Int
ern
at
i
onal
Journ
al of Ele
ctrical
an
d
C
om
put
er
En
gin
eeri
ng
(IJ
E
C
E)
Vo
l.
9
, No
.
6
,
Decem
ber
201
9
, p
p.
4516~
4523
IS
S
N: 20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v9
i
6
.
pp4516
-
45
23
4516
Journ
al h
om
e
page
:
http:
//
ia
es
core
.c
om/
journa
ls
/i
ndex.
ph
p/IJECE
А
ut
o
m
а
t
іϲ
ԍ
ener
а
t
іo
n
c
o
ntr
o
l
bas
ed
w
һ
ale
ор
tim
і
zat
io
n
α
l
gorithm
Wisam
N
aj
m
Al
-
Din
Abed
1
,
Om
ar A.
I
mr
an
2
, Ibr
ah
im
S.
F
atah
3
1,
3
Depa
rt
m
ent of
Elec
tr
ic
on
ic
En
gine
er
ing, Col
l
e
ge
of
Engi
n
ee
rin
g,
Univer
si
t
y
of
Di
y
a
la
,
Ir
aq
2
Depa
rtment of
Chemica
l
Engi
n
ee
ring
,
Co
ll
eg
e of
Engi
n
ee
rin
g,
Univer
sit
y
of
Di
y
a
la,
Ir
aq
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Feb
7
, 2
01
9
Re
vised
Jun
2
4
, 201
9
Accepte
d
J
ul
4
, 201
9
In
the
desi
gn
i
ng
a
nd
operat
ion
of
inte
rconn
ect
e
d
po
wer
syst
e
m
s,
autom
at
ic
-
generati
on
-
co
ntr
ol
(AGC)
re
pre
sent
an
im
po
r
ta
nt
top
ic
.
AG
C
is
respo
ns
ible
f
or
m
ain
ta
inin
g
the
ba
la
nce
bet
wee
n
ge
ne
rati
on
side
an
d
lo
ad
side
via
c
ontr
olli
ng
t
he
f
requen
cy
a
nd
act
ive
pow
e
r
intercha
nge.
A
new
m
et
aheurist
ic
strat
egy
is
propose
d
in
this
wor
k
for
opti
m
a
l
con
tr
oller
tun
i
ng
in
AG
C
syst
e
m
.
Ԝһ
al
e
О
рti
m
іz
at
іоn
Αlgorithm
(
WO
A
)
is
pro
po
s
ed
f
or
op
ti
m
al
tun
i
ng
of
res
et
integral
con
t
ro
ll
er.
T
he
pro
posed
strat
egy
is
us
ed
for
opti
m
a
l
AG
C
in
two
-
areas
int
e
rcon
nected
-
powe
r
syst
em
.
The
pro
po
s
ed
tun
i
ng
strat
egy
is
c
om
par
ed
with
oth
e
r
new
m
et
aheurist
ic
optim
iz
at
ion
strat
egy
te
rm
ed
as
Ha
r
m
on
y
Search
(
HS
)
.
T
he
two
-
are
a
interco
nnect
ed
powe
r
syst
em
are
sim
ulate
d
ba
sed
M
ATL
A
B
-
too
l
box
.
Fr
om
resu
lt
s
ob
ta
ine
d,
it
is
ob
vi
ou
s
that,
the
syst
em
transient
a
nd
ste
ady
-
sta
te
b
e
hav
i
or
a
re enh
anced
great
ly
u
nd
e
r
the sam
e conditi
ons.
This
is
du
e
to
the
us
e
of
the
pr
op
os
e
d
op
ti
m
iz
at
ion
te
chn
iq
ue.
The
pr
opos
e
d
te
chn
iq
ue
has
an
ad
va
nced
a
nd
s
uperi
or
fe
at
ur
e
li
ke,
local
opti
m
u
m
av
oid
i
ng,
fa
st
co
nv
e
r
gen
ce
abili
ty
,
and
lo
wer
searc
h
agen
ts
an
d
it
erati
on
a
re
requi
red.
All
m
entio
ne
d
featu
res,
m
ake
this
strat
egy opti
m
al
f
or
var
i
ou
s
optim
iz
at
ion
p
r
oble
m
s.
Ke
yw
or
d
s
:
Au
t
om
atic
generati
on
co
ntr
ol
In
te
gr
al
c
on
t
rol
le
r
Me
ta
heu
risti
c
al
gorithm
Ԝһ
al
e
o
рtim
іzatі
оn
a
lg
ori
thm
Copyright
©
201
9
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
Al
l
rights re
serv
ed
.
Corres
pond
in
g
Aut
h
or
:
W
i
sam
N
ajm
Al
-
Di
n Abed
,
Dep
a
rtem
ent o
f
Ele
ct
r
on
ic
,
C
ollege
of Engi
neer
i
ng
Un
i
ver
sit
y o
f Diy
al
a,
Ba
quba
h,
Ir
a
q.
Em
a
il
: wisa
m
_alob
ai
de
e@ya
hoo.
c
om
1.
INTROD
U
CTION
I
nter
-
c
onnecte
d
m
ulti
-
area
powe
r
-
plants
is
div
ide
d
i
nto
va
rio
us
c
on
tr
ol
-
areas.
Tie
li
ne
s
are
c
onnect
t
hese
areas
f
or
load
-
s
ha
rin
g
[
1,
2]
.
T
he
co
nt
ro
l
-
a
rea
ge
ne
ra
tor
s
are
s
uppos
ed
co
her
e
ntly
gro
up
i
ng.
N
orm
al
l
y,
any
powe
r
plant
is
sub
j
ect
ed
to
loa
d
va
riat
ion
s
.
P
ow
e
r
sy
stem
fr
eq
uen
c
y
sh
ould
be
c
on
sta
ntly
m
ai
ntaine
d
(freque
ncy
de
viati
on
s
houl
d
be
m
ai
ntained
as
sm
al
l
as
po
s
sible)
f
or
pro
per
operati
on.
