Internati
o
nal
Journal of P
o
wer Elect
roni
cs an
d
Drive
S
y
ste
m
(I
JPE
D
S)
V
o
l.
6, N
o
. 1
,
Mar
c
h
20
15
,
pp
. 17
8
~
18
4
I
S
SN
: 208
8-8
6
9
4
1
78
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
Corroboration of
Norm
alized L
e
as
t Mean S
quare B
a
s
e
d
Adaptive Selective Current Harm
onic Elimination in Voltage
Source I
n
vert
er using DS
P Proces
sor
P Av
irajamanjula*
,
P Pa
laniv
e
l*
*
*Departm
ent
of
EEE
, P
e
r
i
yar M
a
nia
m
m
a
i
Univer
sit
y
, Tam
iln
adu,
India
**
Department o
f
EEE, M.A.M
Co
llege of
Engineering
,
Anna U
n
iversity
, Tamilnadu, Ind
i
a
Article Info
A
B
STRAC
T
Article histo
r
y:
Received Oct 30, 2014
Rev
i
sed
D
ec 23
, 20
14
Accepte
d
Ja
n 18, 2015
A direct Selective curren
t
har
m
onic
elimination pulse width
modulation
techn
i
que
is pr
oposed for indu
ction motor
drive fed
from voltag
e
source
inverter. The developed
ad
aptiv
e filter
i
ng
algori
t
hm
for the
s
e
l
e
ctiv
e cur
r
en
t
harmonic elimination in
a thr
e
e phase
Voltag
e
S
ource Invert
er
is
a dire
ct
method to improve the lin
e cur
r
ent quality
of the Voltage Sour
ce Inver
t
er
base drive
at
an
y lo
ad cond
ition
.
The
self-ad
a
pt
i
v
e algor
ithm
em
plo
y
ed has
the
capab
ility
o
f
managing
the
time var
y
ing nature of
load
(cu
rrent). Th
e
proposed Normalized
Least Mean
Squares algorithm based scheme
elim
inat
es
the s
e
le
cted dom
inan
t harm
onics in load current usin
g only
the
knowledge of
th
e frequ
encies to
be e
liminated. The algorithm
is
simulated
using Matlab
/
Simulink tool for
a thr
ee-ph
ase
Voltage Source Inverter to
elim
inat
e th
e fi
fth and s
e
ven
t
h
harm
onics
. Th
e s
y
s
t
em
perfo
rm
ance is
anal
yz
ed bas
e
d
on the s
i
m
u
lation re
sults c
onside
r
i
ng total harmonic
distortion, magnitude of
elimin
ated
harmonics
and harmonic spectrum.
The
corroboration is
done
in th
e
designed
Voltage Source Inverter f
eedin
g
induction motor
using digital signa
l processor-TMS320L2812.The develop
e
d
algorithm
is tr
an
sferred to
dig
ital signal processor using VisSim
TM
software.
Keyword:
Cu
rren
t h
a
rm
o
n
i
c
elimin
atio
n
DSP Proecss
o
r
Least Mean
Square
algoritha
m
VisSim
m
o
d
e
l
Vol
t
a
ge
s
o
u
r
ce
i
nve
rt
er
Copyright ©
201
5 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
:
P Av
iraj
am
an
ju
la,
Depa
rtem
ent of Electrical a
n
d
El
ect
ro
ni
cs E
n
gi
nee
r
i
n
g,
Periy
a
r Man
i
am
mai University
,
Valla
m
,
Th
an
jav
u
r-61
3 403
,
Tamiln
ad
u
,
Ind
i
a.
Em
a
il: res
m
an
j
u
@g
m
a
il.co
m
1.
INTRODUCTION
I
ndu
ctio
n m
o
t
o
r
fo
r
m
a
n
y
year
s
h
a
s
been re
garde
d
as t
h
e
workhorse
in
i
n
d
u
st
ri
al
a
ppl
i
cat
i
ons.
I
n
t
h
e
last few d
ecades, th
e indu
ctio
n m
o
to
r
h
a
s
ev
o
l
v
e
d fro
m
bei
n
g a
co
nst
a
nt
s
p
eed
m
o
t
o
r t
o
a
vari
a
b
l
e
spee
d
,
v
a
riab
le to
rque m
ach
in
e. Its evo
l
u
tion
was ch
alleng
ed
by th
e easin
ess o
f
con
t
ro
llin
g a DC m
o
to
r
at lo
w
po
we
r ap
pl
i
cat
i
ons
.
Whe
n
ap
pl
i
cat
i
ons
req
u
i
red l
a
r
g
e am
ount
s
of
p
o
w
er
and
t
o
r
q
ue, t
h
e i
n
d
u
ct
i
on
m
o
t
o
r
becam
e
m
o
re
efficient t
o
use
.
Recent
adva
nce
m
ents in
power electronic
s
has
pave
d t
h
e way t
o
provide t
h
e
v
a
r
i
ab
le vo
ltage, v
a
r
i
ab
le fr
eq
u
e
n
c
y dr
iv
es (
VVV
F), th
e
u
s
e
o
f
an
inductio
n
m
o
to
r
has in
cr
eased
[1
]-
[4
].
Th
ree ph
ase dc/ac v
o
ltage so
urce i
n
v
e
rter (VSI) is
u
s
ed
ex
ten
s
i
v
ely in
ind
u
c
tion
m
o
to
r d
r
i
v
es
an
d the
cont
rol
l
a
bl
e
fr
eque
ncy
a
nd a
c
v
o
l
t
a
ge m
a
gni
t
ude
s are
o
b
t
a
i
n
ed em
pl
oy
i
n
g
va
ri
o
u
s
pul
se wi
dt
h m
odu
l
a
t
i
o
n
(
P
W
M
)
str
a
tegies [
5
]-[9
].
