Internati
o
nal
Journal of P
o
wer Elect
roni
cs an
d
Drive
S
y
ste
m
(I
JPE
D
S)
Vol.
4, No. 4, Decem
ber
2014, pp. 508~
516
I
S
SN
: 208
8-8
6
9
4
5
08
Jo
urn
a
l
h
o
me
pa
ge
: h
ttp
://iaesjo
u
r
na
l.com/
o
n
lin
e/ind
e
x.ph
p
/
IJPEDS
Dynamic Power Qualit
y Compensator with
an Adaptive Shunt
Hybrid Filter
Sindhu
M R,
Manjul
a
G
Nair, T
N P
Nambiar
Departm
e
nt o
f
E
l
ec
tric
al
a
nd Electronics
Engin
e
ering,
AMRITA VISH
WA
VIDY
APEETHAM
Article Info
A
B
STRAC
T
Article histo
r
y:
Received
May 24, 2014
R
e
vi
sed Oct
7,
2
0
1
4
Accepted Oct 20, 2014
Major portion o
f
nonlin
ear
load
s in industries are con
t
ributed b
y
v
a
riable
s
p
eed driv
es
be
caus
e
of
the
i
r
des
i
rabl
e fe
atur
es
s
u
ch as
en
e
r
g
y
s
a
v
i
ng,
smooth control, flexib
le
operation and f
a
st res
ponse.
Th
ese electr
i
c dr
ives
introduce large
amount of current and
voltag
e
harmonic distortions at th
e
point of common coupling.
These dist
ortions ar
e propagated thr
oughout the
s
y
s
t
em
and affe
ct al
l other lo
a
d
s
connected
in
the s
y
s
t
em
. H
e
nce th
es
e
distortions are to
be m
itigated with suitable harm
onic filt
ers instal
led near to
the respective load terminals.
This
paper presents an effective
ANN based
digital con
t
roller for shunt h
y
br
id ha
rmonic f
ilter to prov
ide
in
stantan
e
ous
harmonic and
reactive
compensation.
Th
e p
e
rfor
m
ance of
the
ad
aptiv
e shunt
h
y
brid f
ilter is v
e
rified b
y
simulati
on and
experimental
studies under stead
y
state and d
y
n
a
mic conditions. Th
e results s
how that it
is an e
ffective, flexib
le
and low r
a
ted h
ybrid filter
conf
ig
uration
.
Keyword:
Harm
oni
c Di
st
ort
i
o
n
Harm
o
n
i
c Filter
Power Qu
ality
Reactive power com
p
ensati
on
Activ
e Filter
Copyright ©
201
4 Institut
e
o
f
Ad
vanced
Engin
eer
ing and S
c
i
e
nce.
All rights re
se
rve
d
.
Co
rresp
ond
i
ng
Autho
r
:
Sin
dhu
M R,
Depa
rt
m
e
nt
of
El
ect
ri
cal
and
El
ect
roni
cs
E
n
gi
nee
r
i
n
g,
Am
rita Vishwa
Vidy
a
Peetha
m
,
Amrita nagar, C
o
imbatore –
641
112.
Em
a
il:
1
.
IN
TR
OD
UC
TI
ON
Utilities tak
e
effo
rts to
m
a
i
n
tain
goo
d
p
o
wer
q
u
a
lity at th
e p
o
i
n
t
o
f
co
mm
o
n
co
uplin
g
(PCC
).
Majority of industrial loads are induc
tion m
o
tor drives [1,
2]. They are pre
f
erred
because
of sa
ving in energy,
rugged nature, easiness
in
c
o
ntrol
and chea
pness [1,
2]. Va
rious PW
M te
c
hni
que
s suc
h
a
s
Space Ve
ctor pulse
width m
o
dulation,
sinusoi
d
al PW
M etc. a
r
e im
ple
m
en
ted
[3
]-
[5
].
Th
ese ar
e im
p
l
e
m
en
ted
with
d
i
g
ital
cont
rol
l
e
rs a
n
d
i
t
im
prove
d p
e
rf
orm
a
nce of
i
n
d
u
ct
i
on m
o
t
o
r
dri
v
e
s
. Th
es
e equi
pm
ent
resul
t
e
d i
n
ge
ne
r
a
t
i
o
n
of
l
a
r
g
e am
ou
nt
s
of
cu
rre
nt
and
vol
t
a
ge
ha
rm
oni
cs, re
d
u
c
t
i
on i
n
s
o
urce
po
we
r f
act
or
,
vol
t
a
ge
sa
g,
v
o
l
t
a
ge
swell etc. [6
].
Th
e pro
p
a
g
a
tio
n
o
f
t
h
ese power
q
u
a
lity
issu
es are
p
r
ev
ented
b
y
in
stallin
g
h
a
rm
o
n
i
c fi
lter o
f
sui
t
a
bl
e c
o
n
f
i
g
urat
i
o
n at
p
o
i
n
t
of
com
m
on c
o
u
p
l
i
n
g.
Di
f
f
er
ent
p
o
w
er
q
u
al
i
t
y
im
provem
e
nt
sc
hem
e
s such
a
s
p
a
ssiv
e
filters
[6
] an
d activ
e
filters [7
] are
d
e
v
e
l
o
p
e
d
b
y
v
a
ri
o
u
s
research
ers. Pass
iv
e
filters h
a
v
e
drawb
a
cks
su
ch
as
bu
lk
y
in
size, h
i
g
h
n
o
lo
ad
losses, reson
a
n
c
e,
fix
e
d co
m
p
en
satio
n
etc
[6
]. Activ
e filters
p
r
o
v
i
d
e
effectiv
e an
d
dyn
amic co
m
p
en
sation
with
the h
e
lp
of effi
ci
en
t co
n
t
ro
llers
[7
]. Bu
t th
is sch
e
m
e
is ex
p
e
nsiv
e.
Hen
c
e as an
eco
n
o
m
ical so
lu
tio
n
,
h
y
b
r
i
d
filters
were
d
e
v
e
l
o
p
e
d
.
