Int
ern
at
i
onal
Journ
al of
R
obot
ic
s
and
Autom
ati
on
(I
J
RA
)
Vol
.
7
, No
.
2
,
J
un
e
201
8
, pp
.
119
~
128
IS
S
N:
20
89
-
4856, DO
I: 10
.
11
591/ij
ra
.
v7
i
2
.
pp11
9
-
128
119
Journ
al h
om
e
page
: htt
p:
//
ia
escore
.
c
om/j
ourn
als/i
ndex
.
ph
p/IJRA/i
ndex
Compar
ative
Stu
dy Enter
ed New
Ap
proach
FMV a
nd
Co
nt
r
ol
SFR
f
or Acti
ve Comp
ensation of
Harm
onic Current
s i
n Sh
un
t
Activ
e Power
Fil
ter
Loutfi B
en
yetto
u
La
bora
tor
y
of
Elec
tr
ic
a
l Engi
ne
er
ing,
Univ
ersity
o
f
M’s
il
a, Alge
r
ia
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
hi
story:
Re
cei
ved
Feb
16
, 201
8
Re
vised
Ma
y
17
, 2
01
8
Accepte
d
Ma
y
31
, 201
8
In
thi
s
article,
w
e
discuss
the
pr
oble
m
of
d
egr
ad
at
ion
of
cur
ren
t
in
e
lectr
i
ca
l
insta
llati
ons
,
w
hic
h
fo
ll
ows
di
rec
t
l
y
from
th
e
proli
f
era
t
ion
o
f
non
-
li
n
ea
r
loa
ds,
to
solv
e
it
,
we
used
a
t
wo
-
le
vel
inv
ert
e
r
as
a
pa
rallel
ac
t
ive
f
il
t
er,
which
inject
s
cur
ren
t
h
armonics
at
the
co
nnec
t
ion
point
with
two
compensat
ion
m
et
hods
the
m
et
hod
of
insta
n
ta
neous
activ
e
and
r
eact
iv
e
power
(pq
m
et
h
od
and
pq
m
et
ho
d
with
MV
F)
m
et
hod
b
inds
to
th
e
rep
osi
tor
y
s
y
nchr
oni
zation
.
W
e
will
h
ighl
i
ght
two
con
trol
strat
eg
ie
s
b
y
h
ystere
sis
an
d
PW
M
.
Sim
ula
ti
on
result
s
usi
n
g
Blok
set
Pow
er
S
y
st
em
(PBS
)/
Sim
uli
nk
Matl
ab
show
red
uce
d
THD i
n
acc
orda
nce wit
h
sta
ndar
d
IE
EE
-
519
.
Ke
yw
or
d:
Con
tr
ol
hyste
r
esi
s
Con
tr
ol P
WM
MVF
Paral
le
l act
ive
filt
er
Copyright
©
201
8
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
:
Lo
utfi Benyet
tou
,
Lab
or
at
ory
of
Ele
ct
rical
En
gi
neer
i
ng,
Un
i
ver
sit
y o
f M
’sila
,
BP
-
11
1,
Ri
a
d M
’S
il
a, 28
000, Alge
ria
.
Em
a
il
:
ben
ye
tto
u
.
lo
utfi_
l
ge@y
ahoo
.
fr
1.
INTROD
U
CTION
In
rece
nt
ye
ar
s
,
the
e
xp
a
ns
io
n
of
em
plo
yme
nt
in
the
in
dustry
of
non
-
li
ne
ar
-
base
d
pow
er
el
ect
r
on
i
c
loads
ha
s
le
d
t
o
i
ncr
easi
ng
prob
le
m
s
with
i
nt
erf
ere
nce
or
di
stortion
har
m
on
ic
s
of
powe
r
syst
em
s
.
This
aff
ect
s
a
ll
ind
us
tria
l
s
ect
or
s
(u
si
ng
dim
m
ers,
recti
fi
ers,
in
ve
rters
)
,
te
rtia
ry
(co
m
pu
te
r
or
office
li
gh
ti
ng,
c
ommerci
al
)
and dom
est
ic
(t
el
evi
sion
s
, hom
e con
s
um
er devices
)
.
Har
m
on
ic
dist
or
ti
on
is
gen
e
rated
by
non
-
li
near
loa
ds
c
onnected
to
t
he
net
wor
k
a
nd
a
bsor
bs
no
n
-
sin
usoidal
c
urren
ts
.
T
hes
e
har
m
on
ic
c
ur
ren
ts
will
in
tu
rn
gen
e
rate
ha
r
m
on
ic
volt
age
s
at
dif
fer
e
nt
points
of
t
he
netw
ork
co
nn
ect
io
n
.
F
or
ot
her
el
ect
ri
cal
equ
i
pm
ent
connecte
d
at
t
hese
points,
t
he
ha
rm
on
ic
po
ll
ution
has
a
dverse
ef
fects
.
T
hese
ef
fects
m
ay
be
ci
te
d
the
de
for
m
at
ion
of
th
e
gr
i
d
vo
lt
age
at
the
c
onnecti
on
po
i
nt
wh
e
n
the
e
ne
r
gy
distrib
utor
i
s
require
d
to
pro
vid
e
cl
ea
n
powe
r
.
T
his
po
ll
ution
ca
n
al
s
o
le
ad
to
ove
rh
e
at
ing
of
the
cables
an
d
el
ect
rical
equ
ipm
ent
or
eve
n
stoppe
d
sud
de
nly
ro
ta
ti
ng
m
achines
,
eve
n
the
total
destr
uc
ti
on
of
al
l
equ
ipm
ent
.
Ther
e
f
or
e,
s
up
pliers
of
el
ect
rical
energy
are
therefo
re
obli
ge
d
to
im
po
se
sta
nd
a
r
ds
an
d
protect
them
sel
ves
against
these
dis
tur
ban
ces
.
Sta
nd
a
r
ds
on
ha
r
m
on
ic
s
hav
e
been
pro
posed
by
the
I
nter
na
ti
on
al
Ele
ct
ro
te
ch
nic
al
Com
m
issi
on
IEC6
1000,
re
com
m
end
at
io
ns
an
d
I
EEE
St
d
.
519
-
19
92
[
1,
2,
3]
.
To
c
op
e
with
the
phen
om
eno
n
of
ha
rm
on
ic
disturba
nces,
sever
al
s
olu
ti
on
s
ha
ve
bee
n
pro
po
se
d
.
T
he
se
so
luti
ons
a
r
e
based
on
diode
recti
f
ie
rs
sin
gle
an
d
thr
ee
phase
t
o
s
pecial
st
ru
c
tu
res,
P
WM
r
ect
ifie
rs,
passi
ve
filt
ers
a
nd
act
ive
filt
ers
.
Tra
diti
on
al
m
et
ho
ds
for
re
duci
ng
ha
rm
on
ic
s
invo
lve
the
us
e
of
passive
filt
er
s
trap
ping
harm
on
ic
currents
bas
ed
LC
ci
rc
uits
cal
culat
ed
in
li
ne
with
ha
r
m
on
ic
filt
ering
rows
.
T
hey
can
al
so
be
us
e
d
to
co
m
pen
sat
e
th
e
reacti
ve
power
.
Howe
ver,
the
passive
filt
er
s
om
e
prob
le
m
s:
a
la
ck
of
a
dap
ta
bili
ty
to
var
ia
ti
ons
i
n
t
he
im
ped
ance
of
the
net
work
loa
d
a
nd
po
s
s
ible
res
onance
with
the
im
pe
dan
ce
of
the
ne
twork
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2089
-
4856
IJ
RA
V
ol
.
