TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol. 14, No. 3, June 20
15, pp. 402 ~ 4
0
9
DOI: 10.115
9
1
/telkomni
ka.
v
14i3.790
5
402
Re
cei
v
ed Fe
brua
ry 6, 201
5; Revi
se
d April 26, 201
5; Acce
pted Ma
y 18, 201
5
Voltage Regulation in a Microgid with Hybrid PV/Wind
Energy
J.O. Petinrin*
1,2
, J.O. Agbolade
1
, Moha
med Shaaba
n
2
1
Electrical/El
e
c
t
ronic Eng
i
n
eer
ing D
epartme
n
t, School of En
gin
eeri
ng, F
e
d
e
ral Po
l
y
t
e
chn
i
c Ede,
Ede, Osun State, Niger
ia
2
Center of Elec
trical Ener
g
y
S
y
stems (CEES
), F
a
cult
y
of El
ectrical En
gin
e
e
rin
g
/Institute of F
u
ture Ener
g
y
,
Univers
i
ti T
e
knolo
g
i Mal
a
ysia,
Mala
ysi
a
.
*Corres
p
o
ndi
n
g
authror, e-m
a
il:
w
o
l
e
p
e
t0
1
@
hotmai
l
.com, jopeti
n
rin
2
@l
i
v
e.utm.m
y
A
b
st
r
a
ct
Autonomous operati
on of a m
i
cr
ogrid system
hinges on t
he efficient combination of
various
ener
gy resourc
e
s to mai
n
tai
n
self-sustain
abi
lity of ener
gy
supp
ly. F
u
rthermor
e
, it is equ
ally i
m
p
o
rtant t
o
coord
i
nate th
e resourc
e
s to regul
ate the
micr
ogri
d
vo
lta
ge p
r
ofile. T
he pro
b
l
e
m
bec
o
m
es
mor
e
co
mp
licat
e
d
if these res
our
ces hav
e i
n
ter
m
ittent
c
haract
e
ristics suc
h
a
s
solar PV
an
d
w
i
nd turbi
nes.
T
he pot
entia
l
for
usin
g en
ergy s
t
orage
pro
m
is
e
to have a
ma
j
o
r impact
o
n
sche
m
es for vo
l
t
age co
ntrol i
n
a micr
ogri
d
. A
hybri
d
Particle
Sw
arm Optimi
z
a
ti
on/Gravitati
ona
l Sear
ch Al
gorith
m
(PSOGSA) based a
ppro
a
ch is use
d
i
n
this pa
per to
c
onte
m
p
l
ate th
e
opti
m
u
m
s
i
z
e
and
loc
a
tio
n
of ener
gy stor
age to
red
u
ce
voltag
e vari
ati
ons
and
fee
der
los
s
es ca
use
d
by
PV/w
i
nd
ener
gy i
n
tegr
ated
i
n
a
micro
g
ri
d. Effectiven
ess
of th
e pr
op
os
e
d
meth
od
is i
m
pl
emente
d
throu
gh a q
uas
i-static time
seq
u
e
n
c
e an
alysis
ov
er a 24-
ho
urly
simulati
on
peri
o
d
on a
u
ton
o
m
ou
s Microgri
d
sys
tem. T
he c
o
rre
spon
din
g
vo
lt
a
ge pr
ofile
is a
n
a
ly
z
e
d
un
der
d
i
fferent op
erati
n
g
cond
itions,
w
i
th h
i
gh
p
e
n
e
tration
lev
e
l
of
PV/w
i
nd e
nerg
y
. T
e
st results
show
th
at th
e e
nergy
stor
ag
e
causes r
e
d
u
cti
on i
n
syste
m
losses
an
d e
n
hanc
es syst
e
m
c
apa
bi
lity to
mai
n
tai
n
vo
l
t
ages w
i
thi
n
t
h
e
per
missi
bl
e li
mits.
Ke
y
w
ords
: en
ergy storag
e, micr
ogri
d
, PSOGSA,
solar PV, voltage pr
ofil
e, w
i
nd turbine
Copy
right
©
2015 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
Smart g
r
id
s (SGs) will
hav
e a fu
ndam
e
n
tal rol
e
in
tra
n
sforming
tod
a
y’s p
o
wer
grids. Th
e
obje
c
tive is to add
re
ss g
r
owin
g dem
an
d; rene
wa
ble
s
, intermitten
t, and distri
b
u
ted gen
eration
(DG); and
en
vironme
n
tal concern
s
. Microgrid
s a
r
e a
key element in
this tran
sformation (S
ch
mitt,
Kumar
et al. 2013
). A Microgrid
(M
G) i
s
a co
ntiguo
us se
ction
of
th
e
grid whi
c
h co
nsi
s
ts
of one or
multiple
DG units
(in ele
c
trical clo
s
en
ess
to
one an
other) ca
pabl
e of operat
ing
either in p
a
ra
llel
with, or autonomous from
, a power
utility grid,
wh
ile providing rel
i
able
power t
o
multiple loads
and
co
nsume
r
s. A
MG
can
be
conn
ecte
d to/or di
sco
nne
ct from
th
e g
r
id to
e
n
a
b
le o
p
e
r
ation
in
both grid
-con
necte
d mode
or autono
mo
us mod
e
(Sh
ahide
hpou
r a
nd Khodaya
r
2013
). It should
be al
so
ca
p
able of
ridin
g
thro
ugh
b
e
twee
n the t
w
o m
ode
s if
ne
ce
ssary.
A MG
can
be
strategi
cally
placed at an
y site in
a po
wer
sy
st
e
m
,
most
e
s
pe
cia
lly
at the grid
system fo
r g
r
id
reinfo
rcement
,
thereby def
erri
ng or eli
m
inati
ng th
e
need
for
syst
em up
grade
s and
imp
r
ovi
n
g
system reliabi
lity, integrity,
and efficie
n
cy
.
A hybrid sy
stem is an in
tegral p
a
rt i
n
the co
mpo
s
ition of mo
dern
day mi
cro
g
ri
ds
integrated i
n
the
utility grid. In
weak gri
d
s,
the
hybrid PV/wi
nd sy
stem i
s
better than the
indep
ende
nt use of
PV
or wind ene
rgy, sin
c
e
it s
upp
resse
s
rapid
chang
es i
n
the
output p
o
we
r of
the ind
epe
nd
ent sou
r
ce
(Petinrin
and
Shaa
ban
2
013).