Here, shu
n
t
p
a
ssi
v
e
filter co
n
t
ri
b
u
t
es
maj
o
r
p
a
rt of reactive p
o
wer co
mp
ensatio
n
and rem
a
in
in
g
h
a
rm
o
n
i
cs are co
m
p
en
sated
by sh
un
t activ
e filter.
Vari
o
u
s
research
p
a
p
e
rs on
in
tellig
en
t d
i
g
i
t
a
l co
n
t
ro
ller based
h
a
rm
o
n
i
c filters are pub
lish
e
d
[8
-1
2
]
. Th
i
s
pape
r
desc
ri
be
s an
A
N
N
ba
sed a
d
a
p
t
i
v
e s
h
u
n
t
hy
b
r
i
d
fi
l
t
e
r f
o
r
ha
rm
oni
c m
i
ti
gat
i
o
n
an
d
react
i
v
e
po
we
r
com
p
ensat
i
on
of a
d
just
a
b
l
e
spee
d d
r
i
v
es
.
A
N
N
c
ont
rol
l
er hel
p
s t
o
p
r
ovi
de i
n
st
a
n
t
a
neo
u
s
harm
on
i
c
and
reactiv
e
p
o
wer co
m
p
en
sation
u
n
d
e
r stead
y
state an
d d
y
n
a
m
i
c con
d
ition
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
PED
S
I
S
SN
:
208
8-8
6
9
4
Dynamic Po
wer Qu
a
lity Comp
ensa
t
o
r with
a
n
Ad
ap
tive Shu
n
t
Hyb
r
id Filter (S
ind
h
u
M R)
50
9
2
.
TEST
SYSTEM
An ind
u
s
t
r
ial syste
m
is select
ed
for
p
e
rforman
ce st
u
d
y
o
f
th
e
ad
ap
tiv
e
shu
n
t
h
ybrid
filter. A
scaled
do
w
n
l
a
b
o
rat
o
ry
m
odel
of i
n
dust
r
i
a
l
dri
v
e,
t
h
ree
phas
e
3
HP,
4
pol
e,
i
n
duct
i
o
n m
o
t
o
r
dri
v
e i
s
use
d
i
n
t
h
i
s
work. It is ope
rated as an a
d
justable speed
dri
v
e usin
g t
h
r
ee phase
vol
t
a
ge so
urce i
nve
rt
er. T
h
e speci
f
i
cat
i
ons
of
t
h
e i
n
d
u
ct
i
o
n m
o
t
o
r
dri
v
e a
r
e s
h
o
w
n i
n
Ta
bl
e 1
.
Table 1. Induct
ion
M
o
tor Dri
v
e
Speci
fications
Power
rating
3 HP
Switching fr
equen
c
y
10kHz
Ro
to
r resistan
ce
1.
24
Stator resistance
1.
517
Vdc 300V
Base fr
equency
100Hz
Stator Le
akage Re
actance
5.
12
Rotor Leakage Re
actance
120
No.
of poles
4
M
o
m
e
nt of I
n
er
tia
0.
2 kg
m
2
The e
x
peri
m
e
nt
al
set
up
i
s
m
a
de as
sh
o
w
n
i
n
t
h
e
bl
oc
k
di
ag
ram
i
n
Fi
gu
re
1.
Fi
gu
re
1.
Sc
he
m
a
t
i
c
Di
agram
of T
h
ree
Phas
e Test System
The VSI is controlled usi
ng
space vect
or PWM.
The s
p
a
ce vector
PWM based s
p
ee
d control is
im
pl
em
ent
e
d t
h
r
o
ug
h t
h
e f
o
l
l
o
wi
ng
st
eps:
(a)
Measu
r
e th
e m
o
tor
q
u
a
n
tities (sp
e
ed
and
p
h
a
se curren
t
s)
(b
)
Transfo
r
m
th
em
to
two ph
ase syste
m
(
,
β
)
usin
g Clark
e
tran
sfo
r
m
a
tio
n
(c)
Calculate rot
o
r flux s
p
ace
vec
t
or m
a
gnitude
and angle
(d
)
The st
at
o
r
c
u
r
r
e
nt
t
o
rq
ue a
n
d
fl
u
x
c
o
m
pone
n
t
s are se
parat
e
l
y
cont
rol
l
e
d
(i
sd
, i
sq
)
(e)
The
o
u
t
p
ut
st
at
or
v
o
l
t
a
ge s
p
ac
e vect
o
r
i
s
c
a
l
c
ul
at
ed
usi
n
g
de
cou
p
l
i
n
g
bl
oc
k
(
v
sd
, v
sq
)
(f
)
The stator voltage s
p
ace vect
or is tra
n
s
f
ormed by
a
n
invers
e park tra
n
s
f
ormation bac
k
from
d –q
refe
rence
frame into t
w
o phas
e sy
ste
m
fix
e
d
with
th
e stator
(v
s
, v
s
β
)
(g
)
Using s
p
ace
ve
ctor m
o
dulation,
output
three pha
se voltages
is
ge
nerated.
The
wh
ol
e
pr
o
cess o
f
t
h
e
sp
eed c
ont
rol
l
e
r,
im
pl
em
ent
e
d i
n
t
h
e
dsP
I
C
3
0F
40
1
1
, i
s
e
x
pl
ai
ned
us
i
n
g
flo
w
cha
r
t in
F
i
gu
re 2
.
T
h
ree
pha
se
s
o
urce c
u
rrents
, and s
o
urce
voltages
at th
e po
in
t
o
f
co
mm
o
n
co
up
lin
g are
sen
s
ed
and
h
a
rm
o
n
i
c an
alysis is p
e
rform
e
d
with
th
e
h
e
lp o
f
p
o
wer qu
ality an
alyser. Th
e power sy
ste
m
q
u
a
n
tities
–
m
a
g
n
itud
e
s and
wav
e
fo
rm
s
o
f
sou
r
ce curren
t
s
an
d
source v
o
l
t
a
g
e
s for d
i
fferen
t
sp
eed
settin
g
s
are
sho
w
n i
n
Fi
gu
r
e
3.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-86
94
I
J
PED
S
Vo
l.
