In
te
r
n
ation
a
l Jou
rn
al
o
f Po
we
r
Elec
tron
ic
s an
d
D
r
ive S
y
stem
(IJ
PED
S
)
V
o
l.
10, N
o.
1, Mar
ch 20
19,
p
p.
277~
2
8
7
IS
S
N
: 2088-
86
94,
D
O
I
:
10.11
59
1
/ij
ped
s
.
v10
.
i
1.pp
2
77-
28
7
277
Jou
rn
a
l
h
o
me
pa
ge
:
ht
tp:
//i
a
e
score
.
com
/
j
o
u
r
na
l
s
/
i
n
d
e
x
.
p
hp/IJ
PED
S
High-
p
erformance active pow
er fi
lter implementation base
d
on
predictive curren
t control
Bac
h
ar
R
ah
ima,
Gole
a
A
mar,
Ben
c
h
ou
i
a
M
oh
am
ed
Tou
fi
k
, C
h
e
b
aani
M
oh
ame
d
L.G.
E.
B
Laborat
o
ry,
Departm
e
nt
of
Electri
cal E
ngin
eering
, Bisk
r
a
Universit
y
, Alge
r
ia
Art
i
cl
e In
fo
ABSTRACT
A
r
tic
le hist
o
r
y
:
R
e
ce
i
v
e
d
Feb 3,
2018
Re
vise
d O
c
t
2
6
,
2018
A
c
c
e
pte
d
D
ec 13,
2
0
1
8
Th
is
p
aper
p
res
e
n
t
s
t
h
e
us
e
o
f
t
h
e
p
red
i
ct
iv
e
s
t
rateg
y
c
on
c
e
pt
t
o
i
m
p
r
o
v
e
t
h
e
Act
i
v
e
P
ower
F
i
l
t
e
r
(AP
F
)
p
erfo
rman
ce,
b
y
co
m
p
ens
a
tio
n
o
f
t
h
e
r
eactive
po
wer
and
eli
m
inatio
n
of
t
he
h
arm
o
nic
curre
n
t
s
gen
e
rated
b
y
n
o
n
-linear
lo
ads
.
P
red
i
c
t
ive
con
t
ro
l
is
g
enerati
n
g
co
ns
iderab
le
i
n
t
erest
w
hen
i
t
c
om
es
t
o
implemen
t
i
ng
c
urr
e
nt
c
ont
r
ol
s
t
r
at
egi
e
s
in
acti
v
e
power
filter.
T
h
e
p
r
o
p
o
s
e
d
st
rateg
y
p
ro
v
i
des
a
sim
p
l
e
c
on
troller
i
n
co
rpo
r
a
t
in
g
P
h
ase
L
o
ck
e
d
Lo
op
(P
LL
)
in
dep
e
n
d
en
cy.
The
pred
icti
on
i
s
eval
uat
e
d
u
s
i
n
g
a
cost
f
u
nct
i
o
n
t
hat
qu
ant
i
fi
e
s
the
desi
red
s
y
stem
b
e
h
av
ior.
T
h
e
c
os
t
f
u
ncti
on
u
s
e
d
i
n
this
w
ork
eval
uat
e
s
th
e
fil
tered
erro
r
o
f
t
he
c
u
r
ren
t
s
.
T
h
i
s
strat
e
gy
m
i
ni
mi
zed
t
h
e
nu
m
b
er
o
f
s
e
ns
o
r
s,
ease
of
p
ract
ical
i
m
p
lem
e
nt
ation
and
red
u
ced
s
y
s
t
e
m
si
ze
and
cos
t
.
T
h
e
eff
ect
iven
ess
o
f
t
h
e
p
rop
o
sed
con
t
ro
ller
is
c
onfi
rmed
t
hrou
gh
sim
u
l
a
tio
n
and
e
xperi
m
e
nt
a
l
v
a
l
idat
io
n
usin
g
a
h
a
rdw
a
re
p
rot
o
ty
pe
b
as
ed
o
n
d
SPAC
E
-1
10
4
K
eyw
ord
s
:
A
c
ti
ve
pow
er
f
il
t
e
r (
A
P
F
)
Mu
lt
i va
ria
b
le
f
i
lter
(MV
F
)
Ph
as
e l
o
ck
ed
lo
o
p
(PLL)
DS
P
A
CE 1
1
04
Pre
d
i
c
ti
v
e
cont
ro
ll
e
r
Co
pyri
gh
t © 2
019 In
stit
u
t
e
of Advanced
En
gi
neeri
n
g
an
d
S
c
ien
ce.
All
rights
res
e
rv
ed.
Corres
pon
d
i
n
g
Au
th
or:
B
ach
ar R
a
h
i
m
a
,
L.G
.
E.B.La
borator
y,
D
e
pa
rtme
nt
o
f
El
e
c
t
rica
l
Eng
i
ne
eri
ng,
Biskra
U
ni
vers
it
y,
A
l
g
e
r
ia.
Em
ail:
rahim
a
_3
0
0
1
@
yah
o
o
.c
om
1.
I
N
TR
OD
U
C
TI
O
N
Rec
e
n
t
l
y
t
h
e
p
o
we
r
elec
t
r
on
ic
s
ba
se
d
no
nli
n
ea
r
l
o
ad
s
are
i
n
c
r
ea
sed
t
r
e
m
e
n
dou
sl
y
in
t
h
e
d
i
s
t
r
i
but
i
on
netw
ork.
T
he
s
e
i
nc
lud
e
,
sw
it
c
h
e
d
m
o
d
e
p
o
w
e
r
sup
p
lie
s,
v
a
r
ia
ble
sp
e
e
d
d
r
i
ves
a
n
d
un
inter
r
u
p
te
d
p
o
w
e
r
sup
p
l
i
es.
Thes
e
nonl
i
n
ea
r
lo
ads
c
r
ea
t
e
h
ar
m
onic
s
o
r
curre
nt
d
i
s
tor
t
i
o
n
prob
lem
in
t
he
s
u
p
p
l
y
si
de
o
f
t
h
e
di
stri
b
u
t
i
on
s
y
s
tem
.
T
he
h
arm
onic
cur
r
en
ts
i
nduc
e
m
a
l
f
u
n
ct
i
o
n
s
of
s
e
n
si
t
i
ve
e
q
u
i
p
me
nts,
over
v
o
l
t
a
g
e
by
resona
nce an
d ha
rm
on
i
c
v
ol
ta
ge dro
p a
c
ross the
ne
tw
ork
impe
da
nc
e
;
th
e
se
s
ig
ni
fi
c
a
n
tly
d
et
erio
ra
t
e
t
h
e
p
o
w
er
qua
l
ity i
n
t
h
e d
i
s
t
rib
u
tio
n
s
y
st
em
.
S
o
,
i
t
i
s
ne
c
e
ssary
t
o
de
ve
lop tec
h
n
i
ques
to
r
e
duce
al
l
th
e
har
m
on
ics
as
i
t
i
s
rec
o
mm
ende
d
i
n
t
he
I
EEE
5
1
9
-1
99
2.
T
r
a
di
tio
na
l
s
o
lu
t
i
o
n
(
pass
ive
f
ilters)
i
s
associ
at
ed
w
it
h
ce
rt
a
i
n
li
m
i
ta
t
i
o
n
,
like
fi
xe
d
com
p
e
n
sat
i
o
n
,
re
so
na
n
ce
pr
o
b
l
e
m
a
n
d
b
u
lk
y
in
n
a
t
ure.
I
n
orde
r
t
o
o
verc
ome
th
e
abo
v
e
me
ntio
ne
d
l
i
mita
t
i
o
n
s,
s
inc
e
i
t
s
i
n
t
ro
duc
tio
n
som
e
t
w
e
n
t
y
year
s
a
g
o
,
th
e
A
c
tive
P
o
w
e
r
F
i
lter
s
(
A
PF)
h
ave
prese
n
t
e
d
a
g
ood
s
o
l
u
t
i
o
n
.
Bas
i
c
pri
n
c
i
ple
of
t
h
e
A
P
F
i
s
propo
s
e
d
b
y
G
y
ugy
i
and
S
t
ryc
u
l
a
i
n
t
h
e
year 1976
[1]-[3].