Howev
e
r, vari
able
nature
of
so
lar
irra
dian
ce a
n
d
wind
spe
e
d
, resulting i
n
intermi
ttent
output ene
rgy, could lea
d
to voltage rise,
particula
rly when PV/wind
gene
ration i
s
high an
d dem
and is lo
w. Schroed
er in (Schroed
er 20
11)
pre
s
ente
d
th
at ene
rgy (E
S) and
de
m
and
re
spo
n
se (DR) a
r
e
essential
gri
d
technolo
g
i
e
s in
operation of
utility grid by
avoidi
ng
ca
p
a
city sh
orta
g
e
s. Fo
r in
sta
n
ce, the
r
e
ca
n be d
e
ferm
e
n
t o
f
grid
rei
n
force
m
ent at
so
m
e
voltage
lev
e
l with
out
af
f
e
ct
ing
sy
st
e
m
sec
u
rit
y
be
cau
s
e t
he sy
st
em
voltage will
st
ill be kept wit
h
in the p
e
rmi
ssi
ble volt
ag
e
boun
ds. In a
ddition, the e
ffect of DR will
be stro
nge
r with more flexible dema
nd u
s
ing el
ectri
c
vehicl
e (Pou
dineh an
d Jam
a
sb 2
014
).
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Voltage Re
gu
lation in a Microgid
with Hy
brid PV/Wind
Energ
y
(J.O. Petinrin)
403
Depl
oyment of ES throughout the grid from
gen
eration to e
nd-u
s
e
r
s pre
s
ent
s an
oppo
rtunity to transce
nd the real
-time
power
bal
an
ce para
d
igm b
e
twee
n su
ppl
y and deman
d.
ES wa
s repo
rtedly used
with sm
all
sola
r
arrays to
eve
n
out th
e po
wer flo
w
a
s
clo
uds pa
ss ove
r
,
and in an
cilla
ry servi
c
e pro
v
ision, su
ch a
s
frequ
en
cy regulatio
n (Pe
t
inrin and Sh
aaba
n 201
3).
Variou
s ap
proache
s have
been d
e
velop
ed in lit
eratu
r
e to solve th
e pro
b
lem of
voltage
regul
ation in
a grid
syste
m
usin
g ES. A matrix
real
-co
ded g
eneti
c
a
l
gorithm
(GA) techni
que
was
prop
osed to
optimally coordi
nate th
e po
wer
pr
o
ductio
n
of DGs a
nd ES
to minimize
the
operational
costs of a
MG
(Ch
en,
Dua
n
et al., 201
1).
A neu
ral n
e
t
w
ork
wa
s
use
d
to forecast t
he
energy outpu
t of the PV source
s,
and
battery ene
rg
y storag
e wa
s mod
e
led a
ggre
gately. T
he
pape
r, howe
v
er, assume
s identi
c
al b
a
tteries b
a
se
d on the size of the PV
system. In other
words, the
size and lo
catio
n
of the ES were not
studie
d
.
A heu
risti
c
to
ol u
s
ing
the
GA with
si
m
u
lated
ann
ea
ling
wa
s d
e
scrib
e
d
in
(Crosslan
d
,
Jon
e
s et al.
2014
), to locate di
stri
b
u
ted ES in low voltage
(LV) n
e
two
r
ks. Monte Ca
rlo
simulatio
n
s were
utilize
d
to
ra
ndomly
sit
e
the PV
syst
ems at full
po
wer an
d the
highe
st volta
ge
is d
e
termi
ned
. The
heu
risti
c
tool
i
s
the
n
applie
d to fin
d
out
the
storage
nee
ded.
Non
e
thele
s
s,
a
singl
e time step load flo
w
is perfo
rme
d
to evaluate th
e worst-ca
se
scena
rio.
Alam (Alam,
Muttaqi et al., 2012), p
r
opo
se
d di
stributed en
erg
y
storag
e (DES) to
mitigate voltage ri
se p
r
obl
em ca
used b
y
solar PV
in
a distrib
u
tio
n
system
(DS). The meth
od
employed
en
able
DES to
absorb
excess e
nergy at
n
oon
day to mi
tigate the
rev
e
rse p
o
wer fl
ow
as a
re
sult of
high PV outp
u
t. The sto
r
e
d
ene
rgy i
s
th
en di
scha
rge
d
to supp
ort t
he voltage i
n
the
evening
pea
k peri
od. Th
e
method
co
nsi
dere
d
a
n
ov
e
r
-simplified
charg
e
/disch
arge ES cy
cle
with
the assu
mpti
on of n
o
voltage p
r
o
b
lem
except n
oon
day and
eve
n
ing p
e
a
k
pe
riod.
Ho
weve
r, a
preventive
co
ntrol frame
work i
s
requi
re
d to
c
ontin
uo
usly m
onitor
curre
n
t an
d
voltage at th
e
PCC to e
s
tab
lish a real
-time equivale
nt circuit of
the
DS. This
will guide a
gain
s
t
any cha
nge
s in
load
pattern,
mo
st e
s
pe
cially duri
ng
holiday
s,
or else
ap
plicati
on
of su
ch
model might
be
detrime
ntal for the DS.
Optimal
sizin
g
and
siting
o
f
the ES is ne
ce
ssary to im
prove the
voltage p
r
ofile in
the DS
and re
du
ce losse
s
. Different method
s have been e
m
ployed in the literature for optimal si
zin
g
and
siting of
ES/DG to mitigate the
probl
em
s as
so
ciated
with
uncertaintie
s
of re
ne
wa
ble
gene
ration (RG
)
. Gravita
t
ional sea
r
ch
algorit
hm (GSA) and p
a
rticle
swarm optimizati
on-
gravitational
sea
r
ch alg
o
ri
thm (PSOGS
A) are
u
s
e
d
to determin
e
multiple DG cap
a
city a
nd
locatio
n
in
DS in (Kha
n, Gho
s
h et al.,
2013
) a
nd (Tan, Hassa
n
et al., 2013
) re
spe
c
tively. An
OPF-b
a
sed a
l
gorithm for
siting the aggregated
cap
a
city of ES was develope
d to decrea
s
e t
he
wind
ene
rgy
cu
rtailment
and
co
st o
f
energy
su
pply in
(Atwa an
d El-Sa
adany 2
010
). A
coo
r
din
a
ted
control of
DES system
s with
LTC for voltag
e
ri
se mitig
a
tion
und
er
high
PV
penetration i
s
pro
p
o
s
ed i
n
(Liu, Aichho
rn et al., 201
2
)
. Ho
weve
r,
none
of the
reviewe
r
s abo
ve
employed the
hybrid PSO
GSA for their sea
r
ch tech
ni
que on e
nerg
y
storage.