The PW
M
th
eo
r
y
w
a
s adv
a
n
ced in
th
e
1
960
s.
Th
e
po
pu
lar mo
du
latio
n techn
i
qu
e
u
s
ed
in
co
mmu
n
i
cation
field was broug
h
t
in
to
th
e a
p
p
licatio
n
of
pow
er co
nv
er
ters. Th
e Sinu
so
id
al
PW
M
(SP
W
M
)
t
ech
n
i
que
has
bee
n
appl
i
e
d
i
n
i
n
v
e
rsi
o
n si
nce
1
9
7
0
s,
m
a
ki
ng
t
h
e pe
rf
o
r
m
a
nce o
f
i
n
ve
rt
er
great
l
y
im
pro
v
ed a
nd
bei
n
g wi
del
y
sprea
d
. T
h
e p
r
i
n
ci
pl
e o
f
P
W
M
cont
r
o
l
i
s
t
o
co
nt
r
o
l
t
h
e o
n
-
o
f
f
st
at
es of
po
we
r
el
ect
roni
c
swi
t
ches i
n
or
der
t
o
obt
ai
n
a se
ri
es o
f
pul
se
wa
ves
wi
t
h
sam
e
am
pl
i
t
ude
but
di
ffe
re
nt
wi
dt
h
.
Whe
n
an
in
du
ction
m
o
to
r is fed
by su
ch
a PWM co
n
t
ro
lled
VSI, th
e lin
e cu
rren
t b
e
co
m
e
s a d
i
sto
r
ted
wav
e
fo
rm
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
PED
S
I
S
SN
:
208
8-8
6
9
4
C
o
rro
b
o
r
a
t
i
o
n
of
N
o
r
m
al
i
z
ed
Least
Me
a
n
Sq
uar
e B
a
se
d A
d
apt
i
ve
Sel
ect
i
ve C
u
rre
nt
…
(
P
Avi
r
aj
a
m
anj
ul
a)
17
9
[1
0]
. T
h
e
r
ef
or
e t
i
m
e
-harm
oni
cs i
n
fl
ue
nce
t
h
e m
a
gne
to-m
otive force
and i
n
crease
the
harm
oni
cs wit
h
m
i
sm
at
chi
ng
o
f
ai
r-
ga
p pe
rm
eance [
1
1]
. O
w
i
n
g t
o
radi
al
m
a
gnet
i
c
fo
rce
s
cause
d by
t
h
e harm
oni
cs, t
h
e co
re
an
d
ro
to
r system ex
p
e
rien
ce l
a
rg
e
v
i
bration
s
. Th
e
b
a
sic PWM strateg
y
help
s in
filterin
g
th
e h
a
rm
o
n
i
cs easily.
Whi
l
e
an i
nge
ni
o
u
s P
W
M
st
rat
e
gy
can
pre
v
ent
ob
ject
i
o
n
a
bl
e ha
rm
oni
cs fr
om
appeari
ng i
n
l
i
n
e cu
rr
ent
.
T
h
e
unprecede
n
ted
growt
h
of power electronics i
s
m
a
inly due
to de
velopm
ent in control
and digital platform [12].
Th
at is p
e
rfo
r
man
ce i
m
p
r
ove
m
e
n
t
in
po
wer conv
erters are
m
a
in
ly in
co
in
in
g
an
im
p
r
ov
ed
PW
M t
h
eo
ry and
success
f
ully im
ple
m
enting in a digita
l platform
. The im
ple
m
entation is su
pporte
d
by a host
of sim
u
lation a
nd
interface s
o
ftware tools and
proces
sors. Matlab bac
k
e
d
SIMULINK t
o
ps
the powe
r elec
tronics sim
u
lation in
ed
u
cation
a
l institu
tio
n
s
[13
]
, [1
4
]
.
VisSim
is also
in
t
h
is
influ
e
n
tial gro
u
p
.
Th
ere
h
a
v
e
been
q
u
ite a
nu
m
b
er
of success
f
ul work
on sim
u
lating el
ectrical
machines, powe
r syste
m
s
a
nd power electronic networks by
u
s
ing
th
e
soft
ware m
e
n
tio
n
e
d
abo
v
e
,
surpri
sin
g
l
y t
h
e
VisSi
m
h
a
d littl
e atten
tio
n [1
5
]
.
Th
e
po
ten
tiality o
f
faster conv
erg
i
ng
of th
e
n
o
rm
alized
least m
ean
sq
u
a
re (NLMS) algorith
m
o
v
e
r
the LMS al
gorith
m
has
been
indicated
[1
6
]-[
17
].
A
l
so
a ver
y
sim
p
le
m
o
d
e
l fo
r th
e in
pu
t sign
al
v
ector
s th
at
great
l
y
sim
p
l
i
f
i
e
s anal
y
s
i
s
of t
h
e co
n
v
er
g
e
nce be
ha
vi
o
u
r
o
f
t
h
e LM
S
and
NLM
S
al
go
ri
t
h
m
s
has bee
n
pr
o
pose
d
.
A g
e
neral
i
zed
no
r
m
al
i
zed gradi
e
nt
descent
(
G
NG
D) al
g
o
r
i
t
h
m
for l
i
n
ear fi
ni
t
e
-i
m
pul
se respo
n
se
(FIR
) ad
ap
tiv
e filters h
a
s b
e
en
in
trodu
ced
.