G
e
nera
lly,
the
A
P
F
ac
ts
a
s
a
ha
rm
oni
c
curr
ent
s
ource
,
w
h
ich
i
n
j
e
c
t
s
a
n
a
nti
-
p
h
a
se
b
ut
e
qu
al
ma
gni
tude
of t
h
e
ha
rm
on
i
c
a
nd
re
act
i
v
e l
o
a
d
c
urre
nt
.
A
s
a resul
t, com
p
one
n
t
s of harm
o
nic c
u
rre
nt
s c
o
nta
i
ned
in
t
he
l
oa
d
c
u
r
r
ent
a
r
e
c
a
nc
e
l
l
e
d
a
n
d
t
h
e
so
ur
ce
curre
nt
r
e
m
ain
s
s
i
n
u
so
id
al
a
n
d
i
n
pha
s
e
w
it
h
t
h
e
re
sp
ecti
v
e
pha
se
v
o
lta
g
e
.
F
i
gure
1 sh
ow
s the
bas
i
c
conc
e
p
t
an
d
the
pri
n
ci
p
le
o
f
the
APF comp
ens
a
tion
[4]-[6].
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN: 2088-
8694
I
nt
J
P
ow
Elec
& Dr
i
S
y
st, Vol. 10,
N
o.
1, Mar
c
h 2
0
1
9
:
27
7 –
28
7
27
8
F
i
gure
1.
S
hun
t
acti
v
e
pow
er fi
lte
r ope
rat
i
n
g
princ
i
p
l
e
The
P
I
c
on
t
r
o
l
ler
w
h
ic
h
w
a
s
use
d
f
or
c
o
n
tr
ol
l
i
n
g
t
he
A
P
F
requir
e
s
p
rec
i
se
l
i
n
ear
m
athem
a
t
i
ca
l
mode
l
s
,
w
h
ic
h
ar
e
di
ffic
u
l
t
t
o
o
b
ta
i
n
.
S
o
,
sa
tisfa
c
t
ory
pe
rfor
m
ance
under
parameter
varia
t
i
o
ns
a
n
d
loa
d
di
st
urba
nces
c
an
n
ot
b
e
o
b
t
aine
d.
Wi
th
t
h
e
d
eve
l
o
p
m
e
n
t
o
f
p
o
w
e
rf
u
l
a
nd
fast
m
i
c
r
opr
oce
ssor
s
,
seve
ral
tech
n
i
q
u
es
a
d
d
ressed
t
o
c
o
n
tr
ol
t
he
se
k
i
nds
o
f
p
o
w
e
r
co
nver
t
e
r
s
h
a
v
e
be
en
t
he
s
u
b
j
ect
o
f
n
u
m
e
rous
rese
arc
h
es
i
n
r
ece
nt
y
ear
s,
s
u
c
h
as
S
l
i
di
n
g
M
o
d
e
Co
n
t
ro
l
(SMC)
[6],
o
p
t
im
al
c
o
n
t
r
ol
a
nd
M
ode
l
P
r
edict
i
v
e
Control
(MP
C
)
[7]-[14]
.
Currently
,
a
n
d
tha
n
ks
t
o
t
h
e
p
o
t
e
n
tia
l
of
d
i
g
i
t
a
l
s
ig
n
a
l
process
o
rs
(
D
S
P
s
),
t
he
M
P
C
is
a
p
rom
i
si
n
g
con
t
ro
l
m
e
th
od
a
nd
ha
s
se
ve
ral
be
ne
fits
s
uc
h
a
s
fa
st
t
r
a
c
k
in
g
resp
on
se,
a
hig
h
c
o
n
tr
ol
ban
d
w
i
d
t
h an
d
a ve
ry e
asiest w
a
y
to i
n
cl
u
d
e
syste
m
nonl
ine
a
r
iti
e
s
a
nd
c
o
n
s
t
r
ai
nts
[1
5].
Tw
o
di
ffere
n
t
c
ontr
o
l
s
t
rate
g
i
es
r
egar
din
g
t
h
e
M
P
C
a
lg
ori
t
hm
s
c
a
n
b
e
a
d
op
te
d:
C
o
n
tin
u
ous
C
on
tro
l
Se
t
-
Mode
l
P
r
edic
tive
C
o
nt
r
o
l
(CC
S
-MPC)
and
F
i
n
ite
C
o
n
t
r
ol
S
e
t
-
Mo
de
l
Pr
edict
i
ve
C
o
n
tr
ol
(
FCS
-
MPC)
[
9
]
[1
6].
The
f
o
rm
er
i
s
bas
e
d
on
t
he
p
r
e
dic
t
ion
o
f
t
h
e
s
a
t
e
v
a
ria
b
l
es
a
c
c
ord
i
n
g
t
o
a
disc
rete
m
o
d
e
l
o
f
the
con
v
er
t
e
r.
T
he
se
p
re
dic
t
i
o
n
s
a
re
e
val
u
ate
d
i
n
a
cos
t
f
u
n
c
t
i
on
o
ve
r
a
pre
d
ict
i
on
h
o
riz
on
in
o
r
d
er
t
o
o
b
t
a
i
n
a
seq
u
enc
e
o
f
th
e
fu
t
u
re
c
o
n
tr
ol
act
i
ons.
T
h
e
n
,
on
l
y
t
he
f
i
r
st
v
a
lue
o
f
t
h
i
s
se
que
nce
is
c
on
si
dere
d
a
n
d
t
h
e
alg
o
ri
t
h
m
is
c
om
pu
te
d
i
n
e
a
c
h
sam
p
l
i
n
g
pe
ri
o
d
.
Th
is
a
p
p
roa
c
h
h
a
s
sev
e
ral
a
d
va
n
t
ag
e
s
s
uc
h
as
a
n
impro
v
e
m
e
n
t
of
t
h
e
T
H
D
and
t
h
e
fix
e
d
sw
i
t
c
h
i
n
g
fre
q
u
e
ncy.
H
ow
ev
er,
a
h
i
g
h
amou
nt
o
f
cal
cu
l
a
ti
on
s
are
nee
d
e
d
i
n
or
de
r
to
s
o
l
ve
t
he
c
os
t
fu
nc
tio
n
w
h
ic
h
ma
kes
th
i
s
opt
i
o
n
a
m
o
re
c
omple
x
s
o
l
uti
o
n.
I
n
the
l
a
tt
e
r
,
the
op
tim
iza
t
i
o
n
p
r
ob
lem
is
r
ed
u
c
e
d
t
o
a
fi
ni
te
num
ber
of
s
w
itch
i
ng
s
tates.
H
ere
,
t
he
c
os
t
fu
nc
ti
on
is
c
a
l
cula
te
d
f
o
r
e
v
ery
swi
t
c
h
i
ng
s
t
a
t
e
,
ea
ch
o
n
e
r
e
l
a
t
ed
t
o
a
sp
eci
fi
c
v
o
lt
a
ge
vec
tor. Th
e
sw
i
tc
hi
n
g
sta
te
t
ha
t min
i
miz
e
s
t
h
e
e
r
ro
r
b
e
t
w
ee
n
t
h
e
cu
rrent
a
nd
i
t
s
r
e
f
ere
n
ce
i
s
th
e
n
a
p
p
l
i
e
d
to
t
h
e
conv
e
r
te
r.
I
n
t
h
i
s
w
ay,
t
h
e
co
st
f
un
cti
o
n
i
s
eva
l
ua
t
e
d in a
fin
i
t
e num
b
e
r
of
t
im
es e
ac
h sam
p
lin
g
peri
od
[10]
,
[17],
[18].
I
n
t
h
i
s
p
a
per
a
p
r
ed
ic
ti
ve
c
ur
rent
c
o
n
tr
ol
t
e
c
hn
i
que
f
or
a
s
hu
n
t
A
P
F
is
p
re
sen
t
ed.
A
c
t
u
al
v
a
l
ue
s
o
f
l
o
ad
c
u
rren
t
s
are
me
a
s
u
r
ed
a
nd
i
nj
e
c
t
e
d
in
t
he
p
re
di
ct
iv
e
m
o
d
e
l
t
o
gener
a
t
e
p
red
i
c
t
i
o
ns
o
f
f
u
t
u
r
e
c
urr
e
nt
,
one
f
o
r
e
ach
v
o
l
t
a
ge
v
ect
o
r
.