T
he
syst
em
act
ive
powe
r
an
d
f
re
qu
e
ncy
are
relat
ed
beca
u
se
the
f
reque
ncy
aff
ect
ed
by
act
ive
-
powe
r
bala
ncin
g
[
1]
.
The
AG
C
i
s
respo
ns
ible
f
or
m
ai
ntaining
the
bala
nce
bet
ween
gen
e
rati
on
a
rea
an
d
lo
ad
side
at
lowe
r
cost.
It
pla
ys
an
i
m
po
rtant
r
ole
for
fr
e
que
nc
y
con
tr
ol,
eco
no
m
ic
disp
at
ch,
a
nd
act
ive
power
i
nterc
han
ge
[
3]
.
Nowd
ay
s
,
Mult
iobject
ive
Ev
olu
ti
onary
t
echn
i
qes
a
re
use
d
to
s
olv
e
dif
f
eren
t
optim
iz
ation
pro
blem
s
[4
-
6]
.
Re
centl
y
,
m
any
research
e
rs
giv
e
great
de
al
of
at
te
ntaion
on
se
r
ving
diff
e
ren
t
m
eta
-
he
ur
ist
ic
op
ti
m
iz
ation
al
gorithm
s.
Thes
al
go
rithm
s
are
us
ed
f
or
tradit
ion
al
co
ntr
ollers
tun
i
ng
i
n
A
GC
as
il
lustrate
d
in
li
te
ratur
e
s
urve
y.
Ge
netic
Algorith
m
was
pro
po
s
ed
by
Vik
r
r
a
m
et
al
.,
[7]
i
n
(
2012
)
.
Om
a
r
et
al
.,
[
8]
intr
oduce
ACO
te
ch
niqu
e
(
2013
)
.
Sa
r
r
oj
et
al
.,
[
9]
in
tro
du
ce
di
ff
e
re
nt
m
et
a
-
heu
rist
ic
op
ti
m
iz
ati
on
m
et
ho
ds
i
n
(
2014
)
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Аutom
аtіϲ
ԍ
en
erаtіo
n
c
on
tr
ol ваѕẹ
ԁ
wһ
ale
орti
mіza
ti
on
αlg
or
it
hm
(
Wi
sam
Najm
Al
-
Din Abe
d
)
4517
li
ke,
Fire
Fly
(F
A
)
,
(
GA
)
,
P
arti
cl
e
Sw
a
rm
O
pti
m
iz
at
ion
(PSO
),
…
et
c.
i
n
(
2015
)
Ra
bindra
et
al
.,
[10]
introd
uce
a
hy
br
i
d
Fire
fly
op
tim
iz
at
ion
te
ch
nique
as
well
introd
uce
Patt
ern
Searc
h
(hF
A
–
PS).
BF
OA,
G
A
and
Z
N
are
use
d
f
or
com
par
iso
n
.
Lak
shm
i
e
t
al
.,
[1
1]
intro
duce
in
(2
01
6)
the
a
lgorit
hm
of
F
lowe
r
-
Po
ll
inati
on
.
M
us
hta
q
et
al.
,
[12] intr
oduce
(
P
SO
)
and
(
HS
)
techn
i
qu
e
s in
(
2017
)
.
The
m
ajo
r
pro
blem
s
that
faces
the
A
GC
c
on
t
ro
l
syst
em
,
is
the
presence
of
a
prom
inent
fr
e
que
ncy
dev
ia
ti
on.
T
h
i
s
dev
ia
ti
on
is
pr
ese
nted
due
to
the
cha
ng
i
ng
in
the
syst
em
load.
I
ntegral
reset
-
co
ntr
ol
le
r
is
pro
po
se
d
in
t
his
w
ork
to
so
l
ve
this
iss
ue.
T
hi
s
co
ntr
oller
ha
s
the
abili
ty
to
set
tl
e
the
syst
em
dev
ia
ti
on
t
o
zer
o
du
e
to
it
s i
nteg
ral act
ion w
hic
h
inc
rease t
he
s
yst
e
m
ty
pe
by 1
.
The
sec
ond
pr
ob
le
m
in
the
A
GC
co
ntr
ol
de
s
ign
is
t
he
c
hoosi
ng
of
ap
pro
pri
at
e
op
ti
m
iz
ati
on
strat
egy
for
c
on
tr
oller
tun
in
g.
I
n
thi
s
w
ork,
ne
w
natu
re
ins
pire
d
op
ti
m
iz
a
ti
on
al
gorithm
(N
IOA)
c
al
le
d
Wh
al
e
Op
ti
m
iz
ation
Algorithm
(
WO
A
)
is
pro
po
s
ed
f
or
c
on
tr
ol
le
r
tun
i
ng.
W
OA
has
the
c
har
act
erist
ic
of
go
od
balance
betw
een
the
e
xplorati
on
a
nd
exp
l
oitat
ion
ph
a
ses
ove
r
o
the
r
m
et
a
heurist
ic
s
al
gorithm
s.
This
c
har
act
er
ist
ic
m
akes
WOA
e
xplore
the
searc
h
s
pace
ef
fecti
vel
y
with
fast
c
onve
rg
e
nce
as
well
a
s
avo
i
ding
entra
pp
i
ng
i
n
local
op
ti
m
a.
The
resu
lt
s
of
the
propose
d
tu
ning
te
chn
iq
ues
a
re
com
par
ed
with
ot
he
r
op
ti
m
iz
ation
t
echn
i
qu
e
cal
le
d
Har
m
on
y
S
earch
(
HS)
to
pro
ve
t
he
s
uperi
or
it
y
featu
res
of
the
pro
pos
e
d
strat
egy.
2.
MO
DEL
OF
SY
STE
M
Figure
1
il
lust
rate
the
trans
fe
re
f
un
ct
io
n
m
od
el
for
the
tw
o
-
a
rea
po
wer
plant.
T
he
in
di
vidual
areas
include
gove
r
nor
,
tur
bin
e
a
nd
gen
e
rato
r.
