Indonesi
an
Journa
l
of El
ect
ri
cal Engineer
ing
an
d
Comp
ut
er
Scie
nce
Vo
l.
13
,
No.
1
,
Jan
uar
y
201
9
,
pp.
377
~
383
IS
S
N: 25
02
-
4752, DO
I: 10
.11
591/ijeecs
.v1
3
.i
1
.pp
377
-
383
377
Journ
al h
om
e
page
:
http:
//
ia
es
core.c
om/j
ourn
als/i
ndex.
ph
p/ij
eecs
Impact o
f inertia
wei
ght strate
gies
in p
ar
ti
cle s
war
m
optimiz
ation f
or solvin
g economic
d
ispatch
probl
em
Moham
med
A
mi
ne M
ez
iane
, You
s
sef M
oul
ou
di, B
ou
sm
aha B
ou
chib
a,
ab
dell
ah L
aou
fi
Ta
hri
Moham
m
ed
Univer
sit
y
,
BP
417,
08000
Be
c
har
,
Alg
eria
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
A
ug
3
, 2
018
Re
vised
Sep
19
, 2
018
Accepte
d
Oct
1
, 2
018
Parti
cle
Sw
arm
Optimiza
ti
on
(PS
O)
is
a
po
pula
ti
on
b
ase
d
stocha
sti
c
opti
m
iz
ation
tec
hnique
inspire
d
b
y
the
soci
al
l
earning
of
birds
or
fish.
Som
e
of
the
appe
a
li
ng
fac
ts
of
PS
O
are
it
s
conve
ni
ence,
sim
pli
cit
y
and
ea
siness
of
implementa
t
ion
req
uiri
ng
bu
t
fe
w
par
amete
rs
ad
justm
ent
s.
Ine
rt
i
a
W
ei
ght
(ω)
is
one
of
the
e
ss
ent
ia
l
par
amet
ers
in
PS
O,
wh
ic
h
ofte
n
sign
if
ic
an
tly
the
aff
ects
conve
rg
e
nce
and
th
e
ba
lance
b
et
wee
n
the e
xploration
and
expl
oitati
on
cha
ra
cteri
sti
cs
of
PS
O.
Since
th
e
adopt
ion
of
thi
s
par
amet
er,
t
her
e
hav
e
be
en
la
rge
proposals
f
or
det
ermining
t
he
val
ue
of
Ine
rt
ia
W
ei
ght
Strate
g
y
.
In
orde
r
to
show
the
e
ffic
i
ency
of
thi
s
par
amete
r
i
n
the
Ec
onom
ic
Dispatch
proble
m
(ED),
th
is
pape
r
pr
ese
nt
s
a
comprehe
nsi
ve
rev
i
ew
of
on
e
or
m
or
e
tha
n
one
r
ecent
and
popul
ar
ine
r
ti
a
we
ight
stra
tegies
rep
ort
ed
in
the
re
la
t
e
d
li
te
r
at
ur
e.
Am
ong
thi
s
five
rec
en
t
ine
rtia
weight
four
were
ran
dom
l
y
chose
n
for
appl
icat
ion
and
subjec
t
to
empiric
a
l
studies
in
thi
s
rese
ar
ch,
name
l
y
,
Constant
(ω)
,
R
andom
(ω),
Glo
bal
-
Lo
ca
l
Best
(
ω),
Li
n
e
arly
D
e
cre
asing
(ω)
,
which
ar
e
the
n
compare
d
in
t
er
m
of
per
form
an
ce
wi
thi
n
the
co
nfine
s
of
the
discussed
opti
m
i
za
t
ion
probl
em.
Moreve
r,
the
res
ult
s
are
compare
d
to
those
rep
orte
d
in
the
r
ec
en
t
li
t
erature
a
nd
dat
a
from
SO
NELGAZ.
The
s
tud
y
r
esult
s
are
quite
enc
our
agi
ng
show
ing
the
good
appl
ica
bil
ity
of
PS
O
w
it
h
ada
p
ti
v
e
ine
rtia
we
ight fo
r
solving ec
ono
m
ic
dispatch
pro
ble
m
.
Ke
yw
or
d
s
:
Conver
ge
nce
Data F
ro
m
SONEL
GAZ
Eco
no
m
ic
D
ispatc
h
In
e
rtia
W
ei
gh
t
Partic
le
Sw
a
rm
Optim
iz
at
ion
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
:
Moh
am
m
ed
A
m
ine Mez
ia
ne
Tahr
i M
oham
m
ed
U
niv
e
rsity
,
BP 417,
0800
0 B
echar
, Alge
ria
.
Em
a
il
:
a
m
ine
m
oh
m
ezi
ane@gm
ai
l.co
m
1.
INTROD
U
CTION
Am
on
gs
t
the
diff
e
re
nt
issue
s
of
po
wer
s
yst
e
m
s
op
erati
on,
the
eco
no
m
ic
load
dis
pa
tc
h
(E
LD
)
pro
blem
is
on
e
of
the
key
too
ls
i
n
op
e
ra
ti
ng
a
nd
pla
nnin
g
of
m
od
ern
el
ect
ric
util
it
y
gr
id.
Esse
ntial
ly
,
el
ect
rical
gr
id
syst
e
m
s
are
i
nterc
onnected
and
c
onsist
of
powe
r
ge
ner
a
ti
ng
,
tra
ns
m
iss
ion
a
nd
distri
bu
ti
on
util
it
ie
s
in
orde
r
to
pro
duce
el
ect
rical
po
we
r
to
co
ns
um
ers,
at
a
low
pro
duct
ion
cost,
m
axi
m
u
m
reli
abilit
y
an
d
bette
r
operati
ng
co
ndit
ion
s
.
T
he
EL
D
is
a
sta
ti
c
pr
oble
m
,
it
was
first
discu
ssed
by
Ca
rp
e
ntier
in
19
6
2
[
1]
,
the
m
ai
n
pu
r
pose
of
EL
D
is
to
find
the
op
ti
m
al
ou
tp
ut
powe
r
of
ge
ne
rato
rs
to
m
ini
m
iz
e
the
total
gen
erati
on
cost
and
sat
isfy
the
eq
ualit
y
and
i
nequali
ty
const
raints.
To
so
l
ve
this
pr
ob
le
m
m
any
effor
ts
ha
ve
been
m
ade
over
the
ye
ars,
va
ri
ou
s
m
at
he
m
ati
c
al
pro
gr
am
m
i
ng
an
d
optim
izati
on
te
ch
ni
ques
we
re
us
ed
.