Th
e GNGD
rep
r
esen
ts an
exten
s
ion
o
f
th
e
NLMS algo
rit
h
m
b
y
mean
s of an
ad
d
ition
a
l
g
r
ad
i
e
n
t
ad
ap
tiv
e term
in
th
e
d
e
no
m
i
n
a
to
r of t
h
e learn
i
n
g
rate of
NLMS [1
8]. An
al
go
ri
t
h
m
has been s
h
ow
n f
o
r hi
g
h
l
y
n
o
n
-
st
at
i
onary
i
n
t
e
r
f
e
rence si
gnal
s
,
whe
r
e p
r
evi
o
us g
r
a
d
i
e
nt
-a
d
a
pt
i
v
e
learn
i
ng
rate alg
o
rith
m
s
fail. Th
e
propo
sed in
terfere
nce-norm
alized least
m
ean
sq
uare
(I
NLM
S
) alg
o
rithm
extends the gradient-a
da
ptive learning
rate
approac
h
to the case where
th
e si
gnal
s
ar
e no
n-st
at
i
o
nar
y
[19]
.
Perform
a
nce of noise cancell
e
r with DSP proces
sor,
TI
TMS3
20
C
6
713
u
s
ing
th
e LMS, N
L
MS and V
SS-
NLM
S
al
go
ri
t
h
m
s
has bee
n
i
nve
st
i
g
at
ed
[2
0]
. A
m
odi
fi
ed
NLM
S
al
go
ri
t
h
m
has bee
n
pr
op
ose
d
whi
c
h
has t
h
e
ability of handling inse
nsitive to the
tim
e
variations of the input dy
na
m
i
cs [21]. T
h
e MATLAB
has
been a
success
f
ul
numerical tool to sol
v
e
non-line
a
r diffe
re
ntial
equations
and
powe
r
electronic
sim
u
lations can be
carri
ed
o
u
t
by
devel
opi
ng s
y
st
em
di
fferen
t
i
a
l
equa
t
i
ons
.
Sim
u
l
a
t
i
on o
f
p
u
l
s
e wi
dt
h
m
odul
at
i
o
n (
P
W
M
)
i
nve
rt
ers
usi
n
g
M
A
TL
AB
ha
s bee
n
d
o
n
e b
a
sed
on
di
f
f
er
ent
i
a
l
equat
i
o
n
s
usi
n
g
fu
nct
i
o
n “o
de
23
” [
2
2
]
. Th
e
m
a
xim
u
m
po
wer
p
o
int trac
kin
g
(M
PPT
) f
o
r t
h
e m
a
them
a
t
i
cal
l
y
m
odeled s
qui
rrel
-
ca
g
e
i
n
d
u
ct
i
o
n
ge
nerat
o
r
(SC
I
G)
wi
n
d
po
we
r ge
ner
a
t
i
on
sy
st
em
has been
i
m
pl
ement
e
d i
n
Vi
sSi
m
wi
t
h
t
h
e d
o
ubl
e l
o
o
p
of c
ont
ro
l
syste
m
[23].
Th
e con
t
ribu
tio
n of earlier research in
selectiv
e h
a
rm
o
n
i
c eli
m
in
atio
n
is m
a
in
ly o
n
vo
ltage h
a
rm
o
n
i
c
el
im
i
n
at
i
on
w
h
i
l
e
t
h
e
per
f
o
r
m
a
nce en
hanc
em
ent
of
d
r
i
v
e
de
pen
d
s l
a
rge
l
y
/
di
rect
l
y
on
cur
r
ent
ha
rm
oni
cs.
A
adapt
i
v
e sel
ect
i
v
e cur
r
e
n
t
har
m
oni
c el
im
i
n
ati
on p
u
l
s
e wi
dt
h m
odul
at
i
on i
s
devel
ope
d an
d t
e
st
ed f
o
r i
n
d
u
ct
i
o
n
m
o
to
r d
r
i
v
e fed
fro
m
v
o
ltag
e
sou
r
ce inv
e
rter (VSI). Th
e
dev
e
lop
e
d ad
ap
t
i
v
e
filtering
al
g
o
rith
m
req
u
i
res on
l
y
t
h
e val
u
es
of
fr
eque
nci
e
s t
o
b
e
el
im
i
n
at
ed. The c
o
r
r
o
b
o
rat
i
on
o
f
t
h
e
p
r
o
p
o
se
d sc
hem
e
i
s
do
ne i
n
t
h
e
p
r
ot
ot
y
p
e
VSI
fee
d
i
n
g i
n
duct
i
o
n m
o
t
o
r
usi
n
g
di
gi
t
a
l
si
gnal
pr
ocess
o
r
(D
SP)
TM
S3
20
L2
8
12
[2
4]
.
The ef
f
o
rt
i
n
wri
t
i
n
g
DSP
co
de is
largely
re
d
u
ced
t
h
r
o
ug
h
VisSim
TM
so
ftwa
re.
2.
VISS
IM
VISSIM m
ean
s v
i
su
al sim
u
latio
n
.
It is in
t
e
rfaci
n
g
s
o
ft
w
a
re bet
w
een
D
SP co
nt
r
o
l
l
e
r
and
pe
rs
onal
co
m
p
u
t
er. It was fo
und
in
1
989
and
d
e
velo
p
e
d
in
co
ll
ab
oration
with Un
ited
Tech
no
log
i
es. VisSi
m
is a
m
a
t
h
em
at
i
c
al
m
odel
i
ng e
n
v
i
ro
nm
ent
for
devel
opi
ng
no
n-l
i
n
ea
r
dy
na
m
i
c sy
st
em
sim
u
l
a
t
i
ons.