T
he
input
sig
nal
to
co
ntr
oller
(∆
P
ref
)
,
power
e
rror
of
tie
-
li
ne
(
∆P
12
)
and
loa
d
distu
r
ban
ce
(∆
P
L
)
re
pr
ese
nt
i
nd
i
vidual
area
in
puts.
Wh
il
e
the
f
re
qu
e
ncy
of
area
(∆ω
)
a
nd
the
area
con
t
ro
l
er
ror
(
ACE)
re
presen
t
ind
ivid
ual
ar
ea
ou
tp
ut
[
9]
.
Fo
r
eac
h
area
ACE
re
pr
ese
nt
the
con
tr
oller
inp
ut
and it
sho
uld
be
r
e
du
ce
d
t
o
ze
ro
[
12
]
.
ACE
gi
ven
by,
=
.
(
∆
)
−
∆
12
= e(t)
(1)
Wh
e
re:
B
,
∆P
m
,
∆P
v
,
,
,
D
a
nd
H
re
pr
ese
nt
s
f
reque
ncy
-
bi
as
pa
ram
et
ers,
m
echan
ic
al
outp
ut
powe
r
of
tur
bin
e,
go
vernor
ou
t
put
po
wer,
ti
m
e
con
sta
nt
of
tu
rb
i
ne,
tim
e
con
sta
nt
of
gove
r
nor,
dam
pin
g
pa
ram
et
er
a
nd
const
ant
of
i
ne
rtia
r
es
pecti
vely
[
9
]
.
Figure
1.
Tw
o
-
area c
om
plete
m
od
el
w
it
h
A
GC
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
6
,
Dece
m
ber
2
01
9
:
4516
-
4523
4518
3.
Ԝ
OA
O
VER
VIEW
Ԝ
OA
im
it
at
e
the
huntin
g
pr
ocess
to
the
pr
ey
of
hu
m
pb
a
ck
w
hales.
Th
e
hunting
proc
ess
incu
ding
encircli
ng,
sear
chin
g
an
d
at
ta
chin
g
the
prey
.
It
is
a
no
vel
m
et
a
-
he
ur
ist
ic
optim
i
z
at
ion
al
go
rithm
us
ed
to
so
lve
op
ti
m
iz
ation
pro
blem
s.
It
was
introd
uced
re
centl
y
in
2016
by
Lewis
an
d
Mi
rj
al
il
i
[13]
.
A
m
on
g
o
t
her
var
i
ou
s
bio
-
ins
pire
d
optim
iz
at
ion
te
chn
i
qu
e
s,
Ԝ
OA
hav
e
dif
fer
e
nt
su
pe
rio
r
ow
n
featur
e
s.
It
re
quire
d
low
num
ber
of
par
am
et
ers
an
d
ha
ve
fa
st
co
nverg
e
nce
a
bili
ty
[14]
.
Also,
thi
s
strat
egy
ca
n
at
ta
in
global
optim
u
m
so
luti
on
a
nd
avo
i
de
e
ntra
pping
i
n
loc
al
op
tim
u
m
.
All
m
e
ntion
e
d
featu
re
s
m
ake
this
str
at
egy
can
be
a
pp
li
ed
ef
fecti
ve
ly
in
diff
e
re
nt opti
m
iz
at
ion
area
[
15]
.
Kr
il
l
school
an
d
sm
a
ll
-
fishes
near
to
the
sur
face
represe
nt
the
pr
e
ferred
prey
to
hu
m
pb
a
ck
w
hales.
The
hu
nt
proce
ss
sta
rt
by
crea
ti
ng
a
sp
eci
al
pa
th
of
bubble
s
li
ke
9
sh
a
pe
or
al
ong
a
ci
rcle.
These
bubble
s
ca
n
be
disti
nguish
by
hum
pb
ack
wh
al
es
only
as
il
lustrate
in
Fi
gure
2
[16]
.
W
OA
co
ns
ist
s
of
two
disti
nct
phases;
the
fi
rst
ph
ase
is
the
ex
plo
it
at
ion
phase
a
nd
t
he
sec
ond
is
the
e
xplo
rati
on
pha
se
.
T
he
ex
plo
it
at
ion
phas
e
consi
sts
of
pr
e
y
encircle
m
ent
an
d
the
n
bu
bble
net
at
ta
ck.
Wh
il
e
the
e
xplo
r
at
ion
phas
e
re
pr
ese
nt
sea
rch
i
ng
the prey
r
a
ndom
ly
[17]
.
Figure
2.
H
umpb
ac
k b
ubble
-
net
at
ta
ck
[18]
3.1.
Expl
oitatio
n
-
pha
se
3.1.1.
Enci
rcl
ing p
r
ey
This ste
p
sta
rts
b
y
recog
nize the
pr
ey
posit
ion
a
nd en
ci
rcli
ng the
prey
. T
he
n
the
whale
’s
locati
on is
m
od
ifie
d
to
wards
best a
gen
t.
This
process
il
lustrate
s m
at
hem
at
ic
ally as:
⃗
⃗
=
|
.
∗
⃗
⃗
⃗
⃗
(
)
−
(
)
|
(2)
(
+
1
)
=
∗
⃗
⃗
⃗
⃗
(
)
−
.
⃗
⃗
(3)
=
curre
nt
-
it
er.
∗
=
v
ect
or
f
or
be
st
po
sit
io
n
s
olu
ti
on
.
=
po
sit
i
on
vect
or
.
=
r
andom
vector
betwee
n
[
0,1]
.
C
&
A
re
pr
esent
vectors
of
c
oeffici
ent
an
d ob
ta
ine
d
a
s foll
ows,
=
2
.
.
(4)
=
2
.
(5)
3.1.2.
Bubble
‐
net
at
ta
c
k b
eh
avio
u
r:
Th
is st
rate
gy
of the
hum
p
-
bac
k wh
al
es
is
ex
presse
d
as:
1.