A
s
urvey
of
li
t
eratur
e
on
the
m
et
ho
ds
propose
d
to
so
lve
EL
D,
w
hich
ca
n
be
di
vid
e
d
into
tw
o
cat
ego
ries
,
th
e
cl
assic
(trad
it
ion
al
)
m
et
ho
ds
a
nd
t
he
sm
art
(h
euri
sti
c)
m
e
tho
ds.
It
is
ob
ser
ved
that
the
tradit
ion
al
m
et
ho
ds
and
heurist
ic
m
et
ho
ds
hav
e
so
m
e
limit
at
ion
s
to
so
l
ve
ELD
pro
blem
s.
The
tradit
ion
al
m
et
ho
ds
s
uffer
with
la
r
ge
execu
ti
on
ti
m
e
and
would
no
t
be
us
ef
ul
w
hen
t
he
cost
f
unct
ion
s
a
re
nonli
ne
ar.
S
o
in
som
e
cases,
it
w
il
l
be
ver
y
dif
ficult
to
achieve
op
ti
m
al
so
luti
on
s
.
F
or
this
reas
on,
recently
,
the
he
ur
ist
ic
m
e
thods
ha
ve
bee
n
us
e
d
to
overc
om
e
thi
s
pro
blem
[2
-
3]
.
Ther
e
f
or
e
in
recent
ye
ars
,
diff
e
re
nt
sm
art
an
d
in
nova
ti
ve
al
gorith
m
s
su
ch
as:
Gen
et
ic
Algorithm
(GA)
[
4]
,
Partic
le
Sw
arm
Optim
iz
at
ion
(PS
O)
[
5
-
6]
,
E
voluti
on
a
ry
Pro
gram
m
ing
Al
gorithm
(EP)
[
7]
, C
uckoo Sea
rch (CS
)
[
8]
, … ha
ve b
een
pro
po
se
d
t
o
s
olv
e
this
pro
blem
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
13
, N
o.
1
,
Ja
nu
a
ry 20
19
:
377
–
383
378
Re
centl
y,
m
a
ny
resea
rc
hes
ha
ve
bee
n
directed
to
wa
rd
s
the
ap
plica
ti
on
of
pa
r
ti
cl
e
swar
m
op
ti
m
iz
ation
te
chn
i
qu
e
to
s
olve
ELD
pro
ble
m
[9]
.
Th
e
m
os
t
i
m
po
rtant
a
dvanta
ges
of
t
he
PSO
are
t
hat
PSO
is
easy
to
im
ple
m
ent
and
the
re
are
few
pa
ra
m
et
ers
to
ad
ju
st.
I
n
this
a
rtic
le
,
an
at
te
m
pt
has
been
m
ade
to
s
olv
e
econom
ic
load
dis
patch
pro
bl
e
m
us
in
g
par
ti
cl
e
swar
m
opti
m
iz
at
ion
by
m
eans
of
m
ini
m
iz
at
io
n
of
f
uel
costs
wh
il
e
sat
isfyi
ng
ph
ysi
cal
a
nd
operati
onal
li
m
it
a
ti
on
s.
H
oweve
r,
the
pro
m
inent
m
od
el
to
be
discuss
e
d
in
thi
s
pap
e
r,
a
re
I
ner
t
ia
W
ei
gh
t
Str
at
egies,
an
d
thei
r
eff
ect
in
PS
O
for
so
l
ving
the
ELD.
In
orde
r
to
furthe
r
il
lus
trat
e
the
ef
fect
of
s
uch
m
echan
ism
in
PSO
f
or
so
lvi
ng
ELD
,
diff
e
re
nt
inerti
a
weig
ht
m
echan
ism
is
rev
ie
wed
an
d
exp
e
rim
ents
are
carried
out
over
si
ng
le
obj
e
ct
ive
m
ini
m
iz
a
ti
on
case
in
th
e
Re
al
W
est
A
lgeria
22
-
bus
s
yst
e
m
to
c
om
par
e
di
ff
e
ren
t
strat
eg
ie
s
of
set
ti
ng
I
ner
ti
a
Weig
ht.
M
or
e
over,
the
obta
ined
opti
m
a
l
resu
l
ts
al
so
com
par
ed
with
the
s
om
e
rep
ort
ed
resu
lt
f
ound
in
li
te
ratu
re
a
nd
with D
at
a
f
r
om
SO
NEL
G
AZ.
I
t
f
ound
th
at
the
PSO
ca
pa
ble
to
obta
in
lo
wes
t
cost
as
com
par
ed
t
o
oth
e
rs.
Thus,
it
has
great
pote
ntial
to
be
im
ple
m
ented
in
dif
fe
re
nt ty
pes of p
ower syste
m
o
pti
m
iz
at
ion
problem
.
2.
PAR
TI
CLE S
WA
RM OPTI
MIZ
ATION
PSO
is
a
po
pula
ti
on
-
base
d
op
ti
m
iz
ation
te
chn
i
qu
e
w
hich
was
first
int
rod
uced
by
K
enn
e
dy
an
d
Eberha
rt
in
19
95
[
10
]
,
in
sp
ir
ed
by
s
ocial
be
hav
i
or
of
bir
d
floc
king
or
fis
h
sc
hoolin
g
in
search
of
f
ood.
P
S
O
com
par
ed
to
ot
her
e
xisti
ng
he
ur
ist
ic
op
ti
m
izati
on
strat
egies
su
ch
as
ge
netic
al
go
rithm
,
is
easi
er
to
i
m
plem
ent
involvin
g
only
fe
w
par
am
et
ers
to
a
dju
st
wit
h
acc
ur
at
e
res
ul
ts
in
te
rm
of
c
al
culus.
I
n
a
P
SO
syst
em
,
pa
rtic
le
s
fl
y aro
und i
n
a
m
ul
ti
di
m
ension
al
searc
h spac
e.
Durin
g
flig
ht,
each
pa
rtic
le
a
dju
sts
it
s
trajec
tory
towa
rd
s
it
s
own
pr
e
vi
ous
best
po
sit
io
n
(Th
is
valu
e
is
cal
le
d
P
best
),
a
nd
to
wards
the
best
pr
e
vi
ou
s
posit
ion
at
ta
ined
by
an
y
m
e
m
ber
of
it
s
neig
hborh
ood
or
globa
ll
y,
the
w
ho
le
s
wa
rm
(Th
is
val
ue
is
ca
ll
ed
G
best),
[
11
-
17
]
.
The
tw
o
eq
uations
w
hi
ch
are
use
d
i
n
PSO
are v
el
ocity
up
date eq
uatio
n (
1)
a
nd
po
sit
io
n u
pd
at
e e
quat
io
ns
(2).