This
pap
er
pre
s
ent
s a
comp
re
hen
si
ve archite
c
ture th
at do
not only t
a
ke i
n
to
con
s
id
eratio
n
the coordina
tion of
hybri
d
PV/wind en
e
r
gy, but al
so
manag
es sto
r
age fa
cilities
in
an hou
rly op
eration fa
shi
o
n. This give
s the MG ope
rator
option
s
in sele
cting
a
ppro
p
ri
ate an
d
effective voltage
control m
easure
s
. Hy
brid PSO
G
SA is empl
oye
d
for
sizin
g
a
nd lo
cation
of ES.
The impo
rtan
ce of ES in voltage regulat
ion in a
MG
with high
pen
etration of P
V
/wind en
erg
y
is
demon
strated
.
2. Problem Formulation
Multifaceted
oppo
rtunitie
s
are p
r
ovided
with
deploym
ent of ES in
a MG for sig
n
ificant
benefits to M
G
system, el
ectri
c
ity suppl
y, utility cu
sto
m
ers, ancill
ary servic
e
s
, a
nd integration
o
f
rene
wa
ble
s
(Ro
b
e
r
ts a
n
d
Sand
berg
2011
). Th
eir
key fun
c
tion in
mo
dern
MG
s i
s
to
cou
n
terb
alan
ce th
e i
n
term
ittency introd
uce
d
by
inte
grating
vari
a
b
le
ren
e
wa
bl
es
at the
poi
nt of
comm
on co
u
p
ling (PCC). Such effect
of the ES is
likely to minimize the overall lo
sses
and
improve th
e
system volta
ge p
r
ofile. Th
erefo
r
e,
the
probl
em of fi
nding th
e opt
imal lo
cation
and
size of the E
S
can b
e
ca
st as a math
e
m
atical
p
r
og
ramming p
r
o
b
l
e
m, whe
r
e th
e obje
c
tive is to
minimize h
o
u
r
ly en
ergy l
o
sses an
d volt
age
devia
tion
s a
c
ro
ss all
netwo
rk no
d
e
s. T
h
is can
be
expre
s
sed m
a
thematically as:
N
it
L
l
N
it
ref
t
i
v
t
P
w
V
V
w
F
t
i
1
24
1
1
2
24
1
,
,
Min
(
1
)
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 14, No. 3, June 20
15 : 402 – 40
9
404
Whe
r
e
w
v
and
w
l
a
r
e th
e
weig
hted-co
e
f
ficients of vol
t
age an
d lo
ss minimization
re
spe
c
tively
.
V
i,t
and
V
ref
are
the voltage o
f
bus
i
at time
t
and m
agnitud
e
of voltage refere
nce
re
spe
c
tively
,
obtaine
d from
the powe
r
flow
.
V
ref
is con
s
idere
d
unity in this pap
er
.
P
L,t
is the system loss at time
t
.
∆
t
is 1 hour time interval and
N
i
s
the numbe
r of bu
se
s.
The first term of (1), co
rresp
ond
s to the
sq
uared n
o
rm of voltag
e deviation
s over the
given peri
od
of time, whereas the
se
co
nd term r
epre
s
ent
s the total system lo
sses at time t.
The total power losse
s
du
e
to the ES is
given as (S
ha
aban a
nd Petinrin 20
13
):
N
i
Q
P
P
Q
Q
Q
P
P
P
N
j
t
j
t
i
t
j
t
i
t
ij
t
j
t
i
t
j
t
i
t
ij
t
i
L
1
,
,
,
,
,
,
,
,
,
,
,
(
2
)
Whe
r
e,
t
j
t
i
t
j
t
i
ij
ij
t
j
t
i
t
j
t
i
ij
ij
V
V
r
V
V
r
,
,
,
,
,
,
,
,
sin
,
cos
, a
nd
ij
ij
ij
x
r
Z
is the ijth
element of [Zbus] matrix.
P
i,t
=
P
Gi,t
P
ES
i,t
P
Di,t
,
Q
i,t
=
Q
Gi,t
Q
ESi,t
Q
Di,t
.
P
i,t
and
Q
i,t
are the n
e
t acti
ve and
rea
c
ti
ve power i
n
je
ction at th
e b
u
s
i
at
time
t
,
P
Gi,t
and
Q
Gi,t
are the
active and
re
active po
we
rs gen
erat
ed from the PV/wind e
ner
gy system
s at time
t
.
P
ESi,t
and
Q
ESi
,t
are
the
acti
ve and
re
acti
ve po
wer cha
r
ged/di
scha
rg
ed by th
e ES
at bu
s
i
at ti
me
t
, while
P
Di,t
and
Q
Di,t
are th
e load a
c
tive and re
active
powers at bu
s
i
at time
t
res
p
ec
tively.
The co
nst
r
ain
t
s inclu
de po
wer flo
w
equ
ality constrain
t
s rep
r
e
s
ente
d
as:
N
j
t
j
t
i
j
i
t
j
t
i
j
i
t
j
t
i
t
i
B
G
V
V
P
1
,
,
,
,
,
,
,
,
,
sin
cos
(
3
)
N
j
t
j
t
i
j
i
t
j
t
i
j
i
t
j
t
i
t
i
B
G
V
V
Q
1
,
,
,
,
,
,
,
,
,
cos
sin
(
4
)
W
h
er
e
V
i,t
is t
he voltag
e at
bus
i
at
time
t
,
G
i,j
and
B
i,j
are the
co
ndu
ct
ance a
nd
su
scepta
n
ce of t
he
line between
buses
i
a
nd
j
res
p
ec
tively, whereas
δ
i,t
i
s
the voltage a
ngle at bu
s
i
at time
t
.