Vi
sSi
m
provides a
m
e
ans
by
whic
h
syste
m
s can be created wit
h
bloc
k
diagrams, c
o
nnected by wires, in a
way tha
t
would
be done
on a piece of
paper, but then processe
s the
m
a
the
m
a
tica
l
operations re
prese
n
ted
within the
bl
oc
k di
ag
ram
i
t
e
rat
i
v
el
y
over a t
i
m
e
range.
Vi
sSi
m
al
l
o
ws
t
h
e devel
ope
r t
o
st
ruct
ure
m
odel
s
hi
erarc
h
i
cal
l
y
whi
c
h rea
d
i
l
y
l
e
nd
s i
t
s
el
f t
o
creat
i
ng "t
op
do
wn" m
odel
s
,
u
s
i
ng
fu
nct
i
o
n b
l
ocks t
o
re
pres
ent
com
pone
nt
s an
d
su
bsystem
s
. T
h
ese can
th
en
b
e
broug
h
t
togeth
er to
pro
duce larg
er, m
o
re co
m
p
lex
m
o
d
e
ls. VisSim
p
r
ov
id
es
bl
oc
k "
p
ri
m
i
ti
ves" f
o
r
b
u
i
l
d
i
n
g sy
st
em
s.
Vi
sSi
m
i
s
a sim
u
l
a
t
i
on en
vi
r
onm
ent
w
h
i
c
h
ha
ndl
es m
i
xed c
ont
i
n
u
ous
-
and
di
sc
ret
e
-t
i
m
e el
em
ent
s
w
i
t
h
a
g
r
a
p
h
i
c
a
l
u
s
e
r
i
n
t
e
r
f
a
c
e
.
W
i
t
h
t
h
e
a
i
d
o
f
M
S
V
i
s
u
a
l
B
a
s
i
c
p
r
o
g
r
a
m
,
a
m
e
n
u
i
s
p
r
e
p
a
r
e
d
t
o
cl
assi
fy
t
h
e
po
wer
el
ect
ro
ni
c
net
w
o
r
k
s
. T
h
i
s
m
e
nu i
s
di
sp
layed
on
t
h
e scr
e
en
; t
h
us user
s
can easily
select the type
of the power el
ectro
nic system
.
VisSim
is a
n
easy-to-use
, yet p
o
w
erfu
l so
lu
tion
fo
r accu
r
ately
m
odel
i
ng a
n
d
sim
u
l
a
t
i
ng m
o
t
i
on a
n
d m
o
t
o
r c
ont
rol
sy
st
em
s. Vi
sSim
/
M
ot
i
o
n
c
onsi
s
t
s
of
a c
o
m
p
rehe
nsi
v
e
m
o
t
i
o
n
con
t
ro
l
b
l
o
c
k
lib
rary, wh
ich
in
clud
es
m
o
to
rs, am
p
lifiers, filters, co
n
t
ro
llers,
lo
ad
s, sen
s
o
r
s, sou
r
ces,
and t
r
a
n
s
f
o
r
m
s
. The de
si
g
n
o
f
a sy
st
em
i
nvol
ves i
n
si
m
p
l
y
sel
ect
i
ng an
d t
h
en co
nn
ect
i
ng
sy
st
em
co
m
ponent
s
.
Lat
e
r,
wi
t
h
t
h
e
p
u
s
h
of
a
b
u
t
t
o
n
,
t
h
e si
m
u
l
a
ti
on
can
be
r
u
n
and
t
h
e
sy
st
em
be
ha
vi
or
m
a
y
be
n
o
t
i
ced.
Th
ere i
s
anot
her a
d
vant
age
of
havi
n
g
dy
nam
i
c sim
u
lat
i
on s
o
ft
ware
t
h
at
t
h
e real
t
i
m
e cont
r
o
l
can
be ac
hi
eve
d
t
h
r
o
ug
h
the interface hardware without any low-le
vel language
. In Figure 1 the VisSim
m
odel of the three
phase
in
du
ctio
n
m
o
to
r d
r
iv
e
is p
r
esen
ted
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-86
94
I
J
PED
S
Vo
l. 6
,
No
. 1
,
Mar
c
h
2
015
:
17
8
–
18
4
18
0
Fi
gu
re
1.
S
p
ee
d C
ont
r
o
l
of
I
n
duct
i
o
n M
o
t
o
r
3.
PROP
OSE
D
NLMS
BA
SED AD
APTI
V
E
SHE
As s
h
ow
n
i
n
t
h
e a
b
ove
Fi
g
u
r
e 2
t
h
e i
n
vert
e
r
had
P
I
c
ont
ro
l
l
e
r U
_
re
g
f
o
r
dc
b
u
s
vol
t
a
ge
co
nt
r
o
l
a
n
d
t
w
o P
I
re
g
u
l
a
t
o
rs
I
q1 a
n
d I
d
1 i
m
pl
em
ent
e
d i
n
sy
nc
hr
on
o
u
s re
fere
nce
fr
am
e for c
u
r
r
e
n
t
cont
r
o
l
.