T
hese
a
re to
be
m
od
ifi
ed
at
eac
h
ti
m
e
step,
of PS
O
al
gorithm
to
conve
rge the
op
ti
m
u
m
so
l
utio
n.
t
X
i
t
G
b
es
t
i
r
c
t
X
i
t
P
b
es
t
i
r
c
t
V
i
t
V
i
[
2
2
1
1
1
1
1
1
t
V
i
t
X
i
t
X
i
2
Wh
e
re,
i
:
is
the
par
ti
cl
e
in
de
x;
:
is
the
in
erti
a
coeffic
ie
nt;
c
c
2
,
1
are
acce
le
ra
ti
on
c
oeffici
en
ts
2
2
,
1
0
c
c
;
r
r
2
,
1
are
rando
m
values,
r
r
2
,
1
0
reg
ene
ra
te
d
ever
y
vel
oc
it
y
up
date;
V
i
is
the
par
ti
cl
e’s
velocit
y
at
tim
e
t
;
X
i
is
the
pa
rtic
le
’s
posit
ion
at
tim
e
t;
P
b
e
s
t
is
the
par
ti
cl
e’s
i
nd
i
vi
du
al
best
s
olut
ion
as
of
tim
e t;
G
b
e
s
t
is t
he
s
war
m
’s
best
sol
ution
a
s
of
ti
m
e t.
Since
1995
m
any
at
te
m
pts
hav
e
bee
n
m
ade
to
im
pr
ove
t
he
perform
ance
of
the
ori
gi
na
l
PSO.
F
or
instance,
the
m
axi
m
u
m
velo
ci
ty
V
m
a
x
was
intr
od
uced
to
ar
bitra
rily
lim
i
t
the
ve
locit
ie
s
of
the
pa
rtic
le
s
an
d
i
m
pr
ove
the
re
su
lt
of
the
sea
r
ch.
T
he
i
ner
ti
a
weig
ht
(ω
)
is
on
e
of
P
SO
pa
ram
et
ers
or
igin
al
ly
pr
opos
e
d
by
Sh
i
and
E
berhart
[
18
]
t
o
br
i
ng
a
bout
a
balance
betwee
n
t
he
e
xplo
rati
on
an
d
e
xp
l
oitat
ion
c
ha
racteri
sti
cs
of
PSO.
Since
the
intr
oductio
n
of
this
par
am
et
er,
th
ere
ha
ve
bee
n
a
num
ber
of
pro
posal
s
of
dif
fer
e
nt
strat
egie
s
f
or
determ
ining
th
e v
al
ue
of i
ner
t
ia
w
ei
ght d
uri
ng a c
ourse
of
r
un.
3.
DIFFE
RENT
INERTI
A W
EIGHT
ADA
PTATIO
N M
ECHANI
SMS
The
balance
be
tween
global
a
nd
l
ocal
searc
h
thr
oughout
th
e
course
of
a
r
un
is
c
riti
cal
to
the
su
cce
ss
of
a
n
optim
iz
a
ti
on
al
gorithm
[
19
]
.
In
e
rtia
Weig
ht
play
s
a
key
ro
le
in
the
proces
s
of
balance
betwe
en
th
e
exp
l
or
at
io
n
a
nd
e
xp
l
oitat
ion
char
act
e
risti
cs
of
PS
O.
In
19
98
S
hi
a
nd
Ebe
r
har
t
[
18
]
prese
nted
for
t
he
fi
r
st
tim
e
the
co
nce
pt
of
i
ner
ti
a
w
ei
ght
by
intr
oducin
g
Con
st
a
nt
I
ner
t
ia
Weig
ht
in
w
hich
t
he
vel
ocity
of
each
pa
rtic
le
is
updated
acc
or
ding
to
the
e
quat
ion
(
1)
.
T
he
y
cl
aime
d
that
a
la
rg
e
inerti
a
weig
ht
facil
it
at
es
a
glo
bal
search
wh
il
e a sm
all I
ner
ti
a
W
ei
gh
t f
aci
li
ta
te
s a
local
search
. T
he follo
wing para
gr
a
phs r
e
pr
e
se
nt a r
evie
w
of
var
i
ous
inerti
a w
ei
gh
ts
in
P
SO ch
r
onol
og
ic
al
ly
.
Sh
i
a
nd
E
berh
art
[
20
]
pro
pos
ed
a
Co
ns
ta
nt
value
of
In
e
rtia
W
ei
gh
t
a
nd
exp
e
rim
ental
l
y
sho
w
that
w
from
[0
.8
,
1.2]
PSO
pro
vid
e
the
global
op
ti
m
u
m
in
a
reas
on
a
bly
of
it
era
ti
on
.
T
he
Ra
ndom
In
erti
a
Weigh
t
strat
egy
[
21
]
is
us
e
d
in
dy
nam
ic
env
iro
nm
ent
to
ena
bl
e
PSO
to
tra
ck
the
opti
m
a
an
d
inc
rease
s
the
conve
rg
e
nce
of the alg
ori
thm
i
n
ea
rly
it
erati
on
s.
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Impact
o
f i
nertia wei
ght st
ra
t
egies in
parti
cl
e swar
m op
ti
m
izati
on for
…
(
Mo
hamm
e
d A
mine M
ezi
ane
)
379
2
()
5
.
0
r
a
n
d
3
Wh
e
re
()
r
a
n
d
is a ra
ndom
n
um
ber
i
n [
0,
1];
is t
he
n a u
nif
or
m
r
an
dom
v
ariable i
n
the r
a
nge [
0.5,
1].
In
Tim
e
Var
yi
ng
In
e
rtia
W
ei
gh
t
Str
at
egies
[
22
]
the
val
ue
of
ω
is
deter
m
ined
base
d
on
it
erati
on
nu
m
ber
.
Th
e
se
m
e
tho
ds ca
n b
e eit
her
a
li
nea
r or
non
-
li
near
and inc
reasin
g or dec
reasin
g.
A
li
near
ly
Dec
reasin
g
I
ner
ti
a
W
ei
ght
[
23
-
25
]
was
intr
oduc
ed
to
im
pr
ove
the
perform
a
nce
of
PS
O
.
They
sug
gest
that
with
a
ω
from
the
ran
ge
0.9
to
0.4
the
P
SO
prov
i
des
e
xcell
ent
res
ults.
In
t
his
m
et
ho
d,
th
e
value o
f
ine
rti
a w
ei
ght
was de
creased
fro
m
)
m
a
x
(
to
)
m
i
n
(
accor
ding t
o
th
e f
ollow
i
ng equati
on
:
m
i
n
m
i
n
m
ax
m
ax
m
ax
i
t
er
i
t
er
i
t
er
i
t
er
4
Wh
e
re
i
t
e
r
the c
urr
ent it
erati
on of
the alg
or
it
hm
an
d
i
t
e
r
m
a
x
is
the m
axi
m
u
m
n
u
m
ber
of it
erati
ons.