a) Voltage
limits:
ma
x
,
min
V
V
V
t
i
(
5
)
b)
Storage p
h
ysi
c
al an
d ope
ra
ting limits:
0
max
24
1
,
i
i
t
i
ES
ES
i
t
Es
E
E
t
P
(
6
)
0
24
1
,
,
t
i
t
i
ES
i
t
ES
E
t
P
(
7
)
max
min
,
i
t
i
i
ES
ES
ES
E
E
E
(
8
)
max
min
,
i
t
i
i
ES
ES
ES
P
P
P
(
9
)
c)
Powe
r loss constraint:
ES
without
t
i
L
ES
with
t
i
L
P
P
,
,
(
1
0
)
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Voltage Re
gu
lation in a Microgid
with Hy
brid PV/Wind
Energ
y
(J.O. Petinrin)
405
Whe
r
e
i
s
th
e en
ergy
storage
rou
nd-t
r
i
p
efficie
n
cy,
E
ES
is the
ene
rgy sto
r
e
d
in t
he ES, and
E
0
ES
i
is the initial
e
nergy
stored
at bus
i
,
whe
r
e the ES is l
o
cated.
E
mi
n
and E
ma
x
are the minimu
m a
nd
maximum en
ergy ca
pa
city of the storag
e respe
c
tively.
Equation
(6
) and
(7) de
note the
ma
ximum an
d
minimum
am
ount of
the
energy
absorb
ed o
r
injecte
d
from
the ES resp
ectively. Similarly, (8) a
n
d
(9) a
r
e the
maximum an
d
minimum ES capa
city with the releva
nt active
power rating. Equati
on (10)
guarantee
s that
integratio
n of the ES improves the network-wide lo
sses. The form
ulation from (1) to (10
)
gives a
compl
e
te de
scriptio
n for th
e modelin
g o
f
energy
sto
r
age requi
red
to mitigate the impa
ct of the
hybrid PV/win
d
gene
ration
on voltage
de
viation and n
e
twork lo
sses.
d) Weig
hting
Fa
ctor
A com
posite
obje
c
tive function is
molde
d
as
the
weig
hted
sum of t
he obj
ective
s
to avoid
multiobje
c
tive programmin
g
an
d
pro
d
u
c
e an
e
quivale
nt sin
g
le
-obje
c
tive optimi
z
a
t
ion p
r
oble
m
.
A
weig
ht for an
objective is
dire
ctly prop
o
r
tional
to the
pen
cha
n
t wei
ghted facto
r
allocated to that
spe
c
ific o
b
je
ctive. Thus, se
cula
rizi
ng an
obje
c
tive
vector into one compo
s
ite obj
ective functio
n
cha
nge
s the
multi-obj
ectiv
e
optimisatio
n pro
b
le
m int
o
one o
b
jecti
v
e optimizati
on problem
(Deb
2001
). Whe
n
su
ch a comp
osite obj
ectiv
e
function i
s
optimize
d
, in most ca
se
s it is po
ssi
ble to get
one pa
rticul
a
r
trade
-off solution. Each
objective
fu
nction i
s
mul
t
iplied by scalar coefficie
n
ts
calle
d weig
hting facto
r
s. Th
e weig
ht
ing factors a
r
e u
s
ually norm
a
lized as:
1
1
K
k
k
W
(
1
1
)
Therefore, w
v
+ w
l
= 1.
The
co
nverg
ence
crite
r
ion
of the
maxi
mum n
u
mb
er of g
ene
ratio
n
is
che
c
ked
after t
h
e
fitness of ea
ch individual in
a populatio
n is
evaluate
d
by the followi
ng fitness fun
c
tion.
T
t
N
i
L
L
T
t
N
i
ref
t
i
v
t
i
P
w
V
V
w
Fitness
11
11
2
,
,
(
1
2
)
3. Test Resu
lts
The p
r
op
ose
d
metho
d
is tested
on
aut
onomo
u
s microg
rid
system
of an a
c
tual
5MVA,
115 kV/4.16
kV 50-Hz wh
ere bu
s 15
0 is used as
p
o
i
nt of commo
n cou
p
ling to
the utility grid.
The total load
is distrib
u
ted
among com
m
erci
al an
d resid
ential en
ergy co
nsum
ers.
The M
G
a
s
sho
w
n i
n
Fi
gure
1, con
s
ists of th
ree
-
pha
se ove
r
h
ead o
r
u
nde
rgroun
d
prima
r
y feed
ers an
d do
u
b
le-p
ha
se
or singl
e-
pha
se line
se
ctio
ns
nea
r the
end of th
e f
eede
r
lateral
s
. The
MG ha
s 91 l
o
ads
of differe
nt types, incl
u
d
ing
con
s
tant
curre
n
t, con
s
tant impeda
n
c
e
and
con
s
tant
power. T
h
e
voltage at
bus
450, lin
e 99 i
s
mo
n
i
tored
on h
o
u
rly ba
si
s. T
hat
particula
r bus is sele
cted d
ue its high vol
t
age se
nsitiv
ity. It is a point on t
he feede
r that respon
ds
quickly to any chan
ge
s in system co
nditions.
The wei
ghtin
g factors we
re determi
ned
as w
v
= 0.55 and w
l
= 0.4
5
for voltage deviation
and po
we
r lo
sses respe
c
tively, after numero
u
s
simul
a
tion studi
es.
Lowe
r
emph
asi
s
wa
s give
n to
the ene
rgy lo
ss,
as
co
mp
ared
with t
h
e
voltage,
du
e
to its d
epe
n
den
cy on vol
t
age deviatio
n
s.
These
weig
hting fa
ctors i
m
prove
the
o
v
erall
system
perf
o
rma
n
ce
. The
pro
p
o
s
ed meth
od
h
a
s
been
implem
ented i
n
MAT
L
AB, and
exa
m
ined
on th
e
MG
system f
o
r 2
4
h
o
u
r
s u
s
ing
qu
asi
-
st
atic
time sequ
en
ce analysi
s
.
The
size
an
d location of
the ES are
found u
s
in
g
the PSOGS
A-ba
sed
opti
m
ization
approa
ch, a
s
listed in T
a
ble 1. It is e
v
ident from t
he Tabl
e tha
t, the propo
sed ap
pro
a
ch
is
cap
able of e
s
timating the storage
size at
a single lo
ca
tion or multipl
e
locatio
n
s
wi
th comp
ara
b
le
siz
e
s.