4
,
No
.
4
,
D
ecem
b
er
2
014
:
50
8 – 516
51
0
Fi
gu
re
3.
S
o
u
r
ce v
o
l
t
a
ge a
n
d
sou
r
ce c
u
rre
nt
at
t
h
e p
o
i
n
t
o
f
com
m
on cou
p
l
i
ng
f
o
r
va
ri
abl
e
spee
d i
n
d
u
ct
i
o
n
m
o
tor dri
v
e
with s
p
eed
o
f
(a)
1
3
5
ra
d/s
ec (C
ase I) (b)320 ra
d/sec
– (Case
II)
Fi
gu
re
2.
Fl
o
w
cha
r
t
o
f
t
h
e
s
p
eed c
ont
rol
l
e
r
Tabl
e 2
sh
o
w
s
l
a
rge am
ount
o
f
ha
rm
oni
cs i
n
cur
r
ent
s
at
t
h
e
poi
nt
o
f
com
m
on c
o
upl
i
ng
u
n
d
er st
ea
dy
state co
nd
itio
ns.
Wh
en
t
h
e lo
ad
is ch
ang
e
d
sudd
en
ly from 7
5
%
rated
lo
ad
t
o
25
%
rated
lo
ad, th
e so
urce
current,
reacti
v
e
power de
m
a
nd a
n
d ha
rm
onic com
pone
nts a
r
e al
so c
h
ange
d c
o
rres
pondingly. The
fu
n
d
am
ent
a
l
com
pone
nt
of
sou
r
ce c
u
r
r
ent
,
TH
D i
n
s
o
u
r
ce cu
rre
nt
, %
of c
u
r
r
ent
ha
rm
oni
c com
ponent
s
,
displacem
ent powe
r
factor,
distortion po
we
r
factor, sourc
e
powe
r factor, r
eal powe
r
,
reac
tive power, appare
nt
po
we
r a
n
d
di
st
ort
i
o
n
p
o
we
r i
n
eac
h cy
cl
e i
s
sho
w
n i
n
Ta
bl
e 3.
Tabl
e
2(a
)
.
Fu
ndam
e
nt
al
com
p
o
n
e
n
t
o
f
s
o
ur
ce cu
rre
nt
a
n
d
THD
i
n
so
u
r
ce
cu
rre
nt
at
t
h
e
poi
nt
o
f
c
o
m
m
o
n
co
up
ling
with
v
a
riab
le sp
eed
in
du
ctio
n m
o
to
r
d
r
iv
e – und
er stead
y state con
d
ition
s
Case
Fundam
e
ntal
co
m
ponent o
f
source current
(p.u.)
THD in
Sour
ce
Cu
rren
t
(%
)
% of
5
th
har
m
onic
% of
7
th
har
m
onic
% of
11
th
harm
onic
% of
13
th
harm
onic
Phase a
Phase b
Phase c
Phase a
Phase
A
Phase
A
Phase
a
Phase
a
I
0.
92
0.
92
0.
92
34.
5
28.
2
12.
6
4.
1
2.
1
I
I
0.
51
0.
51
0.
51
57.
13
53.
5
19.
4
5.
2
2.
9
Tab
l
e
2
(
b
)
. Est
i
m
a
t
i
o
n
o
f
Power system
p
a
ra
m
e
ters for the sp
ecified
load
co
nd
itio
ns
u
nder stead
y
state
Case
Displace
m
e
nt
power f
actor
Distor
tion power
f
actor
Sour
ce power
f
actor
Real Power
(
pu)
Reactive
Power
(
pu)
Appar
e
nt
Power
(
pu)
Distor
tion
Po
wer (p
u
)
I
0.
98
0.
9453
0.
9293
0.
91
0.
16
0.
92
0.
08
I
I
0.
53
0.
87
0.
4342
0.
46
0.
7967
0.
92
0.
085
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
PED
S
I
S
SN
:
208
8-8
6
9
4
Dynamic Po
wer Qu
a
lity Comp
ensa
t
o
r with
a
n
Ad
ap
tive Shu
n
t
Hyb
r
id Filter (S
ind
h
u
M R)
51
1
Tabl
e
3(a
)
.
Fu
ndam
e
nt
al
com
p
o
n
e
n
t
o
f
s
o
ur
ce cu
rre
nt
a
n
d
TH
D in source
curre
nt in e
a
c
h
cycle at the
point
of
co
mm
o
n
co
up
l
i
n
g
with
v
a
riable sp
eed
indu
ctio
n
m
o
to
r drive –
un
d
e
r
d
y
n
a
mic co
n
d
ition
s
Cy
cle
Fundam
e
ntal
co
m
ponent o
f
source current
(p.u.)
THD in
Sour
ce
cu
rren
t
(%)
% of
5
th
har
m
onic
% of
7
th
har
m
onic
% of
11
th
harm
onic
% of
13
th
harm
onic
Phase a
Phase b
Phase c
Phase a
Phase
A
Phase
A
Phase
A
Phase
a
I
0.
92
0.
92
0.
92
34.
5
28.
2
12.
6
4.
1
2.
1
I
I
0.
84
0.
84
0.
84
39.
2
34.
6
14.
12
4.
29
2.
25
I
I
I
0.
78
0.
78
0.
78
43.
8
39.
87
15.
78
4.
47
2.
42
I
V
0.
70
0.
70
0.
70
47.
96
44.
3
17.
24
4.
7
2.
63
V 0.
62
0.
62
0.
62
53.
2
49.
7
18.
3
4.
95
2.
81
VI
0.
51
0.
51
0.
51
57.
13
53.
5
19.
4
5.
2
2.
9
Tabl
e
3(
b)
. Est
i
m
a
ti
on
of
P
o
wer
sy
st
em
param
e
t
e
rs i
n
eac
h cy
cl
e wi
t
h
va
ri
abl
e
s
p
eed
i
n
duct
i
o
n m
o
t
o
r
dri
v
e -
u
n
d
e
r
d
y
n
a
m
i
c
lo
ad cond
itio
ns
Cy
cle
Displace
m
e
nt
power f
actor
Distor
tion power
f
actor
Sour
ce power
f
actor
Real Power
(
pu)
Reactive
Power
(
pu)
Appar
e
nt Power
(
pu)
Distor
tion
Po
wer (p
u
)
I
0.
