a
n
s. o
n
P
o
w
e
r Sy
st
.
20
00; 1
5
:
1
2
9
-
1
3
6
.
[4]
L
a
rs
e
n
EV, Cl
a
r
k
K, Misk
e SA
, Ur
banek J. Cha
r
a
c
terist
ic a
n
d
ra
ti
ng
c
ons
i
d
e
r
a
t
io
ns
of t
h
y
r
isto
r
contr
o
lled se
ri
e
s
com
p
e
n
s
a
tio
n
.
IEEE T
r
a
n
sac
t
ion on Power Delivery
. 1
994
; 9: 9
9
2
-
1000.
[5] Gy
ugy
i
L
.
Un
if
i
ed p
o
w
e
r fl
ow
control
e
r conce
p
t f
o
r
flexib
le
A
C
t
r
a
n
s
m
issio
n
s
y
ste
m
.
IEE
proce
edi
ng
s
.
199
2; 13
9(
4): 323
-
3
3
1
.
[6]
Yang
XS. Natu
re-Inspir
ed Met
a
-He
u
ristic Alg
o
rithms. 2nd e
d
. Beckingto
n
: Lun
iver Press. 201
0.
[7] Yang
X
S
.
F
i
re
fly al
gorith
m
s
for multi
m
o
d
a
l
opti
m
i
z
a
t
io
n
. Stochastic al
g
o
rithms:
F
o
u
n
datio
ns
a
nd
app
licati
on. SA
GA 2009, LNC
S
, Berlin, Germa
n
y
: S
p
rin
ger
-Verla
g. 200
9; 579
2.69-
178.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 8, August 2014: 597
6 –
5984
5984
[8]
Gerbe
x
S, Che
r
kaom R, Germond AJ. Opti
mal loc
a
tio
n
of multi t
y
pe F
A
C
T
S devices in
po
w
e
r s
y
stems
b
y
me
ans of g
enetic a
l
g
o
rith
ms.
IEEE
Transaction on Power System
s
. 20
01; 16(3): 5
37-
544.
[9]
Sarava
na
n M, Sloch
ana
l MR,
Venkates
h P
Abrah
a
m JPS. Applic
atio
n of
par
ticl
e s
w
arm
optimizati
o
n
techni
qu
e for
optima
l
l
o
cat
i
on
of F
A
CT
S devic
es
co
n
s
ideri
n
g
cost
of insta
llati
on
an
d s
y
stem
loa
dab
ilit
y.
Ele
c
trical Pow
e
r System Res
earc
h
.
2007;
7
7
: 27
6-28
3.
[10]
Mond
al
D, Ch
akrab
a
rti A,
Se
ngu
pta A. Opti
mal p
l
ac
ement
an
d p
a
ramet
e
r setting
of SV
C an
d T
C
SC
usin
g PSO to mitigate sma
l
l sign
al sta
b
il
ity pr
obl
em.
Internati
ona
l Jo
u
r
nal of El
ectri
c
al Pow
e
r &
Energy Syste
m
s
. 2012; 42(
1): 334
–3
40.
[11]
Mohami
d
Idr
i
s
R, Kha
i
ru
ddi
n A, Mustafa
MW
. Optimal
all
o
cati
on
of F
A
CT
S devices for AT
C
enh
anc
ement usin
g
B
ees al
g
o
rithm.
W
o
r
l
d
Acade
my
of sc
ienc
e E
ngi
ne
ering
an
d T
e
c
h
n
o
lo
gy.
2
009
;
30: 313-
32
0.
[12]
T
r
ipath
y
M, Mi
shra S. Bacter
i
a
forag
i
n
g
b
a
s
ed so
luti
on to
optimiz
e bot
h r
eal
po
w
e
r l
o
ss
and v
o
lta
ge
stabilit
y lim
it.
IEEE Trans. on Power Syst.
2007
;
22(1): 2
40-
248.
[13]
Senthi
l Kum
a
r
M, Ren
uga
P.
A
ppl
icatio
n
of
UPF
C
for
enh
a
n
ceme
nt of v
o
lt
age
profi
l
e
an
d
minim
i
zatio
n
of losses
usin
g fast volt
a
ge stabilit
y ind
e
x
(FVSI).
Archives of Electric
al
Engi
ne
erin
g J
ourn
a
l.
2
012
;
61(2): 23
9-2
5
0
.
[14]
Apostol
o
p
oul
o
s
T
,
Vlachos A
.
Appl
icatio
n
of the F
i
r
e
fl
y a
l
g
o
rithm for s
o
lv
i
ng th
e ec
on
om
ic emiss
i
o
n
s
loa
d
disp
atch p
r
obl
em.
Interna
t
iona
l Journ
a
l o
f
Comb
in
atoric
s.
2011; 20
11(
523
80
6).23.
[15]
T
aher Nikn
a
m, Raso
ul
Azizi
p
ana
h-Ab
argh
o
oee,
and
Alir
e
z
a Ro
osta. R
e
serve C
onstra
i
ned
D
y
n
a
mic
Econom
ic Disp
a
tch: A Ne
w
F
a
st Self-A
dapti
v
e Modifie
d
F
i
refl
y
Al
gorit
hm.
IEEE System
Journal.
201
2;
6(4).
[16]
Yang
XS, H
o
s
s
eini
SS, Ga
n
domi
AH. F
i
ref
l
y al
gor
ithm fo
r solvi
n
g
no
n-
conve
x
eco
n
o
m
ic dis
patch
prob
lems
w
i
th
valve l
oad
in
g e
ffect.
Applied S
o
ft Computin
g.
201
2; 12(3): 18
0- 118
6.
[17]
Cha
ndras
ekar
an K, Simon
SP.
Net
w
ork
a
nd reli
ab
ilit
y c
onstrai
ned u
n
it
commitment prob
lem usi
ng
bin
a
r
y
rea
l
co
d
ed firefl
y
al
gori
t
hm.
Internatio
nal J
our
nal
of
Electrical
Pow
e
r & En
ergy S
ystems
. 20
12
;
43(1): 92
1–
932
.
[18]
Selvar
asu
R, S
u
r
y
a K
a
lav
a
thi
M,
Christo
ber
Asir Ra
ja
n C.
SVC p
l
acem
en
t for volta
g
e
co
nstrain
ed
los
s
minimiz
a
tio
n
u
s
ing Se
lf Adapt
ive F
i
refl
y
alg
o
r
i
thm.
Journa
l o
f
Archives of Electrical En
gin
e
e
rin
g
.
2013;
62(4): 64
9-6
6
1
.
[19]
Selvar
asu R,
Sur
y
a Ka
lav
a
thi M. Opti
mal pl
acem
en
t of UPF
C
for volta
ge co
nstrain
ed l
o
ss
minimiz
a
tio
n
u
s
ing se
lf ada
p
t
ive F
i
refl
y
alg
o
rithm.
Interna
t
iona
l Revi
ew
of Electrical E
ngi
neer
in
g.
201
3; 8: 1296-
130
1.
[20]
Po
w
e
r S
y
stems
T
e
st Case, T
he Universit
y
of
W
a
shin
gton Archive,
[Onli
ne].Avail
abl
e
:
http://
w
w
w
.
e
e
.w
a
s
hi
ngto
n
.ed
u
/researc
h
/pstca/, 2000.
Evaluation Warning : The document was created with Spire.PDF for Python.