In
[
26
]
,
Global
-
Local
Be
st
Inerti
a
W
ei
gh
t
is
pr
op
os
e
d
by
Arum
ug
am
and
Ra
o.
They
use
the
rati
o
of
the
local
best
and
gl
ob
al
bes
t
of
the
pa
rtic
le
s
in
eac
h
ge
ne
rati
on
to
dete
rm
ine
the
ada
pt
ive
inerti
a
we
igh
t
in
each it
erati
on
.
p
b
es
t
i
a
ve
r
a
g
e
g
b
es
t
1
.
1
5
Feng
et
al
.
[
27
-
28
]
pro
po
se
d
Cha
otic
I
nert
ia
W
ei
gh
t
us
ing
t
he
m
erit
s
of
c
ha
otic
op
ti
m
iz
at
ion
.
It
fou
nd
t
hat
the
CR
I
W
e
nhan
ces
the
perfor
m
ance
of
PS
O
in
c
om
par
iso
n
with
RI
W.
The
pro
po
se
d
w
is
a
s
fo
ll
ows:
z
i
t
e
r
i
t
e
r
i
t
e
r
i
t
e
r
m
i
n
m
a
x
m
a
x
m
i
n
m
a
x
6
The
s
umm
ary
of v
a
rio
us
i
ner
t
ia
w
ei
ght st
rate
gies are
d
is
pla
ye
d
in
Table
1.
Table
1.
Dif
fere
nt Inertia
Wei
gh
t
A
dap
ta
ti
on
Strategies
No
.
I
W
S
N
AME O
F INER
TIA
W
EI
GH
T S
TRATE
GI
ES
Fo
r
m
u
la of
inertia
weig
h
t
Ref
erence
1
Co
n
stan
t inertia w
eig
h
t
c
[
2
0
]
2
Ran
d
o
m
inert
ia we
ig
h
t
2
()
5
.
0
r
a
n
d
[
2
1
]
3
Linear
Decr
e
asin
g
inertia weigh
t
m
in
)
m
in
m
a
x
(
m
a
x
m
a
x
ite
r
ite
r
ite
r
ite
r
[
2
3
-
2
5
]
4
Glo
b
al
-
Local Bes
t
in
ertia
weig
h
t
p
b
e
s
t
i
a
v
e
r
a
g
e
g
b
e
s
t
1
.
1
[
2
6
]
5
Ch
ao
tic inertia wei
g
h
t
z
ite
r
ite
r
ite
r
ite
r
m
in
m
a
x
m
a
x
m
in
m
a
x
z
z
z
1
4
[
2
7
-
2
8
]
4.
OBJECTI
VE
The
inerti
a
w
ei
gh
t
strat
egie
s
hav
e
been
s
uggeste
d
to
im
pr
ov
e
both
exp
l
or
at
io
n
an
d
ex
plo
it
at
io
n
abili
ty
or
on
e
of
them
in
PSO
.
E
xp
l
oitat
ion
m
eans
that
al
l
par
ti
cl
es
conve
rg
e
to
the
sam
e
peak
of
the
obj
ect
ive
f
unct
ion
an
d
rem
ai
ns
there.
F
ur
the
rm
or
e,
the
expl
or
at
io
n
char
ac
te
risti
c
sh
ow
s
t
he
capa
bili
ty
o
f
th
e
al
gorithm
to
le
ave th
e
curr
ent
p
ea
k
a
nd lo
ok
i
ng for bett
er
sol
ution
s
.
Con
si
der
i
ng
th
e
above
cl
arifica
ti
on
s,
the
in
vestigat
or
aim
at
exp
lo
rin
g
th
e
i
m
pact
of
inerti
a
weigh
t
on
the
e
xplor
at
ion
a
nd
ex
pl
oitat
i
on
capa
bili
ti
es
in
PS
O
a
nd
sug
ges
t
a
bette
r
stra
te
gy
for
us
er
s
of
this
al
gorithm
within
the
a
rea
of
the
EL
D
prob
l
e
m
.
Ex
per
im
e
nts
ha
ve
bee
n
carried
out
on
fou
r
I
ne
rtia
Weig
ht
Strate
gies:
Co
ns
ta
nt
(ω),
Ra
ndom
(ω
),
Gl
ob
al
-
Local
Be
st
(ω
),
Li
near
l
y
Decr
easi
ng
(
ω)
in
t
he
co
nfi
ne
of
econom
ic
d
isp
at
ch
opti
m
iz
a
tio
n p
r
ob
le
m
f
or 22
bu
s
in
po
w
er
netw
ork real
, W
e
st Al
ger
ia
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
13
, N
o.
1
,
Ja
nu
a
ry 20
19
:
377
–
383
380
4.1
.
Ec
onomic
D
isp
at
c
h
The
ec
onom
ic
disp
at
c
h
pr
ob
le
m
, w
hich
is us
ed
to m
ini
m
iz
e
the cost
of pr
oductio
n of real
powe
r,
ca
n
gen
e
rall
y be st
at
ed
as
fo
ll
ows
:
n
i
P
i
F
i
M
i
n
1
7
P
L
P
D
n
i
P
i
1
8
P
n
P
i
P
n
m
a
x
m
i
n
9
Wh
e
re,
g
e
neral
ly
,
P
i
F
i
is a
qu
a
dr
at
ic
cu
r
ve
;
c
i
P
i
b
i
P
i
a
i
P
i
F
i
2
10
Her
e:
b
i
a
i
,
and
c
i
are
the
kn
own
c
oeffici
ent;
n
:
nu
m
ber
of
ge
ner
at
ors;
P
i
:
real
power
ge
ner
at
io
n;
P
D
:
rea
l
powe
r
loa
d;
P
L
: real
losses.
4
.2
.
E
xp
eri
m
ent
p
roce
dure
s
In
or
der
to
te
st
and
com
pare
so
m
e
diff
er
ent
inerti
a
we
igh
t
strat
egies
in
PSO
re
viewed
in
this
researc
h,
im
portant
opti
m
iz
ation
pro
blem
su
ch
as
sta
ti
c
eco
no
m
ic
disp
at
ch
for
22
bu
s
i
n
powe
r
net
work
real,
Wes
t
Alge
ria
are
us
e
d.
In
e
rt
ia
weigh
t
m
ec
han
ism
’s
influe
nce
on
the
E
LD
pr
ob
le
m
i
s
te
ste
d
in
te
r
m
s
of
conve
rg
e
nce s
peed an
d sol
ution q
ualit
y i
n
th
e PS
O
al
go
rith
m
.
