TE
LKOM
NI
KA
Te
le
c
om
munica
tion,
C
omp
u
tin
g,
El
e
ctroni
cs and
Contr
ol
Vo
l.
19
,
No.
3
,
June
2021
,
pp.
1001
~
1009
IS
S
N: 16
93
-
6930, acc
red
it
ed
First G
ra
de by
Keme
nr
ist
ek
di
kti, D
ec
ree
N
o: 21/E/
KP
T/
2018
DOI: 10.
12
928/
TELK
OMN
I
KA.v1
9i3
.
18777
1001
Journ
al h
om
e
page
:
http:
//
jo
ur
nal.
uad.ac
.id
/i
nd
ex.
php/TE
LKOMNIKA
Extend
ed stat
e obs
erver
based lo
ad freq
uenc
y contr
oller fo
r
three ar
ea inte
rconnecte
d pow
er syst
em
Van
Van H
uynh
1
,
P
hong
Th
an
h
Tran
2
, Tu
an
Anh Tr
an
3
,
Dao Huy Tu
an
4
,
V
an
-
Du
c P
ha
n
5
1
,2,
3
,4
Fa
cul
ty
of Electrical
and
E
l
ec
tron
ic
s E
ng
ineeri
ng,
Ton
Duc Thang
Univ
ersity
,
Ho Chi
Minh
Cit
y,
Viet
n
am
5
Facul
ty
of
Auto
mobi
le Technol
o
gy,
Van
La
ng
U
nive
rsity
,
Ho Ch
i
Minh C
it
y,
Vie
tna
m
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
J
un
15, 2
020
Re
vised
N
ov 23, 2
020
Accepte
d
Dec
5,
2020
In
thi
s
p
ape
r
,
w
e
dev
el
op
a
new
extende
d
state
var
ia
b
le
observe
r
base
d
loa
d
fre
quenc
y
cont
r
oll
er
(
LFC
)
sc
hem
e
for
thr
ee
-
are
a
int
er
connect
ed
power
sys
te
ms.
T
he
e
xte
nded
sta
te
o
bserve
rba
sed
lo
ad
fr
eque
n
cy
co
ntrol
lers
a
r
e
deve
lop
ed
whic
h
ut
il
i
ze
distur
banc
e
esti
m
at
io
n
t
ec
hni
ques.
T
he
propos
e
cont
rol
appr
o
ach
assures
t
hat
th
e
flu
ct
u
at
ing
th
i
ngs
of
th
e
loa
d
fre
quencie
s
rea
ch
es
to
a
saf
e
r
ran
ge
and
the
l
oad
fre
qu
encie
s
ca
n
al
so
be
m
ad
e
a
t
a
ver
y
mi
nimal
not
to
h
ave
an
eff
ect
on
power
qua
li
ty
an
d
power
flow
in
mul
ti
-
area
int
er
conne
c
te
d
power
sys
te
m
.
Th
e
r
esult
s
of
th
e
si
mul
a
t
ions
using
MA
TL
AB/S
IM
ULINK
done
did
not
only
add
r
ess
tha
t
th
e
pro
posed
newly
cont
rol
method
works
eff
ectivel
y
but
al
so
cha
n
g
e
powerful
ly
th
e
par
amete
r
var
iations
of
th
e
interc
onne
cted
are
as
of
th
e
p
ower
sys
te
m.
E
spec
ially,
it
works
ver
y
well
to
l
im
i
t
disturb
anc
es
im
p
act
on
in
te
rco
nn
ecte
d
are
as
in
th
e
sys
te
m.
The
r
efo
re,
the
p
erf
ormance
of
int
er
con
nec
t
ed
power
s
ystem
und
er
diffe
r
en
t
mu
lt
i
-
condi
ti
ons
is
simul
ated
with
t
he
new
con
trol
me
thod
t
o
dem
onstra
te
the
fea
sibi
li
ty
of the s
ystem
.
Ke
yw
or
d
s
:
Ar
ea
contr
ol erro
r
Fu
ll
-
order st
at
e obser
ve
r desig
n
Loa
d
f
re
qu
e
nc
y
c
on
t
ro
l
M
ulti
-
a
rea
power syste
m
Tie
-
li
ne
area
in
pow
e
r
s
ys
te
m
This
is an
open
acc
ess arti
cl
e
un
der
the
CC
BY
-
SA
l
ic
ense
.
Corres
pond
in
g
Aut
h
or
:
Dao H
uy T
ua
n
Faculty
of Elec
tric
al
and
Ele
ct
ronics E
nginee
rin
g
To
n Du
c
Th
a
ng
Un
i
ver
sit
y,
19 Ng
uy
e
n H
uu T
ho Street,
T
an
P
hong
Wa
r
d,
Distric
t 7
, H
o
C
hi M
i
nh Ci
ty, Vie
tnam
Emai
l:
daoh
uy
t
uan@tdt
u.
e
du.
vn
1.
INTROD
U
CTION
In
m
ulti
-
area
interc
onnected
powe
r
s
ys
te
m
,
sta
bili
ty
of
frequ
e
nc
y
is
a
very
sig
nif
y
i
nd
ic
at
or
of
powe
r
qual
it
y
as
the
load
s
of
the
ti
e
-
li
ne
mu
lt
iple
are
as
kee
p
incre
asi
ng
a
nd
cha
ng
i
ng
pro
gr
e
s
sively.
M
ea
nwhile
, loa
d
distu
r
ban
ce
s can
occ
ur
s
udde
nly
w
hich
c
an
cause
de
viati
on
s in
ti
e
-
li
ne powe
r
area e
xc
hange
and
nomi
nal
f
r
equ
e
ncies
i
ns
ta
bili
ty
[
1
,
2].
Loa
d
f
re
qu
e
nc
y
c
on
tr
ol
for
ye
ars
now
has
been
one
of
th
e
basic
rob
us
t
co
ntr
ol
mecha
nisms
i
n
la
r
ger
sc
al
e
powe
r
el
ect
ric
sy
ste
ms
with
interco
nnect
e
d
area
an
d
t
he
mai
n
crit
eria
of
t
he
load
f
reque
ncy
co
ntr
ol
is
t
o
keep
a
nd
mai
nt
ai
n
the
syst
em
f
re
qu
e
nc
y
un
if
or
m
at
it
s
nomina
l
value
duri
ng
a
nd
w
he
n
the
re
is
load
cha
nge.
A
la
rg
e
po
wer
el
ect
ric
sy
ste
m
can
be
di
vide
d
a
nd
se
par
at
ed
int
o
sever
al
loa
d
f
r
equ
e
nc
y
a
rea
c
on
t
ro
l
that
is
i
nterc
onnected
by
ti
e
-
li
nes
a
nd
it
relat
es
to
operati
onal
proc
edures
to
be
f
ollow
e
d
in
the
e
ve
nt
of
majo
r
fau
lt
s
or
of
tie
-
li
ne
powe
r.
Ge
ner
al
l
y,
t
he
mai
n
go
al
and
duty
of
load
fr
e
qu
e
nc
y
c
on
t
ro
l
is
sim
ply
to
a
dju
st
the
f
r
equ
e
ncies
of
the
se
par
at
e
d
a
reas
a
nd
t
o
si
mu
lt
ane
ously
modu
la
te
powe
r
fl
ow
i
ng
across
the
ti
e
-
li
nes
acco
rd
i
ng
with
the
ag
r
eement
of
an
i
nter
a
rea
powe
r
s
ys
te
m
.
Mor
eov
e
r,
load
fr
e
quenc
y
co
ntr
ol
nor
ma
li
zes
fr
e
qu
e
nc
y,
mainta
in
s
dynamics,
an
d
m
akes
qual
it
y
as
su
ra
nce
of
the
powe
r
su
ppl
y,
require
s the
us
e
of
loa
d fr
e
qu
e
nc
y
c
ontr
oller (
L
FC
)
scheme
in
[3
-
6].
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
1693
-
6930
TELK
O
M
N
IKA
Tel
ec
om
m
un
C
ompu
t El
C
on
t
ro
l
,
V
ol.
19
,
No.