Th
e
s
e
p
r
edi
c
t
i
on
s
a
r
e
ev
a
l
u
a
t
e
d
wi
th
a
qua
l
ity fun
cti
o
n
a
nd t
h
e ve
ctor tha
t
m
i
ni
m
i
zes
th
i
s
f
unc
t
i
on
i
s
ap
pl
i
e
d
d
u
rin
g
t
he
n
e
x
t
sam
p
li
ng
i
n
t
erva
l
.
A
c
a
p
a
c
i
t
o
r
v
o
lta
ge
c
o
n
t
rol
t
echn
i
que
p
r
e
vi
ou
sl
y
use
d
a
s
a
harm
onic
c
u
rr
ent
e
x
trac
ti
o
n
t
e
c
h
ni
q
u
e.
P
redic
t
i
v
e
cur
r
e
nt
c
on
tro
l
b
a
s
e
d
o
n
t
h
e
su
ppl
y
cur
r
ent
i
s
pro
pose
d
i
n
F
i
gure
2.
The
pro
pose
d
w
ork
i
s
d
e
v
e
l
o
p
ed
i
n
t
h
e
fo
ll
ow
ing
sect
i
o
ns.
F
i
r
s
t
ly,
the
g
e
nera
tio
n
of
t
h
e
r
efe
r
enc
e
curr
ent
is
p
r
e
sen
t
ed.
S
econ
d
l
y
,
the
m
o
de
li
sat
i
o
n
o
f
the
grid-c
o
n
n
ecte
d
i
nverter
i
s
de
ve
l
o
p
e
d.
F
inal
l
y
,
S
i
m
u
l
a
t
i
o
n
r
esul
ts
u
s
i
ng
ma
tl
ab/
S
i
m
u
l
i
n
k a
nd ex
pe
rime
n
t
a
l
r
e
s
u
lt
s aro
und
Dsp
a
c
e
-
1
104 are
p
r
e
s
en
t
e
d
.
F
i
gure
2.
P
re
dic
tive
cur
r
en
t c
o
n
t
r
o
l
b
l
oc
k d
i
agra
m
Evaluation Warning : The document was created with Spire.PDF for Python.
Int J
P
o
w
El
e
c
&
D
ri S
yst
I
S
S
N
:
2088-
86
94
H
i
g
h
-per
fo
rm
a
n
ce
act
i
ve
p
o
w
e
r
filter
i
m
p
l
e
m
e
n
t
a
ti
on b
a
se
d
on pr
ed
ic
tive
curren
t
c
o
n
t
ro
l (
B
ach
a
r Ra
hi
m
ap)
27
9
2.
REFE
RENCE CURR
E
N
T
GEN
E
R
A
T
I
ON
A
m
ong the
pre
s
en
ted
m
e
tho
d
s
f
or
i
de
nt
ify
i
n
g
h
a
r
m
onic
c
o
m
pone
n
t
s
o
f
the
lo
a
d
c
urrent,
a
simplified
and
ap
p
l
i
c
a
b
le me
t
hod
w
i
t
h
h
i
g
h
pe
rform
anc
e
i
s prese
n
te
d
i
n
[
1
9
]-[2
2
].
T
he
i
m
p
orta
nt
q
u
e
st
io
n
of
t
he
A
P
F
i
s
how
t
o c
a
lcu
l
a
t
e
the a
ppr
opri
a
te
c
om
pen
s
at
i
o
n cur
r
ent,
w
hic
h
w
i
l
l
be
de
tai
l
ed
i
n
t
h
e ne
x
t
s
ecti
o
n.
The
bl
oc
k
d
i
a
g
ra
m
of
t
he
p
r
o
p
o
se
d
harm
onic
e
x
trac
tio
n
me
thod
is
p
e
r
f
o
r
m
e
d
u
s
i
n
g
t
h
e
c
a
p
a
c
i
t
o
r
vo
lta
ge
c
on
tro
l
s
h
o
w
n
i
n
F
i
g
u
r
e
3.
T
he
m
ai
n
p
u
r
p
o
s
e
of
t
he
s
u
g
g
e
s
t
e
d
appro
a
ch
i
s
to
r
egu
l
at
e
th
e
po
we
r
f
l
o
w
o
f
t
h
e
w
ho
l
e
s
y
s
t
e
m
a
s
w
e
l
l
a
s
t
o
co
mp
e
n
sat
e
h
armo
n
i
c
an
d
re
ac
t
i
v
e
c
u
r
r
ents,
usi
n
g
t
h
e
P
r
e
d
ict
i
ve
C
urrent
Con
t
ro
l (P
CC) in a
l
l
o
p
era
t
ion
mode
s.
I
t
co
ns
is
t
s
o
f t
w
o p
a
rt
s;
the firs
t
o
ne
is the re
fere
nc
e
cur
r
ent ge
nera
tio
n
a
n
d
t
h
e
secon
d
o
n
e
i
s
t
h
e
regu
l
a
ti
on
o
f
th
e
ca
p
a
c
i
t
o
r
vol
t
a
ge
,
usi
n
g
a
P
I
c
on
tro
l
ler.
T
he
o
u
t
pu
t
is
p
r
o
p
o
r
t
i
ona
l
to
t
he
i
nsta
n
t
a
n
e
o
us
c
ha
nge
s
in
p
ow
er
b
a
l
anc
e
.
M
u
l
tip
l
y
i
n
g
t
h
is
o
utp
u
t
b
y
t
h
e
p
e
r
un
it
vo
ltage
s
of
t
he
c
o
mmo
n
c
oup
li
ng
poi
nt
(
ou
tp
ut
o
f
t
h
e
P
LL).
Th
e
n
,
b
y
su
bt
ra
c
t
in
g
s
o
u
rc
e
curre
nts
fro
m
f
ilte
r
c
u
rre
nt
s,
t
h
e
refere
nce
fi
lter
curre
nts
w
ill
b
e
e
xt
r
acte
d
. F
i
gure
3 sh
ow
s t
h
e
p
ro
pose
d
b
l
o
ck
d
i
a
gram
.
F
i
gure
3. Capa
c
it
or vo
l
tage
c
on
tro
l
2
.
1
.
PI-co
ntro
l
l
er
F
i
gure
3
sh
ow
s
the
b
l
oc
k
d
i
a
g
r
a
m
of
t
he
P
I
con
t
ro
l
sc
hem
e
f
or
t
he
a
c
t
ive
pow
er
f
ilte
r.
T
he
D
C
-
bus
vo
lta
ge
i
s se
nse
d
a
n
d
c
om
par
e
d w
i
t
h
a
re
f
er
ence
vo
l
tage
a
nd
a
ls
o ca
lc
ul
a
t
i
n
g the
er
ror si
gna
l.
The
er
r
or si
gna
l
is
u
se
d
a
s
i
n
p
u
t
f
or
t
he
P
I
co
ntr
o
l
l
er.
The
P
I
c
on
t
r
o
l
ler
es
tim
ates
t
he
m
ag
ni
tude
o
f
t
h
e
r
e
fe
renc
e
c
u
rr
ent
pea
k
and
c
o
n
t
ro
ls
t
h
e
d
c
-
bus
v
o
l
ta
ge
o
f
t
h
e
i
nver
t
er
.
Whe
r
e,
[
=
0
.
2]
i
s
t
h
e
pro
por
ti
o
n
al
c
on
st
ant
tha
t
d
eter
min
e
s
the
d
y
n
am
ic
r
esp
onse
o
f
t
he
D
C-bus
v
o
l
ta
ge
a
n
d
[
=
15]
i
s
the
in
tegr
at
ion
co
ns
t
a
n
t
t
h
a
t
determ
i
n
es
t
h
e
set
tli
ng t
i
m
e
.
2.2.
Ph
a
se
Loc
ked
Loop
(
PLL)
Ei
ther
t
he
m
ea
sured
v
o
l
ta
ge
s
on
a
thre
e-pha
se
n
etw
o
r
k
v
,v
a
n
d
v
s
ub
j
e
c
t
e
d
t
o
a
di
r
ect
Conc
ord
i
a
C3
2
t
r
ansform
a
ti
on
t
o
w
ork
in
a
r
efe
r
enc
e
s
ystem
αβ
t
w
o
d
ime
n
sions.