Table 1. Re
sults for the si
zing a
nd lo
ca
tion of ES
Centralized Ener
g
y
Stor
age (
C
ES)
Distributed Ener
g
y
Stor
age (
D
ES)
Bus
Rating (MW)
Bus
Rating (MW)
81 4.473
81
70
78
43
2.264
1.035
0.485
0.693
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 14, No. 3, June 20
15 : 402 – 40
9
406
Figure 1. Autonomo
u
s mi
crog
rid bu
s fee
der
a) Ca
se I Hyb
r
id Solar PV/Wind fo
r Voltage Regulati
on in a Micro
g
rid
Due to va
riab
le sun
s
hi
ne h
ours a
s
re
ga
rds to sola
r PV, and relativ
e
ly fickle
cut-i
n
wind
spe
e
d
s
, sola
r PV or wind t
u
rbin
es m
a
y not pro
d
u
c
e
usa
b
le en
erg
y
for con
s
ide
r
able
portio
n
of
time duri
ng t
he yea
r
. Each of the
RG
s is integ
r
ate
d
into the fee
der i
ndep
end
ently and p
o
w
er
flow si
mulatio
n
is
ca
rri
ed o
u
t. Finally, the hybri
d
PV/wind
ene
rgy
is integ
r
ate
d
into the fee
d
e
r
and po
we
r flow is
carrie
d
out to determine the vo
ltage profile of the MG. There is
high
-p
owe
r
blockin
g
dio
d
e
between
th
e sol
a
r PV
a
nd wi
nd tu
rb
i
ne to p
r
event
bi-di
r
e
c
tional
cu
rre
nt flow
in
the hybrid. Hybridi
z
atio
n
of solar PV with wind
therefo
r
e form a very reliable RG in
this
scenarios, wi
nd can provi
de energy in both day and
night (where there i
s
av
ailability of wind
blowi
ng
) an
d
sola
r PV a
c
t
s
a
s
b
a
ck u
p
ene
rgy du
rin
g
pea
k
hou
r
of the day
(Habee
bullah
Sait
and A
r
ul
Da
n
i
el 20
11).
Wi
nd
spe
ed
ge
nerally t
end
s
to incre
a
se in
the
evening
durin
g the
sa
me
time that sol
a
r PV begi
n
s
to de
crea
se. Thre
e scenari
o
s
are
con
s
id
ere
d
i
n
the propo
sed
approa
ch are
1) Standalon
e PV generati
on; 2). St
andalone
Wind g
eneration an
d
3) Hybrid
sol
a
r
PV/wind gen
eration.
A 30% PV penetratio
n
is
distrib
u
ted in
a modified p
eak lo
ad fee
der of 10 M
W
with the
wind turbine
isolated. Th
e output voltage of the
MG at different hour of t
he day with
the
integratio
n of
sola
r PV is
shown in Fi
gure 2.
The
dotted line, fo
r P
V
, sho
w
s
a m
a
ximum volta
ge
magnitud
e
of 1.01 pu at the 13.00 hou
r
of t
he day and a minimum
voltage magn
itude of 0.95 pu
with co
rrespo
nding lo
sse
s
of 1.28 MW a
s
sh
own in Table 2.
A 30% wind
penetration
without PV is also di
st
rib
u
te
d in the feede
r. The outp
u
t voltage
with the
wind
turbine
conn
ected i
nde
pe
ndently to
the
MG is
sh
own
with da
sh
ed
line in Fig
u
re
2.
Its maximum
voltage is 1.01 pu at the 5.00 hou
r of the day and
the minimum
is 0.97 pu a
nd
corre
s
p
ondin
g
lo
sse
s i
s
1
.
59 MW. It i
s
cle
a
r th
at b
o
th PV and
wind
co
uld
n
o
t maintain
the
voltage at 1.0
pu at every
hour
of the d
a
y, albei
t still within the a
c
cepta
b
le voltage bo
und
ary of
0.95 pu to 1.05 pu. The voltage inje
ction of the so
l
a
r PV appea
rs to be ze
ro in the night as far
as the h
o
u
r
ly voltage profile is con
c
e
r
ne
d; in
com
pari
s
on
with the
one
s re
sultin
g from the
wi
nd
turbine g
ene
rator. The wi
n
d
appe
ars to exhibit more
boun
ded ex
cursi
o
n
s
thro
u
gh the day.
In the third scen
ario, a total of 15% PV
and 15% wind hyb
r
id togethe
r as
shown in
Figure 1 i
s
di
stribute
d
in
th
e feed
er. Th
e
output volt
ag
e of the
MG
a
t
different h
o
u
r of th
e d
a
y, a
t
bus
450,
with
the co
mbinat
ion of the hyb
r
id
sola
r
PV/wind tu
rbin
e i
s
sho
w
n
solid
line in Fig
u
re
2.
The soli
d line
sho
w
s a ma
ximum voltage magnitud
e
of 0.99 pu at the 13.
00 ho
ur of the day and
0.96 pu a
s
minimum voltag
e and co
rresp
ondin
g
losse
s
is 1.05 MW. There is app
recia
b
le sy
ste
m
loss
redu
ctio
n an
d mini
mu
m voltage
de
viation in th
e
hybrid PV/wi
nd tu
rbine
a
s
co
mpa
r
ed
to
the
indep
ende
nt use of
both solar
PV
a
nd wind
tu
rbi
ne.
This
und
erscores the im
provem
ent of
the
voltage p
r
ofil
e offere
d by t
he hyb
r
id
system ov
er th
e
indep
ende
nt
use
of PV an
d win
d
en
ergi
es
.