3
,
June
20
21
:
10
01
-
100
9
1002
Loa
d
fr
e
quenc
y
c
ontrol
is
al
s
o
gr
eat
l
y
si
gn
if
ic
ant
in
po
wer
el
ect
ric
sy
ste
m
operati
on
al
f
or
deliveri
ng
reli
able
a
nd
e
ffi
ci
ent
po
wer
w
it
ho
ut
poor
qu
al
it
y.
T
her
e
f
ore,
a
novel
co
nt
ro
l
te
ch
nique
ne
eds
to
be
de
ve
lop
e
d
in
ot
her
t
o
achi
eve
LFC
ai
m
s,
there
by,
maint
ai
nin
g
an
d
s
us
t
ai
nin
g
reli
abili
ty
of
the
el
ect
ri
cal
power
sy
st
em
in
mu
lt
i
-
areas
.
I
n
order
to
s
olv
e
ab
ove pro
blem
s,
resea
rc
her
s
a
nd
co
ntr
ol
e
ng
i
neer
s
ha
ve
pro
po
s
ed
wide
ra
nge o
f
methods
of
st
at
e
-
of
-
the
-
art
l
oad
fr
e
quenc
y
c
on
tr
ollers
to
b
e
a
ppli
ed
in
m
ulti
-
area
int
ercon
nected
powe
r
sy
ste
m
[7
-
15
].
That
is,
the
me
thod
of
a
dap
ti
ve
con
t
ro
l
propose
d
to
ta
ke
ca
re
pa
rameter
va
riat
ion
in
[7
-
9].
Bu
t
it
cannot
le
ad
strongl
y
t
o
a
ge
ner
al
so
l
utio
n
to
the
pr
ob
le
m
faced
by
L
F
C
in
po
wer
s
yst
em.
M
a
ny
kin
d
s
of
con
t
ro
l
te
ch
ni
qu
e
s
uc
h
as
pr
oport
ion
al
inte
gr
al
(
PI
)
or
pr
oport
ion
al
inte
gr
al
de
rivati
ve
(
PID
)
c
ontr
oller
f
or
LFC
we
re
us
e
d.
I
n
the
presc
ribe
d
e
nv
i
ron
ment,
some
fa
ct
or
s
li
ke
un
ce
rtai
nties
ma
ke
it
diff
ic
ult
to
a
pp
l
y
th
e
above
-
me
ntioned LFC
tech
niq
ue
s i
n p
racti
ce, which
impli
es [10
-
12
]
to i
nt
rodu
ce
some i
nter
nal m
od
el
con
t
ro
l
to
desig
n
PID
ty
pe
L
FC
c
ontr
ollers.
I
n
[10]
pr
opos
e
d
f
uzzy
P
rop
or
ti
onal
I
nteg
ral
c
ontr
ollers
for
L
FC
of
powe
r
el
ect
ric
sy
ste
ms
.
W
hile
the
di
ff
e
ren
t
methods
s
how
that
it
is
po
s
sible
to
e
nh
a
nce
the
pe
rfo
rma
nc
e
of
LFC
in
s
pecifi
ed
en
vir
onme
nt
s.
The
P
I
D
co
ntr
oller
pro
pos
ed
in
[11]
co
m
bin
e
d
with
ne
w
str
uctu
re
show
n
to
be
rob
us
tl
y
a
nd
to
en
ha
nce
t
he
dam
ping
of
the
powe
r
el
e
ct
ric
sy
ste
m
tr
aci
ng
with
a
s
mall
sign
i
fican
t
ste
p
changes
i
n
loa
d.
T
he
st
rateg
y
of
tu
ni
ng
is
base
d
s
olely
on
t
he
ma
xim
um
pea
k
res
onance
of
the
s
ys
te
m
sp
eci
ficat
io
ns
in
[
12].
Howe
ver,
m
os
t
c
on
ven
ti
onal
PID
co
ntr
ollers
with
gai
ns
fixe
d
was
desi
gned
unde
r
nominal
sy
ste
m
op
e
rati
ng
c
onditi
ons,
the
se
le
ct
ed
m
ode
of
it
s
gains
i
s
usual
ly
al
ways
on
t
rial
an
d
er
ror
wit
h
no
anal
ytica
ll
y
meth
ods
of
de
te
rmin
in
g
it
s
par
a
mete
rs
opti
mall
y
w
hich
s
om
et
imes
fail
s
to
giv
e
out
th
e
be
s
t
and
acc
ur
at
e
con
t
ro
l
sc
hem
es
and
perf
ormance
ov
e
r
a
wider
range
of
s
ys
te
m
op
e
rati
ng
c
onditi
ons
an
d
exh
i
bits,
t
her
e
by,
a
poor
dyna
mic
res
ponse
and
performa
nc
e
du
rin
g
oper
at
ion
s.
S
om
e
of
th
e
c
ontrol
m
et
hods
need
to
a
pp
l
y
the
f
ull
-
sta
te
of
t
he
c
ontro
l
area
a
s
feedback
in
put
whil
e
so
me
will
le
ad
to
higher
-
orde
r
con
t
ro
ll
ers
,
th
ese
facto
rs
we
re
to
o
co
mp
le
x
to
be
under
s
tood
an
d
c
ompre
he
nd
e
d
by
con
t
ro
l
a
nd
el
ect
rical
eng
i
neer
s
in [
13
,
14]
.
Among
va
rio
us
co
ntr
ol
te
ch
ni
qu
es
me
ntioned,
t
he
opti
mal
co
ntr
ol
with
s
ta
te
feedback
t
echn
i
qu
e
is
on
e
of
t
he
best
op
ti
ons.
Als
o,
a
robu
st
dece
nt
rali
zed
li
near
con
t
ro
ll
er
was
app
li
ed
in
[
15
-
19].
Anothe
r
c
on
t
ro
l
te
chn
iq
ue
to
lo
ok
at
is
the
“
sl
iding
m
ode
c
ontr
ol
(
SM
C
)”.
Sli
din
g
m
ode
con
t
ro
l
te
ch
niques
is
a
no
t
her
bette
r
way
a
nd
ap
pro
ach
t
o
s
olv
e
L
FC
pro
blem
.
S
M
C
has
ofco
ursed
bee
n
a
ppli
ed
for
LFC
in
powe
r
el
ect
ric
sy
ste
m
in [19
-
25]
to
a
chie
ve fast
re
s
pons
e a
nd
rob
ust
ly p
e
rforman
ce in the
po
wer netw
ork, t
his
method is a
no
nlinear
con
t
ro
l
strat
eg
y
with
a
fam
ous
r
ule.
It
is
al
so
i
ns
e
ns
it
ive
perha
ps
t
o
c
ha
ng
e
s
of
the
plant
par
a
mete
rs
and
as
well
impro
ves
sy
ste
m
tra
ns
ie
nt
co
ntr
ol
pe
rformance
.
T
he
above
a
ppr
oac
hes
a
re
achie
ve
d
under
assu
mp
ti
on
that
al
l
s
ys
te
m
sta
te
va
riables
are
t
o
be
me
as
ur
a
ble
and
re
a
di
l
y
avail
able
for
fee
db
ac
k.
I
n
fact,
no
t
al
l
s
ys
te
m
sta
te
var
ia
bles
are
meas
ur
a
bl
e
for
fee
dback
s,
an
d
the
n
we
need
t
o
est
im
at
e
the
sta
te
va
riables
that
ar
e
no
t
unmeas
ur
a
ble.
Esti
mati
ng
un
measu
rab
le
sta
te
var
ia
bles
is
of
te
n
cal
le
d
obser
vatio
n
in
[
21
-
25].
T
his
s
cheme
can
a
dap
t
t
he
unknow
n
uppe
r
bo
unds
of
m
at
ched
nonline
arit
y
an
d
distu
rb
a
nce.
It
gets
not
only
t
he
s
ys
te
m
sta
te
traj
ect
or
ie
s
accom
plish
ment
but
al
so
sat
isfie
s
in
pa
rameters
of
t
he
syst
em
sta
te
error
s
.