7
,
No
.
2
,
June
201
8
:
119
–
128
120
and
in
so
m
e
unfavor
a
ble
case
s
w
her
e
the
res
on
a
nce
is
excit
ed,
it
ca
n
a
dri
ving
vo
lt
a
ge
a
nd
a
higher
ha
r
m
on
ic
current
in
sig
ni
ficant
har
m
on
i
c
filt
er
ca
pacit
y
an
d
t
he
netw
ork
.
T
hu
s
,
t
his
so
luti
on
has
a
m
ajo
r
dr
a
wb
ac
k
t
hat
can
be
int
olera
ble
in
t
hese
part
ic
ular
[
1,
3]
ci
rcu
m
sta
nces
.
Anothe
r
s
olu
ti
on
is
to
im
ple
m
ent
an
act
ive
filt
er
in
order
to
a
vo
i
d
the
dr
a
wbacks
of
pa
ssive
filt
ers
.
Ma
ny
s
ol
ution
s
for
act
ive
rem
ediat
ion
of
el
ect
rical
net
works
filt
ers
ha
ve
be
en
pro
posed
in
t
he
li
te
ratu
re
.
T
hose
that
be
st
m
eet
the
act
ive
pa
rall
el
-
s
eries
(als
o
known
a
s
Un
i
fied
Powe
r
Q
ualit
y
Cond
it
ion
er
-
UPQC)
.
I
n
t
he
case
wh
e
re
t
he
s
ource
c
urren
ts
ar
e
non
-
li
nea
r,
pa
rall
el
act
ive
power
f
il
t
er
(Sh
un
t
A
ct
ive
P
ower
F
il
te
r
-
SA
P
F)
is
c
on
si
der
e
d
th
e
best
s
olu
ti
on
to
reduce
ha
r
m
on
ic
currents
in
ap
plica
ti
on
s
of
l
ow
to
m
edium
po
wer
.
T
he
act
ive
filt
er
is
m
or
e
ad
va
ntageous
w
he
re
ra
pi
d
respo
ns
e
is
require
d
in
t
he
presence
of
dy
na
m
ic
loads
.
In
a
dd
it
io
n,
it
repr
esents
a
powe
r
fu
l
to
ol
for
versat
il
e
pack
a
ging
bec
ause
it
is
al
so
a
ble
to
com
pen
s
at
e
reacti
ve
po
wer
a
nd
the
unbalance
d
loa
d
.
The
act
ive
pr
i
nc
iple
of
the
par
al
le
l
filt
er
is
to
generate
har
m
on
ic
currents
in
phase
opposit
ion
to
tho
se
e
xisti
ng
on
the
net
work
.
The
cu
rr
e
nt
ab
so
r
be
d
by
the
po
l
luta
nt
load
s
is
non
-
sin
usoidal
,
wh
il
e
th
e
current
gen
e
rated
by
the
pa
rall
el
act
ive
filt
er
is
su
ch
t
hat
the
cu
r
ren
t
netw
ork
is
sin
usoidal
.
T
o
com
ply
with
sta
ndar
ds
im
po
sed
st
rin
gen
t
qu
al
it
y
el
ect
ric
pro
vid
er
s
an
d
industrial
co
nsum
ers
and
t
o
stem
the
incre
asi
ng
pro
blem
s
of
distu
rbanc
es
on
powe
r
syst
em
s
,
act
ive
filt
ers
m
us
t
adap
t
an
d
m
eet
these
r
equ
i
rem
ents
and
th
us
optim
ize
their
to
polo
gi
es
and
con
t
ro
l
te
ch
niques
.
F
or
this
pur
pose,
se
ver
al
researc
h
c
on
ti
nu
e
t
o
be
publ
ished
on
t
he
pa
rall
el
act
ive
filt
er,
consi
der
i
ng
th
r
ee
m
ai
n
areas
.
The
fi
rst
is
t
o
est
i
m
at
e
the
offset
c
urre
nt,
t
he
sec
ond
is
t
o
eval
uate
al
te
r
native
t
opologies,
an
d
t
he
thir
d
are
a
deals
with
c
on
t
ro
l
strat
egi
es
that
gen
e
rate
the
c
on
t
ro
l
s
ign
al
s
f
or
t
he
powe
r
switc
hes
.
O
ur
w
ork
is
a
com
par
at
ive
st
udy
unde
r
the
sam
e
op
e
rati
ng
co
ndit
ion
s b
et
wee
n
tw
o
co
ntr
ol m
et
ho
ds
pro
po
se
d f
or
t
he
iden
ti
ficat
io
n o
f
re
fer
e
n
ce
c
urren
ts
dis
rupting a
nonlinea
r l
oad
FAP
.
2.
PARALL
EL
STRUCT
URE
OF AN
AC
TIVE FILT
E
R
The
gen
e
ral
s
tructu
re
of
a
three
-
phase
volt
age
ty
pe
FAP
is
pr
ese
nte
d
by
Fig
ur
e
1
wh
e
re
w
e
disti
nguish
th
e
in
ver
te
r
a
nd
t
he
filt
er
outp
ut
powe
r
as
well
as
the
diff
e
re
nt
bl
ocks
of
t
he
par
ty
c
ontr
ol
s
yst
e
m
s
par
t
.
T
he
po
wer
sect
io
n
usual
ly
co
ns
ist
s
of
a
volt
age
in
ver
te
r
base
d
powe
r
swit
ches
w
hose
sta
te
s
of
the
in
ver
te
r
s
witc
hes
a
re
c
ontr
olled
by
t
he
hyste
resis
c
on
t
ro
ll
er
or
P
WM
,
a
ci
rcu
it
e
nergy
st
or
a
ge
(often
capaci
ti
ve
)
an
d
a
n
outp
ut
fi
lt
er
of
th
e
fir
st
orde
r
t
o
m
i
ti
gate
the
c
om
po
nen
ts
due
to
t
he
s
witc
hi
ng
of
the
inv
e
rter
a
nd
has
c
onne
ct
the
cro
ss
ov
er
to
the
pow
er
gr
id
.
[
4
,
5
]
The
co
ntr
ol
-
com
m
and
port
ion
is
com
po
sed
of
a
first
ste
p
w
ho
s
e
ro
le
is
to
ide
ntify
the
ha
rm
on
ic
fil
te
r
whose
qual
it
y
la
rg
el
y
de
pends
o
n
the
ide
ntific
at
ion
of
the
har
m
on
ic
s
re
fer
e
nce
m
et
ho
d
,
an
d
a
sec
ond
ste
p
th
at
pe
rfor
m
s
th
e
re
gula
ti
ng
th
e
DC
v
oltage
.
A
t
hir
d
a
nd
final
ste
p
ge
ner
at
es
t
he
in
ver
te
r
c
on
tr
ol
.
T
his
stu
dy
is
lim
it
ed
to
ca
ses
w
he
re
t
he
so
urc
e
vo
lt
age
is
sin
usoidal
,
a
nd
where
t
he
c
urre
nt
dra
wn
by
the
load
is
t
ai
nte
d
with
ha
rm
on
ic
com
ponen
ts
.
Unde
r
these
c
onditi
ons,
the
total
ha
rm
on
ic
disto
r
ti
on
T
H
D
is
well
su
it
e
d
to
desc
ribe
the
degree
of
harm
on
ic
po
ll
utio
n o
n
t
he
grid
.