98
0.
9453
0.
9293
0.
91
0.
16
0.
92
0.
08
I
I
0.
9397
0.
945
0.
888
0.
7893
0.
287
0.
84
0.
03
I
I
I
0.
866
0.
9289
0.
8044
0.
6755
0.
39
0.
78
0.
026
I
V
0.
766
0.
9143
0.
7004
0.
5362
0.
45
0.
7
0.
0036
V
0.
6428
0.
8955
0.
5756
0.
3985
0.
4749
0.
62
0.
0082
VI 0.
53
0.
87
0.
4342
0.
46
0.
7967
0.
92
0.
085
Results show
an effective harm
onic filter is n
ecessary
to provide
variable reactive powe
r and
harm
onic compensation to meet IEEE
standards
of good powe
r quality.
Ac
tive filters are comm
only u
s
ed
for
d
y
n
a
m
i
c p
o
w
er q
u
a
lity co
m
p
en
sation
.
Bu
t th
ese activ
e filte
rs are h
i
gh
ly ex
p
e
n
s
i
v
e. Hen
ce as an
econo
m
i
ca
l
so
lu
tion
,
h
ybrid
filters were in
trodu
ced.
Here, shun
t passiv
e
filter co
n
t
ribu
tes m
a
j
o
r
p
a
rt
o
f
reactiv
e
com
p
ensation
and selected harm
onic co
m
p
ensat
i
o
n.
O
n
l
y
rem
a
i
n
i
ng
har
m
oni
c com
p
ensat
i
on i
s
t
o
be
m
e
t
by
th
e sh
un
t active filter.
3
.
PROPOSE
D
ADAPTIVE
SHUNT HYBRID
FILTER FOR POWER QUALIT
Y
COMPENSATION
IN
I
N
DU
CTI
O
N MOT
O
R
DRI
VES
Th
e ad
ap
tiv
e sh
un
t h
ybrid
fil
t
er co
n
s
ists of th
r
ee ph
ase adap
tiv
e shun
t p
a
ssiv
e
filters an
d
ad
ap
tive
sh
un
t activ
e
filter. Ad
ap
tiv
e sh
un
t p
a
ssiv
e
filters are d
e
v
e
l
o
p
e
d
b
y
series co
m
b
in
atio
n
o
f
t
h
yristo
r
switch
e
d
series co
m
b
in
atio
n
o
f
ind
u
c
t
o
r an
d cap
acitor. Th
eir
v
a
lu
es
are designed
s
u
ch
as
t
o
pr
ov
i
d
e
by
pass pat
h
fo
r
fi
ft
h a
nd se
ve
nt
h
harm
oni
cs
. It
al
so p
r
o
v
i
d
es f
u
ndam
e
nt
al
freq
u
e
n
cy
r
eact
i
v
e po
we
r
t
o
t
h
e sy
st
em
. Th
e
capacitor val
u
es are
selected as
210
F and inductor
values a
r
e
15m
H and
7m
H res
p
ectively. Re
maining
am
ounts of ha
rm
onics
and reactive
co
m
p
ensation curre
nt
s are provi
ded by the a
d
apti
ve s
h
unt active filter.
The s
h
unt active filter is c
o
ntrolled to i
n
ject rem
a
ining am
ount
of ha
rm
onics
and re
active
com
p
ensation
cur
r
ent
s
suc
h
t
h
at
so
u
r
ce
nee
d
t
o
su
p
p
l
y
o
n
l
y
real
pa
rt
o
f
f
u
n
d
am
ent
a
l
com
ponent
of c
u
rre
nt
. T
h
e sc
he
m
a
t
i
c
d
i
agram
o
f
th
e
filterin
g
system
is sh
o
w
n
i
n
Fig
u
re
4
.
Th
e h
ybrid
filter
el
e
m
en
ts
are
contro
lled
b
y
ANN b
a
sed
in
tellig
en
t d
i
g
i
tal co
n
t
ro
ller. Th
e respo
n
s
i
v
en
ess of f
ilter
is i
m
p
r
o
v
e
d
by h
y
steresis con
t
ro
ller b
a
sed
clo
s
ed
lo
op
co
n
t
ro
l. PI co
n
t
ro
ller is also
u
s
ed to
com
p
en
sate for i
n
v
e
rter l
o
sses
an
d regu
latio
n
o
f
d
c
lin
k vo
ltag
e
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-86
94
I
J
PED
S
Vo
l.
4
,
No
.
4
,
D
ecem
b
er
2
014
:
50
8 – 516
51
2
Figure
4. Thre
e phase syst
em feed
ing
v
a
riable sp
eed
indu
ctio
n
m
o
to
r drive with
Ad
ap
tive Shun
t Hy
b
r
i
d
Filter
3.
1. Me
th
od
ol
og
y
Th
e l
o
ad curren
t
in
i
n
du
ction
m
o
to
r d
r
i
v
e is
represen
ted
b
y
:
)
sin(
)
sin(
)
(
,
5
,
3
1
1
h
Lh
Lh
L
L
t
h
i
t
i
t
i
L
= R
eal
pa
rt
of
fu
n
d
am
ent
a
l
com
pone
nt
of
l
o
a
d
c
u
r
r
ent
+R
eact
i
v
e pa
rt
of
f
u
n
d
am
ent
a
l
com
pone
nt
of
l
o
ad c
u
rre
nt
+
H
a
r
m
o
n
i
c
co
mp
on
en
ts
o
f
lo
ad
cu
rr
e
n
t
(1
)
Ad
ap
tiv
e shu
n
t p
a
ssiv
e
filter co
m
p
en
sates fo
r m
a
j
o
r
p
a
rt
o
f
fund
am
en
ta
l frequ
en
cy reactiv
e p
o
wer
an
d
h
a
rm
o
n
i
cs. Ad
ap
tiv
e sh
un
t activ
e filter in
j
ects rem
a
in
in
g
sm
all a
m
o
u
n
t
o
f
h
a
rm
o
n
i
cs an
d
h
e
lp
s to
o
b
t
ain
unity power fa
ctor si
nusoi
d
al
curren
t at t
h
e
po
in
t
o
f
co
mm
o
n
co
up
ling
.