The
work
il
lustrate
d
a
bove,
achie
ve
a
sign
ific
a
nt
re
s
ult
relat
ed
to
LFC’s
of
inte
r
connecte
d
power
syst
ems
a
pp
l
ying
var
i
ou
s
contr
ol
techn
i
qu
e
s.
Howe
ver,
the
r
e
are
some
li
mit
at
ion
s
of
the
a
bove
a
pp
r
oach
e
s.
Firstl
y,
the
dist
urbance
s
are
not
tru
ncated
from
the
outp
ut
poi
nts
in
ste
ad
y
s
ta
te
.
Seco
nd
l
y,
the
co
ntr
oller
gains
a
re
not
set
to
be
e
xtre
mely
high
to
at
te
nu
at
e
disturba
nce
s
of
unkn
own
bounda
ries.
T
hi
rd
ly
,
the
c
on
t
r
oller
is
desi
gned
in
acc
orda
nc
e
to
the
nomin
al
transf
e
r
f
unct
io
n
of
the
plant
wh
e
n
a
no
-
loa
d
dist
urban
ce
is
con
si
der
e
d.
In
orde
r
to
s
olv
e
t
he
above
li
mit
at
ion
s,
in
t
he
pa
pe
r
we
dev
el
op
ne
wly
e
xten
ded
sta
te
ob
se
rv
e
r
-
base
d
L
FC
scheme
for
th
ree
-
are
a
interco
nnect
ed powe
r
el
ect
ric
sy
ste
m.
T
he
m
ai
n
co
ntri
bu
ti
ons
of the
pa
per are as
foll
owing:
−
Exten
de
d
sta
te
obser
ver
is
fir
st
desig
n
t
o
es
ti
mate
the
un
measu
rab
le
syst
em
sta
te
va
riables
a
nd
al
s
o
th
e
load u
ncer
ta
int
ie
s.
−
The
e
xten
de
d
sta
te
ob
s
er
ver
-
base
d
loa
d
fr
e
qu
e
nc
y
c
ontrol
le
rs
are
de
vel
op
e
d
w
hi
ch
ut
il
iz
e
disturban
ce
est
imat
ion
te
ch
niques.
T
hus,
t
he
co
ntr
oller ga
ins
are not
set
to
be
e
xtre
mely
hi
gh
t
o
at
te
nuat
e
distu
r
ban
c
es
of un
known
bo
unda
ries,
wh
ic
h
is
very
us
e
fu
l
in
loa
d fr
e
que
ncy co
ntr
oller
desig
n.
−
The
simulat
io
n
res
ults
i
nd
ic
at
e
that
t
he
pr
op
os
e
d
ne
wly
m
et
hod
im
pro
ve
s
the
s
ys
te
m
dyna
mic
respo
nse
wh
ic
h pro
vid
es
a s
ys
te
m c
on
t
r
ol to
a
da
pt a
nd meet u
p
t
he
L
FC re
qu
i
reme
nt
.
The
oth
er
pa
rts
of
t
he
pro
po
sed
pa
pe
r
are
structu
re
d
in
t
he
f
ollo
wing.
A
mathe
mati
cal
dynamics
model
of
t
hr
e
e
-
area
power
el
e
ct
ric
sy
ste
m
is
pr
ese
nte
d
in
sect
io
n
2.
The
f
ollo
wi
ng
pro
posed
e
xten
ded
ob
s
er
ver
c
on
tr
oller
is
sho
wed
in
sect
io
n
3.
T
he
res
ults
of
th
e
va
rio
us
simu
la
ti
on
s
of
th
ree
-
area
po
wer
s
yst
em
app
l
ying
the
pro
posed
new
l
y
ba
sed
c
on
t
ro
l
ap
proac
hes
ar
e
de
scri
bed
in
sect
io
n
4.
La
stl
y,
c
oncl
us
i
ons
ar
e
discusse
d i
n
se
ct
ion
5.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELK
O
M
N
IKA
Tel
ec
om
m
un
C
ompu
t El
C
on
t
ro
l
Ext
end
e
d
st
ate
ob
s
erver
based
load
fre
quenc
y co
ntro
ll
er f
or
…
(
V
an Van
H
uynh
)
1003
2.
THE
MATHE
MA
TI
C
AL MODEL
O
F
T
HREE
-
AR
E
A
INTER
CON
NECTED
PO
WER S
YS
TE
M
MO
DEL
At
first,
we
analyze
L
FC
of
powe
r
s
ys
te
m
of
m
ul
ti
-
areas.
Fi
gur
e
1
il
lustrate
s
three
-
area
interco
nnect
ed
sy
ste
ms
[16]
a
s
desm
os
trat
e
d.
The
s
ole
miss
ion
in
t
he
rese
arch
wor
k
is
simply
t
o
exa
mine
th
e
var
i
ou
s
ti
e
-
li
ne
area
s
i
n
pow
e
r
s
ys
te
ms
i
n
ot
her
to
c
ontrol
t
he
f
reque
ncies
of
the
f
ollo
wi
ng
mu
lt
i
-
area
and
t
o
regulat
e
sim
ultaneo
us
l
y
powe
r
fl
ow
i
ng
th
rought
the
ti
e
-
li
ne
s
acco
rd
i
ng
t
o
a
n
inte
r
-
a
rea
op
e
rati
ng
a
gr
e
ement.
In
th
ree
-
a
rea
netw
orks,
eac
h
c
ontrol
area
is
in
dicat
ed
by
a
t
urbi
ne,
ge
ner
at
or
,
a
nd
gove
rnor
s
ys
te
m.
T
he
tie
-
li
ne
po
wer
in
the
inte
rcon
nected
th
ree
-
a
r
ea
m
us
t
be
c
onside
red
to
ge
ner
at
e
the
incr
ement
powe
r
s
ta
bili
ty
equ
at
io
n o
f
eac
h powe
r
s
ys
te
m ar
ea
, s
inc
e t
her
e
is po
wer f
low
i
n
eac
h
a
re
a thro
ugh
t
he
ti
e li
ne.
Fr
om
[3
,
4], th
e ti
e li
ne
po
we
r
inc
reme
nt tha
t i
s out of a
rea
is:
(
.
)
=
2
∑
(
∫
−
∫
)
∈
≠
(
1
)
To
a
chie
ve
ba
la
nce
betwee
n
i
nterc
onnect
ed
c
ontr
ol
are
as,
the
fr
e
qu
e
ncy
dev
ia
ti
on
an
d
ti
e
-
li
ne
powe
r
fluctuati
on
are
detect
ed
in
othe
r
to
dete
rmin
e
the
area
co
ntr
ol
error
(A
CE
)
of
eac
h
co
ntr
ol
area.
The
AC
E
can
be
e
xpres
sed
f
or
eac
h
c
ontr
ol
area
as
a
li
ne
ar
it
y
c
ombina
ti
on
of
ti
e
-
li
ne
po
wer
flu
nctu
at
ion
an
d
f
requen
c
y
dev
ia
ti
on.
=
+
(
2
)
Figure
1.
A
simpli
fied
sket
ch of
3
-
area
inte
rcon
nected
po
wer syste
m
A
po
wer
netw
ork
is
sta
ble
if
and
only
if
A
CE
i
=
0.
I
n
ot
he
r
w
ords:
and
te
nd
t
ow
a
r
ds
z
ero
ov
e
r
a
s
hort
ti
me
fr
ame
.
I
n
pr
act
ic
e,
a
t
ypic
al
interco
nne
ct
ed
gen
e
rati
on
s
ys
te
m
is
nonlinea
r
ity
an
d
al
s
o
dynamic
s
;
a
pp
li
cat
ion
of
the
li
near
it
y
mod
el
is
al
lowa
ble
in
the
LFC
powe
r
pro
blem
s
that
is,
i
n
m
oder
n’s
powe
r
s
ys
te
m,
small
ch
a
ng
e
s
in loa
d
is al
wa
ys
a
ntici
pated unde
r normal
operati
on
s i
n [
18,
26
-
28
].