Secu
rity of su
pply is di
spla
yed a
s
the
wi
nd g
ene
ratio
n
offer
su
pply
wh
en th
e
sol
a
r PV
coul
d n
o
t
and the solar
PV acts as a
back-up whe
r
e
the wind g
e
neratio
n drop
s in the day.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Voltage Re
gu
lation in a Microgid
with Hy
brid PV/Wind
Energ
y
(J.O. Petinrin)
407
Figure 2. Hyb
r
id PV/wind voltage outp
u
t
In
these sce
nario
s, wind
ca
n
provide
energy in
bot
h day
and
ni
ght (while
wi
nd bl
ows)
and sola
r
PV
panel syste
m
s can provi
de
ba
ck-up
energy for p
eak
hou
rs of
the day
when
applia
nce
s
a
r
e
on
a
nd
work i
s
i
n
p
r
og
re
ss in
offices (Ha
b
e
ebulla
h Sait
and A
r
ul
Da
n
i
el
2011
). Althou
gh, the t
w
o
RGs
ca
n b
e
m
ade to
op
erat
e in
100%
re
dund
an
cy; while the
solar
PV
is mad
e
to op
erate d
u
rin
g
the day, the wind turbi
ne is
made to op
erate in the nig
h
t. It is evident
in Figu
re 2 th
at the appli
c
a
t
ion of a hyb
r
i
d
sy
ste
m
offe
rs
app
re
ciabl
e voltage reg
u
lation, re
du
ces
(if not elimin
ate) the cost
of ES system and t
he st
orag
e capa
city of ES system, comp
are
d
to
stand
alon
e
wind tu
rbine
o
r
PV sy
stem
(Katiraei
and
Iravani
200
6,
Katiraei, Iravani et
al. 2
0
0
8
).
The hybrid
system p
r
ovides the be
n
e
fits of
peak load shavin
g, mitigation of peak-vall
e
y
differen
c
e, im
prove
s
the voltage profile
quality and
offers a
c
tive power adj
ust
m
ent ca
pa
city for
the MG (Qin
g
,
Nanhu
a et al. 2013) which ES could h
a
ve provide
d
if employed.
The
po
wer d
e
livered
by
e
a
ch
of th
e
RG is p
r
e
s
ent
ed in
Fig
u
re
3. Thi
s
hi
ghli
ghts th
e
fickle
cha
r
a
c
t
e
risti
cs
of the rene
wabl
e e
nergy
sy
stem
; either sol
a
r
or win
d
sy
ste
m
. Noneth
e
le
ss,
Figure 3
suggest
s
that the vari
ability of sol
a
r PV
is
somewhat
less than t
hat of the
wind
gene
ration; a
l
beit drop
pin
g
to zero at night. While t
he deg
ree
of variation is
site-d
epe
nde
nt,
s
h
or
t-
ter
m
fluc
tu
a
t
io
ns
ar
e
s
m
a
ller
in
the c
a
s
e
of PV. Therefore th
e
total losse
s
a
nd voltage
s a
r
e
slightly better than the win
d
gene
ration
as de
picte
d
in Table 2.
Figure 3. PV
and wi
nd ge
n
e
ration p
o
wer injection
Table 2. Voltage ra
nge
s a
nd incurred lo
sses
RG
Min pu
Voltage
Max pu
Voltage
Voltage
deviation
Losses,
MW
%
Losses
Reduction
Base case
0.9466
0.9466
-
2.182
-
PV 0.9466
1.0100
0.0634
1.278
41.43
Wind 0.9684
1.0089
0.0405
1.586
27.31
PV/Wind 0.9640
0.9929
0.0289
1.053
51.74
0
5
10
15
20
25
0.
9
4
0.
9
5
0.
9
6
0.
9
7
0.
9
8
0.
9
9
1
1.
0
1
1.
0
2
H
our
Vo
l
t
ag
e (
p
u
)
PV
WT
PVW
T
0
5
10
15
20
25
0
500
1
000
1
500
2
000
2
500
3
000
3
500
4
000
4
500
H
our
Po
w
e
r
(
k
W
)
PV
WT
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 14, No. 3, June 20
15 : 402 – 40
9
408
The above
result
s und
erscore the imp
r
oveme
n
t of the voltage control offered
by th
e
hybrid
syste
m
over the i
ndep
ende
nt
use
of PV
an
d win
d
en
erg
i
es. Th
e coordination
between
RG
s can ma
ke the
be
st use
of the la
tter in
the lo
ng-time
peri
o
d and
ca
n contend
with t
he
inherent inte
rmittency of e
a
ch
of them
acting
alo
ne.
This
ca
n, in t
u
rn, e
nha
nce
the reliability
of
sup
p
ly with minimum op
era
t
ion co
sts.
b) Ca
se II Energy Storag
e for Voltage S
uppo
rt in Hyb
r
id PV/Wind
Energy Syste
m
In this sce
n
a
r
io, both the
sola
r PV and
the win
d
ge
nerato
r
d
e
livered th
eir full
cap
a
city
(30% pen
etra
tion each
)
du
e to load increase in t
he MG from 10M
W to 14MW. It is expected
that
bus voltage
magnitud
e
in
the fee
d
e
r
i
s
maintain
ed
within
acce
ptable limit
of
0.95 p
u
a
nd
1.05
pu. Figu
re
4 i
llustrate
d the
effect of the
hybrid
PV/wi
nd
system
on
the voltage
p
r
ofile of th
e
MG.
Here, the in
d
epen
dent u
s
es of
sola
r P
V
and
wi
nd t
u
rbin
e en
erg
y
have the v
o
ltage b
e
lo
w th
e
accepta
b
le
v
o
ltage boun
d
s
.
Except at 11th
to 14t
h
hour of the
d
a
y for the
PV (dotted
line
)
and
5th to 7th an
d 21
st hou
r o
f
the day for
wind
ene
rgy (dashed
-dotte
d-da
sh
ed lin
e
)
. The voltag
es
are
within
th
e range
of
0.918
pu
– 0.9
61 p
u
fo
r
sol
a
r PV
and
0.
935
pu
-
0.9
60 p
u
fo
r
wi
nd
energy. Non
e
t
heless, the
hybrid PV/wi
nd turbi
ne
(solid line)
ma
nage
d to increase the voltage
(0.939
7 pu
–
0.981 p
u
)
an
d nea
rly keep
s it within
the
accepta
b
le b
ound
ary. The
r
e i
s
ap
pre
c
ia
ble
voltage in
cre
a
se
in the
h
y
brid PV/win
d
turbi
ne
as
comp
ared to
the ind
epe
n
dent u
s
e
of
both
sola
r PV and wind turbine.
Ho
wever, 9 h
ours are
still belo
w
the sta
t
utory limits in the night.