Sh
ri
nkin
g
-
e
nci
rcl
ing
-
te
c
hn
i
que
;
ove
r
the
c
ourse
of
it
erati
on,
the
vect
or
is
dec
rased
li
near
ly
f
r
om
2
to
0
f
or
both
ph
ases
of
WOA
f
or
at
ta
inin
g
t
his
be
hav
i
or
ref
e
r
(
4)
.
B
y
set
ti
ng
the
vecto
r
va
lues
ra
ndom
l
y
in
range
[
‐
1,
1]
,
the
ne
w
l
ocati
on
of
disco
ver
i
ng
a
ge
nts
in
a
nyplace
ca
n
be
determ
ined.
This
posit
io
n
can
be
sp
eci
fied
in bet
ween t
he o
rigi
nal ag
e
nt
posit
ion o
f
c
urre
nt best
ag
e
nt.
2.
Sp
ir
al
-
updati
ng
posit
io
n
:
Th
e
helic
al
-
sh
a
pe
d
m
ov
em
ent
of
the
hum
p
-
ba
ck
w
hales
tow
ard
s
pr
ey
ca
n
be
expresse
d
m
ath
em
atical
ly
b
y
the s
piral e
qua
ti
on
as
,
(
+
1
)
=
′
⃗
⃗
⃗
⃗
.
.
cos
(
2
)
+
∗
⃗
⃗
⃗
⃗
(
)
(6)
Hu
m
p
-
back
w
hales
swim
encircli
ng
the
prey
and
al
te
rn
at
es
instanti
nous
ly
al
ong
a
sp
iral
‐
sh
ape
d
pa
t
h
and
withi
n
a
sh
ri
nk
i
ng
ci
rcle
.
This
sud
de
n
be
ha
vior
ca
n
be
m
od
el
ed
by
ch
oosin
g
50%
pr
ob
a
bili
ty
fo
r
m
echan
ism
o
f
sh
ri
nk
i
ng en
ci
r
cl
ing
a
nd
s
pira
l
m
od
el
, th
is
ca
n be e
xpresse
d m
at
he
m
at
ic
a
lly as,
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Аutom
аtіϲ
ԍ
en
erаtіo
n
c
on
tr
ol ваѕẹ
ԁ
wһ
ale
орti
mіza
ti
on
αlg
or
it
hm
(
Wi
sam
Najm
Al
-
Din Abe
d
)
4519
(
+
1
)
=
{
∗
⃗
⃗
⃗
⃗
(
)
−
.
if
<
0
.
5
′
⃗
⃗
⃗
⃗
.
.
cos
(
2
)
+
∗
⃗
⃗
⃗
⃗
(
)
if
≥
0
.
5
(7)
The
be
st
so
luti
on
obta
ined
ca
n
be
represe
nts
by
the
ℎ
whal
e
distance
to
the
pr
ey
.
T
his
sp
ec
ifie
s
by
′
⃗
⃗
⃗
⃗
=
|
∗
⃗
⃗
⃗
⃗
(
)
−
(
)
|
.
=
co
ns
ta
nt
of
log
a
rithm
ic
s
piral
-
s
ha
pe
.
=
m
ult
ipli
cat
ion
of
el
em
ent
by
el
em
ent
wh
ic
h
re
pr
ese
nt
a
ran
dom
nu
m
ber
between
[
−
1
,
1
]
[13
,
17
,
19]
.
Encircli
ng
stra
te
gy
is
si
m
ulated
by
fir
st
part
of
(
7)
w
hile
bubble
-
net
at
ta
ck
m
echan
is
m
is
rep
rese
nt
ed
by
the
se
cond
pa
rt
of
the
sam
e
equat
ion
.
The
va
riable
p
al
te
rn
at
es
with
an
e
qu
al
pr
obabili
ty
betwe
en
these
tw
o
m
od
es
.
The
se
arch
a
gen
ts
possible
locat
i
ons
(
X,
Y)
base
d
(7)
a
re
m
od
ifie
d
t
owar
ds
best
c
urren
t
best
posit
ion
(
X*,
Y
*).
T
he
bubble
net
strat
egy
are
e
xp
la
ine
d
i
n
Fi
g
ure
3.
(a)
(b)
Figure
3. Bu
bble
-
net
at
ta
ck (
or
sea
rch
i
ng
)
str
at
egy.
(a
)
s
hr
i
nkin
g
-
e
ncircli
ng
(b)
s
piral
upda
ti
ng
posit
ion
3.2.
Expl
oration
pha
se
(
prey
se
archin
g)
This
phase
ba
sed
on
vector
var
ia
ti
on
ra
ndom
l
y.
To
ens
ure
t
he
searc
h
wh
al
es
m
ov
e
away
from
the
posit
ion
i
ng
w
hale,
t
he
vecto
r
val
ues
m
us
t
be
m
or
e
than
1
or
bel
ow
−
1
.
T
his
is
due
to
t
he
fact
that
the
rand
om
l
y
pr
ey
searc
hi
n
g
of
the
hu
m
pb
ack
w
hale’s
de
nds
on
the
l
ocati
on
of
each
oth
e
r.
This
is
expresse
d
as
,
Ɗ
⃗
⃗
=
|
.
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
−
|
(8)
(
+
1
)
=
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
−
А
⃗
⃗
.
Ɗ
⃗
⃗
(9)
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
⃗
= locat
io
n
vec
tor
c
hosed
r
a
ndom
ly
[15
,
13
,
17
,
19]
.
Fig
ur
e
4
il
lustrate
WOA
ps
e
udo
c
ode.
Figure
4.
WOA
ps
eu
do c
ode
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
6
,
Dece
m
ber
2
01
9
:
4516
-
4523
4520
4.
RESU
LT
S
AND DI
SCUS
S
ION
4.1.
Pro
po
se
d wo
rk
desi
gn
MATLAB
-
S
I
MULI
NK
to
ol
box
is
us
e
d
to
si
m
ulate
and
desig
n
t
he
unde
r
-
stu
dy
m
od
el
of
A
GC
.