Tables
and Fig
ures a
r
e
presente
d
ce
nter, as s
how
n belo
w
a
nd cit
ed
in
the m
anu
s
cript
.
Figure
1
.
Ge
ne
ral Str
uc
tu
re
of a F
AP
volt
age
stru
ct
ur
e
3.
CA
L
CU
L
ATI
ON OF
C
URRENT
HAR
MO
NICS
The
c
urre
nt
c
on
s
um
ption
of
the
poll
utant
load
is
c
om
po
se
d
of
a
f
un
dam
ental
and
har
m
on
ic
s
.
The
act
i
ve
f
il
te
r
is
us
ed
to ge
ne
rate t
he harm
on
ic
c
urre
nts of th
e
sam
e
m
ag
nitud
e
but in
phase
opposit
io
n wit
h
tho
se
e
xisti
ng
in
the
feed
.
T
o
do
this,
we
m
us
t
ide
ntify
th
e
har
m
on
ic
c
urre
nts
of
the
lo
ad
.
S
eve
ral
m
e
thods
exist
f
or
ide
nti
fyi
ng
[
1, 2,
3
]
: The f
irst
is
bas
ed
on
t
he
s
pect
ral
a
naly
sis
of p
olluta
nt
strea
m
.
The
sec
ond uses
a
band
pass
filt
er
f
or
filt
erin
g
t
he
f
undam
ental
.
The
thir
d
us
es
the
co
ncep
t
s
of
in
sta
ntane
ou
s
real
powe
r
and
i
m
aginati
on
.
T
he
la
tt
er
is
m
os
t
com
m
on
ly
us
e
d
in
m
os
t
cro
ss
overs
beca
us
e
it
a
chie
ves
the
best
a
gree
m
ent
betwee
n
the
st
at
ic
an
d dyn
am
ic
p
er
form
ance
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
RA
IS
S
N:
20
89
-
4856
Com
par
at
ive S
tud
y E
ntere
d N
ew Ap
proac
h
FMV
And
C
on
trol S
FR F
or Ac
ti
ve
…
(
L
ou
tf
i Benyet
to
u
)
121
3
.
1
.
Me
thod
of instan
t
an
e
ous ac
tive an
d re
act
i
ve
p
ower
The
m
et
ho
d
of
insta
ntane
ou
s
act
ive
a
nd
re
act
ive
po
wer
(
com
m
on
ly
denoted
pq
m
et
hod)
oper
at
es
the
Con
c
ordia
trans
form
ation
of
vo
lt
ages
a
nd
li
ne
c
urre
nts
[
6
,
7
]
to
cal
culat
e
the
instanta
ne
ous
act
ive
an
d
reacti
ve
powe
rs
.
The
f
unda
m
ental
com
po
nen
t
is
c
onve
rted
i
nto
a
D
C
com
po
ne
nt
an
d
t
he
ha
r
m
on
ic
com
po
ne
nts
in
the
AC
com
pone
nts
.
By
r
e
m
ov
in
g
the
DC
com
ponent
of
t
he
insta
nt
aneous
act
ive
p
ow
e
r
(corr
es
pondin
g
to
the
fun
da
m
ental
co
m
ponen
t
of
the
c
urre
nt
of
t
he
lo
ad)
us
in
g
a
sim
ple
low
-
pass
filt
er
(F
PB)
,
the
ha
r
m
on
ic
com
po
ne
nts
can
be
i
de
ntifie
d
.
The
pri
nciple
of
t
his
c
onve
ntion
al
m
et
hod
is
now
bri
efly
descr
i
bed
.
Ar
e
res
pecti
vely
the
vo
lt
age
s
f
r
om
the
po
i
nt
of
co
nnect
(
Pcc
),
fr
ee
do
m
fro
m
zero
(con
ne
ct
ed
to
a
po
ll
utant
loa
d)
a
nd
the
t
hree
load
c
urre
nt
s,
de
no
te
d
3
2
1
,
,
s
s
s
v
v
v
and
3
2
1
,
,
c
c
c
i
i
i
.
The
t
ran
s
f
orm
at
ion
can
reduce t
his Co
ncor
dia a
balan
ced tw
o
-
phase
syst
e
m
in
wh
ic
h
the
a
xe
s ar
e
qu
a
drat
ic
phas
e syst
e
m
(1
-
2)
:
−
−
−
=
3
2
1
.
2
3
2
3
0
2
1
2
1
1
3
2
s
s
s
s
s
v
v
v
v
v
(1)
−
−
−
=
3
2
1
.
2
3
2
3
0
2
1
2
1
1
3
2
c
c
c
c
c
i
i
i
i
i
(2)
Neg
le
ct
in
g
t
he vo
lt
age
h
a
rm
on
ic
s,
t
he
act
ual
pow
e
r
c
p
an
d
im
agina
ry po
wer
c
q
are e
xpresse
d by
(
3)
:
−
=
c
c
s
s
s
s
c
c
i
i
v
v
v
v
q
p
.
(3)
Com
po
ne
nts
of
real
powe
r
a
nd
insta
ntane
ou
s
i
m
aginar
y
e
xpress
ed
a
s
the
su
m
of
a
DC
c
om
po
ne
nt
an
d
an
AC
com
po
ne
nt
(
4)
:
+
+
=
c
c
c
c
c
c
q
q
p
p
q
p
~
~
(4)
with:
-
c
p
: The c
on
ti
nuou
s
po
wer ass
oc
ia
te
d
to t
he
f
undam
ental
acti
ve
c
om
po
ne
nt
of the c
urre
nt a
nd volt
age;
-
c
q
: The c
on
ti
nuou
s
po
wer ass
oc
ia
te
d
to t
he
f
undam
ental
co
m
pone
nt of t
he
r
eact
ive curre
nt
and
vo
lt
age;
-
c
p
~
et
c
q
~
:
Alte
r
nativ
e
po
wer
c
orres
pondin
g
t
o
the
s
um
o
f
the
in
te
rf
ere
nce
com
pone
nts
of
t
he
cu
rr
e
nt
an
d
vo
lt
age
.
To
isolat
e
the
conve
ntio
nal
act
ive
an
d
rea
ct
ive
powe
r,
it
is
necessa
ry
to
know
acc
urat
el
y
the
fr
eq
uenc
y
pu
lsa
ti
ons
in
sta
ntane
ous
po
wer
form
ed
from
equ
at
ion
(
4)
.
It
sho
uld
be
note
d
t
hat
now
c
on
si
de
rs
that
the
st
ud
ie
d
sy
stem
is
com
po
se
d
of
th
ree
son
that
pre
ven
ts
zer
o
se
qu
e
nce
com
ponen
ts
are
ci
rc
ulati
ng
.
Af
te
r
ide
ntifyi
ng
the
pulse
i
ns
ta
nta
neous
powe
r,
the
pow
er
filt
er
c
harg
ed
is
olate
c
onven
ti
onal
act
iv
e
an
d
reacti
ve
po
we
r
can
be
siz
e
d
.
A
ci
rcu
it
c
om
pr
isi
ng
a
l
ow
pass
filt
er
with
a
subtr
act
or
can
b
e
us
ed
,
as prese
nted
i
n
F
ig
ur
e
2
.
F
igure
2
.
Dia
gra
m
sh
owin
g
t
he
principle
of s
epar
at
io
n of p
owers
The
orde
r
of
th
is
low
-
pa
ss
filt
er
def
i
nes
t
he
dynam
ic
s
and
eff
ect
ive
ness
of
the
ide
ntific
at
ion
m
et
ho
d
.