The
vol
t
a
ges
o
b
ta
i
n
ed
a
re
v
θ
andv
θ
(
w
ith
θ
i
ns
t
a
n
t
a
n
eo
us
pha
se
o
f
the
actua
l
forw
ard
vol
t
a
g
e
).
T
he
r
ota
t
i
on
a
n
g
l
e
i
s
ob
ta
ine
d
b
y
i
n
t
e
gr
at
ion
of
t
he
e
sti
m
ate
ang
u
l
a
r
f
r
e
quenc
y
d
eter
mi
ned
b
y
t
he
P
I
control
l
er.
In
o
rder
f
or
t
h
e
PLL
t
o
b
e
c
l
o
s
e
d
,
t
h
e
est
i
m
ate
d
r
o
t
ati
o
n
an
g
l
e
m
ust
be
e
qua
l
t
o
t
he
a
ctua
l
an
gl
e
of
t
he
f
orw
a
rd
vo
lta
ge
,
thi
s
can
b
e
ac
hie
v
ed
w
he
n
the
va
lu
e
of
i
s
z
e
ro
[
1].
Let’s
the
me
asured
v
o
lta
g
e
s
at
t
he
con
n
ec
tio
n
poi
nt o
f t
h
e
A
P
F
c
ons
i
d
ere
d
a
s
sinus
o
i
dal a
n
d b
a
la
n
c
e
d
,
gi
ve
n by
the
fol
l
o
w
i
n
g
equa
tio
ns
:
⎩
⎨
⎧
√
2
√
2
√
2
(
1
)
The
ca
lcu
l
a
t
io
n
pr
oce
dur
e
af
te
r
app
lica
t
i
on
of
t
he
t
ra
nsfor
m
ati
on
of
C
o
n
c
ord
i
a
a
nd
P
a
rk
i
s
g
i
ve
n
b
y
t
h
e
fol
l
o
w
ing
r
e
l
a
ti
o
n
ship
:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN: 2088-
8694
I
nt
J
P
ow
Elec
& Dr
i
S
y
st, Vol. 10,
N
o.
1, Mar
c
h 2
0
1
9
:
27
7 –
28
7
28
0
√
3
(
2
)
w
h
er
e C32 i
s
t
he
C
onc
ord
i
a m
a
trix
g
i
v
e
n
b
y:
1
0
√
√
(
3
)
The
o
p
er
ator
P
s
ymbo
l the
P
a
rk tr
a
nsform
ati
on, de
f
i
n
ed
b
y
:
(
4
)
The
phas
e
s
ub
j
ecti
o
n
w
ill
be
a
chie
ve
d
w
h
e
n
.
I
n
thi
s
c
a
s
e,
t
he
actua
l
a
n
g
l
e
o
f
t
he
n
e
t
w
o
rk
vo
lta
ge a
n
d
t
he
e
sti
m
ate
d
a
ng
l
e
a
re equa
l. S
o,
w
e
can
w
rite:
0
√
3
(
5
)
To
s
ea
rch
o
t
he
r
c
o
mpr
o
mi
se
b
etw
e
e
n
t
he
d
ynam
i
c
be
ha
v
i
or
o
f
c
o
n
v
e
n
t
i
ona
l
P
LL
and
its
in
se
ns
iti
v
ity
t
o
vari
o
u
s
d
i
st
ur
ba
nce
s
,
a
ne
w
appr
oac
h
i
s
to
i
n
t
ro
d
u
ce
a
M
u
lt
i
V
a
r
i
a
b
le
F
il
ter
(
M
V
F
)
be
t
w
e
e
n
two
q
u
a
d
r
a
ti
c
pi
ns of
t
h
e co
n
v
en
t
i
o
n
a
l
v
e
r
si
on
of t
h
e
P
LL
in or
der
to
o
ve
rcom
e th
e
s
e d
i
st
urbanc
e
s
b
e
h
avi
o
ral
pro
b
lem
s
F
ig
ure
4.
T
he
t
r
a
nsfer
fun
c
t
i
o
n
o
f
t
h
e
MV
F
as
s
how
n
i
n
F
ig
ure
5
is
e
x
p
r
e
sse
d
b
y
t
he
fo
l
l
ow
i
n
g
ex
pre
s
si
on
:
∗
(
6
)
F
i
g
u
r
e
4
.
S
t
r
u
c
t
u
r
e
o
f
conv
e
n
ti
on
al
P
.L
.
L
w
i
t
h
M
VF
F
i
gure
5.
S
c
h
em
atic d
iagr
am
of
t
h
e P
LL incl
ud
i
ng a
MV
F
Evaluation Warning : The document was created with Spire.PDF for Python.
Int J
P
o
w
El
e
c
&
D
ri S
yst
I
S
S
N
:
2088-
86
94
H
i
g
h
-per
fo
rm
a
n
ce
act
i
ve
p
o
w
e
r
filter
i
m
p
l
e
m
e
n
t
a
ti
on b
a
se
d
on pr
ed
ic
tive
curren
t
c
o
n
t
ro
l (
B
ach
a
r Ra
hi
m
ap)
28
1
3.
MODDEL
I
N
G
OF A
GRI
D
-CON
NECTED
I
N
V
E
R
TER WITH L-FILTER
The
three
-
p
h
as
e
sh
un
t
A
P
F
circ
uit
a
s
s
how
n
in
F
i
g
ure
6.
F
rom
t
h
is
c
irc
u
it,
t
he
f
ol
l
o
w
i
n
g
e
q
u
a
tio
ns
for
ea
ch pha
se
-leg
can
b
e
de
fine
d
as
fo
l
l
ows
:
F
i
gure
6.
S
hun
t ac
tive
p
o
w
e
r
fil
t
er
A
ssum
i
ng
t
h
e
var
i
a
b
l
e
s
,
a
nd
r
e
p
rese
nt
t
h
e
s
w
i
t
c
h
i
ng
s
t
a
t
es
o
f
eac
h
pha
se
,
w
h
er
e
“
1
''
s
ta
nd
s
for
t
h
e
i
nve
rte
r
outpu
t
c
o
n
n
e
c
ted
to
t
he
posit
i
v
e
dc
-l
ink
v
o
lta
ge,
and
``0''
m
ea
ns
t
hat
t
h
e
inve
r
t
er
o
u
t
pu
t
is
con
n
ec
ted
t
o
t
he
n
e
g
a
t
i
v
e
d
c
-li
nk
vo
l
t
a
g
e.
S
w
i
tc
hin
g
sta
t
e
s
o
f
the
two
-
leve
l
thr
ee-
ph
ase
i
n
ve
rter
c
an
b
e
re
p
r
e
s
e
n
t
e
d
by
,
w
ith
j
=
0
,.
..,
7.
A
cc
ording
t
o
out
pu
t
sw
i
t
c
h
i
n
g
c
o
mb
ina
t
i
ons
,
tota
l
l
y
8
v
o
l
t
a
g
e
vec
t
ors
ca
n
be
g
e
n
era
t
e
d
[
17]
,
[1
8],
[22]
.
The
sw
itch
i
n
g
s
ta
tes
o
f
t
he
c
on
v
erter
ar
e
de
t
e
rm
i
n
e
d
b
y
t
h
e
ga
t
i
n
g
s
ig
na
l
,
a
nd
a
s fol
l
ows:
1,
0,
(
7
)
1,
0,
(
8
)
1,
0,
(
9
)
A
nd c
a
n be
e
x
p
r
e
ssed
in
v
ec
t
o
ria
l
f
orm
by
.
.
(
1
0
)
Whe
r
e
/
The
o
u
tp
ut spa
c
e
volta
ge
ve
c
t
ors gene
rate
d
b
y
the
i
n
v
e
r
ter
are
de
fi
n
ed
b
y
.
.
(
1
1
)
The
o
u
tp
ut p
ha
se
vo
lta
ges
a
nd
c
an
b
e
e
x
p
r
ess
e
d
as
:
(
1
2
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN: 2088-
8694
I
nt
J
P
ow
Elec
& Dr
i
S
y
st, Vol. 10,
N
o.
1, Mar
c
h 2
0
1
9
:
27
7 –
28
7
28
2
The
n
,
the
out
p
u
t v
o
lta
ge
v
ec
t
o
r
ca
n
be
r
elate
d
to t
h
e
sw
itch
i
ng
s
tate
vec
tor
by
;
.