The
pr
opos
e
d
WOA
tu
ning
s
trat
egy
is
base
d
on
m
file
cod
ing
.
T
he
unde
r
-
stu
dy
m
od
el
i
s
done
base
d
di
ff
ere
nt
op
e
rati
ng
co
nd
it
ion
s.
T
he
int
egr
al
re
st
co
ntro
ll
er
is
pr
opose
d
as
c
on
tr
oll
er
f
or
eac
h
are
a
in
the
un
der
-
stud
y
syst
e
m
.
The
pro
posed
c
ontr
oller
is
tun
e
d
base
d
WO
A
f
or
c
on
t
ro
ll
in
g
the
ti
e
li
ne
po
we
r
an
d
f
re
quency
dev
ia
ti
on
op
ti
m
al
l
y.
The
re
s
ults
that
obta
ined
from
the
pro
po
se
d
new
strat
egy
(
WOA)
a
re
c
om
par
ed
with
oth
e
r
good
ne
w
m
et
aheu
risti
c
op
ti
m
iz
a
ti
on
al
go
rithm
te
rm
ed
as
Har
m
on
y
Searc
h
(
HS).
T
he
com
par
is
on
of
resu
lt
s
betwee
n
the
tw
o
opti
m
iz
at
ion
techni
qu
es
, is to p
rove th
e
supe
riori
ty
ad
van
ce
d fe
at
ur
es
of the
pr
opos
e
d
al
gorithm
(W
O
A)
ve
rsus
HS
t
echn
i
qu
e
s.
4.2.
Analysis
of
re
sults
In
t
his
w
ork
,
t
wo
reset
c
ontr
ollers
ar
e
use
d.
Each
c
ontr
oller
is
use
d
f
or
on
e
area.
A
ste
p
i
nput
is
us
e
d
to
m
od
el
le
d
the
loa
d
cha
ng
e
disturba
nce.
The
loa
d
disturbance
is
use
d
to
pro
ve
the
integ
ral
co
ntr
oller
rob
us
tness
when
tun
e
d
usi
ng
the
propose
d
te
chn
i
qu
e
(
WOA)
.
T
he
pa
ram
et
ers
of
the
under
-
stu
dy
s
yst
em
are
m
entioned
in
T
able
1
.
I
nteg
ral
Squr
e
Er
ror
(
IS
E
)
is
us
e
d
as
a
perfor
m
an
ce
ind
e
x
i
n
the
tu
ni
ng
proces
s.
Fo
r
prop
e
r
c
om
par
ison
of
the
tw
o
t
unin
g
te
chn
i
ques,
t
he
al
go
rithm
s
par
am
et
ers
are
set
eq
ually
f
or
both
te
chn
iq
ues
.
E
qual
it
erati
on
s
num
ber
is
u
se
d
for
both
te
ch
niq
ue
s
(=
40
it
er
at
ion
)
a
nd
eq
ua
l
search
a
ge
nts
(=20
search
ag
e
nt)
a
s w
el
l.
Ta
ble
2 i
ll
us
trat
e the
both c
ontr
ollers
par
am
et
ers
bas
ed
the
tw
o op
ti
m
iz
at
ion
strategie
s
.
Table
1.
Syst
em
p
ara
m
et
ers
of t
wo
-
a
rea
power pla
nt
P
ara
m
eter
V
alu
e
P
ara
m
eter
V
alu
e
T
g1
0
.5
0
T
g
2
0
.6
0
T
T1
0
.2
0
T
T
2
0
.3
0
D
1
0
.6
0
D
2
0
.9
0
R
1
0
.05
0
R
2
0
.06
2
5
0
B
1
0
.9
0
B
2
0
.9
0
H
1
5
H
2
4
Table
2.
Param
et
ers
of i
ntegra
l con
t
ro
ll
er
K
1
K
2
Co
st
-
f
u
n
ctio
n
W
OA
0
.50
8
7
7
6
1
0
.22
0
0
0
4
2
0
.08
5
5
2
2
1
HS
0
.49
2
9
1
3
2
0
.00
0
1
2
1
1
0
.08
6
0
2
3
1
Fr
om
Fig
ur
e
5
wh
ic
h
represe
nt
the
plo
t
of
co
st
functi
on
,
it
is
ob
vious
t
hat
the
propose
d
optim
iz
at
ion
strat
egy
re
qu
ir
ed
lo
wer
it
erat
ion
nu
m
ber
w
hich
dem
on
str
at
e
the
fast
co
nv
e
r
gen
ce
a
bili
ty
.
WO
A
ha
ve
lowe
r
cost
functi
on
wh
ic
h
m
eans
a
best
so
luti
on
is
attai
ned
as
com
par
ed
with
HS
.
Fig
ur
e
6
and
Fi
gure
7
s
hows
ACE
plo
t
f
or
bo
th
ar
eas.
F
igure
s
8
-
9
s
hows
the
fr
eq
ue
ncy
dev
ia
ti
on
fo
r
bo
t
h
area
s
.
Figure
10
s
hows
the po
wer
exc
ha
ng
e
of
tie
-
li
ne
b
et
wee
n
t
he
t
wo
-
area
.
T
he
l
oad d
ist
urban
c
e f
or
pro
posed
work is c
hoo
se
n 0.1.
Figure
5.
Cost
functi
on
plo
t
f
or
both
tech
niques
0
5
10
15
20
25
30
35
40
10
-1
.
0
6
7
10
-1
.
0
6
6
10
-1
.
0
6
5
10
-1
.
0
6
4
ite
rati
on
best c
ost
WOA
HS
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Аutom
аtіϲ
ԍ
en
erаtіo
n
c
on
tr
ol ваѕẹ
ԁ
wһ
ale
орti
mіza
ti
on
αlg
or
it
hm
(
Wi
sam
Najm
Al
-
Din Abe
d
)
4521
Figure
6. ACE
1 for a
rea
1
Figure
7. ACE
2 for a
rea
2
Figure
8.