In
this
pa
pe
r,
a
lo
w
-
pass
sec
ond
order
filt
er
is
sel
ect
ed
f
or
the
e
xtracti
on
of
ha
rm
on
ic
s
.
The
cut
off
f
re
qu
e
ncy
fc=25
Hz
is
c
ho
s
en
s
o
that
the
filt
er
ca
n
blo
c
k
any
disturbin
g
co
m
po
nen
t
of
i
ns
ta
nta
neous
powe
r
.
Fr
om
equati
on
(3),
we
ca
n fin
d
c
urren
t c
om
p
on
e
nt
(5)
:
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2089
-
4856
IJ
RA
V
ol
.
7
,
No
.
2
,
June
201
8
:
119
–
128
122
−
+
=
−
=
−
c
c
s
s
s
s
s
s
c
c
s
s
s
s
c
c
q
p
v
v
v
v
v
v
q
p
v
v
v
v
i
i
.
1
.
2
2
1
(
5)
By
introdu
ci
ng (4)
i
n
(5),
t
he c
urren
ts
in
t
he a
xes
)
,
(
beco
m
es
(6)
:
−
+
+
−
+
=
c
c
s
s
s
s
s
s
c
c
s
s
s
s
s
s
c
c
q
p
v
v
v
v
v
v
q
p
v
v
v
v
v
v
i
i
~
~
.
1
.
1
2
2
2
2
)
(6)
Dep
e
ndin
g
on
the
f
unct
ion
t
hat
we
wa
nt
t
o
giv
e
t
o
t
he
FA
P
,
we
can
si
m
ultaneou
sly
com
pen
sat
e
the
c
urre
nt
ha
rm
on
ic
s
an
d
react
ive
power
or
only
one
.
Ta
ble
1
s
umm
arizes
the
possi
ble
m
od
es
of
com
pen
sat
i
on
[
8
]
,
[
9
]
.
If
we
wa
nt
f
or
exam
ple
to
c
om
pen
sat
e
curr
ent
har
m
on
ic
s
an
d
reacti
ve
powe
r
si
m
ultaneou
sly
,
then
we
el
im
inate
the
DC
com
po
ne
nt
of
c
p
us
in
g
a
sim
ple
LPF
.
I
n
this
case
an
d
after
add
i
ng
to
t
he
AC
com
ponent
of
t
he
insta
nt
aneous
act
i
ve
powe
r,
act
ive
powe
r
dc
p
fo
r
re
gu
la
ti
on
of
the
D
C
vo
lt
age
dc
v
.
Table
1
.
Me
t
hods
o
f
Co
ntr
olli
ng
t
he
C
om
pensat
ion
I
nst
anta
neous
Powe
r
Ty
pe
o
f
co
mpe
ns
ation
Con
tro
l
param
ete
r
Com
pensa
ti
on
o
f
cur
r
ent
h
armonics
dc
c
r
e
f
f
p
p
p
+
=
c
r
e
f
f
q
q
~
=
Com
pensa
ti
on
o
f
reactive ene
rg
y
dc
r
e
f
f
p
p
+
=
0
c
r
e
f
f
q
q
=
Com
pensa
ti
on
o
f
cur
r
ent
h
armonics
and
re
ac
t
ive
power
dc
c
r
e
f
f
p
p
p
+
=
~
c
r
e
f
f
q
q
=
The refe
rence
currents
, den
ot
ed
r
e
f
f
i
et
r
e
f
f
i
, ar
e e
xpre
ssed
al
ong
t
he a
xes
)
,
(
by:
Finall
y, it
is easy
to ob
ta
in
the
r
efe
re
nce c
ur
r
ents acc
ordin
g t
o
the a
xes
ab
c
by the
tra
ns
f
orm
at
ion
(
7)
:
−
−
−
=
r
e
f
f
r
e
f
f
r
e
f
r
e
f
r
e
f
i
i
i
i
i
f
f
f
.
2
3
2
1
2
3
2
1
0
1
3
2
3
2
1
(7)
3
.
2
.
Pri
ncip
le
of multi
-
vari
abl
e filter
MVF
Mr
.
Be
nh
a
bib
sug
gested
in
his
thesis
[
1]
a
ne
w
filt
er
cal
le
d
M
VF
extracti
on,
f
or
ext
racti
ng
the
f
undam
ental
com
po
ne
nt
of
el
ect
rica
l
si
gn
al
s
(
vo
lt
age
or
cu
rr
e
nt)
di
rectl
y
al
on
g
t
he
a
xes
(α
-
β)
.
The
trans
fer
f
un
ct
io
n of t
his
filt
er is
(
8)
:
H
(
s
)
=
v
̂
αβ
(
s
)
v
αβ
(
s
)
=
K
(
s
+
K
)
+
j
ω
c
(
s
+
K
)
2
+
ω
c
2
(8)
(
ω
c
)
re
pr
ese
nt
s
t
he
f
undam
ental
pu
lsa
ti
on
(f
c=
50
Hz
),
(
K)
a
posit
ive
c
o
nst
ant,
(v)elec
tric
al
input
vo
lt
ag
e
MVF
an
d
(
v
̂
)
is
the
volt
age
c
orr
esp
onding
to
(
v)
outp
ut
MV
F
.
T
he
f
undam
ental
com
po
ne
nt
of
the
el
ect
r
ic
al
sign
al
(
vo
lt
age
or
c
urre
nt)
ca
n
be
extr
act
ed
dir
ect
ly
by
t
he
axe
s
(α
-
β
),
without
phase
change
or
am
plit
ud
e
.
Fr
om
eq
uation
(
8),
we
obta
in
the
f
ollow
i
ng
two
ex
pr
e
ssio
ns
as
s
how
n
in
(
9
-
10)
.
MV
F
use
d
at
t
he
t
wo
-
phase
vo
lt
age
s,
ca
n
eff
ect
ively
filt
er
the
ha
r
m
on
ic
co
m
ponen
ts
of
t
he
power
sup
pl
y
vo
lt
ages
.
Th
us
,
it
s
i
m
ple
m
enta
ti
on
ca
n
im
pr
ov
e
the
perfor
m
ance
of
the
filt
er
.
Fi
gures
are
pr
e
sente
d
center,
as
s
hown
in
Figure
3
a
nd ci
te
d
in t
he
m
anu
scri
pt
.
̂
(
)
=
(
s
+
K
)
K
(
s
+
K
)
2
+
ω
c
2
(
)
−
k
ω
c
(
s
+
K
)
2
+
ω
c
2
(
)
(9)
̂
(
)
=
(
s
+
K
)
K
(
s
+
K
)
2
+
ω
c
2
(
)
+
k
ω
c
(
s
+
K
)
2
+
ω
c
2
(
)
(10)
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
RA
IS
S
N:
20
89
-
4856
Com
par
at
ive S
tud
y E
ntere
d N
ew Ap
proac
h
FMV
And
C
on
trol S
FR F
or Ac
ti
ve
…
(
L
ou
tf
i Benyet
to
u
)
123
Figure
3
.
Bl
oc
k diag
ram
MVF
3
.
3
.
Me
thod
c
alled t
he refe
r
ence rel
ated
t
o
s
yn
chr
on
iz
ati
on
This
m
et
hod,
intr
oduce
d
by
[
8
]
,
[
9
]
,
t
he
pr
ocessin
g
al
so
op
e
rates
C
onc
ordia
but
a
pp
l
ie
d
on
ly
to
the
li
ne
cu
rr
e
nt
s
of
t
he
l
oad
2
1
,
c
c
i
i
an
d
3
c
i
.