Th
ree ph
ase i
n
stan
tan
e
ou
s so
urce
v
o
ltag
e
s, fu
nd
am
en
tal
lo
ad
cu
rren
ts an
d filter com
p
en
satio
n
currents are se
lected as input
qua
n
tities for training ada
p
tiv
e ANN. The refere
nce co
m
p
ensation curre
n
ts are
devel
ope
d by
anal
o
g
c
ont
rol
l
er
ci
rcui
t
.
Th
e
act
ual
c
o
m
p
ensation c
u
rre
n
ts are
com
p
ared with re
ference
com
p
ensat
i
on
cur
r
ent
s
t
o
ge
nerat
e
s
w
i
t
c
hi
ng
p
u
l
s
es t
o
I
G
B
T
s
of t
h
ree
pha
se v
o
l
t
a
ge
so
urce i
nve
rt
er. T
h
e
current transducers and volta
ge tran
sdu
cers sa
m
p
le in
stan
tan
e
ou
s lo
ad
c
u
rrents
, com
p
ensa
tion current
s
and
so
urce vo
ltag
e
s.
Sam
p
led
v
a
lu
e o
f
in
stan
ta
ne
o
u
s
lo
ad
cu
rr
en
t is
ex
pr
e
s
s
e
d
a
s
:
i.e. i
L
[k]
=
)]
cos(
)
sin(
[
2
,.....
3
,
2
,
1
1
t
nk
W
t
nk
W
n
N
n
n
=
)]
cos(
)
sin(
[
)]
sin(
)
cos(
[
2
,.....
3
,
2
1
21
11
t
nk
W
t
nk
W
t
nk
W
t
nk
W
n
N
n
n
= i
L1, re
[k]
+ i
L1, im
[k]
+
h
[k]
ih
(
2
)
whe
r
e W
1n
an
d
W
2n
a
r
e am
pl
i
t
ude
s
of si
ne a
n
d c
o
si
ne
com
pone
nt
s
of
t
h
e
m
easured
l
o
a
d
cur
r
ent
.
Th
e equ
a
tio
n is rep
r
esen
ted in
v
ectorial fo
rm
as:
i
L
[
k
] =
[W
11
W
21
--
--
--
--
-
W
1N
W
2N
]
)
cos(
)
sin(
)
2
cos(
)
2
sin(
)
cos(
)
sin(
t
Nk
t
Nk
t
k
t
k
t
k
t
k
The
a
d
a
p
t
i
v
e AN
N net
w
or
k i
s
im
pl
em
ent
e
d usi
n
g Wi
dr
ow
-H
of
f w
e
i
g
h
t
s up
datin
g algor
ith
m
[
6
], [7
]
.
The w
e
i
g
ht
s o
f
t
h
e co
n
n
ect
i
on l
i
n
ks a
r
e u
pdat
e
d u
s
i
n
g e
r
r
o
r i
n
t
h
e est
i
m
at
ed fu
ndam
e
nt
al
com
pone
nt
an
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
PED
S
I
S
SN
:
208
8-8
6
9
4
Dynamic Po
wer Qu
a
lity Comp
ensa
t
o
r with
a
n
Ad
ap
tive Shu
n
t
Hyb
r
id Filter (S
ind
h
u
M R)
51
3
act
ual
real
part
of f
u
n
d
am
ent
a
l
co
m
pone
nt
o
f
l
o
ad cu
rre
nt
. The bac
k
p
r
o
p
a
gat
i
on
neu
r
al
net
w
or
k was t
r
ai
ned
wi
t
h
M
A
TL
A
B
usi
ng
1
5
0
0
t
r
ai
ni
n
g
pat
t
e
r
n
s f
o
r
25
0
0
e
poc
hs
wi
t
h
g
o
a
l
of 0.
0
0
1
.
T
r
ai
ni
ng m
e
t
hod
used i
s
Leve
nbe
rg
–
M
a
rq
uar
d
t
al
g
o
ri
t
h
m
.
Trai
ne
d art
i
f
i
c
i
a
l
neu
r
al n
e
two
r
k
con
s
ists o
f
two
l
a
yers: in
pu
t layer (6
neu
r
ons
), o
u
t
p
ut
l
a
y
e
r (7 neu
r
o
n
s) a
nd
hi
d
d
e
n l
a
y
e
r (6 ne
u
r
o
n
s)
. The A
N
N
net
w
or
k i
s
sho
w
n i
n
Fi
gu
re
5. The
per
f
o
r
m
a
nce charact
eri
s
t
i
c
s
o
f
neu
r
al
net
w
o
r
k
t
r
ai
ni
ng
i
s
s
h
o
w
n i
n
Fi
gu
r
e
6.
T
h
e
AN
N
i
s
p
r
og
ram
m
e
d a
n
d
im
pl
em
ent
e
d
u
s
i
ng ds
PIC
30F
40
1
1
m
i
croco
n
t
r
ol
l
e
r.