(
1
3
)
Ac
cordi
n
g
t
o
t
he
r
e
f
er
ence
c
urr
e
nt
d
irec
ti
o
n
i
n
Fig
u
re
1
,
the
o
u
t
pu
t
curre
nts
of
t
he
t
w
o
-leve
l
t
hre
e
-
pha
se
i
nve
rter
c
a
n
be
expre
sse
d a
s
:
⎩
⎪
⎨
⎪
⎧
(
1
4
)
The
ma
the
m
a
tic
m
ode
l
of
t
he
t
hre
e
-phas
e
i
n
v
e
r
ter
i
n
t
h
e
a
bc
r
ef
e
r
e
n
c
e
f
r
a
m
e
n
e
e
d
s
t
o
b
e
trans
f
or
me
d
to
t
he
m
ode
l
i
n
t
he
s
ta
t
i
o
n
a
ry
r
efere
n
ce
fram
e
[
22]
[
9]
[
23],
usin
g
the
corr
esp
o
n
d
i
n
g
trans
f
or
ma
t
i
o
n
m
atrix
(in
v
aria
nt a
mpl
i
t
ude tr
a
nsf
o
rm
atio
n)
gi
v
e
n by.
/
1
1
/
2
1
/
2
0
√
3
/2
√
3
/2
1/
√
2
1/
√
2
1/
√
2
(
1
5
)
The
com
b
i
n
a
t
i
on o
f
(14)
and
(15)
can
b
e e
xpressed
as
f
o
l
l
o
ws:
(
1
6
)
I
n
o
rder
t
o
a
c
h
ieve
t
he
i
nve
rt
er
c
urre
nt
s
tra
c
ki
n
g
w
ith
r
esp
e
c
t
t
o
t
h
e
r
e
fe
renc
e
i
n
t
he
n
e
x
t
sam
p
l
i
n
g
t
i
me
,
t
h
e
i
nve
rt
er
d
e
s
i
r
ed
o
utp
u
t
v
o
lt
ag
es
i
n
th
e
st
at
io
n
a
ry
r
e
fere
nc
e
fram
e
c
a
n
b
e
o
b
t
a
i
ne
d
b
y
t
he
di
sc
retiz
a
t
i
on o
f
(16)
as
fol
l
o
w
s.
∗
∗
∗
∗
(
1
7
)
Whe
r
e
∗
and
∗
are
the
pr
ed
ict
e
d
o
u
t
p
ut
v
o
l
t
a
ge
r
e
f
ere
n
ce
s
respec
tive
l
y,
a
nd
i
s
t
h
e
s
a
m
p
l
i
n
g
t
i
me
.
Usi
n
g
Ki
rchh
of
f
’
s
cu
rren
t
l
a
w
a
t
th
e
Co
mmon
C
ou
pli
n
g
Poi
n
t
(C
CP
),
t
he
f
ol
l
o
w
i
ng
e
x
pr
essi
o
n
i
s
ob
ta
ine
d
:
(
1
8
)
Whe
r
e
a
nd
ar
e
the
l
o
a
d
a
n
d
s
u
p
p
l
y
phas
e
c
ur
rents
resp
ec
t
i
ve
l
y
,
a
t
t
he
s
am
plin
g
i
n
s
t
a
n
t
(k)
.
S
i
nce
t
h
e
sam
p
l
i
ng
in
sta
n
t
(k
+1)
i
s
n
ot
a
va
i
l
a
bl
e
,
∗
1
i
s
rep
l
ac
ed
b
y
∗
.
Th
i
s
i
nt
rod
u
c
e
s
one
sam
p
lin
g
tim
e
de
l
a
y
w
h
ic
h
be
com
e
s
less
si
gn
ifica
n
t
as
t
he
s
a
m
pl
i
n
g
fr
eque
nc
y
i
n
cre
a
ses.
T
he
r
efe
r
enc
e
curr
ent o
f
t
he s
hu
n
t
APF
can be
expre
s
se
d as:
∗
∗
(
19)
The
su
bs
ti
tut
i
o
n of
(
18)
a
n
d
(19)
in (
17) gives:
∗
∗
∗
∗
(
2
0
)
I
n
(
20)
r
e
pre
s
e
n
t
s
t
he
p
red
i
ct
e
d
i
n
v
e
r
ter
ou
t
p
u
t
v
o
l
t
a
ge
e
xpr
essed
i
n
t
e
r
ms
o
f
t
h
e
ref
e
ren
c
e
a
n
d
th
e
ac
t
u
a
l
s
u
p
p
l
y
c
urr
e
nts
[12],
[13],
[23].
T
h
e
ma
in
c
har
acter
i
s
ti
c
o
f
t
hi
s
c
o
ntr
o
l
s
t
r
a
teg
y
i
s
t
he
s
elec
t
i
o
n
of
t
he
Evaluation Warning : The document was created with Spire.PDF for Python.
Int J
P
o
w
El
e
c
&
D
ri S
yst
I
S
S
N
:
2088-
86
94
H
i
g
h
-per
fo
rm
a
n
ce
act
i
ve
p
o
w
e
r
filter
i
m
p
l
e
m
e
n
t
a
ti
on b
a
se
d
on pr
ed
ic
tive
curren
t
c
o
n
t
ro
l (
B
ach
a
r Ra
hi
m
ap)
28
3
v
o
l
t
a
g
e
v
e
c
t
o
r
t
h
a
t
m
i
n
i
m
i
z
e
s
a
q
u
a
l
i
t
y
f
u
n
c
t
i
o
n
g
[
9
]
,
[
2
3
]
.
T
he
q
ua
lit
y
func
t
i
o
n
i
s
ex
presse
d
in
o
r
t
h
o
g
o
n
a
l
coor
di
na
tes
in
t
he f
o
l
l
o
w
i
n
g
fo
r
m.
g
∗
∗
(
2
1
)
Whe
r
e
∗
an
d
∗
a
re
t
he
r
ea
l
and
im
agi
n
ary
p
a
rt
s
of
t
h
e
f
u
t
ure
re
fer
e
nce
cur
r
e
nt.
a
nd
a
r
e
t
h
e
r
e
a
l
a
n
d
ima
g
ina
r
y pa
r
t
s
of
t
he
pre
di
c
t
e
d
c
urrent.
4.
SIMU
L
A
TION
R
ESULT
S
F
i
gure
7
show
s
the
b
l
o
c
k
dia
g
ra
m
of
t
he
s
h
u
n
t
A
P
F
a
ssoc
iate
d
w
i
th
a
m
ode
l
pre
d
ic
t
i
ve
c
on
t
r
o
l
le
r.
A
c
tua
l
v
a
l
ue
s
of
s
o
u
rc
e
c
u
rrent
s
a
re
m
e
a
sure
d
at
i
ns
t
a
n
t
k
a
nd
use
d
a
s
in
pu
t
for
a
predic
t
i
v
e
m
odel
tha
t
com
p
u
t
e
s
t
he
v
al
ues
of
t
he
c
urre
nt
a
t
t
h
e
ne
xt
s
a
m
p
l
in
g
t
i
m
e
f
o
r
ea
ch
o
f
the
pos
si
ble
sw
itc
hi
n
g
s
ta
tes
of
t
h
e
in
su
la
te
d
ga
te
b
i
p
ol
a
r
t
r
a
ns
ist
o
r
(IG
B
T
)
inv
e
rte
r
.
These
pre
d
i
c
tio
ns
a
re
e
va
l
u
a
t
ed
w
ith
t
he
q
ua
li
t
y
f
unc
t
i
o
n
and
the
ve
c
t
or
t
ha
t
mi
nimiz
e
s
t
his
fu
nc
ti
on
i
s
a
pp
lie
d
dur
i
n
g
t
he
n
e
x
t
sa
mpl
i
ng
inter
v
a
l
.
The
sw
itc
hi
n
g
s
ta
te
cou
l
d be
cha
n
g
e
d on
l
y
onc
e a
t
e
a
c
h sa
mpli
ng
ins
t
a
n
t
.