̇
1
(
)
=
−
1
1
1
(
)
+
1
1
1
(
)
−
1
1
1
(
)
−
1
1
3
(
)
−
1
1
1
(
3
)
̇
1
(
)
=
−
1
1
1
(
)
+
1
1
1
(
)
(
4
)
̇
1
(
)
=
−
1
1
1
1
(
)
−
1
1
1
(
)
+
1
1
1
(
5
)
̇
1
(
)
=
1
1
(
)
+
1
(
)
+
3
(
)
(
6
)
̇
1
(
)
=
2
12
1
(
)
−
2
12
2
(
)
+
2
13
1
(
)
−
2
13
3
(
)
(
7
)
̇
2
(
)
=
−
1
2
2
(
)
+
2
2
2
(
)
−
2
2
1
(
)
−
1
1
2
(
)
−
1
1
2
(
8
)
̇
2
(
)
=
−
1
2
2
(
)
+
1
2
2
(
)
(
9
)
̇
2
(
)
=
−
1
2
2
2
(
)
−
1
2
2
(
)
+
1
2
2
(
10
)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
1693
-
6930
TELK
O
M
N
IKA
Tel
ec
om
m
un
C
ompu
t El
C
on
t
ro
l
,
V
ol.
19
,
No.
3
,
June
20
21
:
10
01
-
100
9
1004
̇
2
(
)
=
2
2
(
)
+
1
(
)
+
2
(
)
(
11
)
̇
2
(
)
=
2
12
2
(
)
−
2
12
1
(
)
+
2
23
2
(
)
−
2
23
3
(
)
(
12
)
̇
3
(
)
=
−
1
3
3
(
)
+
3
3
3
(
)
−
3
3
2
(
)
−
3
3
3
(
)
−
3
3
3
(
13
)
̇
3
(
)
=
−
1
3
3
(
)
+
1
3
3
(
)
(
14
)
̇
3
(
)
=
−
1
3
3
3
(
)
−
1
3
3
(
)
+
1
3
3
(
15
)
̇
3
(
)
=
3
3
(
)
+
2
(
)
+
3
(
)
(
16
)
̇
3
(
)
=
2
23
3
(
)
−
2
23
2
(
)
+
2
13
3
(
)
−
2
13
1
(
)
(
17
)
The
mat
rix
form
s
hows
i
n
t
he
dy
namic
e
qu
at
io
ns
from
(3)
to
(
17),
t
he
the
re
-
a
rea
i
nterc
onnected
powe
r
sy
ste
m
desc
rib
ed by
Fig
ure
2 wh
ic
h
ca
n
be writt
en
a
nd e
xpress
ed
in
sta
te
-
sp
ace
r
e
pr
e
se
ntati
on
bel
ow
:
̇
(
)
=
̃
(
)
+
̃
(
)
+
̃
(
)
(
18
)
wh
e
re
(
)
∈
is
the
sta
te
vector,
(
)
∈
is
t
he
co
ntr
ol
vect
or,
an
d
̃
,
̃
,
̃
is
con
st
ant
matri
x
e
quivale
nt
i
th
of
eac
h
a
rea.
(
̃
∈
×
,
̃
∈
×
,
∈
×
).
̃
=
[
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
3
0
0
]
̃
=
[
−
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
−
2
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
−
3
3
0
0
0
0
]
̃
=
[
−
1
1
1
1
0
0
−
1
1
0
0
0
0
0
0
0
0
0
−
1
1
0
−
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
−
1
1
1
0
−
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
1
0
0
0
0
0
0
0
0
0
1
2
(
12
+
13
)
0
0
0
0
−
2
12
0
0
0
0
−
2
13
0
0
0
0
0
0
0
0
2
2
−
1
2
2
2
0
0
−
2
2
0
0
0
0
0
0
0
0
0
0
0
−
1
2
1
2
0
0
0
0
0
0
0
0
0
0
0
0
−
1
2
2
0
−
1
2
1
2
0
0
0
0
0
0
0
0
0
0
1
2
0
0
0
1
0
0
0
0
0
−
2
21
0
0
0
0
2
(
21
+
23
)
0
0
0
0
−
2
23
0
0
0
0
0
0
0
0
0
0
0
0
0
−
3
3
−
1
3
3
3
0
0
3
3
0
0
0
0
0
0
0
0
0
0
0
−
1
3
1
3
0
0
0
0
0
0
0
0
0
0
0
0
−
1
3
3
0
−
1
3
1
3
0
0
0
0
0
0
0
0
0
0
1
3
0
0
0
1
−
2
13
0
0
0
0
−
2
23
0
0
0
0
2
(
31
+
32
)
0
0
0
0
]
̇
(
)
=
[
1
(
)
1
(
)
1
(
)
1
(
)
1
(
)
2
(
)
2
(
)
2
(
)
2
(
)
2
(
)
3
(
)
3
(
)
3
(
)
3
(
)
3
(
)
]
Be
cause
it
is
te
dio
us
to
determine
the
va
lues
of
the
s
yst
em
exa
ct
pa
rameters
of
̃
,
̃
,
̃
du
e
t
o
nonlinea
rity
a
nd
dyna
mics
of
a
po
wer
el
ec
tric
sy
ste
m,
t
he
dyna
mic
model
(
18)
is
re
vi
sed
to
the
no
minal
par
a
mete
rs
a
nd
p
a
rameter
v
a
riat
ion
s se
pa
rati
on
s
in
the
fo
ll
owin
g:
Evaluation Warning : The document was created with Spire.PDF for Python.
TELK
O
M
N
IKA
Tel
ec
om
m
un
C
ompu
t El
C
on
t
ro
l
Ext
end
e
d
st
ate
ob
s
erver
based
load
fre
quenc
y co
ntro
ll
er f
or
…
(
V
an Van
H
uynh
)
1005
̇
(
)
=
[
+
(
,
)
]
(
)
+
[
+
(
,
)
]
(
)
+
̃
(
)
=
(
)
+
(
)
+
(
,
)
(
)
=
(
)
(
19
)
wh
e
re
,
is
the
e
xact
values
of
̃
,
̃
;
the
unknow
n
matri
ces
(
,
)
an
d
(
,
)
de
no
te
s
by
their
ti
me
-
var
ia
nt
syst
em
of
pa
rametric
va
riat
ion
s;
a
nd
(
,
)
is
cal
le
d
the
l
umpe
d
un
ce
rtai
nties
an
d
we
c
an
al
s
o
denote
by
(
20).
(
,
)
=
(
,
)
(
)
+
(
,
)
(
)
+
̃
(
)
(
20
)
Figure
2. Bl
oc
k diag
ram of
3
-
area
LFC s
ys
t
em
3.
PROP
OSE
D
EXTE
ND
E
D OB
SER
VER
-
BASED
LOA
D
F
REQ
UEN
CY CONT
ROL
LE
R
The
pro
pose
d
new
l
y
c
ontr
ol
te
chn
iq
ue
that
is,
sta
te
s
obser
ver
performs
t
he
f
un
ct
io
ns
by
e
sti
mati
ng
the
sta
te
va
riables
of
th
e
s
yst
ems
ty
pical
ly
the
ou
t
pu
t
a
nd
co
ntr
ol
va
riab
le
s.
Stat
e
obse
rv
e
rs
can
be
de
sig
ne
d
and
a
pp
li
e
d
on
ly
wh
e
n
the
ob
serv
a
bili
ty
require
d
c
onditi
on
is
sat
isfie
d.
First,
we
exte
nd
l
umped
unce
rtai
nty
as
an
ad
diti
on
a
l
sta
te
var
ia
ble
to
desig
n
a
nd
init
ia
te
an
extend
e
d
ob
se
r
ver
-
base
d
load
fr
e
qu
e
nc
y
co
ntr
ol
le
r
in
the s
ys
te
m as
(
21).
+
1
(
)
=
(
,
)
(
21
)
The
n,
t
he
th
ree
-
area
powe
r
s
yst
em in (1
9) ca
n be
wr
it
te
n
as:
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
1693
-
6930
TELK
O
M
N
IKA
Tel
ec
om
m
un
C
ompu
t El
C
on
t
ro
l
,
V
ol.
19
,
No.