The
n,
a
seco
nd p
r
ocessin
g
is
pe
rfo
rm
ed
to sw
it
ch
t
he
li
ne
cu
r
ren
ts
in
dq,
wh
ic
h
t
ran
s
f
or
m
s
the
fun
dam
ental
c
om
po
ne
nt
of
c
urren
t
in
a
D
C
com
po
ne
nt
and
ha
rm
on
ic
current
com
po
ne
nts
in
the
AC
com
pone
nts
.
T
his
al
lows
us
to
e
lim
inate
us
in
g
a
si
m
ple
low
pass
filt
er,
t
he
DC
com
po
ne
nt
of
t
he
c
urren
t
.
T
he
m
ajo
r
a
dv
a
nt
age
of
this
m
eth
od
c
om
par
ed
to
pre
vious
li
es
in
t
he
fact
t
ha
t
any
te
ns
io
ns
har
m
on
ic
s
ha
ve
no
m
or
e
influ
e
nce
on
t
he
i
de
ntifie
d
c
urren
t
an
d
t
her
e
fore
t
he
filt
er
will
b
e
bette
r
.
Its
pr
i
nciple
i
s
set
ou
t
bel
ow
.
A
re
the
l
ine
curre
nts
of
a
three
-
phas
e
syst
e
m
witho
ut
ze
ro
se
qu
ence
.
The
tran
sf
or
m
at
ion
ca
n
re
du
ce
this
C
onco
r
dia
ph
a
se
syst
e
m
to
a
tw
o
-
phase
e
qu
il
ibri
um
syst
e
m
,
as
a
bove
by
the foll
owin
g
r
el
at
ion
sh
i
p
(11
)
:
−
−
−
=
3
2
1
.
2
3
2
3
0
2
1
2
1
1
3
2
c
c
c
i
i
i
i
i
c
c
(11)
By
gen
erati
ng
sign
al
s
with
a
P
.
L
.
L
)
c
os
(
an
d
)
s
i
n
(
der
i
ved
from
t
he
f
undam
ental
te
ns
ion
netw
ork,
we ob
ta
in
the
e
xpressi
on of th
e cu
rr
e
nts in
th
e dq syste
m
(12)
[
9]
:
−
=
−
c
c
cq
cd
i
i
i
i
.
)
s
i
n
(
)
c
os
(
)
c
os
(
)
s
i
n
(
1
(12)
These c
om
po
ne
nts m
ay
then
be
e
xpresse
d
a
s the sum
o
f
a
DC c
om
po
ne
nt and a
n AC c
om
po
nen
t
(13)
:
+
+
=
cd
cd
cd
cd
cq
cd
i
i
i
i
i
i
~
~
(13)
with:
cd
i
and
cq
i
: T
he DC c
om
po
ne
nts
of
cd
i
and
cq
i
,
cd
i
~
and
cq
i
~
: ac com
ponen
ts
of
cd
i
an
d
cq
i
Eq
uation (
12),
we
ca
n deduc
e
the co
m
pone
nts of c
urren
t α
-
β
(
14)
:
−
=
−
=
−
cq
cd
cq
cd
c
c
i
i
i
i
i
i
.
)
s
i
n
(
)
c
os
(
)
c
os
(
)
s
i
n
(
.
)
s
i
n
(
)
c
os
(
)
c
os
(
)
s
i
n
(
1
(14)
So
int
(
15)
:
−
+
−
=
cq
cd
cq
cd
c
c
i
i
i
i
i
i
~
~
.
)
s
i
n
(
)
c
os
(
)
c
os
(
)
s
i
n
(
.
)
s
i
n
(
)
c
os
(
)
c
os
(
)
s
i
n
(
(15)
Nex
t
we
giv
e
the
f
unct
io
n
a
ct
ive
po
wer
fil
te
r
[
8
]
,
we
ca
n
al
s
o
c
om
pensat
e
with
this
com
m
and
is
the
c
urren
t
ha
rm
on
ic
s
an
d
r
eact
ive
ene
r
gy
or
o
nly
one
of
the
tw
o
.
T
able
2
s
umm
a
rizes
the
m
od
es
of
com
pen
sat
ion
po
s
sible
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2089
-
4856
IJ
RA
V
ol
.
7
,
No
.
2
,
June
201
8
:
119
–
128
124
Table
2
.
Me
t
hods
C
om
pen
sat
ion C
ontr
ol Re
po
sit
or
y B
in
ds
t
o
Sync
hro
nousl
y
Ty
pe
o
f
co
mpe
ns
ation
Con
tro
l
param
ete
rs
Com
pensa
ti
on
o
f
cur
r
ent
h
armonics
dc
cd
r
e
f
fd
i
i
i
+
=
~
cq
r
e
f
fq
i
i
~
=
Com
pensa
ti
on
o
f
reactive ene
rg
y
dc
r
e
f
fd
i
i
+
=
0
cq
r
e
f
fq
i
i
=
Com
pensa
ti
on
o
f
cur
r
ent
h
armonics
and
re
ac
t
ive
power
dc
cd
r
e
f
fd
i
i
i
+
=
~
cq
r
e
f
fq
i
i
=
The refe
rence
currents
are
e
xpress
ed
b
y
(16
)
:
−
=
r
e
f
fq
r
e
f
fd
r
e
f
f
r
e
f
f
i
i
i
i
.
)
s
i
n
(
)
c
os
(
)
c
os
(
)
s
i
n
(
(16)
Af
te
r
the
c
ho
i
ce
of
co
ntr
ol
par
am
et
ers,
th
e
in
ver
se
tra
nsfo
rm
at
ion
ca
n
be
trace
d
ba
ck
t
o
t
he
C
onco
r
dia
ref
e
ren
ce
curre
nts
(
17)
:
−
−
−
=
r
e
f
f
r
e
f
f
r
e
f
r
e
f
r
e
f
i
i
i
i
i
f
f
f
.
2
3
2
1
2
3
2
1
0
1
3
2
3
2
1
(17)
4.
CUR
RENT
C
ONTROL
B
Y
HYSTE
RESI
S
The
m
et
ho
d
is
base
d
on
t
he
com
par
iso
n
of
the
d
i
ff
e
ren
c
e
Ɛ
betw
een
the
c
urre
nt
re
f
eren
ce
an
d
the
m
easur
e
d
c
urren
ts
with
a
fixe
d
ba
nd
.
Ea
ch
vi
olati
on
of
this
band
gi
ve
s
a
s
witc
hing
c
omm
and
switc
hes
to
4
.
Fig
ur
es
are
pr
ese
nted
cent
er,
as
s
how
n
in
Fig
ur
e
4
an
d
c
it
ed
in th
e m
anu
scri
pt
.
Figure
4
.
Pr
i
n
c
iple o
f
c
urren
t
con
t
ro
l
by
hyste
resis
5
.
CUR
RENT
C
ONTROL
B
Y
PW
M
The
P
W
M
co
nt
ro
l
pr
i
nciple
i
s
des
cribe
d
in
Figure
5
.
I
n
thi
s
case,
the
dif
f
eren
ce
bet
wee
n
t
he
c
urren
t
ref
e
ren
ce
i
f
a
nd
the
act
ua
l
curre
nt
ifre
f
is
ap
plied
to
the
in
pu
t
of
a
co
ntr
oller
.