R
e
fe
rence
angl
e
fo
r ge
ne
r
a
t
i
on o
f
si
ne a
nd c
o
si
n
e
f
unc
t
i
ons wi
t
h
fre
q
u
ency
of
fu
n
d
a
m
ent
a
l
co
m
p
o
n
ent
a
nd
fre
q
u
e
nci
e
s
of fi
ft
h an
d sev
e
nt
h ha
rm
oni
cs i
s
creat
ed by
a phase l
o
ok l
o
op (
P
LL
) bl
oc
k. Si
ne an
d cos
i
ne com
pone
nt
s wi
t
h
fund
am
en
tal freq
u
e
n
c
y are ph
ase lo
ck
ed
w
i
th
u
tility
v
o
ltag
e
and
are u
s
ed
for statio
n
a
ry to
syn
c
h
r
onou
s (an
d
vi
ce versa
)
re
f
e
rence f
r
am
es
t
r
ans
f
o
r
m
a
ti
on
s. Si
ne an
d co
si
ne com
pone
n
t
s wi
t
h
fi
ve an
d seve
n t
i
m
e
s
hi
g
h
er
fre
que
nci
e
s a
r
e
use
d
fo
r sel
e
c
t
i
v
e ha
rm
oni
c el
im
i
n
at
i
on.
Sa
m
p
l
e
current
s
Ia,I
b,
Ic
fr
om
the st
at
i
o
nary
(
a
,b,
c
)
referen
ce
were tran
sform
e
d
in
to
two
phase q
,
d
statio
nary
re
fere
nce
fram
e
(bloc
k
3/2) a
n
d the
n
into
syn
c
hr
ono
us fra
m
e
I
q
,I
d (b
lo
ck
s/e) [25
]
-
[
2
6
]
.
Th
e co
nv
en
tion
a
l p
a
rt o
f
con
t
ro
l wo
rk
s as fo
llows: vo
ltag
e
regu
lato
r
U_
reg
d
e
p
e
nd
ing
on
d
c
bu
s
voltage e
r
ror c
r
eates an active curre
nt re
ference Iq*. Fo
r
unity
po
we
r fa
ctor reactive c
u
r
r
ent re
fere
nc
e Id
* is
kept
zer
o. PI
cur
r
ent
re
gul
at
ors m
a
i
n
tain a
n
avera
g
e val
u
e of fee
d
bac
k
curre
n
ts Iqe
and Ide equal
to the
avera
g
e val
u
e
s
of c
o
r
r
esp
o
ndi
ng
refe
renc
es. O
u
t
p
ut
s o
f
cur
r
e
n
t
reg
u
l
at
ors are t
r
a
n
sf
orm
e
d fi
rst
from
sy
nch
r
o
n
o
u
s
to statio
nary
re
fere
nce
fram
e
(bl
o
ck
e/s) a
n
d the
n
fr
om
two
-
p
h
ase
(
q
,
d
)
to th
ree
p
h
ase
(a,
b
,c
)
sy
st
em
and wri
t
t
e
n i
n
t
o
P
W
M
cont
r
o
l
t
h
e i
n
vert
er
. The c
o
m
ponent
s co
nt
ri
b
u
t
e
d t
o
P
W
M
from
ASHE
bl
ock
s
will create v
o
ltag
e
at th
e ou
tpu
t
of th
e i
n
v
e
rter
w
ith
am
p
l
itu
d
e
s and p
h
a
se ang
l
es as n
e
ed
ed
t
o
can
cel
h
a
r
m
o
n
i
c
co
mp
on
en
ts
fro
m t
h
e
lo
ad
cur
r
e
n
t
s
.
Fig
u
r
e
2
.
Propo
ased
N
L
MS basedA
SHE
sche
m
e
f
o
r
VSI
Fi
gu
re
3.
The
s
t
eps i
n
v
o
l
v
e
d
i
n
t
h
e
p
r
op
oase
d S
H
E
-
PW
M
U_reg
Iq
1
Id
1
e
s
2
3
q,
d
a,b,c
2x
MF
_ASHE
P
L
L
P
W
M
e
s
u
a
u
b
u
c
i
a
i
b
i
c
R
,L
U
dc
U
dc
i
e
d
i
e
q
i
s
q
i
s
d
U
q
U
d
+
+
_
_
i
*
d
i
*
q
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
PED
S
I
S
SN
:
208
8-8
6
9
4
C
o
rro
b
o
r
a
t
i
o
n
of
N
o
r
m
al
i
z
ed
Least
Me
a
n
Sq
uar
e B
a
se
d A
d
apt
i
ve
Sel
ect
i
ve C
u
rre
nt
…
(
P
Avi
r
aj
a
m
anj
ul
a)
18
1
3.1.
Hardwar
e
Implementation
Fi
gu
re
4.
Lay
o
u
t
f
o
r
ha
r
d
wa
r
e
im
pl
em
ent
a
t
i
on
Fi
gu
re
5.
P
hot
og
ra
ph
o
f
e
xpe
ri
m
e
nt
al
set
up
Fig
u
r
e
6
.
N
L
M
S
algo
r
ith
m
in
V
I
SSI
M
w
i
ndow
Fi
gu
re
7.
W
e
i
g
ht
u
p
d
at
e i
n
N
L
M
S
al
g
o
ri
t
h
m
Fi
gu
re
8.
Er
ro
r
fr
om
NLM
S
a
l
go
ri
t
h
m
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-86
94
I
J
PED
S
Vo
l. 6
,
No
. 1
,
Mar
c
h
2
015
:
17
8
–
18
4
18
2
The l
a
y
o
u
t
of t
h
e
pr
o
pose
d
s
y
st
em
i
s
show
n i
n
Fi
g
u
r
e
4
.
The e
x
peri
m
e
nt
al
set
up
fo
r t
h
e Ha
rd
war
e
i
m
p
l
e
m
en
tatio
n
o
f
SHEPWM strateg
y
is
sh
own
in
Figure
5
.