T
w
o bi
p
o
l
a
r ADC
s
are
u
s
ed f
o
r
c
o
n
v
e
r
si
o
n
of
a
n
al
o
g
q
u
a
n
tities to
dig
ital q
u
a
n
titie
s. Sam
p
lin
g
freq
u
e
n
c
y
o
f
3kHz is selected co
nsid
ering
time fo
r sen
s
i
n
g in
pu
t
sam
p
l
e
s, di
gi
t
a
l
con
v
er
si
o
n
,
p
r
o
g
ram
execut
i
on a
n
d
gene
rat
i
on
of
swi
t
c
hi
n
g
si
g
n
al
s. T
h
e
swi
t
c
hi
n
g
pul
s
e
s are
a
m
p
lified
and
g
i
v
e
n
to
IGBTs in
shu
n
t
active filter. Th
e isolatio
n
b
e
tween
p
o
wer circu
it an
d
con
t
ro
ller ci
rcu
i
t
is d
o
n
e
u
s
i
n
g an
o
p
t
o
c
o
upler 6
N
13
6. Th
e shun
t acti
v
e po
wer filter circu
it in
clu
d
e
s po
wer in
v
e
rter
(SKM
5
0
GB
12
B) with
th
ree
pha
se I
G
BT
br
idg
e
an
d two
cap
acito
r
s
o
f
22
00
F. T
h
e c
o
m
p
ensat
i
ng c
u
rre
nt
s
in
j
ected
b
y
th
r
e
e ph
ase inv
e
r
t
er thr
oug
h cou
p
l
i
n
g indu
ctor
s t
o
th
e po
in
t
o
f
commo
n
coup
lin
g. For
s
A
dt
di
F
/
10
max
, 400
V
5
A
10
m
H
cou
p
ling ind
u
c
t
o
r is
used
. Th
e ex
p
e
rim
e
n
t
al resu
lts with the
in
sertion
o
f
ANN Con
t
ro
ller
b
a
sed
sh
un
t
h
y
b
r
i
d
filter are
discu
ssed
in fo
llo
wi
n
g
section
.
Fig
u
re
5
.
Artificial Neural Net
w
ork fo
r
c
ont
r
o
l
l
i
ng a
d
a
p
t
i
v
e
sh
unt
hy
bri
d
fi
l
t
e
r
Fig
u
re 6
.
Perform
an
ce
p
l
o
t
o
f
train
i
ng
n
e
ural n
e
two
r
k
for
con
t
ro
llin
g
t
h
e
adap
tiv
e
shu
n
t
h
y
b
r
i
d
filter
4
.
PE
RFORMANCE OF ANN
CONTROLLED
SHUNT ACTIVE FILTER
UNDER
DYNAMIC
CO
NDITI
ON
S–
RESULTS
AN
D DIS
C
US
SION
Th
e o
p
e
rating
p
e
rform
a
n
ce
of
th
e
ad
ap
tiv
e sh
un
t h
ybr
id
filter is stu
d
i
ed
u
n
d
e
r d
y
n
a
m
i
c
con
d
ition
s
–
wi
t
h
t
h
e a
ppl
i
c
at
i
on o
f
2
5
% l
o
ad a
n
d 7
5
% l
o
ad at
t
= 0.
5s.
Fu
ndam
e
nt
al
com
pone
nt
o
f
sou
r
ce c
u
r
r
ent
,
THD
in source c
u
rrent and
powe
r syste
m
parameters at th
e
poin
t
o
f
co
mm
o
n
co
up
ling
with
th
e in
stallatio
n of
ad
ap
tiv
e shu
n
t
h
y
b
r
i
d
filter is sh
own
in
Tab
l
e 4
and
5
.
Fi
g
u
re 7, and
8
sh
ow wav
e
form
s
o
f
so
urce vo
ltag
e
,
lo
ad
curren
t
,
filter cu
rren
t an
d source cu
rren
t
with th
e
i
n
stallatio
n
o
f
ad
ap
tiv
e
sh
un
t
h
ybrid
filter u
n
d
e
r
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-86
94
I
J
PED
S
Vo
l.
4
,
No
.
4
,
D
ecem
b
er
2
014
:
50
8 – 516
51
4
dynam
i
c condi
tions. T
h
e pe
rform
a
nce of a
d
aptive s
h
unt
h
ybrid filter is verifie
d
with
reduction in s
o
urce
cur
r
ent
ha
rm
oni
cs a
nd
si
n
u
s
o
i
d
al
s
o
urce c
u
rre
nt
s i
n
p
h
ase
wi
t
h
s
o
urce
v
o
l
t
a
ges.
W
a
vef
o
rm
s al
so co
n
f
i
r
m
t
h
e
p
e
rform
a
n
ce o
f
th
e
filter und
er
d
y
n
a
m
i
c co
nd
itio
n
s
.
Tabl
e
4(a
)
.
Fu
ndam
e
nt
al
com
p
o
n
e
n
t
o
f
s
o
ur
ce cu
rre
nt
a
n
d
THD
i
n
so
u
r
ce
cu
rre
nt
at
t
h
e
poi
nt
o
f
c
o
m
m
o
n
co
up
ling
with
v
a
riab
le sp
eed
in
du
ctio
n m
o
to
r
d
r
iv
e with Adap
tiv
e Shu
n
t
Hyb
r
id Filter – Stead
y state
co
nd
itio
ns
Case
Fundam
e
ntal
co
m
ponent o
f
source current
(p.u.)
THD in
Source
Cu
rren
t
(%
)
% of
5
th
har
m
onic
% of
7
th
har
m
onic
% of
11
th
harm
onic
% of
13
th
har
m
onic
Phase a
Phase b
Phase c
Phase a
Phase
A
Phase
A
Phase
a
Phase
a
I
0.
84
0.
84
0.
84
3.
15
2.
2
1.
6
0.
7
0.
4
I
I
0.
23
0.
23
0.
23
3.
19
2.
5
1.
4
1.
2
0.
9
Tab
l
e
4
(
b
)
. Est
i
m
a
te o
f
Po
wer syste
m
p
a
rameters fo
r th
e
specified
lo
ad
con
d
ition
s
–
with Ad
ap
tiv
e Sh
un
t
Hyb
r
i
d
Filter
– Stead
y state co
nd
itio
ns
Case
Displace
m
e
nt
power f
actor
Distor
tion power
f
actor
Sour
ce power
f
actor
Power De
livered b
y
Source
Real Power
(
pu)
Reactive
Power
(
pu)
Appar
e
nt
Power
(
pu)
Distor
tion
Po
wer (p
u
)
I
1
0.
9991
0.
9991
0.
84
0
0.
84
0
I
I
1
0.
9995
0.
9995
0.
23
0
0.
23
0
.