The param
et
ers
set
ti
ng
s
of th
e ex
per
im
ent are
as
f
ollow
s:
Popu
la
ti
on
siz
e
(Swarm
siz
e)
is
100
pa
rtic
l
es.
T
he
m
axi
m
um
it
erati
on
al
lowe
d
nu
m
ber
of
f
unct
ion
evaluati
ons
is
200.
The
valu
e
of
acce
le
rati
on
pa
ram
et
ers
c1
an
d
c
2
are
ta
ken
e
qu
al
t
o
2.
Th
e
ex
pe
rim
ent
cond
ucted
i
n
t
he
EL
D
i
nv
est
i
gation
was
set
in
22 bus
syst
e
m
of
powe
r
net
work
real,
W
es
t
Alger
i
a
.
T
his lat
te
r
consi
sts
of
7
t
her
m
al
un
it
s,
15
l
oad
buses
and
31
tra
nsm
issi
on
li
nes,
03
c
om
pen
sat
or
var
sta
ti
c
SV
C
[
3*
(+40Mva
r
et
)
10
M
var)].
T
he
total
syst
e
m
de
m
and
is
856
M
W
.
F
or
im
pl
e
m
enting
thes
e
differe
nt
strat
egies
in
PSO,
the
pro
gram
m
ing
of
th
e
ELD
pro
blem
us
ing
t
he
P
SO
m
et
ho
d
ha
s
bee
n
de
velo
ped
a
nd
a
pp
li
e
d
usi
ng
MATLAB
soft
war
e
envir
on
m
ent, test
ed
on a
CORE i
5,
pers
on
al
c
om
pu
te
r wit
h 2.20 G
Hz
and
4 GO R
A
M.
5.
SIMULATI
O
N RESULT
A
ND D
I
SCUS
S
ION
The
fou
r
strat
egies
a
dopted
for
c
om
par
ison
s:
C
onsta
nt
In
e
rtia
W
ei
ght,
Ra
ndom
In
erti
a
Weig
ht,
Global
-
L
ocal
Be
st
In
erti
a
W
ei
gh
t
an
d
Li
ne
ar
Dec
reasin
g
In
e
rtia
W
ei
ght
are
sho
wn
i
n
T
able
2
pro
vidi
ng
the
best s
olu
ti
ons
of the E
LD
problem
.
Table
2
.
O
pti
m
iz
at
ion
Re
s
ults o
f
Di
ffren
t
I
ne
rtia
W
ei
ght St
r
at
egies in
PS
O for Ec
onom
ic
D
ispatc
h
Criterion
Co
n
stan
t
Ran
d
o
m
Glo
b
al
-
Local Bes
t
Linear
Decr
e
asin
g
P
1
[
MW]
320
320
320
1
8
2
.826
P
2
[
MW
]
140
140
140
1
9
2
.257
P
3
[
MW
]
100
1
0
2
.5166
1
0
0
.6703
1
5
4
.319
P
4
[
MW
]
1
0
4
.6458
1
0
2
.0303
1
0
3
.9486
150
P
5
[
MW
]
110
110
110
6
3
.71
9
8
P
6
[
MW
]
50
50
50
50
P
7
[
MW
]
80
7
9
.99
9
8
80
7
9
.99
8
6
Tra
n
smis
sio
n
Los
s
4
8
.64
5
8
4
8
.54
8
2
4
8
.61
8
9
1
7
.12
To
ta
l ou
tp
u
t
9
0
4
.6458
9
0
4
.5482
9
0
4
.6189
8
7
3
.1204
Load
de
m
an
d
856
856
856
856
Total Co
st [
$
/h
]
9
5
4
8
.9
9549
9
5
4
8
.9
8
9
9
9
.3
4
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Impact
o
f i
nertia wei
ght st
ra
t
egies in
parti
cl
e swar
m op
ti
m
izati
on for
…
(
Mo
hamm
e
d A
mine M
ezi
ane
)
381
Accor
ding to
t
he
a
bove
ta
ble,
w
e
no
ti
ce that
Con
sta
nt and
Global
-
L
ocal
Be
st In
e
rtia
Weigh
t
giv
es
us
the
sam
e
pr
oduction
c
os
t
a
nd
a
sli
gh
tl
y
lowe
r
of
0.1
$/
h
in
com
par
i
so
n
with
Ra
ndom
In
erti
a
Weig
ht,
transm
issi
on
l
os
ses
gi
ven
by
Ra
ndom
(ω)
is
l
ow
e
r
t
ha
n
th
a
t
gi
ven
by
Co
ns
ta
nt
(
ω)
a
nd
Globa
l
-
Local
Be
st
(ω
).
I
n
c
on
t
rast,
a
Line
arly
Decr
easi
ng
I
ner
ti
a
W
ei
ght
giv
es
a
m
uch
bette
r
producti
on
c
os
t
of
549.
66
[$
/
h]
and
m
ini
m
u
m
trans
m
iss
ion
loss
of
31.
5258
[M
W
]
,
in
com
par
ison
to
oth
er
strat
e
gies.
The
di
ff
e
rence
i
n
ge
ne
rati
on
cos
t
between
the
se
m
echan
ism
s
and
in
real
powe
r
loss
cl
early
sh
ows
the
ad
van
ta
ge
of
this
m
echan
ism
.
Figure
1
il
lustrat
es co
nver
gen
ce
ch
a
racteri
sti
cs of PSO
us
i
ng the
four I
ner
ti
a
Weig
ht Strate
gi
es.
(a)
(b)
(c)
(d)
Figure
1.
Co
nverg
e
nce c
ha
rac
te
risti
c o
f PS
O usin
g fou
r
ine
r
ti
a w
ei
gh
t
ad
ju
sti
ng
m
et
ho
ds
on (
a
)
C
on
sta
nt
(
ω)
;
(b)
Ra
ndom
(
ω)
; (c
) Glo
bal
-
L
ocal Be
st (
ω)
;
(
d) Linea
rly
d
ec
reasin
g (ω)
These
gra
ph
s
cl
early
ind
ic
at
e
that
PSO
co
nv
e
r
ges
ra
pid
l
y
to
a
hig
h
qual
it
y
so
luti
on
at
the
early
it
erati
on
s.
T
he
m
ini
m
iz
e
cost
and
powe
r
loss
ob
ta
ine
d
by
the
pro
posed
al
gorithm
is
le
ss
than
val
ue
re
porte
d
in
[
29
-
31]
usi
ng
the
e
voluti
onary
c
opulati
on
te
ch
niques,
gen
et
ic
al
gorit
hm
,
An
t
col
ony
op
ti
m
iz
a
ti
on
for
th
e
so
m
e test
syst
e
m
s.