It m
a
in
ly
co
n
s
ists o
f
an u
n
c
o
n
t
ro
lled
rectifier,
DC lin
k
filter, App
licatio
n
Sp
ecific In
tellig
en
t
Power Mod
u
l
e (ASIPM) an
d
Tex
a
s TMS3
20
LF281
2 DSP
Pro
cesso
r.
Gat
i
ng
pul
ses f
o
r t
h
e i
nve
rt
er s
w
i
t
c
hes are ge
ner
a
t
e
d by
DSC
cont
rol
l
e
r an
d 0
.
2
5
k
W
,
41
5
V
,
50
H
z
th
ree
ph
ase
Ind
u
c
tion
m
o
to
r
is u
s
ed
as lo
ad. Th
e
NLMS
based
ad
ap
tiv
e
alg
o
rith
m
is sch
e
m
a
t
i
zed
in
VISSIM
and t
h
e
n
d
o
w
nl
oa
ded t
o
pe
rso
n
al
com
put
er. The
devel
ope
d schem
a
t
i
c i
s
di
agram
m
ed i
n
Fi
g
u
re
6. The
represen
tativ
e
weigh
t
upd
ate
is p
r
esen
ted
in
Fig
u
re
7
while the error is i
ndi
cated in
Figure
8.
3.
2.
W
ave
f
o
r
m
s
a
nd C
o
mp
ari
s
o
n
Fi
gu
re
9.
DC
l
i
nk
v
o
l
t
a
ge a
n
d
l
i
n
e v
o
l
t
a
ge
(
ab
V
)
Fi
gu
re 9 s
h
o
w
s t
h
e dc l
i
nk vol
t
a
ge a
nd
o
u
t
p
ut
l
i
n
e-l
i
n
e
vol
t
a
ge al
on
g wi
t
h
t
h
e dc
l
i
nk v
o
l
t
a
ge.
Fig
u
re
1
0
d
e
tails ab
ou
t th
e R-Phase lin
e curren
t
wh
ile
NL
M
S
al
go
ri
t
h
m
i
s
i
n
pr
ocess
.
Fi
gu
re
11 s
h
o
w
s R
-
Phase line current whe
n
NL
MS algor
ithm
reache
d
optimum
point and
corr
es
ponding harm
onic spec
trum
is
illu
strated
in Fi
g
u
re
12
. Tab
l
e
1
sh
ortens th
e resu
lts
of
bo
th
si
m
u
latio
n
and
h
a
rdware
fo
r co
m
p
ariso
n
.
Fi
gu
re
1
0
.
O
u
t
put
c
u
rre
nt
w
h
i
l
e LM
S al
g
o
ri
t
h
m
i
s
i
n
pr
oces
s
Figure
11. R
-
phase
line c
u
rre
n
t whe
n
LMS
algorithm
reach
es optim
u
m
point
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
PED
S
I
S
SN
:
208
8-8
6
9
4
C
o
rro
b
o
r
a
t
i
o
n
of
N
o
r
m
al
i
z
ed
Least
Me
a
n
Sq
uar
e B
a
se
d A
d
apt
i
ve
Sel
ect
i
ve C
u
rre
nt
…
(
P
Avi
r
aj
a
m
anj
ul
a)
18
3
Figure
12.
R-phase line
curre
nt whe
n
NLM
S
algorithm
re
ach
es op
ti
m
u
m
po
in
t
Tabl
e
1. C
o
m
p
ari
s
o
n
of
si
m
u
lat
i
on a
n
d
ha
r
d
ware
res
u
l
t
s
Si
m
u
lation
Hardware
I
5
I
7
THD
I
5
I
7
THD
% of I
1
%
% of I
1
%
W
ithout ASHE
19.
82
14.
41
109.
06
22.
12
14.
81
113.
26
W
ith
ASHE
1.
29
1.
50
75.
86
2.
09
1.
77
77.
44
4.
CO
NCL
USI
O
N
The c
o
ncept
o
f
p
u
l
s
e
wi
dt
h
m
odul
at
i
on
ha
s bee
n
bo
rr
o
w
ed f
r
o
m
co
m
m
uni
cat
i
o
n e
ngi
neeri
n
g
an
d
i
n
v
o
l
v
e
d
i
n
p
o
w
er c
o
nve
rt
ers
,
pa
rt
i
c
ul
arl
y
i
n
v
o
l
t
a
ge
so
ur
ce i
nve
rt
ers.
A
ho
st
o
f
P
W
M
t
echni
q
u
es
ha
v
e
bee
n
devel
ope
d a
n
d
i
nvest
i
g
at
e
d
.
These t
e
c
hni
q
u
es
have
t
h
ei
r
speci
fi
c
ob
ject
i
v
e an
d
pri
n
ci
pl
e t
o
war
d
s sa
t
i
s
fy
i
n
g
t
h
ei
r ap
pl
i
cat
i
o
n
.
C
u
rre
nt
ha
rm
oni
c
elimination techniques are class of
PW
M techniques whic
h are
direct
way to enha
nce the
pe
rformance of
dri
v
es.