(
∆
1
)
for
a
rea
1
Figure
9.
(
∆
2
)
for
a
rea
2
Figure
1
0.
Tie
li
ne
powe
r for
bo
t
h
a
rea
The
syst
e
m
tr
ansient
an
d
ste
ady
-
sta
te
behavio
r
are
i
m
pr
ove
d
prom
inently
wh
en
us
ed
integral
con
t
ro
ll
er
base
d
WOA
eve
n
unde
r
loa
d
di
sturb
a
nce.
I
n
AG
C
it
is
pe
r
efera
ble
to
use
integral
c
ontrolle
r
op
ti
m
al
l
y
tun
ed
base
d
WOA
du
e
to
it
s
abi
li
ty
to
set
tl
e
t
he
fr
e
que
ncy
dev
ia
ti
on
an
d
k
ept
it
zero
a
prrox.
So
,
this
co
ntr
ol
le
r
play
s
an
i
m
po
rtant
ro
le
in
power
syst
e
m
con
trol.
T
he
po
r
pose
d
WO
A
tu
ning
te
chn
i
que
sh
ows
a
supe
rior
featu
re
in
th
e
fiel
d
of
c
ontr
oller
opti
m
izing
ver
s
us
HS
st
rategy.
WOA
has
fast
co
nver
gen
c
e
abili
ty
,
lower
par
am
et
ers
require
d,
a
nd
a
vo
i
d
e
ntra
pp
i
ng
in
local
opti
m
u
m
.
Th
e
obta
ine
d
resu
lt
s
base
d
WOA
sh
ows
a
ve
ry
sm
a
ll
os
ci
ll
at
i
on
as
c
om
par
ed
with
res
ults
-
base
d
H
S.
T
hi
s
m
eans
ver
y
sm
a
ll
ov
ers
ho
ot
an
d
unde
r
s
hoot.
0
10
20
30
40
50
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
tim
e
(s)
ACE1
WOA
HS
0
10
20
30
40
50
-0.015
-0.01
-0.005
0
0.005
0.01
0.015
tim
e
(s)
ACE2
WOA
HS
0
10
20
30
40
50
-8
-6
-4
-2
0
2
4
x
10
-3
tim
e
(s)
f
1
(Hz)
WOA
HS
0
10
20
30
40
50
-2
-1.5
-1
-0.5
0
0.5
1
x
10
-3
tim
e
(s)
f
2
(Hz)
WOA
HS
0
10
20
30
40
50
-18
-16
-14
-12
-10
-8
-6
-4
-2
0
2
x
10
-3
tim
e
(s)
P
Tie
(p.u.)
WOA
HS
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
6
,
Dece
m
ber
2
01
9
:
4516
-
4523
4522
5.
CONCL
US
I
O
N
M
et
a
-
heurist
ic
op
ti
m
iz
at
ion
strat
egies
play
a
vital
ro
le
in
th
e
field
of
co
ntro
ll
er
tun
i
ng
i
n
the
interc
onee
ct
ed
-
power
pl
ant.
T
hese
st
r
at
egies
ha
ve
t
he
a
bili
ty
to
im
pr
ov
e
th
e
co
ntr
oller
pe
rfo
r
m
ance.
This
fact
le
ad
s
to
ap
pear
va
rio
us
ty
pes
of
m
et
a
-
heurist
ic
op
tim
iz
at
io
n
al
gorithm
s.
The
m
et
a
-
heu
risti
c
o
ptim
iz
ation
strat
egies
are
di
vid
e
d
to
tw
o
di
sti
nct
ph
ases
r
egardless
it
s
na
ture.
T
he
first
ph
a
se
is
ex
plorat
ion
ph
a
se w
hic
h
m
eans
in
vestigat
ing
gl
ob
al
ly
the p
roblem
search
sp
ace
. Th
e s
econd phase is the ex
poil
ti
ng
ph
a
se
wh
ic
h
m
eans
exp
l
or
i
ng
the
search
s
pa
ce
prom
isi
ng
reg
i
ons
that
f
ound
in
the
ex
plorat
ion
ph
a
se.
He
nc
e
the
prom
isi
ng
reg
i
on
that
f
ound
by
exp
lo
rati
on
phase
are
r
efinin
g
by
the
n
ext
phase
(
exp
l
oitat
ion
phase)
.
So
,
in
desig
ni
ng
any
m
et
a
-
heu
risti
c
strat
eg
y,
it
is
i
m
po
rtant
to
ac
hieve
a
pro
per
bala
nc
e
betwee
n
t
hes
e
two
ph
a
ses
due
t
o
the
ra
ndom
iz
e
natu
re.
T
his
f
act
represents
a
real
chall
en
ge
in
de
sig
ning
any
m
et
a
-
heurist
ic
al
gorithm
du
e
to
it
s
stochasti
cal
featur
es
t
ha
t
m
ake
them
avo
i
ding
local
op
ti
m
u
m
entrap
m
ent.
This
superi
or
featur
e
is
pro
m
inent
in
th
e
WO
A
pro
po
sed
te
c
hn
i
qu
e
.
As
well
WO
A
ha
ve
fast
co
nv
e
r
gen
ce
abili
ty
,
lowe
r
pa
ram
eter
s
re
quire
d.
WOA
do
well
in
the
fiel
d
of
co
ntr
oller
t
unin
g
es
pecial
ly
in
the
fiel
d
of
A
GC
in po
wer pla
nts.
REFERE
NCE
S
[1]
K.
R.
Sudha
an
d
R.
Vij
a
y
a
San
thi
,
"Robus
t
d
ecent
ra
li
z
ed
lo
ad
f
req
uency
cont
ro
l
of
interc
onn
ec
t
ed
power
s
y
s
tem
with
Gene
r
at
ion
Rat
e
Constrai
n
t us
ing
T
y
pe
-
2
fu
zzy
appr
oa
ch,
"
J
EP
E
Inte
rnat
ion
al
Journal
of El
e
ct
rical P
ow
er
an
d
Ene
rgy
S
yste
ms
,
vol.