In
or
der
to
de
m
on
strat
e
the
eff
ic
ie
ncy
an
d
the
rob
us
tness
of
the
propose
d
PS
O
an
d
the
per
f
orm
ance
of
usa
ge
the
in
erti
a
weigh
t
st
rategie
in
PS
O
fo
r
the
s
olu
ti
on
of
eco
no
m
ic
disp
at
ch
.
The
resu
lt
s
obta
ine
d
f
or
the
po
wer
net
work
real,
W
e
st
Alge
ria
22
0
kV
of
t
he
22
-
bus
a
re
c
om
par
ed
to
th
os
e
obta
ined
us
i
ng
Dat
a
from
SONEL
GA
Z
a
nd prese
nt in T
able 3.
0
20
40
60
80
100
120
140
160
180
200
9500
9600
9700
9800
9900
10000
10100
10200
I
t
e
r
a
t
i
o
n
B
e
s
t
C
o
s
t
[
$
/
h
]
0
20
40
60
80
100
120
140
160
180
200
9500
9600
9700
9800
9900
10000
10100
10200
10300
I
t
e
r
a
t
i
o
n
B
e
s
t
C
o
s
t
[
$
/
h
]
0
20
40
60
80
100
120
140
160
180
200
9500
9600
9700
9800
9900
10000
10100
I
t
e
r
a
t
i
o
n
B
e
s
t
C
o
s
t
[
$
/
h
]
0
0
.
5
1
1
.
5
2
2
.
5
3
x
1
0
4
10
3.
95
5
10
3.
95
7
10
3.
95
9
10
3.
96
1
10
3.
96
3
10
3.
96
5
10
3.
96
7
I
t
e
r
a
t
i
o
n
B
e
s
t
C
o
s
t
[
$
/
h
]
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
13
, N
o.
1
,
Ja
nu
a
ry 20
19
:
377
–
383
382
Table
3.
C
om
par
iso
n
R
es
ults
Criterion
Data F
ro
m
S
ONE
LGAZ
Linear
Decr
e
asin
g
P
1
[
MW]
200
1
8
2
.826
P
2
[
MW
]
200
1
9
2
.257
P
3
[
MW
]
300
1
5
4
.319
P
4
[
MW
]
80
150
P
5
[
MW
]
100
6
3
.71
9
8
P
6
[
MW
]
100
50
P
7
[
MW
]
10
7
9
.99
8
6
Tra
n
smis
sio
n
Los
s
2
1
.40
1
7
.12
To
ta
l ou
tp
u
t
890
8
7
3
.1204
Load
de
m
an
d
856
856
Total Co
st [
$
/h
]
9
1
0
4
.4
2
8
9
9
9
.3
4
Fr
om
the
ab
ov
e
ta
ble,
i
t
ap
pe
ars
that
P
SO
al
gorithm
wh
en
us
in
g
Li
near
ly
Decr
easi
ng
I
ne
rtia
W
ei
ght
giv
es
m
uch
be
tt
er
res
ults
tha
n
the
Data
f
rom
So
nelgaz.
T
he
differe
nce
i
n
gen
e
rati
on
c
os
t
a
nd
in
Re
al
pow
e
r
loss clearl
y s
hows
the
adva
ntage of
this m
eth
od.
6.
CONCL
US
I
O
N
In
this
pap
e
r,
a
com
par
at
ive
stu
dy
on
fou
r
sug
gested
i
ne
rtia
weig
ht
str
at
egies
was
c
onduct
ed
to
i
m
pr
ove
thei
r
i
m
pact
on
e
xp
l
or
at
io
n
a
nd
e
xploit
at
ion
abili
ti
es
in
pa
rtic
le
swar
m
op
ti
m
izati
on
al
gorith
m
ov
er
econom
ic
dispa
tc
h
pr
ob
le
m
.
These
strat
egie
s
are
C
onsta
nt
I
ner
t
ia
Weig
ht
,
Ra
ndom
In
e
rtia
W
ei
ght,
G
lob
al
-
Local
Be
st
In
e
rtia
W
ei
gh
t
an
d
Linear
D
ecr
easi
ng
I
ner
ti
a
Weig
ht.
The
r
esults
ver
ifie
d
and
prov
e
d
th
e
m
ai
n
obj
ect
ive
of
t
his
stu
dy
a
bout
t
he
im
pact
of
in
erti
a
weig
ht
on
the
pe
rfor
m
ance
of
PSO
f
or
op
ti
m
al
disp
at
ch.
A
s
an
ove
rall
ou
tc
om
e
of
the
ex
pe
rim
ents
resu
lt
s
carrie
d
out
as
sign
m
ent,
Line
ar
Dec
reasin
g
I
ner
ti
a W
ei
gh
t
i
s
the
best strate
gy fo
r
a
bette
r pro
duct
ion
c
os
t a
nd
a low t
ra
ns
m
is
sion l
os
ses
.
ACKN
OWLE
DGME
NT
Au
t
hors
w
ould
li
ke
to
than
k
the
hea
ds
of
Lab
or
at
or
y
of
An
al
ysi
s,
C
on
t
ro
l
an
d
O
pt
i
m
iz
at
ion
of
Ele
ct
ro
-
Ene
rg
e
ti
c
Syst
e
m
s
(CAO
S
EE)
a
nd
Sm
art
Gr
ids
the
ren
e
wabl
e
ener
gies
(
E
NERG
ARI
D)
at
the
un
i
ver
sit
y T
A
HRI
M
oh
am
m
ed of
Becha
r
(
Alge
ria).
REFERE
NCE
S
[1]
Hong
y
e
W
ang,
Carl
os
E.
Murillo
-
Sanche
z
,
Ra
y
D.
Zi
m
m
erman
and
Robe
rt
J.
Th
om
as.
On
Co
m
p
uta
ti
on
al
Iss
ues
of
Marke
t
-
Based
O
pti
m
al
Pow
er
Fl
ow.
IEEE Trans
ac
t
ions o
n
Pow
e
r
S
y
st
ems
,
Aug
2007
;
Vol
.
22(3
)
:
1185
-
1193
.
[2]
G.Sree
niva
san
,
Dr.
C.
H.Sa
iba
b
u,
Dr.S.Siv
ana
g
ara
ju
.
Solut
ion
of
D
y
n
amic
Econom
ic
Loa
d
Dispatc
h
(DE
L
D)
Problem
with
V
al
ve
Point
Lo
ading
Eff
ec
ts
and
Ramp
Rat
e
Li
m
it
s
Us
ing
P
SO
.
I
nte
rnational
Jou
rnal
of
El
ectric
a
l
and
Computer
E
ngine
ering
(
IJECE)
.
2011;
Vol
.
1
(1):59
-
70.