NLMS
al
g
o
rith
m
b
a
sed ad
ap
tiv
e
on
lin
e cu
rren
t h
a
rm
o
n
i
c
el
im
i
n
at
i
on t
e
chni
que
s i
s
p
r
op
ose
d
a
nd
si
m
u
l
a
t
e
d i
n
M
A
TL
AB software. It is e
v
idenced that sel
ected
harm
oni
cs c
u
r
r
ent
ha
rm
oni
cs (fi
ft
h a
n
d se
v
e
nt
h
)
a
r
e s
u
p
p
r
esse
d
bel
o
w
2%
of
f
u
ndam
e
nt
al
. T
h
e sy
st
em
i
s
im
pl
em
ent
e
d i
n
har
d
ware
usi
n
g
V
I
SS
IM
s
o
f
t
ware a
n
d
DSP
TI
TM
S3
2
0
L
2
81
2.
REFERE
NC
ES
[1]
Bos
e
BK
.
P
o
w
e
r
El
ec
tronics
V
a
riab
le
F
r
eque
nc
y D
r
ives
.
IEEE
P
r
e
s
s
. New Y
o
rk. 1997; 400-4
53.
[2]
AMHava, R Kerkman, TA Lipo. A High-Pe
rform
ance Generalized Discon
tinuous PWM
Algorithm.
IE
EE
Transactions on
Industry Applica
tions
. 1998
; 34(5
)
.
[3]
G Nara
y
a
n
a
n,
VT Ranganath
a
n
. Triang
le-
c
om
pa
rison approac
h
and space-ve
c
t
or m
odulation.
Journal of Indian
Institute
of
Sc
ien
ce.
2000; 80: 40
9-427.
[4]
F
G
Turnbull,
S
e
le
cted
harm
oni
c redu
ction
in s
t
ati
c
d
c
-a
c inv
e
rters
.
I
E
EE Transactions on C
o
mmunication a
nd
Electronics.
196
4; 83: 374-378.
[5]
HS Pa
t
e
l
,
R
G
Ho
f
t
.
Ge
ne
r
a
li
ze
d t
e
c
h
ni
q
u
e
s o
f
ha
r
m
o
n
i
c
el
i
m
ination and
voltage control in th
y
r
istor inverters: P
a
rt
II-Voltage con
t
r
o
l Techniqu
es.
I
EEE Transactio
ns on
Industry Applications
. 197
4; IA(10): 6
66-6
73.
[6]
HS Patel,
RG H
o
ft. Gen
e
ra
liz
ed
techn
i
ques of
ha
rm
onic el
im
in
ati
on and vo
ltag
e
c
ontrol in
th
yr
isto
r inver
t
ers:
Part
I-
Harm
onic el
im
ination
.
I
EEE Transactions on In
d
u
stry Applicatio
ns.
1973; IA(9): 310-317.
[7]
J Pitel, SN Talukdar
,
P Wo
od. Char
acterization
of progr
amme
d-waveform pulsewi
dth modulation.
IEEE
Transactions on
Indus
try Applica
tions.
1980
; IA(
16): 707-715.
[8]
Vassilios G Agelidis, Anast
a
sios I. Bal
ouktsis,
and Calum
Cossar. On Attaini
ng
t
h
e Multip
le Solu
tions of Select
iv
e
Harmonic Elimination PWM Three-
Lev
e
l Wa
veforms Through Function Minimization.
I
E
EE Tractions on
Industrial Electronics.
2008
; 55(
3): 996-1004.
[9]
Joachim Holtz.
Pulse width mo
dulation–A survey
.
IEEE Transactions on Industrial Electronics
, 1992; 39(5): 410-
420.
[10]
Hava AM, R
Kerkman, TA Lipo.
A High-Performance Generalized
Discontinuous PWM Algorithm.
IEEE
Transactions on
Industry applica
tions.
1998
; 34(5
)
: 1059-1071.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-86
94
I
J
PED
S
Vo
l. 6
,
No
. 1
,
Mar
c
h
2
015
:
17
8
–
18
4
18
4
[11]
C Risnidar, I
Daut, H S
y
afru
ddin, N
Hasim. Influen
ce of
Harmonics in
Laborator
y
due to
nonlinear
Lo
ads.
International Jo
urnal of
Power
Electronics and
Drive Systems
. 2
012; 2(2) 219
-22
4
.
[12]
S Jeevananth
an, R Nandhakumar and P Dananjay
an. Inv
e
rted
s
i
ne car
rier for f
undamental for
t
ification in PWM
inverters and
FPGA ba
sed implementations.
Serb
ian Journal of Electri
cal Engin
e
ering
(
S
JEE)
.
200
7; 4(2).
[13]
Logue D, Krei
n T Philip.
Simulation of Electric Machinery and Po
wer electronics Interfacing Using
MATLAB/SIMULINK
. Proceed
in
gs of COMPEL’2000. Bl
acksbur
g Virgini
a
, USA. 2000;
34-39.
[14]
VisSim User Gu
ide,
Visual Solu
t
ion Inc.
2005
.
[15]
V Vis
w
anathan,
S
J
eevanantha
n
. S
i
m
u
lation of Elec
tr
ic Driv
es using VisSim Software: A Stud
y
on Toolbo
x
Development.
P
r
oceedings of
Int
e
rnational Conf
e
r
ence on In
tel
lig
ent Design and
Analysis of
Engi
neering Produc
t
s
,
Systems and Co
mputation (
I
DA
PSC-10)
. Sri Kri
s
hna College of Engineering and
Technolog
y
,
Co
imbatore, INDIA
.
2010; 50
[16]
Dirk TM Slock
.
On th
e Conv
ergence B
e
havior
of th
e
LM
S
and
the
Norm
ali
zed
LM
S
Algorithm
s
.
IEEE
Transactions
on Signal Processin
g
. 1993; 41(9): 2
811-2825.