33
,
pp
.
699
-
707,
2011
.
[2]
K.
Manic
kav
asa
gan,
"F
uzzy
bas
ed
power
flow
cont
rol
of
two
ar
ea
power
s
y
st
em
,
"
IJE
CE
Int
ernati
onal
Journal
of
El
e
ct
rica
l
and
C
omputer
Engi
n
e
ering
(
IJE
CE)
,
v
ol.
2
,
no
.
1
,
2012
.
[3]
H.
Bevr
ani
and T.
Hi
y
am
a,
Intel
li
gent A
u
tomati
c
Gene
ration
Con
trol
,
CR
C
Press
,
2017.
[4]
W
.
N.
A.
L.
D.
Abed,
A.
H.
Sale
h,
and
A.
S.
Ham
ee
d,
"S
pee
d
C
c
ontrol
of
PM
D
CM
b
as
ed
GA
and
DS
t
ec
hniques
,
"
IJP
EDS
In
te
rnat
ional
Journal
of
Powe
r E
le
c
troni
cs
and
Dr
ive
Sys
te
ms
(
IJP
EDS)
,
vol.
9
,
p
p
.
1467
,
2018.
[5]
O.
A.
Im
ran
,
W
.
N.
A.
D
.
Abed
,
an
d
A.
N.
Jbar
ah,
"S
pee
d
control
of
univ
ersa
l
m
otor,
"
Inte
rnat
ional
Journal
o
f
Powe
r E
le
c
troni
cs
and
Dr
ive
Sys
te
ms
(
IJP
EDS)
,
vol.
10
,
pp
.
41
-
4
7,
2019
.
[6]
W
.
N.
A.
D.
Abed,
O.
A.
I
m
ran
,
and
A.
N.
Jbara
h,
"V
olt
age
cont
ro
l
of
buck
conv
erter
-
base
d
Ant
Col
on
y
Optimiza
ti
o
n
for
self
-
reg
u
la
t
ing
power
supplie
s,"
J.
Eng
.
App
l.
S
ci
.
Journal
of
E
ngine
ering
and
Appl
ie
d
Sc
ie
n
ces
,
vol.
13
,
pp
.
4463
-
4467,
2018
.
[7]
V.
K.
Kam
bo
j,
K.
Arora,
and
P.
Khurana
,
"A
uto
m
at
ic
gen
era
t
ion
cont
rol
for
in
terconne
c
te
d
h
y
dr
o
-
the
rm
al
s
y
st
e
m
with
th
e
he
lp
of
conve
nt
iona
l
co
ntrol
lers
,
"
Int
ernati
onal
Journal
of
Elec
tric
al
an
d
Computer
Eng
ine
ering
(
IJE
CE
)
,
vol.
2
,
no
.
4
,
201
2.
[8]
M.
Om
ar,
M.
Solim
an,
F
.
Bend
ar
y
,
and
A.
M
.
Abdel
Ghan
y
,
"
Optimal
tuni
ng
of
PID
cont
roll
e
rs
for
h
y
droth
er
m
al
loa
d
fre
quen
c
y
cont
rol
using
ant
col
on
y
optim
iz
at
ion
,
"
Inte
rnational
Journa
l
on
El
ec
tri
cal
Engi
nee
ring
a
nd
Informatic
s
,
vo
l. 5, pp. 348
-
360,
2013.
[9]
S.
Padhan,
R.
K.
Sahu,
and
S.
Panda,
"A
ppli
c
at
ion
of
fire
f
l
y
al
gorit
hm
for
loa
d
fre
quency
co
ntrol
of
m
ult
i
-
ar
e
a
int
er
conne
c
te
d
p
ower
s
y
st
em,"
E
le
c
tric
Powe
r C
omponents
and
Syste
ms
,
vol
.
42
,
pp.
1419
-
1430,
2014.
[10]
R.
K.
Sahu,
S.
Panda,
and
S.
Padhan,
"A
h
y
b
rid
fire
fl
y
al
gor
it
hm
and
pat
t
er
n
sea
rch
techni
que
for
aut
om
atic
gene
ra
ti
on
co
ntr
ol
of
m
ult
i
ar
ea
power
s
y
stems
,
"
Inte
rnational
Journal
of
E
le
c
t
rical
Pow
er
and
Ene
rgy
Syst
ems
,
vol.
64
,
pp
.
9
-
23
,
2015
.
[11]
D.
La
kshm
i,
A.
P.
Fathi
m
a,
and
R.
Muthu,
"A
no
vel
flowe
r
poll
in
at
ion
al
gor
it
hm
to
solve
loa
d
fre
quency
cont
ro
l
fo
r
a
h
y
dro
-
th
ermal
der
egulate
d
pow
er
s
y
s
te
m
,
"
CS
C
ircui
ts and
System
s
,
vol.
7,
pp.
1
66
-
178,
2016
.
[12]
M.
Naje
eb
,
H.
F
e
y
ad,
M.
Manso
r,
E.
Ta
h
a,
and
G.
Abdulla
h,
"A
n
opti
m
al
LFC
i
n
two
-
are
a
pow
er
s
y
stems
using
a
m
et
a
-
heur
isti
c
o
pti
m
iz
ation
algorithm,"
Inte
rn
ati
onal
Journal
of
El
e
ct
rica
l
and
Computer
Engi
nee
ring
,
vol.
7
,
no.
6
,
pp
.
3217
-
3225,
2017
.
[13]
S.
Mirja
lili,
A.
Le
wis,
and
S.
Mirja
lili
,
"The
whale
opti
m
izat
ion
al
gori
thm
,
"
Adv
anc
es
in
Engi
nee
ring
Sof
tw
are
,
vol.
95
,
pp
.
51
-
6
7,
2016
.
[14]
H.
M.