3
,
June
20
21
:
10
01
-
100
9
1006
̄
̇
(
)
=
̄
̄
(
)
+
̄
(
)
+
ℎ
(
)
̄
(
)
=
̄
̄
(
)
(
22
)
ℎ
(
)
=
(
,
)
(
23
)
wh
e
re
:
̄
(
)
=
[
(
)
+
1
(
)
]
;
̄
=
[
×
×
0
×
0
×
]
;
̄
=
[
×
0
×
]
;
=
[
0
×
×
]
;
̄
=
[
×
,
0
×
]
.
With
re
gards
to
t
he
sta
te
obs
erv
e
rs
disc
us
s
ed,
we
will
ap
ply
the
nota
ti
on
̄
̂
to
in
dicat
e
the
vecto
r
ob
s
er
ved
sta
te
.
The
vect
or
̄
̂
o
f
t
h
e
obs
e
rv
ed
sta
t
e
is
us
ed
and
a
ppli
ed
in
the
sta
te
fee
db
ac
k
to
i
niti
at
e
the
desire
d
a
nd r
e
quire
d
c
on
tr
ol
ve
ct
or
.
If
we
cal
l t
he
sta
te
̄
is ap
pro
ximate
d
t
o st
at
e,
̄
̂
the dy
na
mica
l mo
del:
̄
̂
̇
(
)
=
̄
̄
̂
(
)
+
̄
(
)
+
(
̄
(
)
−
̄
̂
(
)
)
̄
̂
̇
(
)
=
̄
̄
̂
(
)
(
24
)
The
sta
te
s obse
rv
e
d
ha
ve
u
a
nd
̄
as
t
he
i
nput
an
d o
utput
si
gnal
.
The
g
ai
n
L
of
sta
te
observ
er
is
c
hosen
s
o t
hat
the eige
nval
ue of
̄
−
̄
li
e in the
de
sired
l
ocati
ons
in the le
ft
-
half s
-
pla
ne.
The
contr
ol in
pu
t i
s
c
ho
s
en
as
;
(
)
=
−
̄
̂
=
[
̃
̂
]
̄
̂
(
25
)
wh
e
re
,
̃
is
the
f
eedb
ac
k
c
ontro
l
gain
to
be
c
hose
n
so
t
hat
th
e
ei
genvalues
of
−
̃
li
e
in
sp
eci
f
ic
locat
ion
s
in the le
ft
-
half s
-
pla
ne
a
nd the
lum
ped unce
rtai
nty
c
ompe
nsa
ti
on
gain
̃
is de
sign
e
d:
̂
=
[
(
−
̃
)
−
1
]
−
1
(
−
̃
)
−
1
(
26
)
Com
bin
e
(22
) a
nd (2
4)
,
the e
sti
mati
on
e
rror o
f
sta
te
obse
rvers
(
)
=
̄
(
)
−
̄
̂
(
)
can
be re
vised b
y:
̇
(
)
=
̄
(
)
−
(
̄
(
)
−
̄
̂
(
)
)
+
ℎ
(
)
=
(
̄
−
̄
)
(
)
+
ℎ
(
)
(
27
)
Denote
ℎ
(
)
by
(
)
and u
si
ng f
inal
-
val
ue
the
orem,
w
e
h
a
ve:
→
∞
(
)
=
→
∞
(
(
−
(
̄
−
̄
)
)
−
1
(
)
=
→
∞
(
−
(
̄
−
̄
)
)
−
1
×
→
∞
(
)
=
→
∞
(
−
(
̄
−
̄
)
)
−
1
×
→
∞
(
)
(
28
)
Sine
→
∞
(
−
(
̄
−
̄
)
)
−
1
is
bounde
d
a
nd
→
∞
(
)
=
0
.
The
refor
e
,
est
imat
ion
error
of
sta
te
obser
vers
is:
(
)
=
̄
(
)
−
̄
̂
(
)
is asy
mp
t
otica
ll
y
sta
ble.
Re
mark
1:
I
f
t
he
s
ys
te
m
sta
te
s
are
not
mea
su
ra
ble,
the
n
t
he
est
imat
io
n
of
the
lum
pe
d
uncertai
nt
y
a
nd
the
par
a
mete
rs
of
sy
ste
m
sta
te
s
c
an
be
ap
ply
i
ng
in
t
he
de
si
gn
con
t
ro
l.
T
her
e
f
or
e
,
the
c
omposi
te
con
t
ro
l
la
w
will
be desig
ne
d
as
in (2
1)
.
Re
mark
2:
It
is
note
d
t
hat
t
he
lu
mp
e
d
unce
rtai
nty
can
not
be
at
te
nuat
ed
c
omplet
el
y
an
d
t
otall
y
from
th
e
sta
te
equ
at
io
n
no
matt
er
wh
at
c
on
t
ro
ll
er
was
desig
ne
d.
I
n
this
a
ppr
oac
h,
on
e
of
th
e
m
os
t
rece
nt
ac
hi
evab
le
ob
je
ct
ives
is
si
mp
ly
t
o
tr
un
ca
te
the
disturba
nces
at
the
out
pu
t
po
i
nt
in
ste
ady
sta
te
by
t
he
ap
plica
ti
on
of
th
e
com
posit
e
co
nt
ro
l
la
w
.
T
her
e
fore,
t
he
li
mit
at
ion
s
recor
ded
by
oth
e
r
c
on
t
ro
l
st
rategies
i
n
this
pa
pe
r
[
21
-
25]
h
as
bee
n
s
olv
e
d.
4.
SIMULATI
O
N RESULTS
In
t
he
case
to
evaluate
the
ne
wly
e
xten
ded
sta
te
ob
se
rv
e
r
appr
oach,
tw
o
sim
ulati
on
s
by
us
i
ng
t
he
M
A
TLAB/S
I
M
U
LI
NK
s
of
t
war
e
are give
n as f
ollow
i
ng
:
Simulat
io
n 1:
The para
mete
r
s of the
th
ree
-
a
rea inter
co
nnec
te
d
po
wer sy
ste
m w
e
re s
ee
n
a
s g
i
ven in
[16
,
17].
Ca
se
1.
F
or
th
e
sim
ulati
on
at
this
insta
nce
in
c
ase
1,
t
he
pa
rameters
with
the
re
nomi
nal
val
ues
of
the
3
-
a
rea
powe
r
netw
ork
are
a
ppli
ed.
At
this
point,
we
ass
um
e
zer
o
disturba
nces
occ
ur
i
ng
on
t
he
giv
e
n
s
ys
te
m,
i.e.
,
(
,
)
=
0
.
The
refor
e
,
th
e
fr
e
qu
e
nc
y
fl
un
ct
uation
or
dev
ia
ti
ons
of
t
he
3
-
area
i
nter
connecte
d
pow
er
net
work
accor
ding
t
o
t
he
insta
nce
of
case
1
w
he
re
by,
a
pp
l
ying
t
he
new
l
y
e
xtend
e
d
sta
te
obs
erv
e
r
c
on
t
ro
ll
e
r
are
disp
la
yed
in
th
e
res
ults
of
sim
ulati
on
in
Fi
gure
3
to
Fig
ure
4.
I
n
Fig
ur
e
3,
the
fr
e
quenc
y
dev
ia
ti
on
ap
proach
e
s
to
zer
o
mar
k
at
exactl
y
1.5
s.
Con
se
quently
,
Figure
4
sho
ws
ti
e
li
ne
power
dev
ia
ti
on
getti
ng
to
zer
o
mar
k
wit
h
the
desig
ne
d
exten
ded
sta
te
obse
r
ver
c
ontrolle
r.
By
c
ompa
rin
g
t
he
s
imulat
ion
resul
ts
from
t
he
new
l
y
exten
ded
obse
rv
e
r
c
on
t
ro
ll
er
with
res
ults
giv
e
n
s
how
n
in
[
16
,
17],
the
new
l
y
e
xtend
e
d
sta
te
ob
serv
e
r
con
t
ro
ll
er
was
able to a
ssure
f
ast
r
es
pons
e t
o t
he
s
ys
te
m a
nd also ca
pa
ble of tr
un
cat
e
smal
le
r
over
sho
ots.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELK
O
M
N
IKA
Tel
ec
om
m
un
C
ompu
t El
C
on
t
ro
l
Ext
end
e
d
st
ate
ob
s
erver
based
load
fre
quenc
y co
ntro
ll
er f
or
…
(
V
an Van
H
uynh
)
1007
Figure
3.