Fig
u
re
7
.
So
urce vo
ltag
e
, Lo
ad
cu
rren
t, Filter cu
rren
t, So
urce curren
t
waveform
s
fo
r indu
ctio
n m
o
to
r
driv
e
lo
ad
at
25
%
rat
e
d
lo
ad
un
d
e
r
b
a
lan
c
ed
system
with
ad
ap
tive shun
t h
y
b
r
id
filter –d
yn
am
ic
con
d
ition
s
Fi
gu
re
8 E
x
per
i
m
e
nt
al
resul
t
s
- S
o
u
r
ce c
u
rre
nt
wa
ve
fo
rm
s for
t
h
e i
n
d
u
ct
i
o
n m
o
t
o
r
dri
v
e l
o
ad
u
n
d
er
bal
a
nced
system
with adaptive s
h
unt
hy
bri
d
filter
–
dy
nam
i
c c
onditions; (a
)
Without
filter,
(b)
With ada
p
tive s
h
unt
h
ybrid
filter
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
PED
S
I
S
SN
:
208
8-8
6
9
4
Dynamic Po
wer Qu
a
lity Comp
ensa
t
o
r with
a
n
Ad
ap
tive Shu
n
t
Hyb
r
id Filter (S
ind
h
u
M R)
51
5
Tabl
e
5(a
)
.
Fu
ndam
e
nt
al
com
p
o
n
e
n
t
o
f
s
o
ur
ce cu
rre
nt
a
n
d
THD
i
n
so
u
r
ce
cu
rre
nt
at
t
h
e
poi
nt
o
f
c
o
m
m
o
n
co
up
ling
with
v
a
riab
le
sp
eed in
du
ctio
n
m
o
to
r d
r
iv
e
af
ter the in
stallatio
n
of Ad
ap
tiv
e
Shun
t Hy
b
r
i
d
Filter
–
Dy
nam
i
c condi
t
i
ons
Cy
cle
Fundam
e
ntal
co
m
ponent o
f
source current
(p.u.)
THD in
Sour
ce
Cu
rren
t
(%
)
% of
5
th
har
m
onic
% of
7
th
har
m
onic
% of
11
th
harm
onic
% of
13
th
harm
onic
Phase a
Phase b
Phase c
Phase a
Phase
A
Phase
A
Phase
a
Phase
A
I
0.
84
0.
84
0.
84
3.
15
2.
2
1.
6
0.
7
0.
4
I
I
0.
72
0.
72
0.
72
3.
24
2.
31
1.
59
0.
6
0.
31
I
I
I
0.
60
0.
60
0.
60
3.
56
2.
43
1.
87
0.
78
0.
35
I
V
0.
48
0.
48
0.
48
3.
32
2.
36
1.
63
0.
63
0.
32
V 0.
36
0.
36
0.
36
3.
26
2.
34
1.
60
0.
58
0.
29
VI 0.
23
0.
23
0.
23
3.
19
2.
5
1.
4
1.
2
0.
9
Tabl
e
5(
b)
. Est
i
m
a
t
e
of P
o
we
r
sy
st
em
param
e
t
e
rs i
n
ea
ch
c
y
cl
e un
der
dy
n
a
m
i
c l
o
ad c
o
n
d
i
t
i
ons
wi
t
h
t
h
e
in
stallatio
n
of A
d
ap
tiv
e Shunt
H
ybrid
Filter
Cy
cle
Displace
m
e
nt
power f
actor
Distor
tion power
f
actor
Sour
ce power
f
actor
Power De
livered b
y
Source
Real Power
(
pu)
Reactive
Power
(
pu)
Appar
e
nt
Power
(
pu)
Distor
tion
Po
wer (p
u
)
I
1
0.
9991
0.
9991
0.
84
0
0.
84
0
I
I
1
0.
998
0.
998
0.
72
0
0.
72
0
I
I
I
1
0.
997
0.
997
0.
60
0
0.
60
0
I
V
1
0.
997
0.
997
0.
48
0
0.
48
0
V 1
0.
998
0.
998
0.
36
0
0.
36
0
VI 1
0.
999
0.
999
0.
23
0
0.
23
0
Th
e sim
u
latio
n and
ex
p
e
rim
e
n
t
al resu
lts related
to
t
h
e ad
ap
tiv
e sh
un
t
h
y
b
r
i
d
filter illu
strated
:
a)
A sing
le co
n
t
ro
ller is used
for con
t
ro
llin
g
bo
th
ad
ap
tiv
e sh
un
t
p
a
ssiv
e
fi
lter an
d ad
ap
tiv
e sh
un
t
activ
e filter
b)
Th
e
d
i
g
ital ANN
b
a
sed
co
n
t
roller is flex
i
b
le
an
d easy to imp
l
em
en
t in
large qu
an
tities.
c)
In
th
e case of ad
ap
tiv
e shu
n
t
hyb
rid
filter, th
e k
V
A d
e
liv
ered
b
y
sou
r
ce for th
e lo
ad
is
m
u
ch
less
co
m
p
ared
with shu
n
t
ad
ap
tiv
e p
a
ssi
v
e
filter or
ANN co
n
t
ro
ll
er
b
a
sed sh
un
t
activ
e filter.
d)
The fundam
ental
reactiv
e
power drawn from
source is
much
r
e
du
ced an
d h
e
n
c
e source po
wer
factor is im
proved.
e)
Th
e p
e
rform
a
n
ce o
f
th
e
ANN con
t
ro
ller is satisfacto
r
y with
b
a
lan
c
ed
/ un
b
a
lan
ced
sou
r
ce and
balance
d
/unbalanced nonlinea
r loa
d
unde
r st
eady state and
dynam
i
c conditions.
5
.
CONC
LUSION
Ad
j
u
st
abl
e
s
p
e
e
d i
n
duct
i
on
m
o
t
o
r dri
v
es
cause l
a
r
g
e a
m
ount
s of
cu
r
r
ent
a
nd
v
o
l
t
a
ge ha
rm
oni
c
di
st
ort
i
o
ns
an
d
react
i
v
e
p
o
w
e
r a
b
so
r
p
t
i
o
n
at
t
h
e
p
o
i
n
t
of
c
o
m
m
on co
upl
i
n
g
.