F
i
gure
7. S
hun
t
A
P
F
w
i
th pr
e
dic
t
ive c
u
rre
nt
c
on
tro
l
I
n
o
rder
t
o
i
n
ve
stiga
t
e
i
t
s
pe
r
f
or
ma
nce,
t
he
p
r
opose
d
s
y
s
te
m
sh
ow
n
in
F
igure
7
is
f
irs
t
l
y
e
va
lua
t
e
d
by
s
i
m
u
l
a
t
ion
us
in
g
MA
TLA
B
/
S
i
mul
i
n
k
.
T
he
n
o
n
-l
i
n
ea
r
l
o
ad
i
s
rep
r
e
sen
t
ed
b
y
a
thre
e-pha
se
d
i
o
de
r
e
c
ti
fier
fe
ed
i
n
g
a
n
ind
uc
ti
ve l
oad a
c
t
i
ng as a harm
o
n
i
c c
u
rre
nt
p
ro
d
u
ci
n
g so
urc
e
(R=
50
Ω
;
L=
0
.5mH).
Th
e
i
nduc
t
a
n
ce
of
t
he
s
hu
n
t
A
P
F
c
oup
l
i
n
g
i
n
duc
t
o
r,
i
s
3
m
H
.
The
i
nver
t
e
r
c
a
p
ac
ito
r
a
nd
t
h
e
r
ef
e
r
e
n
c
e
v
ol
t
a
g
e
a
re
204
0
μF
and
17
3
V
r
e
spec
ti
ve
ly.
The
si
m
u
lat
i
o
n
re
s
u
l
t
s
for
the
system
b
ef
o
r
e
and
a
f
t
e
r
co
mp
e
n
sat
i
o
n
a
r
e
sh
o
w
n
in
F
i
gure
s
8,
9,
10 and
1
1
re
s
pe
cti
v
e
l
y.
The
per
f
orm
a
nce
and
t
h
e
effe
cti
v
e
n
e
s
s
o
f
t
he
s
hu
nt
A
P
F
a
ssoc
i
at
e
d
w
i
t
h
pr
ed
ict
i
ve
c
ur
rent
c
on
t
r
ol
have
b
e
e
n
v
er
i
f
ie
d
b
y
s
i
m
u
l
atio
n
usi
n
g
ma
tl
a
b
/
s
im
u
l
i
nk.
F
i
g
ur
e
8
s
h
o
w
s
t
h
e
si
m
u
la
t
e
d
w
a
ve
form
s
of
t
he
source
c
urre
nt,
t
h
e
l
o
ad
c
urr
e
nt,
the
fi
l
t
e
r
c
u
rre
nt
a
nd
t
he
D
C
b
us
v
o
lta
ge
b
e
f
ore
a
n
d
a
f
te
r
com
p
ens
a
t
i
o
n.
T
he
re
su
lt
s
sho
w
t
ha
t
af
t
e
r
c
o
mp
en
sa
t
i
on
, th
e
s
ou
rce
c
u
rrent
i
s n
e
a
rl
y s
i
n
u
so
id
al w
it
h
l
o
w
harm
onic
co
n
t
en
t w
h
ic
h
is
b
el
ow
t
he
h
a
r
m
oni
c
l
i
mit
im
po
sed
b
y
t
h
e
I
EEE
51
9-
1
992
s
t
a
nda
r
d
a
nd
t
he
D
C-bu
s
v
o
lta
ge
i
s
reg
u
la
te
d
arou
nd
i
t
s
r
e
fe
renc
e.
F
i
gur
e
9
show
s
t
h
e
D
C
-bus
v
o
l
tage
,
t
h
e
so
u
r
ce
c
u
rre
nt
a
n
d
t
he
s
o
u
r
ce
volta
ge
i
n
stea
d
y
st
a
t
e.
T
he
s
imula
t
io
n
resu
l
t
s
show
t
ha
t
t
h
e
source
c
urre
nt
a
nd
t
h
e
so
urc
e
vo
l
t
a
g
e
a
r
e
in
pha
se
,
w
i
th
a
p
ow
er
fa
ct
or
n
ear
l
y
u
ni
ty.
The
harmon
ic
s
pec
t
rum
befor
e
c
on
nec
tin
g
t
h
e
A
P
F
i
s
s
h
o
w
n
i
n
F
i
g
u
r
e
1
0
.
T
h
e
h
a
r
m
o
n
i
c
spec
trum
of th
e
sour
ce
curr
ent a
f
ter
con
n
ec
ti
ng
the
AP
F is
s
how
n
in F
ig
ure
11,
w
ith a
g
o
o
d
TH
D
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N: 2
0
8
8
-
86
94
I
nt
J
P
ow
E
l
e
c
&
Dr
i
S
y
st,
Vol.
10,
N
o.
1
,
Mar
c
h
2
0
1
9
:
27
7
–
287
28
4
F
i
gur
e
8.
S
i
m
u
lat
i
on
r
e
su
l
t
s:
(
a)
D
C
bus
v
o
l
t
a
ge
(
V
)
,
(
b
)
sour
ce
c
ur
rent
(
A
)
,
(
c)
l
oa
d
cur
r
ent
(
A
),
(d)
f
il
ter
cur
r
ent (
A
)
F
i
gur
e
9.
S
i
m
ulat
i
on
r
e
su
l
t
s:
(
a
)
D
C
bus
v
o
l
t
a
ge
(
V
)
,
(
b
)
sour
ce
c
ur
rent
a
n
d
v
ol
ta
ge
c
ur
r
e
nt
F
i
gur
e
1
0
.
H
a
rm
onic
spe
c
tr
u
m
b
ef
or
e
conn
ec
tin
g
th
e
A
P
F
F
i
g
u
r
e
11.
H
arm
onic
s
p
ec
tr
u
m
a
f
t
e
r
c
onnec
t
i
n
g
th
e
AP
F
5.
EX
PERIMENT
AL R
E
S
ULTS
Th
e
con
t
r
o
l
strat
e
gy
i
s
i
m
pl
eme
n
t
e
d
u
s
ing
a
d
S
PAC
E
card
DS
1
1
0
4
u
nde
r
Ma
t
l
ab
/S
i
m
ul
ink
T
M
R
T
W
e
n
vi
ronmen
t
. Th
e
sam
p
lin
g
ti
me
usi
ng
in pra
c
tica
l
t
e
s
ts
i
s
1e-
4
s
a
s
s
how
n
i
n
F
ig
u
r
e
12.
F
i
gur
e
1
2
.
Exp
e
r
i
m
e
nta
l
s
e
t
up
o
f
shu
n
t
A
PF
0.
1
0.
2
0.
3
0.
4
0.
5
-1
0
0
10
Ti
m
e
(
s
)
(d
)
0.
1
0.
2
0.
3
0.
4
0.
5
-1
0
0
10
Ti
m
e
(
s
)
(c)
0.
1
0.
2
0.
3
0.
4
0.
5
-1
0
0
10
Ti
m
e
(
s
)
(b
)
0.
1
0.
2
0.
3
0.
4
0.
5
10
0
20
0
13
0
17
3
Ti
m
e
(
s
)
(a
)
0.
3
4
0.
3
6
0.
3
8
0.
4
0.
4
2
0.
4
4
0.
4
6
0.
4
8
0.
5
0.
5
2
-5
0
0
50
Ti
m
e
(
s
)
(b
)
0.
3
4
0.
3
6
0.
3
8
0.
4
0.
4
2
0.
4
4
0.
4
6
0.
4
8
0.
5
0.
5
2
0
20
0
17
3
Ti
m
e
(
s
)
(a
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t
J
P
o
w
Elec
&
D
r
i
S
y
st
I
S
S
N
:
2088-
86
94
Hig
h
-pe
r
fo
rma
n
ce
ac
tive
p
o
w
er fi
lt
e
r
im
p
l
e
m
ent
a
t
i
on b
a
se
d on
pr
ed
ic
tive
curren
t c
o
n
t
r
o
l (B
a
c
h
a
r Ra
hi
map)
28
5
Ex
perim
e
n
t
al r
esu
lts a
im
ed to
val
i
d
a
t
e the
s
i
mulat
i
o
n
r
e
su
lts h
a
v
e bee
n
ob
t
ai
ne
d f
o
r
st
e
a
dy-
s
t
ate a
n
d
tr
a
n
sit
o
r
y
o
per
a
t
i
o
n
m
ode,
a
s
s
how
n
in
F
i
g
u
r
e
13
a,
b
a
nd
c
.