[17]
PE An, M Bro
w
n, CJ Harris. On the Conver
g
ence R
a
te
Per
f
ormance of th
e Norm
alized Least-
Mean-Square
Adaptation.
IEEE Transactions
on Neural N
e
tw
orks
. 1997; 8(5): 1211-1214.
[18]
Danilo P Mandic.
A Gene
ral
i
ze
d Norm
alized Gradien
t
Des
cent
Algorithm
.
IEEE Signal Proces
sing Letters
. 2004;
11(2): 115 -
118.
[19]
J
ean-M
arc Va
lin
, Iain
B Coll
ings
. Inte
rferen
ce-N
o
rm
alized
Le
as
t
M
ean S
quare Al
gorithm
.
IEEE S
i
gnal Processing
Letters
. 2007
; 14
(12): 988-991
.
[20]
Boo-Shik R
y
u, J
ae-K
y
un Lee, Jo
onwan Kim, Chae-Wook Le
e. The Performance
of an ad
ap
tive n
o
ise can
celler with
DSP processor.
Pr
oceed
ings
of
40th IE
EE Sou
th
-eas
ter
n
Sympos
ium on Sys
t
em T
h
eor
y
.
University
of New Orleans
New Orleans, LA, USA. 2008; 4
2
-45
[21]
Monia Turki-Hadj Alouane. A Square
Root Norm
alized
LMS Algorithm
for Adaptive Id
entif
ica
tion with No
n-
Stationar
y
Inputs.
Journal of Co
mmunications a
nd Networks.
20
07; 9(1): 18-27.
[22]
LK Wong Frank, HF Leung
, P
e
ter
KS Tam. F
a
st
Simulation
o
f
PWM Inverter
s using MATLAB.
Proceeding
s of
IEEE Internatio
nal Confer
ence on
Power Electr
onics
and Drive
Systems (
PEDS'99)
, Hong Kong. 1999.
[23]
Wu Dinghui, Li Yuanlong, Ji Zhicheng
. Modeling and MPPT C
ontrol of Squi
rrel-cage Inductio
n
Generator Wind
Power Generation Sy
ste
m
via VisSi
m
.
Proceedings of IEE
E
Internati
onal
Chinese Control and Decision
Conference (
CCDC 2009)
. 2009; 48-53
[24]
K Vinoth kuma
r
, S Suresh ku
mar, Kishore Redd
y
.
Impl
emen
tation of Scalar
Control Techn
i
que in SVPWM
Switched Three-Level Inverter
Fed I
nduction motor Using
DSP Controller.
International
Journal of Power
Electronics and
Drive Systems
. 2
011; 1(2): 83-93.
[25]
S
S
a
ngeetha, C
H
Venkates
h
, S
J
eevanan
than
. S
e
le
ctiv
e Current
Harm
onic Elim
i
n
ation in a Curr
ent Controll
ed DC-
AC Inverter Drive S
y
stem using LMS Algorithm.
Proceed
ings of International
co
nf
erenc
e
on Computer Applica
t
i
o
n
in El
ectri
cal En
gineering
Rec
e
n
t
Advan
ces (
C
ERA-2009)
. Indian Institute of Technolog
y
,
ROO
R
KEE, 19th
to 2
1
st.
2010. 333
[26]
T Suresh Padma
n
abhan, M Sudhakaran
,
S Jeevan
anthan
. Se
lectiv
e Current Harmonic Elim
in
ation
in an AC Voltage
Controller Driv
e Sy
stem using LMS Algorithm.
Pr
oceed
ings
of 46th Inter
nation
a
l Univer
s
iti
es
’ Power
Engine
er
ing
Conference (
U
PEC 2011)
. South
Westphalia University
of
Applie
d Scien
ces, Soest, German
y
.
201
1; 169.
BIOGRAP
HI
ES OF
AUTH
ORS
P.Avirajamanju
la obtain
e
d her B.E degree in El
ectronics and Communication En
gineer
ing in 1998
from Bharathidasan University
,
Trich
y
, India, M
.
Te
ch. d
e
gre
e
in
P
o
wer Elec
troni
cs
and Drives
in
2002 from
SASTRA University, Thanjavur, In
dia.
She is curr
ently
working as an Assistant
Profe
ssor a
t
Pe
riy
a
r Ma
nia
mma
i Unive
r
sity
, Th
anjavur
, India.
She has publish
e
d two r
e
sear
ch
papers in Intern
ation
a
l Journals. He has pres
en
ted two pap
e
rs in National C
onferences. Her
research
in
terest
s are
in FACTS,
m
u
ltileve
l
invert
ers and
H
y
br
id E
n
erg
y
S
y
s
t
em
s
P.Palaniv
e
l obtained his B.E d
e
gree
in Electr
ical
and Electro
nics Engin
eerin
g in 1998 fro
m
University
of M
a
dras, M.E degr
ee in Power
Elect
ronics and Driv
es in 2004 from Anna University,
Chennai, Ind
i
a
and Ph.D.
in 201
2 from SRM Univer
s
i
t
y
,
Chenn
a
i.
He is
current
l
y
working as
a
P
r
ofes
s
o
r at the
M
.
A.M
.
Coll
eg
e of Engin
eer
in
g,
Tiru
chir
appalli, Ind
i
a. He has published ten
research p
a
pers
in Interna
tion
a
l Journals. H
e
has presented five papers
in Internation
a
l
conferen
ces. His
research in
teres
t
s are in power
qualit
y,
FACTS, m
u
ltileve
l inver
t
ers and resonant
inverters.
Evaluation Warning : The document was created with Spire.PDF for Python.