Fr
e
quenc
y de
viati
ons
(H
z
) of t
he
t
hr
ee
-
area
without
disturb
ances
Figure
4.
Tie
li
ne powe
r devia
ti
on
for powe
r
sy
ste
m
with
ou
t
d
ist
urba
nces
Ca
se
2:
Desig
ning
a
c
ontrol
le
r
with
t
he
a
im
an
d
ca
pacit
y
to
perf
or
m
excell
ent
within
a
n
uncert
ai
nt
y
env
i
ronme
nt
o
f
po
wer
netw
orks
a
re
al
wa
ys
the
main
goal
of
s
eve
ral
el
ect
rical
and
c
on
trol
en
gin
ee
rs.
In
t
his
case,
the
pro
po
s
ed
e
xten
de
d
sta
te
obse
r
ver
c
ontr
oller
was
a
ppli
ed
unde
r
un
ce
rtai
nties
with
matc
hed
par
a
mete
rs
an
d
l
oad
disturb
ance
i
n
oth
e
r
to
e
xamine
th
e
net
wor
k
performa
nce
unde
r
m
at
che
d
pa
ramet
e
r
un
ce
rtai
nties
a
nd
loa
d
distu
r
ban
ce
s.
The
load
dist
urba
nc
e;
1
(
)
=
0
.
02
pu
,
2
(
)
=
0
.
015
pu
,
an
d
3
(
)
=
0
.
01
pu
wer
e
pres
um
e
to
ta
ke
place
at
area
1,
a
rea
2,
an
d
area
3
acc
ordin
gly.
T
he
r
esp
on
ses
in
t
he
c
losed
-
lo
op
f
or
every
one
of
t
he
co
ntr
ol
area ap
pl
ying
the
e
xten
ded
sta
te
obser
ve
r
c
on
tr
ol
le
r
an
d
th
e
co
nt
ro
ll
er
giv
e
n
in
[1
6
,
17]
are
sho
wn in F
i
gure
5
a
nd
Figure
6.
Figure
5,
Fi
gure
6
a
nd
Ta
ble
1
s
how
cl
ea
rly
t
hat
the
res
ponse
s
of
the
s
ys
te
m
a
re
not
on
l
y
gr
eat
t
o
deal
with
ove
rsho
ot
-
prob
le
ms,
but
al
s
o
ens
ur
es
quic
k
an
d
fast
set
tl
ing
pe
rio
d
a
s
matc
hed
-
to
th
e
rece
nt
appr
oach
in
[
16
,
17]
.
In
the
same
c
onditi
on,
it
is
see
n
t
ha
t
fr
e
quenc
y
de
viati
on
co
nve
rg
e
d
t
o
ze
r
o
in
ab
out
1.6s wit
h
t
he n
ewly p
rop
os
ed
ex
te
nded
stat
e
ob
s
er
ver that
is
, s
at
isfie
d re
qu
i
ment
of LFC
pro
blems.
Figure
5.
Fr
e
quenc
y de
viati
ons
(H
z
) of t
he
three
-
a
rea
unde
r
matc
hed unc
ertai
nties an
d
l
oad
disturba
nces
Figure
6.
Tie
li
ne powe
r devia
ti
on
unde
r
mat
ched
un
ce
rtai
nties a
nd loa
d dist
urb
ances
Table
1.
Sett
ing
ti
me
-
and Ma
x.O.S (
maxim
un
-
over
-
sho
ot
-
cal
culat
ion)
of ELFC a
nd
DL
FC
Kin
d
s o
f
co
n
troller
Exten
d
ed
state ob
serv
er
-
b
ased
load
fr
eq
u
en
cy
con
troller
(
EL
FC
)
Decentraliz
ed
load
f
requ
en
cy
co
n
troller (
D
LFC)
[16
]
Para
m
eters
(s)
Ma
x.O.
S
(pu
)
(s)
Ma
x.O.
S
(pu
)
1
1
.5
2
.1
×
1
0
−
3
7
.5
3
.7
×
1
0
−
3
2
2
.0
1
.45
×
1
0
−
3
7
.5
3
.8
×
1
0
−
3
3
1
.8
2
.0
×
1
0
−
3
7
.5
4
.0
×
1
0
−
3
Simulat
io
n
2:
The
pract
ic
al
powe
r
sy
ste
m wi
th
loa
d
distu
rbance
is
c
on
si
de
red
in
this
exa
mp
le
.
T
he
c
on
diti
on
s
and
par
am
et
ers
us
i
ng
in
this
s
imulat
ion
are
the
sa
me
with
the
recent
resea
rch
in
[
18].
Fi
gure
7
an
d
Fig
ur
e
8
sh
ow
that
t
he
pro
pose
d
e
xtend
e
d
sta
te
ob
serv
e
r
c
ont
r
ol
sche
me
has
f
ast
er
re
spo
ns
e
an
d
le
sse
r
sign
ic
a
nt
ov
e
rs
hoot in
c
ompari
ng the
previ
ou
s
contr
ol
in [1
8].
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
1693
-
6930
TELK
O
M
N
IKA
Tel
ec
om
m
un
C
ompu
t El
C
on
t
ro
l
,
V
ol.
19
,
No.
3
,
June
20
21
:
10
01
-
100
9
1008
Re
mark
3:
B
y
matc
hing
-
up
t
he
re
su
lt
s
of
the
sim
ulati
on
for
the
tw
o
si
mu
la
ti
ons
a
bove,
t
he
new
l
y
exten
ded
sta
te
ob
se
rv
e
r
-
base
d
loa
d
f
re
qu
e
nc
y
c
on
tr
ol
le
r
s
disp
la
yed
rob
us
tness
an
d
fast
res
ponse
to
dist
or
ti
on
s
an
d
disturba
nce
oc
currin
g
on
t
he
s
ys
te
m
co
rr
el
at
ed
with
var
ia
nt
of
t
he
matc
he
d
unc
ertai
nties
an
d
loa
d
disturba
nces
use
d for sim
ulati
on
s
.
Figure
7.
Fr
e
quenc
y de
viati
ons
of the
t
hr
ee
-
area
unde
r
matc
hed uncertai
nties a
nd loa
d dist
urb
ances
Figure
8.
Tie
li
ne powe
r devia
ti
on
unde
r
mat
ched
un
ce
rtai
nties a
nd loa
d dist
urb
ances
5.
CONCL
US
I
O
N
In
this
pa
per,
t
he
ne
wly
e
xten
ded
sta
te
ob
se
r
ver
f
or
LFC
f
or
a
n
i
ntercon
ne
ct
ed
s
ys
te
m
is
performe
d.
In
real
en
vir
on
ment,
s
om
e
va
rio
us
sta
te
va
riables
are
not
measu
rab
le
i
n
load
fr
e
quenc
y
con
t
ro
l
s
ys
te
m
f
or
instance
a
rea
c
on
t
ro
l
e
rror
or
com
bin
at
io
n
of
area
co
ntr
ol
er
rors.
T
o
res
olve
this
unmeas
urable
sta
te
va
ri
ables
pro
blem,
the
exten
ded
sta
te
ob
se
r
ver
is
pro
posed
for
e
sti
mati
ng
the
unmeas
ur
a
ble
sta
te
var
ia
bles
.
The
exten
ded
sta
te
ob
s
er
ver
-
ba
sed
load
f
reque
nc
y
c
on
tr
ollers
ut
il
iz
e
distur
ba
nc
e
est
imat
ion
t
echn
i
qu
e
s;
th
us,
the
con
t
ro
ll
er
gai
ns
are
not
set
t
o
be
extre
mely
high
to
at
te
nua
te
disturba
nces
of
unknow
n
boun
dar
ie
s,
w
hi
ch
is
very
use
f
ul
i
n
load
f
re
qu
e
nc
y
co
ntr
oller
de
s
ign
.