[3]
Hos
sein
Shahinz
ade
h
,
Sa
y
ed
Mohs
en
Nasr
-
Aza
dani,
Naz
er
eh
Janne
sari,
"
Applic
a
ti
ons
of
Parti
cl
e
Sw
ar
m
Optimiza
ti
o
n
Al
gorit
hm
to
Solv
ing
the
Ec
onom
ic
Lo
ad
Dispat
c
h
of
Units
in
P
ower
S
y
stems
with
Valve
-
Poin
t
Eff
ects"
,
Int
erna
ti
onal Journal
o
f
Elec
tri
cal and Com
pute
r E
ngin
ee
ring (
IJE
C
E)
.
2014;
Vol.
4,
pp.
858~867.
[4]
W
al
te
r
DC,
Sheble
GB.
Gen
e
ti
c
a
lgori
thm
soluti
on
of
e
con
om
ic
dispat
ch
with
val
ve
po
i
nt
loa
ding
.
IEEE
Tra
nsac
ti
ons on
Pow
er
S
y
stems
.
1993;
8(3):
1325
-
1332.
[5]
Jong
-
Bae
Park,
Ki
-
Song
Lee,
Jo
ong
-
Ri
n
Shin,
L
ee
K
Y
.
A
parti
cle
swar
m
opti
miz
ati
on
fo
r
ec
ono
mic
dispatc
h
wit
h
non
-
sm
ooth
cost
functions.
IEEE
Tr
ansacti
ons on
Powe
r S
yste
ms
.
2005;
20(
1)
:
34
-
42.
[6]
Ji
ang
S,
J
i
Z,
Shen
Y
,
“
A
novel
h
y
brid
par
ti
c
le
sw
arm
opti
m
iz
at
ion
and
gra
vitat
i
onal
sea
rch
al
go
rit
hm
for
solving
ec
onom
ic
emiss
ion
loa
d
dispatc
h
p
roble
m
s
wit
h
var
ious
pr
ac
t
i
ca
l
constraints
”,
Int
J
Elec
tr
Po
wer
Ene
rgy
Sys
t,
2014;
55:
628
–
4
4.
[7]
Venka
te
sh,
P
Gnana
dass
R,
Pa
dh
y
NP
.
Com
pa
rison
and
app
li
c
at
ion
of
evol
u
tionar
y
progra
m
m
ing
te
chni
qu
es
to
combined
ec
ono
m
ic
emiss
ion
dispat
ch
with
li
ne
flow
constra
in
t
s.
IEE
E
Tra
nsa
ct
ions
on
Pow
er
S
y
stems
.
2003
;
18(2):
688
-
697
.
[8]
N.
Kart
hik
,
A.K.
Parva
th
y
RA,
“
Non
-
conve
x
E
co
nom
ic
Loa
d
Dis
pat
ch
using
Cu
c
koo
Sear
ch
Algo
rit
h”
.
Indone
si
an
Journal
of El
ec
tr
ic
a
l
Eng
ineeri
ng
and
Com
puter S
ci
en
ce (IJEE
CS
)
,
2017;
5:
48
–
57.
[9]
R.
Chakr
aba
r
ti
,
P
K
Chat
top
adh
y
a
y
,
M
B
asu,
C
K
Panigra
hi
,
“
Pa
rti
cle
Sw
arm
Optimiza
t
ion
Te
chn
ique
For
D
y
nam
ic
Ec
onom
ic
Dispa
tc
h”
,
Jul
y
26,
20
05.
[10]
J.
Kenne
d
y
,
R.
C
.
Eb
erh
ar
t,
Particle
swar
m
opti
mization
.
IEEE
In
t
ern
ational
Confe
ren
ce
on
Neur
an
Networks
.
1995
;
pp
:
1942
–
1948.
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Impact
o
f i
nertia wei
ght st
ra
t
egies in
parti
cl
e swar
m op
ti
m
izati
on for
…
(
Mo
hamm
e
d A
mine M
ezi
ane
)
383
[11]
Djil
ani
Ben
Att
ous,
Yac
ine
La
b
bi
.
Parti
c
le
Swa
rm
Optimizati
on
based
Optimal
Powe
r
Fl
ow
for
Units
wit
h
Non
-
Smooth
Fue
l
Co
st
Func
ti
ons
.
Int
ern
ational
Conf
e
ren
ce
on
Elec
tr
i
ca
l
and
Elec
tron
ic
s
Engi
n
ee
ring
.
Bursa,
Turkey
.
2009
:
377
-
381.
[12]
Ebe
rha
r
t,
R
.
C.
,
and
Shi,
Y.
Co
mpar
ing
ine
rtial
wei
ghts
and
C
onstrict
ion
fa
ct
o
r
in
partic
l
e
Sw
arm
opti
mization
.
proc
ee
d
ing
of
t
he
2000
In
te
rn
a
ti
onal
Congr
ess
on
Eva
lu
at
ing
C
om
puta
ti
on,
San
Diego,
Ca
li
forn
ia
.
I
EE
E
Serv
ice
Cent
er
,
Pis
c
at
aw
a
y
;
2000
:
84
-
88
.
[13]
Sajj
ad
Ahm
adnia,
Ehsan
Ta
f
ehi.
Us
ing
Parti
cle
Sw
arm
Optimi
za
t
ion,
Gene
tic
Algorit
hm
,
Hone
y
B
ee
Mat
in
g
Optimiza
ti
o
n
a
nd
Shuffle
Fr
og
Leapi
ng
A
lgori
thm
for
S
olvi
ng
OP
F
Problem
with
t
hei
r
Com
par
iso
n.
TE
LKOM
NIK
A
Indone
sian
Jour
nal
of
E
le
c
tri
c
al
Engi
ne
eri
ng.
20
15.
Vol
.
15(3):4
45
-
451.
[14]
Kenne
d
y
.
J.
and
Ebe
rha
rt
,
R.
C.
Parti
cle
Swarm
Optimzation
.
Proce
ed
ing
of
the
1997
Inte
rna
t
ion
al
Confer
en
ce
o
n
Eva
lu
at
ion
ar
y
C
om
puta
ti
on,
IE
E
E
servi
ce Ce
nt
er
,
Pis
cata
wa
y
.
19
97;
pp:
303
-
308.
[15]
Yous
sef
M
OU
L
OU
DI,
Mohammed
A
m
ine
M
EZ
IAN
E,
Abde
ll
ah
LAOU
FI,Bousm
aha
BOUCHIBA,
Othm
a
ne
HA
RISI
.
A
Swa
rm
Algorit
hm
Inte
lligen
t
Opti
m
iz
at
ion
PS
O
i
n
Pow
er
Network
Rea
l
,
W
est
Alger
ia
2
20
kV
.