A total energ
y
storag
e (ES
)
of 4.473M
W as
dete
r
min
e
d
by PSOGSA base
d
opti
m
ization
is di
stribute
d
in the MG at
their respe
c
tive opt
imal lo
cation
s. The
ES systems
are divid
ed i
n
to
two
sectio
ns becau
se
the maximum ch
argin
g
an
d
di
scharging
rat
e
is 6
hou
rs
each. Section
‘A’
comp
ri
se
s only the ES o
n
bus 81 of
2.264M
W ca
pacity while
se
ction ‘B’ compri
se
s of ES
integrate
d
at buses 7
0
, 78
and 43 with
total size
of
2.214M
W. Th
e integratio
n of the ES is to
inject p
o
wer
into the MG
as voltag
e
sup
port a
nd
absorb
po
we
r du
ring
high
gene
ration f
o
r
voltage levelling.
Figure 4. Energy storage fo
r voltage su
p
port
There is a
re
markabl
e imp
r
oveme
n
t in t
he volt
age
profile as comp
ared
with
the
hybrid
PV/wind e
nergy. The volta
ge i
s
m
a
intai
ned
within
0.
962
pu
and
0
.
967 p
u
with
voltage d
e
via
t
ion
of 0.005. T
h
i
s
ca
se e
m
ph
asi
z
e
s
the
rol
e
of ES a
s
a
reme
dy to th
e voltage
dep
ressio
n p
r
obl
em
in a MG, whe
n
the coo
r
din
a
tion of other control d
e
vices fell sho
r
t in resto
r
ing th
e voltage within
its prescribed
bou
nd
s. The
integration
o
f
ES in
the fe
eder was abl
e to a
b
so
rb
a
nd inje
ct p
o
wer
into the bu
ses a
s
d
eem
fit, thereby demon
st
rated
the ben
efits of pea
k loa
d
sh
aving a
n
d
mitigation of peak-valley
difference. This ha
s
effectively assi
sts to harne
ss inte
rmitte
nt
rene
wa
ble e
nergy resource
s, redu
ce
d
energy
lo
ss,
improve the
voltage profi
l
e and bri
ng
the
voltage within
statutory limit.
3. Conclusio
n
Accel
e
rated i
n
stallatio
n
of
variable
ren
e
w
abl
e gen
era
t
ion co
uple
d
with the intro
duction
of
the sma
r
t grid, have cre
a
ted
a
n
i
n
cre
a
se
d
inte
re
st
in mi
cro
g
ri
ds.
Thi
s
p
ape
r
h
a
s
develo
ped
a
frame
w
ork fo
r voltage reg
u
lation in aut
onomo
u
s mi
crog
rid
s
that is ca
pable to
operate u
n
d
e
r
5
10
15
20
25
0.
91
0.
92
0.
93
0.
94
0.
95
0.
96
0.
97
0.
98
0.
99
H
our
V
o
l
t
age
(
pu)
P
V
onl
y
W
i
nd onl
y
PV
/
W
i
n
d
P
V
/
W
i
nd +
E
S
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Voltage Re
gu
lation in a Microgid
with Hy
brid PV/Wind
Energ
y
(J.O. Petinrin)
409
wide
ra
nge
of
ope
ration
m
ode
s a
nd
con
d
itions. Sol
a
r PV and
win
d
turbin
e
size
and l
o
cations in
the mi
cro
g
ri
d
we
re
p
r
e
s
el
ected,
ho
wev
e
r, the
si
ze
and l
o
cation
of the
en
ergy sto
r
ag
e
was
determi
ned
u
s
ing PSO
GS
A optimizatio
n app
ro
ach. Simulation
studie
s
were carri
ed o
u
t on
a
microgri
d
system to te
st
the imp
a
ct
of vari
o
u
s i
ndividual
an
d varia
b
le
rene
wable
en
ergy
(sol
ar/
w
ind
)
combinatio
n. The hybri
d
solar
PV/wind
gene
ration p
r
ovided mo
re
effective voltage
regul
ation to
the microg
rid
system a
s
compa
r
ed
with
each of the
sola
r PV/win
d
turbin
e acti
ng
alone. Furth
e
r
more, when
the
voltage
variation
fe
ll b
e
yond th
e
ca
pabilitie
s of t
he hyb
r
id
sy
stem,
the coo
r
din
a
tion of the hybrid PV/wind
energy sy
stem with ene
rgy storag
e, a feature of the
sma
r
t microg
rid, we
re a
p
t to bring th
e
voltage ba
ck withi
n
stat
utory limits.T
h
is imp
r
ove
s
the
voltage p
r
ofil
e quality a
n
d
offers active
power
adj
u
s
tment ca
pa
city to the DS.
The efficacy
of
real
-time p
r
ici
ng (RTP
) de
mand
re
spo
n
s
e to
ol in
sh
a
p
ing lo
ad
de
mand i
s
su
gg
ested
for fu
rt
her
studie
s
which
will not only greatly minim
i
ze
s the
pea
k load, but also the load de
mand vari
atio
n).
Referen
ces
[1]
Alam M, K Mut
t
aqi, D S
u
tanto
.
Distribute
d
e
n
e
rgy stora
ge fo
r miti
gat
i
on of
voltag
e-rise i
m
pact
caus
e
d
by rooftop sol
a
r PV
. Po
w
e
r a
n
d
Energ
y
Soc
i
e
t
y
Gener
al Me
eting. 20
12: 1-
8.
[2]
At
w
a
YM, E El-Saadany
.
Optimal
allocation
of ESS in dist
ribution s
y
stems
w
i
t
h
a high
penetration of
w
i
nd ener
g
y
.
I
EEE Transactions on Power
System
s.
20
10
; 25(4): 181
5-1
822.
[3]
Che
n
C, S Duan, T
Cai, B Liu, G Hu. Smar
t energ
y
mana
geme
n
t s
y
stem for opt
imal micro
g
ri
d
econ
omic op
er
ation.
R
e
n
e
w
ab
l
e
Po
we
r Ge
ne
ra
ti
on
.
201
1; IET
5
(3): 258-26
7.
[4]
Crossla
nd
AF
, D Jon
e
s, NS
W
ade. Pl
ann
in
g the
loc
a
ti
on
a
nd rati
ng
of d
i
s
t
ributed
en
erg
y
storag
e in
L
V
net
w
o
rks
usi
n
g
a g
e
n
e
tic a
l
g
o
r
ithm
w
i
th
simu
lated
an
ne
ali
n
g
.
Internatio
na
l Journ
a
l of
Elec
trical
Pow
e
r
& Energy Systems
. 2
014; 5
9
(
0
): 103-1
10.