T
he
refor
e
,
it
ca
n
be
c
oncl
ud
e
d
that
t
he
a
pp
li
cat
ion
of
the
pro
po
se
d
exte
nd
e
d
sta
te
obs
erv
e
r
for
l
oad
fr
e
qu
e
nc
y
c
ontrols
of
inte
rconn
ect
e
d
powe
r
sy
ste
m
ca
n
operate
eff
ect
i
vel
y
i
n
pr
act
ic
al
si
gh
t.
By
us
in
g
M
A
TLAB/SI
MUL
INK,
the
sim
ul
at
ion
res
ults
a
bove
pr
ese
nt
t
hat
t
he
new
l
y
meth
od
imp
r
ov
es
t
he
dynamics
r
esp
on
s
es
of
t
he
s
ys
te
m
an
d
pr
ovide
de
sig
ns
f
or
new
LFC
’s
sy
ste
m
that
sat
isfie
s
th
e
LFC
require
ments.
In
the
f
uture
w
ork,
we
te
nd
to
desig
n
e
xten
ded
sta
te
obse
r
ver
f
or
r
obus
t
LFC’s i
n
m
ulti
-
area
powe
r
s
yst
ems combi
ne
d wit
h rene
wa
ble en
e
r
gy syst
ems.
ACKN
OWLE
DGME
NT
This
re
searc
h
i
s
fun
ded
by
F
oundat
io
n
f
or
S
ci
ence
an
d
Tec
hnolog
y
D
evel
opment
of
To
n
Du
c
Th
an
g
Un
i
ver
sit
y
(
FOSTEC
T)
, webs
it
e:
h
tt
p://
fo
ste
ct
.tdtu.
e
du.
vn, unde
r Gra
nt F
OS
TECT
.20
17.BR
.05
REFERE
NCE
S
[1]
Fu
C.
,
&
Ta
n
W.
,
"
De
ce
ntr
ali
sed
Lo
ad
Frequ
enc
y
Control
fo
r
Pow
er
Sys
te
m
s
with
Comm
un
ic
a
ti
on
Del
ays
via
Acti
ve
Disturba
nce
R
ej
e
ction,
"
I
ET
Gene
ration
,
Tr
ansm
i
ss
ion
& D
istribut
ion,
vol
.
12
,
no
.
6
,
pp
.
1
751
-
8687,
2018
.
[2]
Zha
ng
Y.
,
&
Ya
ng
T.,
"
De
ce
ntr
a
li
z
ed
Sw
it
ch
ing
Control
Stra
te
gy
for
Lo
ad
Frequ
enc
y
Con
trol
in
Multi
-
Area
Pow
er
Sys
te
ms wi
th Ti
me
De
la
y
and
Pa
cke
t
Losse
,
"
IEEE
A
ccess
,
vol
.
8
,
pp.
15838
-
1585
0,
Jan
2020
.
[3]
Deve
ndra
K.
Ch
at
urve
d
i
,
"
Techn
ique
s a
nd
it
s
Applicati
ons
in El
e
ct
ri
ca
l
Eng
ine
er
i
ng
"
Springer
,
20
08.
[4]
Vija
y
Vi
tt
a
l
,
Ja
me
s
D.
McCalle
y
,
Paul
M.
Ande
rson
,
A.
A
.
Foua
d
P.
,
"
Pow
er
Sys
te
m
Con
trol
an
d
Stabilit
y
,
"
Wi
l
ey
3rd E
dit
io
n.
,
Oct
ober
2019
.
[5]
Guh
a
D.
,
Roy
P.
K
.
,
&
B
ane
r
jee
S.
,
"Lo
ad
Fre
quenc
y
Contro
l
of
Int
erc
onn
ecte
d
Pow
er
Sys
te
m
Us
ing
Grey
W
olf
Optim
izati
o
n,
"
S
warm
and
Ev
olu
ti
onary
Computa
ti
on,
vol. 27
,
pp
.
97
-
115,
Apr
20
16
.
[6]
Yous
ef
H.
A.
,
AL
-
Kharusi
K.
,
Albadi
M.
H.,
&
Hos
seinz
ade
h
N.
,
"Lo
ad
Frequ
en
cy
Contro
l
of
a
Multi
-
Area
Pow
er
Sys
te
m:
An
Ada
pti
ve
Fuz
zy
Log
ic
Approac
h
,
"
IE
EE
Tr
ansacti
ons
on
Powe
r
System
s,
"
vol.
29,
no.
4,
pp.
1822
-
183
0
,
Jan
2014
.
[7]
Ze
ng
G.
Q.,
Xi
e
X
.
Q
.
,
&
Ch
en
M.
R
.
,
"
An
Adapti
ve
Model
Predi
ct
iv
e
Lo
a
d
Freque
n
cy
Co
ntrol
Method
fo
r
Multi
-
Area
In
te
r
conne
c
te
d
Pow
e
r
Sys
te
ms
with
Photovolt
aic
Ge
ner
ations,
"
Elec
t
rical
Pow
er
and
Ene
rgy
S
yste
m
,
vol.
10
,
no
.
11
,
p
p.
1
-
23
,
Nov 201
7.
[8]
Beni
Reh
ia
r
a
A.
,
Y
orino
N
.
,
Sas
aki
Y.
,
&
Zok
a
Y
.
,
"A
n
Adapt
iv
e
Lo
ad
Frequ
en
cy
Control
Base
d
on
L
ea
st
Squa
re
Method
,
"
Adv
an
ce
s in
Mode
ll
ing
and
Control
of
Wind
and
H
ydroge
nerators
,
vol
.
49,
pp
.
220
,
202
0.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELK
O
M
N
IKA
Tel
ec
om
m
un
C
ompu
t El
C
on
t
ro
l
Ext
end
e
d
st
ate
ob
s
erver
based
load
fre
quenc
y co
ntro
ll
er f
or
…
(
V
an Van
H
uynh
)
1009
[9]
Dong
L
.
L
,
Zhang
Y
.
,
Gao
Z
.
Q.
,
"
A
Robust
Dec
ent
r
al
i
ze
d
L
oad
Frequ
enc
y
Control
le
r
for
I
nte
rco
nn
ec
t
ed
Pow
er
Sys
te
ms,
"
ISA
Tr
ansacti
ons
,
vol
.
51,
no.
3,
pp.
41
0
-
419,
May
201
2.
[10
]
Gheisa
rne
j
ad
M
.
,
&
Khooban
,
M.
H.,
"
Design
an
Opt
im
a
l
F
uzz
y
Fra
ct
ion
al
Proportiona
l
I
nte
gra
l
Deri
v
ati
ve
Control
le
r
with
Deri
vative
Filt
er
for
L
oad
Frequ
enc
y
Con
trol
in
Pow
er
Sys
te
ms,
"
Tr
ansacti
ons
of
th
e
Insti
tute
of
Me
asur
eme
nt
an
d
Control
,
vol
.
4
1,
no
.
9
,
pp
.
1
-
1
9,
Jan
2019.
[11
]
Anw
ar
M.
N.
,
a
nd
Pan
S.,
"
A
N
ew
PID
Loa
d
Freque
ncy
Contro
l
le
r
Design
Meth
od
in
Frequ
enc
y
Doma
in
Throug
h
Dire
ct Synthe
sis
Approac
h
,
"
E
lectric P
ower
and
Ene
rgy
S
yste
ms
,
vol.
67
,
pp
.
560
-
569,
May
2015.
[12
]
Sonkar
P.,
&
R
a
hi
O.
P.
,
"
Tun
in
g
of
Modi
fie
d
PI
D
Loa
d
Freque
n
cy
Contro
ller
for
Int
erc
onne
cted
Sys
te
m
with
Wind
Pow
er
Plant
vi
a
IMC
Tuni
ng
Method
,
"
201
7
4th
IE
EE
Ut
tar
Pradesh
Se
ct
ion
Int
ernati
o
nal
Conf
ere
nc
e
o
n
El
e
ct
rica
l
,
Computer
and
Elec
tr
onic
s
,
Jan
2018
.