The
o
utput
s
ign
al
of
the
c
on
tr
oller,
cal
le
d
m
od
ul
at
or
,
is
the
n
c
om
par
ed
with
a
tria
ngula
r
w
ave
of
fi
xed
f
reque
ncy
(ca
rri
er)
t
o
determ
ine
the
order
of
swit
chin
g
of
the
s
witc
hes
.
T
he
f
reque
ncy
of
t
he
tria
ng
ular
carrier
t
her
e
fore
set
s
the s
witc
hing freq
ue
ncy of
th
e pow
e
r
sem
ico
nd
ucto
rs
.
Figure
5
.
Pr
i
nc
iple o
f
c
omm
a
nd curre
nt
by P
WM
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
RA
IS
S
N:
20
89
-
4856
Com
par
at
ive S
tud
y E
ntere
d N
ew Ap
proac
h
FMV
And
C
on
trol S
FR F
or Ac
ti
ve
…
(
L
ou
tf
i Benyet
to
u
)
125
6
.
SIMULATI
O
N
O
F
RESU
L
TS
The
sim
ulati
on
of
the
direct
c
on
t
ro
l
in
the
phase
m
ark
er
ba
sed
on
the
m
eth
od
binds
t
o
th
e
reposit
ory
synch
ronizat
io
n or t
he
t
heory
of instanta
ne
ous pow
e
r
(pq
) w
as co
nducte
d wit
h
the
foll
ow
ing
pa
ram
et
ers
:
-
T
he
loa
d
c
urr
ent is inc
rease
d
50% to
t =
0,3s
-
T
he value
of
the D
C
volt
age
is eq
ual to
60
0V
-
T
he hyst
eresi
s b
a
nd att
ache
d
to
3A,
for co
ntr
ol b
y
hyste
r
esi
s
-
T
he
s
witc
hing
fr
e
qu
e
ncy is
set
at 1
3
.
5 KHz
.
f
or P
WM c
ontr
ol
.
-
D
r
op f
il
te
r
lo
w pass sec
on
d order
cu
t
off fre
qu
e
ncy e
qu
al
t
o 25Hz
.
-
T
he param
et
e
r
K
of M
VF
w
as ch
os
e
n
e
qu
a
l t
o
80
.
Table
3
incl
udes all
n
et
w
ork set
ti
ng
s a
nd po
ll
utant loa
ds
use
d
in
the
sim
ul
at
ion
.
Table
3
.
Sim
ul
at
ion
Sett
in
gs
Par
ame
t
er
s
nu
me
ri
cal val
u
es
Volta
ge
eff
ec
t
iv
e
E
s
[v]
240
Freque
nc
y
f
50
Li
ne
resista
n
ce
R
s
[m
]
1
.
59
Li
ne
choke
L
s
[
H]
45
.
56
appa
ren
t
power
S KV
A
400
Loa
d
r
esista
n
ce
DC R
c
[m
]
2
.
73
Loa
d
induc
t
ance
DC L
c
[
H]
23
.
19
Filt
er
resista
n
ce
R
f
[m
]
6
.
87
Filt
er
induct
ance
L
f
[
H]
100
Volta
ge
Vdc
[v]
600
7
.
1
.
Res
ults
of simul
at
io
n
of the
ov
er
all n
etwork
lo
ad
b
efo
re
t
he
n
on
l
inear
filter
The
pre
vious
fig
ur
e
s
hows
the
l
oad
c
urre
nt
(ic1
)
t
he
fi
rst
phase
an
d
the
har
m
on
ic
s
pectr
um
.
The
c
urre
nt
T
HD
f
or
this
l
oa
d
is
23,
72
%
.
The
T
H
D
is
c
al
culat
ed
f
or
t
he
first
f
ort
y
ha
rm
on
ic
s
acco
rd
i
ng
to
"IEC"
Fig
ures
are
pr
ese
nted
c
enter,
as s
how
n
in
Fig
ure
6,
7
an
d
ci
te
d
i
n
t
he
m
anu
script
.
Figure
6
.
Loa
d cu
rr
e
nt (p
hase
1)
Figure
7
.
H
a
rm
on
ic
sp
ect
r
um
o
f
the l
oad cu
rrent (
Ph
ase
1)
T
HD
=
23
.
72%
7
.
2
.
Direct
c
on
t
rol me
thod
ba
se
d
on
th
e
inst
an
t
an
e
ous
po
w
er
(PQ
)
7
.
2
.
1
.
Resul
ts
of
simul
ati
on
using
th
e
h
ys
t
eresi
s
contr
ol
Figures
8
Th
e
sourc
e
c
urrent
har
m
on
ic
s
pe
ct
ru
m
after
c
om
pen
sat
ion
(
ph
a
se
1)
TH
D
=
6
.
81%
a
re
pr
ese
nted
cent
er,
as
s
how
n
be
low
a
nd cit
ed
in the m
anu
sc
r
ipt
.
0
0
.
1
0
.
2
0
.
3
0
.
4
0
.
5
0
.
6
-
1
5
0
0
-
1
0
0
0
-
5
0
0
0
500
1000
1500
T
e
m
p
s(
s)
i
c1
(
A
)
5
10
15
20
25
30
35
40
0
10
20
30
40
50
60
70
80
90
100
H
a
r
m
o
n
i
c
o
r
d
e
r
Fu
n
d
a
m
e
n
t
a
l
(
5
0
H
z
)
=
7
8
2
.
1
,
T
H
D
=
2
3
.
7
2
%
A
m
p
l
i
t
u
d
e
(
%
)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2089
-
4856
IJ
RA
V
ol
.
7
,
No
.
2
,
June
201
8
:
119
–
128
126
Figure
8
.
The
s
ource c
urre
nt harm
on
ic
sp
ect
r
um
after co
m
pen
sat
io
n
(
phas
e 1)
T
HD
=
6
.
81%
7
.
2
.
2
.
Simul
ati
on
re
sults
usi
ng
th
e
com
ma
nd
h
ystere
sis
w
ith MVF
Figure
9
is
c
urre
nts
of
non
-
li
near
loa
d
ic
1,
is1
of
s
ource
and
th
at
of
if
1
cro
s
sove
r
an
d
Fig
ur
e
1
0
.
The
s
ource
c
urre
nt
ha
rm
on
ic
sp
ect
r
um
a
fter
com
pen
sat
io
n
(
phase
1)
T
HD
=
1
.
68%
a
r
e
pr
e
sente
d
ce
nter,
as
sh
ow
n belo
w
a
nd cit
ed
in
the
m
anu
scri
pt
.
Figure
9
.
Cu
rr
e
nts
of
non
-
li
ne
ar lo
a
d
ic
1,
is
1 of so
urce a
nd that of if
1
c
r
os
s
ov
e
r
Figure
1
0
.
T
he
sour
ce
curre
nt
h
a
rm
on
ic
sp
ec
trum
a
fter co
m
pensat
ion (
ph
a
se 1) T
HD=
1
.
68%
7
.
2
.
3
.
Simul
ati
on
re
sults
usi
ng
th
e
PW
M
c
omm
an
d
In
Fi
gures
1
1
Currents
of
no
n
-
li
nea
r
loa
d
ic
1,
is
1
of
sou
rc
e
and
t
hat
of
if
1
cr
os
s
over
a
nd
Fig
ure
1
2
.
Th
e
source
c
urre
nt
har
m
on
ic
sp
ect
ru
m
a
fter
c
om
pen
sat
ion
(
phase
1)
THD
=
1
.