[5]
Deb K. Multi-o
b
jectiv
e optim
i
z
ation us
in
g ev
oluti
onar
y a
l
g
o
r
it
hms. John W
ile
y & Sons. 2
0
01.
[6]
Hab
eeb
ull
ah
Sait H, S Aru
l
Dan
i
el. N
e
w
contro
l
par
adi
gm for inte
gra
t
ion of
photov
oltaic
en
er
g
y
sources w
i
th
utilit
y
n
e
t
w
ork.
In
ternatio
nal
Jou
r
nal
of Electric
al Pow
e
r & E
n
ergy Syste
m
s
. 201
1;
33(
1):
86-9
3
.
[7]
Katirae
i
F
,
M Iravani. P
o
w
e
r man
a
g
e
me
nt strategi
es f
o
r a m
i
crogr
i
d
w
i
th
multi
p
l
e
distri
bute
d
gen
eratio
n un
its.
IEEE
Transactions on Power System
s.
200
6; 21(4): 18
21-
183
1.
[8]
Katiraei F, R Iravani, N Hatziarg
yri
ou, A
Dimeas. Micr
ogri
d
s man
a
g
e
ment.
Pow
e
r
and En
ergy
Maga
z
i
ne.
2008; IEEE6(3): 54-65.
[9]
Khan N, S Gh
osh, S Ghosh
a
l,
Binar
y Gra
v
itation
a
l Se
ar
ch base
d
Al
go
rithm for Optimum Siting
an
d
Sizin
g
of DG a
nd Sh
unt C
a
p
a
citors
i
n
Ra
di
al Distri
buti
on
S
y
stems.
En
er
gy an
d Pow
e
r
Engi
neer
in
g
.
201
3; 5: 1005.
[10]
Liu
X, A Aic
h
h
o
rn, L
Li
u, H
L
i
. Co
ordi
nate
d
Cont
ro
l of
Dist
r
ibute
d
En
erg
y
Storag
e S
y
ste
m
W
i
th T
a
p
Cha
nger T
r
ans
formers for Vol
t
age R
i
se Miti
g
a
tion
Und
e
r Hi
gh Ph
otovolt
a
i
c
Penetrati
on.
Sm
art Grid,
IEEE Transactions
. 201
2; 3(2)
: 897-90
6.
[11]
Petinri
n
JO, M Sha
aba
n.
A
h
y
brid s
o
l
a
r PV/
w
i
nd e
nergy
sy
stem for
volt
ag
e re
gul
atio
n i
n
a
micro
g
rid
.
201
3 IEEE Student Co
nfere
n
c
e on Res
earc
h
and D
e
ve
lop
m
ent (SCORe
D). 2013: 54
5-
549
[12]
Petinri
n
JO, M
Shaa
ba
n.
Implementati
on of Energ
y
Stor
age in a Future Smart Grid.
Australia
n Jour
n
a
l
of Basic and A
ppli
ed Sci
enc
e
s
. 2013; 7(4): 2
73-2
79.
[13]
Poud
ine
h
R, T Jamasb. Distri
buted ge
nerati
on,
storag
e,
d
e
man
d
resp
on
se an
d en
erg
y
efficienc
y
a
s
altern
atives to grid ca
pacit
y e
nha
nceme
n
t.
Energy Po
licy
. 2
014; 67: 2
22-2
31.
[14]
Qing Z
,
Y Nan
hua, Z
Xia
opi
n
g
, Y You, D Li
u. Opti
mal sitin
g
& sizin
g
of b
a
tter
y
e
ner
g
y
s
t
orage s
y
ste
m
in active
distrib
u
tion n
e
t
w
ork.
Innovativ
e S
m
art Grid T
e
chn
o
lo
gies E
u
rop
e
(ISGT
EUROPE).
2013: 1-
5.
[15]
Rob
e
rts BP, C Sand
ber
g.
T
he rol
e
of e
nerg
y
storage
in
d
e
vel
o
p
m
e
n
t of
smart gr
ids
. Pr
ocee
din
g
s o
f
the IEEE. 2011
; 99(6): 113
9-1
144.
[16]
Schmitt L, J Kumar, D Sun,
S
Ka
yal, S Ve
nkata. Ecocit
y
upo
n a Hi
ll:
Mi
crogrids
an
d th
e F
u
ture of the
Eu
ro
pe
an
C
i
t
y
.
Pow
e
r and En
ergy Mag
a
z
i
n
e
,
IEEE
. 2013; 11(4): 59-7
0
.
[17]
Schroe
der A.
Mode
lin
g st
ora
ge a
nd
dem
an
d man
a
g
e
ment
in p
o
w
e
r d
i
stributio
n gr
ids.
A
ppli
ed Ener
gy
.
201
1; 88(1
2
): 4700-
471
2.
[18]
Shaa
ba
n M, J
O
Petinri
n
. Siz
i
ng
an
d siti
ng
o
f
di
strib
u
ted
ge
nerati
on
in
dist
ributi
on s
y
stem
s for vo
ltag
e
improvem
ent a
nd loss re
ducti
on.
Internati
o
n
a
l Jour
nal of S
m
art Grid a
nd
Clea
n
Ener
gy
.
2013; 2(3):
350-
356.
[19]
Shah
ide
h
p
our
M, M Khod
a
y
ar. Cuttin
g
Camp
us
Ener
g
y
Costs
w
i
th
Hierarc
hica
l
Contro
l: T
he
Econom
ical a
n
d
Reli
ab
le Ope
r
ation of a Micr
ogri
d
.
Electrific
ation Ma
ga
z
i
n
e
.
2013; 1(1): 40
-56.
[20]
T
an W
S
, MY
Hassa
n, HA R
ahma
n
, MP Abdu
lla
h,
F
Hu
ssin. Multi-d
i
stribute
d
ge
ner
a
t
ion p
l
an
ni
n
g
usin
g h
y
br
id p
a
rticle s
w
a
rm
optimis
ation-
gr
avitatio
nal s
e
a
r
ch alg
o
rithm i
n
clu
d
in
g volt
a
ge rise
issu
e.
IET
Generation
,
T
r
ansmissi
on
& Distributi
o
n
. 201
3; 7(9): 929
-942.
Evaluation Warning : The document was created with Spire.PDF for Python.