[13
]
K.
L
ia
o
and
Y.
Xu,
"
A
Robust
Loa
d
Freque
n
cy
Control
Sch
em
e
for
Pow
er
Sys
tems
Based
on
Se
cond
-
Order
Sl
iding
Mode
and
Extended
Disturba
nce
Obs
erv
er
,
"
IEE
E
Tr
ansac
ti
on
s
on
Industrial
Informatic
s
,
vol.
14
,
no
.
7,
pp.
3076
-
3086
,
J
uly
2018.
[14
]
Zhe
ng
Y
.
,
Li
u
J.
,
L
iu
X.
,
Fang
D
.
,
&
Wu
L.
,
"
Ad
apt
iv
e
Second
-
Order
Slid
ing
Mo
de
Contro
l
Desi
gn
for
a
Cl
ass
o
f
Nonline
ar
Sys
tems wit
h
Unkno
wn Input
,
"
Ma
th
emati
ca
l
Probl
e
ms
in
Engi
n
ee
ri
ng
,
vol
.
2015
,
no
.
1
,
pp
.
1
-
7,
2015
.
[15
]
Y.
Sun,
Y.
Wa
n
g,
Z.
Wei
and
X.
Wu
,
"
Robust
H∞
Lo
ad
Freque
n
cy
Con
trol
of
M
ult
i
-
Are
a
Pow
er
Sys
te
m
wi
th
Tim
e
Dela
y:
A
Slidi
n
g
Mode
Control
Approac
h,
"
IE
EE
/CAA
Journa
l
of
Aut
omat
ic
a
Sini
ca
,
vol.
5
,
n
o.
2,
pp
.
610
-
61
7,
2018.
[16
]
Ya
ng
M
.
,
Yang
F
.
,
Chengshan
W
.,
and
Peng
W.
,
"
Dec
en
tra
l
iz
ed
Sl
idi
ng
M
ode
Loa
d
Fr
eq
uenc
y
Contro
l
f
or
Multi
-
Area
Pow
er
Sys
te
ms,
"
IE
EE
Tr
ansacti
ons
on
Pow
er
Syst
e
m
,
vol
.
28
,
no
.
4
,
pp.
4301
-
4309,
Aug 2013.
[17
]
Muthana
T.
Alr
ifa
i
,
Moha
me
d
F.
Hass
an,
Mohame
d
Zr
ibi
.
,
"
Dec
ent
r
al
i
ze
d
L
oad
Frequ
enc
y
Control
le
r
for
A
Multi
-
Area
Int
er
conne
c
te
d
Pow
er
Sys
te
m,
"
Elec
t
rical
Powe
r and
Ene
rgy
S
yste
ms
,
vol.
33
,
no
.
2
,
pp
.
198
-
209,
2011.
[18
]
Yang
Mi
et
a
l.
,
"
The
Slidi
ng
Mode
Loa
d
Fre
quenc
y
Con
trol
for
Hybrid
Pow
er
Sys
te
m
B
a
sed
on
Disturb
a
nce
Obs
erv
er,
"
E
lect
rical
Powe
r and
Ene
rgy
S
yste
ms
,
vol.
74
,
pp
.
446
-
452,
Jan
2016.
[19
]
D.
Qian
,
and
G.
Fan,
"N
eur
a
l
-
Network
-
Base
d
Te
r
mi
n
al
Sl
id
ing
Mode
Cont
rol
for
Freque
n
cy
Stab
il
i
zation
of
Rene
wabl
e
Pow
er
Sys
te
ms,
"
IE
EE
/CAA
Journal
of Aut
omati
ca
S
i
nic
a
,
vol
.
5
,
no
.
3,
pp
.
706
-
717
,
Apr 2018.
[20
]
S.
Tri
p
e
t
al.
,
"P
assivit
y
-
Based
Design
of
Slidi
n
g
Modes
for
Optim
al
Loa
d
Frequ
enc
y
Contro
l,
"
I
EE
E
Tr
ansacti
o
ns
on
Control
S
ystem
s Tec
hnology
,
vol.
27
,
no
.
5
,
pp
.
1893
-
1906,
201
9.
[21
]
Li
H
.
Y
.
,
Shi
P
.
,
Yao
D
.
Y
.
,
Wu
L
.
G
.
,
"
Obs
erv
e
r
-
Based
Adapt
iv
e
Slidi
ng
Mode
Control
of
Nonl
ine
ar
Markovi
an
Jump System
s,
"
Aut
omatic
a
,
vo
l. 64, pp.
133
-
142
,
Mar
2016
.
[22
]
Khaya
ti
K.
,
"
Multi
var
ia
b
le
Ada
pti
ve
Slid
ing
-
Mode
Obs
erv
er
-
B
ase
d
Control
for
Mec
hanica
l
Sys
te
ms,”
Canad
ia
n
Journal
of
Elec
t
rical
and
Computer
Eng
ine
ering
,
vol.
38
,
no
.
3
,
pp
.
253
-
265,
Nov 2
015.
[23
]
Wa
ng
B
.
,
Shi
P.,
Kari
m
i
H
.
R
.
,
&
L
im
C.
C,
"
Obs
erv
er
-
Based
Slidi
ng
Mode
Co
ntr
ol
for
Stab
il
i
z
at
ion
of
a
Dyn
a
mi
c
Sys
te
m
with
Del
aye
d
Outpu
t
Fe
e
dbac
k,
"
Ma
the
ma
ti
cal P
robl
ems i
n
Engi
n
ee
ring
,
v
ol.
3
,
pp
.
1
-
6,
Se
p
2013.
[24
]
Ouass
ai
d
M.,
Maa
roufi
M.,
&
C
her
kaoui
M,
"
O
bserve
r
-
Based
Nonlinear
Contro
l
of
Pow
er
Sys
tem
Us
ing
Slid
ing
Mode
Control St
rat
egy
,
"
E
lectric
Powe
r Sy
st
ems
Re
search
,
vol
.
8
4,
no
.
1
,
pp
.
135
-
143,
Jan
2012.
[25
]
Yang
B.
,
Yu
T.,
Shu
H
.
,
Yao
W
.
,
&
Ji
ang
L.
,
"
Slidi
ng
-
Mode
Perturbation
Obs
e
rve
r
-
Based
Slid
i
ng
-
Mode
Contro
l
Design
for
Stabilit
y
Enh
anc
e
me
n
t
of
Multi
-
Mac
hi
ne
Pow
er
Sys
tems,
"
Tr
ansacti
o
ns
of
the
Insti
tute
of
M
easurement
and
Control
,
vol
.
41
,
no
.
15
,
pp
.
1418
-
1434,
Jul
2
018.
[26
]
Prasad
S.,
Purw
ar
S.,
&
Kishor
N.,
"
Non
-
Li
n
e
ar
Slidi
ng
Mode
Loa
d
Freque
n
cy
Control
in
Multi
-
Area
Pow
er
Sys
te
m,
"
Contro
l
Eng
ine
ering
Pr
act
i
ce
,
vol
.
61
,
p
p.
81
-
92
,
De
c
20
17
.
[27
]
Dianwe
i
Q.
,
Shi
wen
T.
,
Xiangjie
L.
,
"
Loa
d
Frequ
enc
y
Contro
l
for
Micro
Hydro
Pow
er
Plant
s
by
Slidi
ng
Mode
an
d
Model
Order
Re
duct
ion
,
"
Au
tomati
ka
,
vol
.
56
,
no
.
3
,
pp
.
318
-
330
,
Jan
2017
.
[28
]
Dianwe
i
Q.
,
Shi
wen
T.
,
Hong
L.
,
Xiang
ji
e
L
.
,
"
L
oad
Freque
n
cy
Control
by
N
eur
al
-
Network
-
Bas
ed
Int
egr
al
Slidi
ng
Mode
for
Nonl
in
ea
r
Pow
er
Sys
tems wit
h
Wi
nd
T
urbine
s,
"
N
euroc
omputing
,
vo
l.
1
73,
pp
.
875
-
885
,
Jan
2016
.
Evaluation Warning : The document was created with Spire.PDF for Python.