Hasani
en,
"W
hal
e
opti
m
is
at
ion
al
gori
thm
for
aut
om
atic
g
e
ner
ation
con
trol
of
int
er
connect
e
d
m
oder
n
power
s
y
stems
inc
ludi
n
g
ren
ewa
ble
en
e
rg
y
source
s,"
IE
T
GENER
ATION
TRANSMISSION
AND
DIS
TRIBUTIO
N
,
vol.
12,
pp.
607
-
614
,
20
18.
[15]
G.
Kaur
and
S.
Arora,
"Chaot
i
c
whale
opti
m
i
za
t
ion
al
gori
thm,"
JCDE
Journal
of
Computational
Design
and
Engi
ne
ering
,
20
18.
[16]
H.
Hu,
Y.
Ba
i,
and
T
.
Xu,
"
Im
prove
d
whale
opti
m
izati
on
al
gorit
hm
s
base
d
on
ine
r
ti
a
w
ei
ghts
and
thei
rs
appl
i
ca
t
ions,"
In
te
rnational
Jour
nal
of
Circuits,
S
yste
ms
and
Sign
al
Proc
essing
,
v
ol.
11
,
pp
.
12
-
26
,
2017
.
[17]
M.
Mafa
rja
and
S.
Mirja
lili,
"
W
hal
e
opti
m
izat
ion
appr
oa
che
s
for
wrappe
r
feat
ure
select
ion
,
"
ASOC
Appl
ie
d
Soft
Computing
Jour
nal
,
vo
l. 62, pp.
441
-
453,
2018
.
[18]
I.
Alj
ara
h
,
H.
Faris,
and
S.
Mirja
lili
,
"O
ptim
iz
ing
conn
ec
t
i
on
weight
s
in
neur
al
net
works
using
th
e
wha
le
opti
m
iz
ation
al
g
orit
hm
,
"
Soft
Co
m
put
Soft
Com
puti
ng
:
A
Fus
ion
of
Foundati
ons
,
Methodol
ogi
es
and
Appli
ca
t
ion
s,
vol.
22
,
pp
.
1
-
15
,
2018
.
[19]
M.
M.
Mafa
rja
and
S.
Mirja
l
i
li
,
"H
y
b
rid
wha
le
opti
m
izati
on
al
gorit
hm
with
sim
ula
te
d
ann
ea
l
ing
for
featu
re
sele
c
ti
on,
"
NEU
COM Ne
urocomputing
,
vol
.
260
,
pp.
302
-
312,
20
17.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Аutom
аtіϲ
ԍ
en
erаtіo
n
c
on
tr
ol ваѕẹ
ԁ
wһ
ale
орti
mіza
ti
on
αlg
or
it
hm
(
Wi
sam
Najm
Al
-
Din Abe
d
)
4523
BIOGR
AP
H
I
ES
OF
A
UTH
ORS
The
Lectu
re
r
Wisam
Najm
AL
-
Din
A
bed
rec
eived
a
bac
he
lor
'
s
degr
ee
in
elec
tr
ic
a
l
power
and
m
ac
hi
nes
from
engi
nee
r
ing
colle
ge
-
Di
y
ala
Unive
rsit
y
in
2005
an
d
rec
e
ive
d
a
m
a
ster'
s
degr
ee
in
el
e
ct
ri
ca
l
eng
ineeri
ng
/
power
from
the
Univer
sit
y
of
Technol
og
y
in
2011.
Area
of
rese
arc
h
int
er
est
in
the
e
l
ec
tr
ic
power,
m
ac
hin
er
y
and
co
ntrol
engi
n
ee
rin
g
and
art
if
ic
i
al
i
nte
lligen
ce
and
al
gorit
hm
s
Enginee
ring
Opt
imiz
at
ion
.
He
has
m
ore
tha
n
sc
ie
n
ti
f
ic
rese
arc
h
pub
l
ished
in
lo
ca
l
and
in
te
rna
ti
on
al j
ourna
ls.
Email
:
wisam
_al
obai
d
e
e@
y
ahoo.com
Omar
A.
Imran
rec
ei
ved
a
ba
che
lor
'
s
degr
e
e
in
el
e
ct
r
oni
c
from
engi
nee
ring
col
l
ege
-
Di
y
a
l
a
Univer
sit
y
in
200
6
and
r
ecei
ved
a
m
aste
r
'
s
degr
e
e
in
e
lectr
i
ca
l
en
erg
y
from
Bel
gorod
Governm
ent
Technol
og
y
Unive
rsit
y
Russ
ia'
s
Feder
a
l
in
201
3
.
Area
of
rese
ar
ch
int
er
est
in
the
elec
tr
ic
pow
er
engi
ne
eri
ng
.
He
has
m
ore
th
an
scie
nt
ifi
c
r
ese
arc
h
publ
ished
in
int
ern
at
ion
a
l
journa
ls.
Emai
l
:
Om
ari
m
ran
53@
y
ahoo
.
com
Ib
rahim
Saa
d
oon
Fatah
rec
ei
ved
a
ba
chel
or'
s
degr
e
e
in
el
ectroni
c
en
gine
er
ing
from
the
Univer
si
t
y
of
Sara
je
vo
/
Bosnia
in
1984
and
recei
v
ed
a
m
aste
r
'
s
degr
ee
in
el
e
ct
ron
i
c
engi
ne
eri
ng
fro
m
the
Univer
sit
y
of
B
el
gr
ade
/
Serbia
in
198
7.
Serve
d
as
th
e
direct
or
fo
r
reg
istration
in
t
he
Coll
eg
e
of
E
ngine
er
ing
/
Univer
sit
y
of
Di
y
a
la
for
the
per
i
o
d
from
2010
to
2013.
Th
e
ar
ea
of
rese
ar
ch
in
te
r
est
in
cont
rol
e
ngine
er
ing
and
el
e
ct
roni
cs.
He
has
m
ore
tha
n
scie
ntific
rese
arch publ
ished
in
lo
ca
l
and
interna
t
i
onal
journa
ls
.
E
m
ai
l
:
saa
don@
yahoo.
com
Evaluation Warning : The document was created with Spire.PDF for Python.