ELECTROT
EHNICĂ,
ELECTR
ONICĂ,
AUTO
MATICĂ(
EE
A)
.
2016
;
Vol
64
(
1
)
:
55
-
60
.
[16]
J.
C.
Bans
al
,
&
al
l
.
In
erti
a
W
eight
Strate
g
ie
s
i
n
Parti
c
le
Swar
m
Optimizati
on
.
IEE
E
Thi
rd
W
orld
Congress
o
n
Natur
e and
B
iologicall
y
Inspir
ed
Com
puti
ng
.
Sa
l
amanc
a
,
Spain
.
2011.
[17]
Sam
ar
Bashat
h,
Am
el
ia
Rit
ah
a
ni
Ism
ai
l,
Com
par
ison
of
Sw
arm
Inte
ll
ige
n
ce
Algorit
hm
s
for
High
Dim
ensiona
l
Optimiza
ti
o
n
Problems
.
Indone
s
ian
Journal
of
El
e
ct
rica
l
Eng
in
ee
ring
and
Com
pute
r
Scienc
e
.
2
018;
Vol.
11(1)
:
300
-
307.
[18]
Y.
Shi
and
R.
Ebe
rha
r
t.
A
modifi
ed
part
ic
l
e
swar
m
opti
mizer
.
In
Evol
uti
on
ar
y
Com
puta
t
ion
Proce
edi
ngs,
19
98.
IEE
E
W
orld
Co
ngre
ss
on
Com
puta
ti
on
al Int
e
ll
ig
enc
e
.
2002
;
pag
es
69
–
73.
[19]
Y.
Shi,
R
.
Eb
erh
art
,
Fuzzy
adapt
i
ve
parti
cle
swar
m
opti
mization
.
In
Congress
on
Evol
uti
on
ar
y
Co
m
puta
ti
on
.
Seou
l
,
Korea
;
2001.
[20]
R.
C.
Eb
erh
ar
t
an
d
Y.
Shi
.
Tr
ac
ki
ng
and
opti
mizin
g
dynamic
syste
ms
wit
h
partic
le
swar
ms
.
Proce
edi
ngs
of
the
2001
Congress on
Ev
olut
iona
r
y
Com
puta
ti
on
.
2002
.
V
o
l
1
:
94
–
100
.
[21]
Y.
H.
Shi,
R.
C.
Ebe
rha
r
t
.
A
mod
if
ie
d
parti
cle
swar
m
opti
mizer
.
in:
IEE
E
Int
ern
a
t
iona
l
Confer
ence
on
Evol
uti
on
a
r
y
Com
puta
ti
on,
Anchora
ge
Alaska
,
1998
,
pp
.
69
–
7
3.
[22]
Ahm
ed
Nicka
badi
et
al.
A
nove
l
partic
le
swar
m
opti
mization
alg
orithm
wit
h
adapti
v
e
ine
rtia
weight
.
Applie
d
Sof
t
Com
puti
ng.
201
1:
3658
–
3670.
[23]
J.
Xin,
G.
Chen,
and
Y.
Hai.
A
Parti
cl
e
Swar
m
Optimizer
wit
h
Mult
istage
Linearly
-
Dec
reasi
ng
Ine
rtia
We
ig
ht
.
Inte
rna
ti
ona
l
Joi
nt
Confer
ence
o
n
Com
puta
ti
on
al Sci
en
ce
s
and
O
pti
m
iz
ation.
CS
O 2009.
Vol
1:
5
05
–
508.
[24]
R.
C.
Eb
erh
art,
Y.H.
Shi.
Compar
ing
ine
rtia
we
ig
hts
and
constric
ti
on
fac
tors
in
partic
le
swar
m
opti
mization
.
I
n:
IEE
E
Congress
on
Evol
u
ti
onar
y
Com
puta
ti
on.
20
00:
pp.
84
–
88.
[25]
Y.H.
Shi,
R.
C
.
E
ber
har
t
.
Expe
rim
ent
a
l
stud
y
o
f
pa
rti
cle
sw
arm opt
i
m
iz
at
ion
.
In
:
Co
nfe
ren
c
e, Orl
and
o,
2000
.
[26]
M.S.
Arum
uga
m
and
M
VC
Ra
o.
On
the
performance
of
the
partic
le
swar
m
opti
mization
algo
rithm
wit
h
various
Ine
rtia
We
ight
v
ariants f
or c
omp
uti
ng
opti
mal
co
ntrol
of
a
cl
ass
o
f
hybrid
systems
.
Discre
te
D
y
n
amics
in
Natur
e
and
Socie
t
y
,
2006
.
[27]
Y.
Feng,
G.
T
e
ng,
A.
W
ang,
Y.M.
Yao.
Cha
oti
c
in
erti
a
we
i
ght
in
particle
swar
m
opti
miza
ti
on
.
In:
Se
con
d
Inte
rna
ti
ona
l
Co
nfe
ren
c
e
on
Inno
vat
iv
e
Com
puti
n
g,
Inform
ation
a
nd
Control (ICI
CIC
07), 2007,
p
p:
475
–
1475.
[28]
Y.
Feng,
Y.M.
Yao,
A.
W
an
g.
Compar
ing
wit
h
chaotic
in
erti
a
wei
gh
ts
in
partic
le
swar
m
opti
mi
zation
.
In:
Inte
rna
ti
ona
l
Co
nfe
ren
c
e
on
Mac
hine
Le
a
rning an
d
C
y
b
ern
e
ti
cs
,
A
ugust
2007.
pp
: 329
–
333.
[29]
W
.
Ongs
akul
,
T
.
Ta
nt
imaporn.
O
pti
m
al
powers
fl
ow
b
y
improved
evol
ut
iona
r
y
pr
ogra
m
m
ing.
El
e
ct
.
Powe
r
Com
p
.
and
Syst
.
2006;
Vol.
34:
pp.
79
-
95.
[30]
J.
Yury
evich,
K.
P.
W
ong,
Evol
uti
onar
y
Program
m
ing
Based
Optimal
Pow
er
Flow
Algorit
hm
.
IEEE
Tr
ansacti
on
o
n
power
Syste
ms
.
1999;
Vol.
14,
N
o.
4
.
[31]
C.
Thi
tithamron
gcha
i
,
B.
Eua
-
ar
porn.
Selfa
dap
tive
Diffe
ren
t
ia
l
E
volut
ion
Based
Optimal
Pow
er
Flow
for
Units
with
Non
-
sm
ooth
Fu
el
Cost Func
ti
ons.
J
.
El
e
ct
rica
l
S
yste
ms
.
2007:
88
-
9
9.
Evaluation Warning : The document was created with Spire.PDF for Python.