T
h
e
p
r
o
p
a
g
at
i
o
n
of
di
st
o
r
t
i
ons
th
ro
ugh
ou
t th
e p
o
wer system affects lifetime o
f
all
o
t
her
po
we
r sy
st
em
equi
pm
ent
.
Va
ri
o
u
s p
o
w
er
q
u
al
i
t
y
im
pro
v
em
ent
schem
e
s were s
u
g
g
est
e
d
by
di
ffe
rent
a
u
t
h
or
s
.
T
h
i
s
pa
pe
r s
h
ows
an
A
N
N
b
a
sed a
d
a
p
t
i
v
e
shu
n
t
h
ybrid
filter
fo
r
po
wer qu
ality en
h
a
n
cemen
t in
a
v
a
ri
ab
le sp
eed
d
r
iv
e system
. T
h
e sim
u
latio
n
and
ex
p
e
rim
e
n
t
al resu
lts sho
w
effectiv
e
p
e
rforman
ce o
f
ad
ap
tiv
e sh
un
t hyb
rid
filter und
er stead
y
state an
d
tran
sien
t cond
itio
n
s
.
ACKNOWLE
DGE
M
ENT
The a
u
t
h
or
s w
i
sh t
o
t
h
an
k
A
m
rit
a
Vi
shwa
Vi
dy
apeet
ham
,
C
o
i
m
bat
o
re a
n
d
De
part
m
e
nt
of
Sci
e
nc
e
an
d Tech
no
logy,
N
e
w D
e
l
h
i
f
o
r
th
ei
r
f
i
n
a
ncial su
ppo
r
t
f
o
r
car
r
y
ing
o
u
t
t
h
is w
o
rk
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-86
94
I
J
PED
S
Vo
l.
4
,
No
.
4
,
D
ecem
b
er
2
014
:
50
8 – 516
51
6
REFERE
NC
ES
[1]
Bim
a
l K Bose.
Power Ele
c
troni
cs and Motor
Drives: Advan
ces
and Tr
ends. A
c
ademic pr
ess, 20
10, Ch 7
.
Gopal
K
Dubey
.
Fundamentals of
Electric Drives
. Narosa
Publishing House, 2013
; Ch
6.
[2]
Ga
ic
ea
nu M,
Rosu E,
Pa
duraru R,
Munteanu T.
Vector-controlled optimal drive
system for the in
duction motor
.
4t
h
Interna
tiona
l S
y
m
posium
on Ele
c
tri
cal
and
El
ec
t
r
onics Eng
i
ne
eri
ng (IS
EEE)
, 201
3; 1 -
6
.
[3]
Zakar
i
a
L,
Barr
a
K.
Pred
ictive D
i
rect Torque an
d Flu
x
Control
of
an indu
ction
motor drive
fed
by a Direct Ma
trix
Converter with
reacti
ve power minimization
. 1
0
th IEEE Inte
rn
ation
a
l Confere
n
ce on Networking, S
e
nsing and
Control (ICNSC), 2013; 34 -
39.
[4]
Rajesh,
Te
java
th
u, Panda AK, K
u
m
a
r, S Shiva,
Bonala
, Sath
ya
m
.
High performance direct tor
que and flux
con
t
rol
of indu
ction
motor drive using
fuzzy
logic based
speed
controller Circuits
. Inter
n
ation
a
l Conf
erence on
Power and
Computing Tech
nologies (ICCP
CT). 2013
; 213
-
218.
[5]
Ewald Fuchs and Mohammad
Masoum.
Power Quality
in P
o
wer Systems a
nd Ele
c
trica
l
Machines
. Elseveir
Academic Press, 2008; Ch
.3.
[6]
HL J
ou. P
e
rfor
m
ance com
p
aris
on of the
thre
e
phase activ
e po
wer filter
algorithms.
IEEE
Pro
c
. Gen
e
r. T
r
ansm.
Distrib.
1995: 1
42(6): 646-652
.
[7]
Ginnes K John,
Sindhu MR, Ma
njula G Nair
, “DSP Based digital controller for sh
unt activ
e filter to improve powe
r
quality
.
In
ternational Journal of
Re
cent Trends
in
Electrical
and Electronics Eng
i
n
eering
. 2009; 2(
7): 92-94.
[8]
SMR Rafiei, Reza Ghazi
, Ham
i
d A Toli
y
a
t. IE
EE-519-Base
d
Real-
T
im
e and Optim
al Control
of Active Filters
Under Non sinusoidal Lin
e
Volt
ages
Using Neural Networks,” I
EEE Tr
ansac
tio
ns on P
o
wer Del
i
ver
y
, Vol
.
17,
No.
3, July
2002
,pp.
815-821.
[9]
Patricio Salm
ero
n
, Jesus R Vazquez. Prac
tic
al Design of
a Three
P
h
as
e Active Power Line Conditioner con
t
roll
e
d
b
y
Artifi
c
ia
l Ne
ural Ne
tworks.
I
EEE Trans. on Power Delivery
. 2
0
(2): 2005; 1037
-1044.
[10]
Zhenfeng Xiao
, Yilong Chen, Xiangti
a
n Deng. A Shunt
Ac
tive Power Filter with enhanced d
y
n
a
m
i
c perform
ance
using Dual-Rep
i
titiv
e
control
l
er
and pred
ict
i
ve
c
o
m
p
ensati
on.
Int
e
rnat
ional
Journal of
Power E
l
e
c
t
ronics
and Driv
e
S
y
stems. 2013;
3: 209-217.
[11]
Mridula Jha, SP Dubey
,
Neur
o-Fuzzy
Based
Controller for
a Thr
ee Phas
e Four Wire Shunt Activ
e Filter
,
International Jou
r
nal of
Power Electron
i
cs
and Dr
ive S
y
s
t
ems. 201
1; 7; 11(2):148-
155.
Evaluation Warning : The document was created with Spire.PDF for Python.