73%
a
re
pr
ese
nted
c
enter,
as sho
wn b
el
ow a
nd cit
ed
i
n t
he
m
anu
scri
pt
.
Figure
1
1
.
C
urren
ts
of
non
-
li
near l
oa
d
ic
1, i
s1
of sou
rce a
nd that
of if
1
c
r
os
s
ov
e
r
Figure
1
2
.
T
he
sour
ce
cu
rr
e
nt
h
a
rm
on
ic
sp
ec
trum
a
fter co
m
pensat
ion (
ph
a
se 1) T
HD=
1
.
73%
5
10
15
20
25
30
35
40
0
5
10
15
20
25
30
35
40
45
50
O
r
d
e
r
d
e
s
H
a
r
m
o
n
i
q
u
e
s
T
H
D
=
6
.
8
1
%
A
m
p
l
i
t
u
d
e
(
%
)
0
.
2
6
0
.
2
8
0
.
3
0
.
3
2
0
.
3
4
0
.
3
6
0
.
3
8
-
1
5
0
0
-
1
0
0
0
-
5
0
0
0
500
1000
1500
T
e
m
p
s(
s)
i
(
A
)
5
10
15
20
25
30
35
40
0
10
20
30
40
50
60
70
80
90
100
O
r
d
r
e
d
e
s
H
a
r
m
o
n
i
q
u
e
s
T
H
D
=
1
.
6
8
%
A
m
p
l
i
t
u
d
e
(
%
)
0
.
3
0
.
3
5
0
.
4
-
1
5
0
0
-
1
0
0
0
-
5
0
0
0
500
1000
1500
T
e
m
p
s
(
s
)
i
(
A
)
5
10
15
20
25
30
35
40
0
10
20
30
40
50
60
70
80
90
100
O
r
d
r
e
d
e
s
H
a
r
m
o
n
i
q
u
e
s
T
H
D
=
1
.
7
3
%
A
m
p
l
i
t
u
d
e
(
%
)
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
RA
IS
S
N:
20
89
-
4856
Com
par
at
ive S
tud
y E
ntere
d N
ew Ap
proac
h
FMV
And
C
on
trol S
FR F
or Ac
ti
ve
…
(
L
ou
tf
i Benyet
to
u
)
127
7
.
3
.
Direct
c
on
t
rol me
thod
ba
se
d
on
th
e
sy
nchr
onou
s r
eference
fr
am
e SF
R
7
.
3
.
1
.
Resul
ts of si
mula
tion
using the
h
ys
t
eresi
s co
n
tr
ol
Figure
13
c
urr
ents
of
no
n
-
li
ne
ar
loa
d
ic
1,
i
s1
of
s
ource
a
nd
t
hat
of
i
f1
cro
ss
over
a
nd
Figure
1
4
.
The
source
c
urre
nt
har
m
on
ic
sp
ect
ru
m
after
c
om
pen
sat
ion
(
phase
1)
THD
=
1
.
23%
a
re
pr
ese
nted
c
enter,
as sho
wn b
el
ow a
nd cit
ed
i
n t
he
m
anu
scri
pt
.
Figure
1
3
.
C
urren
ts
of
non
-
li
near l
oa
d
ic
1, i
s1
of sou
rce a
nd that
of if
1
c
r
os
s
ov
e
r
Figure
1
4
.
T
he
sour
ce
curre
nt
h
a
rm
on
ic
sp
ec
trum
a
fter co
m
pensat
ion (
ph
a
se 1) T
HD=
1
.
23%
7
.
3
.
2
.
Simul
ati
on
re
sults
usi
ng
th
e
PW
M
c
omm
an
d
Figure
1
5
.
Cu
r
ren
ts
of
non
-
li
near
loa
d
ic
1,
is1
of
s
ource
and
t
hat
of
i
f1
cr
os
s
ov
e
r
an
d
Fig
ur
e
1
6
.
The
source
c
urre
nt
har
m
on
ic
sp
ect
ru
m
a
fter
c
om
pen
sat
ion
(
phase
1)
THD
=
1
.
36%
a
re
pr
ese
nted
c
enter,
as sho
wn b
el
ow a
nd cit
ed
i
n t
he
m
anu
scri
pt
.
Figure
1
5
.
C
urren
ts
of
non
-
li
near l
oa
d
ic
1, i
s1
of sou
rce a
nd T
hat
of
i
f1 cr
os
s
ov
e
r
Figure
1
6
.
T
he
source
curre
nt
h
a
rm
on
ic
sp
ec
trum
aft
er co
m
pensat
ion (
ph
a
se 1) T
HD=
1
.
36%
0
.
2
2
0
.
2
4
0
.
2
6
0
.
2
8
0
.
3
0
.
3
2
0
.
3
4
0
.
3
6
0
.
3
8
0
.
4
-
1
5
0
0
-
1
0
0
0
-
5
0
0
0
500
1000
1500
T
e
m
p
s
I
(
A
)
5
10
15
20
25
30
35
40
0
10
20
30
40
50
60
70
80
90
100
O
r
d
r
e
d
e
s
H
a
r
m
o
n
i
q
u
e
s
T
H
D
=
1
.
2
3
%
A
m
p
l
i
t
u
d
e
(
%
)
0
.
2
2
0
.
2
4
0
.
2
6
0
.
2
8
0
.
3
0
.
3
2
0
.
3
4
0
.
3
6
0
.
3
8
0
.
4
-
1
5
0
0
-
1
0
0
0
-
5
0
0
0
500
1000
1500
T
e
m
p
s(
s)
I
(
A
)
5
10
15
20
25
30
35
40
0
10
20
30
40
50
60
70
80
90
100
O
r
d
r
e
d
e
s
H
a
r
m
o
n
i
q
u
e
s
T
H
D
=
1
.
3
6
%
A
m
p
l
i
t
u
d
e
(
%
)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2089
-
4856
IJ
RA
V
ol
.
7
,
No
.
2
,
June
201
8
:
119
–
128
128
8
.
CONCL
US
I
O
N
The
ta
ble
belo
w
present
the
su
m
m
ary
resu
lt
s
of
t
he
dif
fe
ren
t
sim
ulati
o
ns
obta
ine
d
for
t
he
direc
t
con
t
ro
l
strat
eg
y
ap
plied
t
o
t
he
t
opologies
of
act
ive
pow
er
filt
er
ha
s
t
wo
-
le
vel
i
nver
te
r
.
F
r
om
the
resu
lt
s
ob
ta
ine
d
by
t
he
sim
ulati
on
m
ade
under
the
sam
e
op
e
r
at
ing
c
onditi
ons
of
the
F
A
P
ne
xt
t
wo
pr
opos
e
d
m
et
ho
ds,
the
com
po
ne
nt
m
e
thod
re
fer
e
nce
relat
ed
to
t
he
tim
ing
and
m
et
hod
of
i
ns
ta
ntane
ous
act
iv
e
an
d
reacti
ve
powe
r
with
M
VF
de
te
ct
ed
dis
rupt
ive
cu
rr
e
nt
s
uc
cessf
ul
they
eff
ect
ively
re
duced
the
TH
D
i
after
com
pen
sat
ion,
we
f
ound
t
hat
the
F
AP
hyste
r
esi
s
co
ntr
ol
m
et
hod
ga
ve
sat
isfact
ory
re
su
lt
s
a
nd
sli
ghtl
y
bette
r
than
t
he
P
WM
m
et
ho
d
a
s s
hows
in
T
a
ble
4
.