These
e
x
p
e
rime
nt
a
l
r
e
s
ul
ts
s
h
o
w
t
h
e
effec
t
i
v
ene
s
s
o
f
t
he
A
PF
a
s
so
ci
at
ed
w
i
t
h
a
p
re
dic
t
i
v
e
c
u
rre
nt
co
nt
rol
l
e
r,
t
o
co
mp
e
n
sa
t
e
h
armo
n
i
c
cu
rre
n
ts
c
rea
t
e
d
b
y
a
no
nli
n
ear
l
oa
d
a
nd
t
o
m
in
imiz
e
the
r
e
ac
t
i
ve
pow
e
r
c
o
n
s
um
p
t
i
o
n.
F
igur
e
13
(
a
)
i
l
l
us
tr
a
t
e
s
t
he
d
y
n
am
ic
r
e
s
ponse
s
o
f
t
h
e
s
ou
rce
cu
rrent
,
th
e
lo
ad
c
u
rrent
,
th
e
in
jec
t
e
d
c
ur
r
e
n
t
a
n
d
t
he
D
C
bus
v
o
lta
ge
.
I
t
i
s
ob
ser
v
e
d
t
h
a
t
t
h
e
D
C
b
us
v
o
l
t
a
ge
p
a
s
s
t
h
r
o
u
gh
a
t
r
ans
i
tio
na
l
pe
r
i
od
b
e
f
or
e
sta
b
i
l
i
zat
i
o
n
a
n
d
r
e
a
c
he
s
i
t
s
r
e
fe
r
e
nc
e
w
i
th
m
ode
rate
p
e
a
k
v
o
ltage
.
Af
te
r
com
p
ensa
t
i
on,
t
h
e
c
u
r
r
e
nt
s
our
ce
i
s
ne
ar
ly
s
inus
oi
da
l.
F
igur
e
1
3
(
b
)
,
s
ho
w
s
t
h
e
s
a
m
e
w
avefo
r
m
s
i
n
stea
dy
s
t
ate o
p
era
t
i
o
n.
F
ig
ur
e
13
(
c
)
dem
o
n
s
t
r
a
t
es
t
ha
t
a
f
ter
the
c
o
nne
c
t
i
o
n
of
t
he
A
P
F
,
the
s
our
c
e
c
ur
r
e
nt
i
s
near
ly
s
i
n
usoi
da
l
an
d
i
n
p
h
a
se
w
i
t
h
t
he
s
our
c
e
v
ol
t
a
ge
w
ith
a
g
oo
d
p
o
w
e
r
f
a
ctor
.
F
i
g
u
r
e
1
4
sh
o
ws
t
he
s
p
ectrum
o
f
t
he
s
our
ce
c
u
r
r
ent
b
e
for
e
c
o
mpe
n
sa
t
i
on.
T
he
s
o
u
r
ce
c
u
r
r
e
nt
s
h
o
w
s
h
i
g
h
h
a
r
m
onic
con
t
e
n
t.
F
igur
e
15
i
llus
t
rate
s
th
e
spec
trum
o
f
t
h
e
sour
c
e
c
ur
r
e
nt af
t
er
com
p
e
nsa
t
i
o
n.
These r
es
ul
ts dem
o
n
str
a
te t
h
a
t
the perf
o
rm
ance
o
f t
h
e
shu
n
t
APF
b
ase
d
o
n
pr
ed
ict
i
ve
c
ur
r
e
nt
c
o
n
tr
ol
i
n
ter
m
s
of
h
ar
mon
i
cs
e
l
i
m
ina
t
i
o
n
a
n
d
r
e
acti
v
e
p
o
w
e
r
c
om
pens
ati
o
n
is
ve
r
y
s
a
tisfa
c
t
or
y.
(
a
)
C
h1,
C
h
2
,
C
h3
sca
l
e
5
A
/
d
i
v:
s
our
c
e
c
ur
r
e
nt
i
(A),
l
o
a
d
cu
r
r
en
t
i
,
injec
t
e
d
cur
rent
i
(
A
)
r
e
spe
c
t
i
v
e
l
y
;
Ch4
sc
a
l
e
15
0
V
/
d
i
v
ca
pac
i
t
o
r
vol
tage
.
Tim
e
sc
a
le:
50
ms/d
iv
(b
)Ch1
,
C
h
2
,
C
h
3
scal
e 5
A/d
i
v: sou
rce
cu
rren
t
(
A
),
l
oad
cur
r
ent
,
injec
t
e
d
cur
rent
(
A
)
r
e
spe
c
t
i
ve
l
y
;
Ch4
sca
l
e
15
0
V
/
d
i
v
c
a
pac
i
t
o
r
vo
lta
ge,
Tim
e
s
cale:
2
0 m
s
/d
iv
(c
) Ch
1 sc
ale
30
V/d
i
v
:
s
o
u
r
ce
volta
ge
(
V)
,
C
h2
so
ur
c
e
c
ur
r
e
nt
(
A)
;
C
h3
scale
1
50
V
/
di
v
c
a
pac
i
t
o
r
vo
l
t
age
;
T
i
m
e
scale
:
10 m
s
/di
v
F
i
gur
e
1
3
.
Exp
e
r
i
m
e
nta
l
r
esul
t
s
b
ef
or
e
a
nd
a
f
t
e
r
con
n
ec
t
i
o
n
of
th
e
APF
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N: 2
0
8
8
-
86
94
I
nt
J
P
ow
E
l
e
c
&
Dr
i
S
y
st,
Vol.
10,
N
o.
1
,
Mar
c
h
2
0
1
9
:
27
7
–
287
28
6
F
i
gur
e
1
4
.
H
a
r
m
onic
spe
c
tr
u
m
b
ef
or
e
conn
ec
tin
g
the
AP
F
F
i
g
u
r
e
15.
H
arm
onic
s
p
ec
tr
u
m
a
f
t
e
r
c
onnec
t
i
n
g
th
e
AP
F
6.
CONCLUSION
In
t
hi
s
pa
p
e
r
t
h
e
i
m
pl
ement
a
t
i
o
n
of
a
s
hu
nt
a
c
t
i
v
e
po
w
e
r
f
ilt
e
r
f
or
t
hr
e
e
phase
t
hr
ee
w
ir
e
sys
t
em
a
s
soc
i
a
t
e
d
w
it
h
a
cur
r
en
t
pr
edic
t
i
ve
c
on
tr
o
l
le
r
has
bee
n
c
a
r
r
i
e
d
o
u
t.
T
he
o
ptim
al
v
o
l
ta
ge
v
e
c
t
o
r
a
p
p
l
ied
t
o
t
he
c
o
n
v
erter
is
s
e
l
ec
te
d
a
c
c
o
r
d
ing
to
a
q
ua
li
ty
f
u
n
c
tio
n,
w
hi
ch
m
i
nim
i
z
e
s
t
he
e
r
r
or
b
e
t
w
een
t
he
a
c
t
ue
l
c
u
r
r
e
n
t
a
nd t
h
e pr
e
d
ic
t
e
d o
n
e.
T
he
s
i
m
ula
t
i
o
n
r
e
su
lts show
t
he r
o
b
u
s
t
ne
ss an
d
ef
f
ecti
v
e
n
e
s
s
of
t
he
p
r
o
p
o
se
d ap
pr
oa
c
h
in
t
e
r
m
s
o
f
e
l
i
m
i
n
a
t
ing
har
m
on
ics
a
nd
mi
ni
mizin
g
t
h
e
r
ea
ctive
p
ow
e
r
,
wit
h
a
g
o
od
pow
er
f
actor
.
The
T
H
D
is
s
i
g
n
i
f
i
c
a
n
t
l
y
r
e
d
u
c
e
d
a
n
d
i
s
i
n
c
o
n
f
o
r
m
i
t
y
w
i
t
h
t
h
e
I
E
E
E
s
t
a
n
d
a
r
d
nor
m
s
.
The
exper
i
m
e
nta
l
r
esu
lts
o
bta
i
ne
d
us
in
g
a
n
e
x
p
e
r
i
m
e
n
t
a
l
se
tu
p
a
r
ou
n
d
a
D
spac
e 11
0
4
,
dem
ons
t
r
ate
t
he
p
e
r
f
o
r
m
a
n
c
e
s
obta
i
n
e
d
in
s
imula
tio
n.