The
resu
lt
s
obta
ined
with
t
h
e
two
com
m
and
s
sho
w
that
th
e
i
m
ple
m
entat
i
on
of
the
act
iv
e
filt
er
can
sign
ific
a
ntly
r
edu
ce
the
ha
r
m
on
ic
con
te
nt
of
the
cu
rrent
source
,
wh
ic
h
res
ults
in
dec
reased
afte
r
fil
te
ring
THD
well
below
5%
.
We
se
e
that
TH
DS
obta
ined
with
the
so
-
cal
le
d
hy
ste
resis
or
de
r
s
are
l
ow
e
r
t
ha
n
t
hat
ob
ta
ine
d
with
the
P
WM
.
As
agai
ns
t,
hyste
resis,
des
pite
it
s
si
m
plici
ty
,
has
t
he
disad
va
ntage
of
a
va
riabl
e
switc
hing
f
re
quency,
w
her
e
the
prese
nce
of
a
la
rg
e
nu
m
ber
of
har
m
on
ic
s
i
n
the
c
urren
t
wave
.
P
W
M
te
chn
iq
ue,
eas
y
to
use
,
t
hen
pr
ovides
a
fixed
f
reque
ncy
operati
on,
ha
rm
on
ic
co
nten
t,
but
al
so
ha
s
f
ully
identifie
d by it
s struct
ur
e
, lim
it
ed
dynam
ic
propertie
s
.
Table
4
Sy
nth
e
sis of the
Re
s
ul
ts
Control
le
d
h
y
ste
resis
C
ontrol
le
d
MLI
cont
rolled
Dir
ect
p
-
q
6
.
81
-
cont
rolled
d
irect
p
-
q
with
MVF
1
.
68
1
.
73
cont
rolled
d
irect
SR
F
1
.
23
1
.
36
REFERE
NCE
S
[1]
M
.
Abdus
alam
,
"S
truc
ture
s
and
Strat
egies
fo
r
C
ontrol
of
Para
l
l
el
and
H
y
brid
Acti
ve
Fil
te
rs
with
Expe
r
imental
Vali
da
ti
ons,"
Ph
D T
hesis,
Henri
Poinca
ré
Univer
sit
y
,
Nan
c
y
-
I, M
a
y
2008
.
[2]
N
.
Bru
y
an
t,
"G
ene
ra
li
z
ed
Study
a
nd
Control
o
f
Para
ll
e
l
Ac
tive
Fil
te
rs,
Glob
al
o
r
Sel
ective
Com
pensa
ti
on
o
f
Harm
onic
s,
Ba
lance
d
o
r
Unbal
an
ce
d
R
egi
m
e
,
"
Ph
.
D T
hesis
,
Univ
e
rsit
y
of
Nan
te
s,
Franc
e, 1999
.
[3]
A
.
Sahban
i,
M
.
Bra
ie
k,
M
.
Da
m
i
and
M
.
J
emli,
"Control
of
a
Thr
ee
-
Ph
ase
Shunt
Fil
te
r
b
y
F
uzzy
Logic
,
"
3r
d
Inte
rna
ti
ona
l
Co
nfe
ren
c
es
SETI
T
,
Tun
isia,
27
-
31
,
Marc
h
2005
.
[4]
A
.
Rahmouni,
"
Perform
anc
e
Co
m
par
ison
of
Two
New
Cont
rol
Methods
of
a
Pa
ral
l
el
Acti
v
e
Fi
lter",
ACTA
,
vo
l5
3,
2012,
pp278
-
28
2
.
[5]
Y
.
Kus
um
a
La
th
a,
Ch
.
Saiba
bu
,
Y
.
P
.
Obulesh,
Control
Strateg
y
for
Three
Phase
Shunt
Acti
v
e
Pow
er
Filt
er
wit
h
Minim
um
Curre
nt
Mea
surem
e
n
t
s,
(IJECE)
,
Vol1,
No1,2011, pp
.
31
-
42
.
[6]
K
.
Shahra
m
,
"Continui
t
y
of
S
erv
ice
o
f
Thr
ee
-
Phase
Pow
er
Convert
ers
a
nd
FP
GA
Protot
y
ping
i
n
t
h
e
Lo
op:
Applic
a
ti
on
t
o
P
ara
l
le
l
Act
ive
Fi
l
te
r,
"
PhD
the
sis,
Henri
Poinc
aré
Univer
sit
y
,
Nan
c
y
-
I, Janua
r
y
20
09
.
[7]
R
.
Ba
la
subram
a
nia
n,
S
.
Pala
n
i,
Sim
ula
ti
on
and
Perform
anc
e
Ev
al
ua
ti
on
o
f
Shu
nt
H
y
brid
Pow
er
Filt
er
for
Pow
er
Quali
t
y
Im
prove
m
ent
Us
ing
PQ
The
or
y
,
(IJECE)
,
Vol6,
No6,2016
,
pp
.
2603
-
2609
.
[8]
B
.
Choukr
i,
"Contribut
ion
to
th
e
Stud
y
of
th
e
Di
ffe
ren
t
Topo
logies
and
Com
m
ands
of
Par
al
l
el
Acti
v
e
Fil
te
rs
W
ith
Struct
ure
T
ensi
on:
Model
ing,
Sim
ula
ti
on
and
Expe
r
imental
Vali
da
ti
on
o
f
t
he
Order
,
"
Doc
tora
l
Th
esis,
H
en
ri
Poinca
ré
Univer
sit
y
,
Nan
c
y
-
I, D
ec
ember
2004
.
[9]
A
.
MO
RS
LI,
"
Depoll
uti
on
of
Low
-
Volta
ge
E
l
ec
tr
ic
a
l
N
et
wor
ks
Us
ing
a
T
wo
-
Le
vel
Para
llel
A
ct
iv
e
Fi
lt
er
Control
le
d
b
y
T
he
PQ
,
"
ACTA
a
lgori
thm,
vo
l
53
,
2012,
pp105
-
11
1
.
BIOGR
AP
H
Y
O
F
AU
TH
OR
B
eny
ett
ou
Lou
t
fi
wa
s born
in
M’s
il
a, Alge
ri
ai
n
1
979
.
He
r
ec
e
ive
d
his E
ngin
ee
r
d
e
gre
e
in
Elec
tronics
and
Magiste
r
d
e
gre
e
in
industrial
con
trol
from
M’s
il
a
Univer
sit
y
in
2002
and
2006
respe
ctively
.
He
re
ce
iv
ed
h
is
Ph
.
D
.
in
elec
tr
i
ca
l
eng
ine
er
ing
from
Bec
har
Un
ive
rsit
y
,
Alg
er
ia,
in
2016
.
He
w
as
rec
rui
te
d
in
200
9
as
elec
tron
ic
s
assistant
profe
ss
or
in
Univer
sit
y
of
M'
sil
a
.
His
sc
ie
nti
f
ic
int
e
rests
are
power
qua
li
t
y
c
o
ndit
ioni
ng
,
DS
P
and
d
igi
t
al
con
tr
ol,
cont
ro
l
and
d
ia
gnostic
.
He
is
m
ember
of
sev
er
al
rese
arc
h
proj
ec
ts
at Unive
rs
i
t
y
of
Ms
il
a
and
Elec
tr
ic
a
l
Eng
ineeri
ng
La
bor
at
or
y
of M
sila
Univer
si
t
y
Evaluation Warning : The document was created with Spire.PDF for Python.