REFERE
NC
E
S
[1]
Ch
enn
a
i
S
a
l
i
m
,
B
ench
ou
ia
M
o
h
am
ed
T
o
u
fi
k
,
"
Int
e
lli
gent
C
ontrolle
rs
f
or
S
hu
nt
A
ct
i
v
e
F
ilter
to
C
ompensate
Cu
rrent
H
arm
onics
B
as
ed
o
n
S
R
F
an
d
S
CR
Contro
l
St
rategi
es,
"
In
ter
natio
nal Jo
ur
nal
on
El
ectrica
l E
n
g
i
neer
in
g
an
d In
fo
rma
tic
s
,
Vo
l
u
me 3
,
Numb
er 3 ,
20
1
1
.
[2]
Ch
enn
a
i
S
a
lim
,
B
encho
u
i
a
M
oh
am
ed
T
oufi
k
,
G
oléa
Am
ar,
"H
arm
o
ni
c
Current
C
o
m
p
e
ns
ati
o
n
bas
e
d
o
n
T
hree-
phase
T
h
ree-lev
e
l
Shunt
Active
Filter
using
Fuz
z
y
L
o
gic
C
u
rrent
C
o
n
t
r
o
l
l
e
r
,
"
Jo
ur
nal
o
f
El
ectr
i
c
a
l
Engin
eeri
ng
&
T
ech
nol
og
y
Vol
. 6
, No
. 5
, p
p.
59
5
~
6
0
4
, 2
01
1
.
[3]
A
n
u
p
K
u
m
ar
P
an
da
a
n
d
S
ur
es
h
M
i
k
k
i
l
i
,
"
FLC
based
sh
un
t
acti
v
e
fi
lter
(p–q
and
Id–Iq)
c
o
n
trol
s
trategies
f
o
r
m
i
ti
gati
on
o
f
ha
rm
on
ics
wit
h
d
iff
e
rent
f
u
zzy
M
Fs
u
s
i
n
g
M
ATLAB
a
nd
r
eal-t
i
me
d
i
g
ital
si
mul
a
tor
,
"
In
tern
at
io
nal
Jo
ur
nal
of
Elect
rica
l P
o
wer an
d
E
n
er
gy S
y
st
ems
,
v
o
l. 4
7,
p
p.
31
3
–
3
3
6
,
2
01
3.
[4]
Bach
ar
R
ah
im
a,
G
o
l
éa
A
m
ar,
Ben
c
hou
ia
M
-T,
"Th
e
E
ffici
ency
o
f
Cu
rrent
P
redict
io
n
Con
t
roll
ed
S
hu
n
t
A
cti
v
e
P
o
w
e
r
F
i
l
t
e
r
B
a
s
e
d
o
n
T
h
r
e
e
L
e
v
e
l
I
n
-
v
e
r
t
e
r
i
n
B
a
l
a
n
c
e
d
a
n
d
U
n
b
al
anced
N
e
t
wo
rk
C
ases
,
"
I
n
ter
n
a
t
io
na
l
J
o
u
r
n
a
l
O
f
E
ngin
eerin
g Scien
ce An
d A
ppli
c
a
t
i
on
Rah
i
ma
E
t.A
l
.
, Vo
l
.1,
N
o
.3, 2
01
7
[5]
N
abil
E
l
h
a
j,
M
.
Brah
im
S
ed
ra,
Hi
nd
D
j
e
ghl
ou
d
,
"
D
F
P
I
-b
ased
C
o
n
t
rol
of
t
he
D
C-bu
s
V
o
ltage
a
n
d
t
he
A
C-si
d
e
Cu
rrent
o
f
a
S
hun
t
A
c
ti
ve
P
o
w
er
F
ilt
e
r
,
"
Bul
l
etin of
E
l
ectric
a
l En
gine
e
r
ing
a
n
d
Info
rmat
ic
s
V
ol.
5,
N
o.
4
,
D
ecemb
e
r 20
16
, p
p. 4
30~44
1, DOI:
10.11
59
1
/
eei.v
5i
4
.
5
7
2
[6]
G
h
a
m
r
i
A
.
,
B
e
n
c
h
o
u
i
a
M
.
T
,
G
o
l
e
a
.
A
,
"
S
l
i
d
i
n
g
-
m
o
d
e
C
o
n
t
r
o
l
B
a
s
e
d
T
h
ree-ph
ase
S
h
u
n
t
A
c
tive
P
o
w
e
r
F
i
lt
er:
Simula
t
i
on
a
nd
E
xperi
m
en
tation
,
"
El
ectr
i
c Power
Co
mp
onen
t
s
a
nd Sys
t
em
s Jour
na
l ,
V
o
l
.
4
0,
N
o
.
4
,
pp.
3
83-3
9
8
,
J
a
nu
ary
2
012
.
[7]
T
a
risci
o
tti,
L
.
F
or
ment
i
n
i,
A
..
Ga
et
a,
A
..
Degano,
M..
Z
anchet
ta,
P
.
Rab
b
eni
,
R
..P
uc
ci
,
M.
,
"
M
o
del
P
r
edi
c
tiv
e
Con
t
ro
l
f
o
r
Shunt Ac
t
i
ve F
ilters with F
ixed Switching
Fr
e
quency, "
IE
EE
Tra
n
s
. In
d.
App
l
.
2
0
1
7
,
5
3
,
2
9
6–30
4.
[8]
T
e
sf
aye Al
am
i
r
ew, V
.
Ba
l
aji,
N
ig
us
G
ab
bey
e
,
"Com
p
a
ri
so
n
of
P
ID
Co
ntro
l
l
e
r
w
it
h
Mo
de
l
Pr
edicti
ve
C
ontroller
f
o
r
M
ilk
P
as
teuri
zati
o
n
P
r
o
cess,
"
Bu
ll
eti
n
of
Elect
ri
cal E
ngineer
in
g an
d Inf
o
r
m
a
tics
V
o
l
.
6
,
N
o
.
1
,
p
p
.
2
4
~
3
5
,
M
a
rch
20
17.
[9]
N
a
n
J
i
n
,
S
h
i
y
a
n
g
Hu
,
C
hu
n
Gan
,
an
d
Z
h
i
b
i
n
L
ing
,
"Fin
it
e
Sta
tes
Model
Predi
c
t
i
ve
C
on
trol
f
or
F
ault-Tolerant
O
p
erati
o
n
o
f
a
T
h
r
ee-P
h
ase
Bi
di
recti
onal
AC/DC
Co
nv
erter
U
n
d
e
r
Un
balan
ced
G
rid
Vo
lt
ages,
"
IEEE
T
r
an
sactio
ns On Ind
u
s
t
ria
l
Electron
i
cs
,
Vo
l
.6
5,
N
o
.1
,
Jan
u
ary
20
18
.
[10]
A
t
if
I
qb
al,
S
h
aikh
M
oi
no
ddin,
K
hali
qu
r
Rahman
,
"
Fi
n
i
te
S
tate
P
redi
cti
v
e
Cu
rrent
a
nd
C
o
m
m
on
M
ode
V
o
l
t
a
ge
Co
ntro
l
o
f
a
S
even-p
has
e
V
olt
a
g
e
S
ou
rce
In
verte
r
,
"
In
ter
n
a
t
io
nal
Jou
r
n
a
l
o
f
Power
Elect
ro
nics an
d
Dri
ve S
y
st
em
(
I
JPEDS)
Vol
.
6
,
No
.
3
,
p
p
. 45
9
~
47
6, Sep
t
e
mb
er
2
0
1
5
.
[11]
Ram
o
n
G
u
zm
án
,
L
u
í
s
G
arcía
de
V
i
c
uñ
a,
M
i
g
u
e
l
Cas
till
a
,
J
a
u
m
e
M
i
r
et
a
nd
A
nt
onio
Cam
acho
,
"Finite
C
ontro
l
Set
M
o
del
P
r
ed
ictive
Co
nt
rol
f
o
r
a
Th
ree-P
h
ase
S
h
un
t
A
c
tiv
e
P
ower
F
il
t
e
r
with
a
K
alman
F
ilter-B
ased
E
st
i
m
a
t
ion
,
"
E
n
er
gies
2
017
.
Evaluation Warning : The document was created with Spire.PDF for Python.