In
tern
ation
a
l
Jou
rn
al of
Po
wer
E
l
e
c
tr
on
i
c
s an
d
D
r
ive
Syste
m
(
IJ
PE
D
S
)
Vol.
10,
N
o.
4
,
Decem
b
e
r
20
1
9
,
pp.
177
2~
17
80
I
S
SN
:
2088-
8694,
DOI: 10.
11591/
ijpeds.v10.i4
.pp1772-1780
1
772
Jou
rn
al h
o
me
page
: ht
t
p
:/
/ia
e
s
c
o
re.c
om
/
j
ou
rna
l
s
/in
dex
.
p
h
p
/IJPED
S
Sensorless control of PMSM w
ith fuzzy model reference
adaptive system
Lar
b
i M’hame
d, Adjir
R
oufaid
a, A
i
t
A
me
r
Mez
i
an
e
Naw
a
L
2
GEGI-L
abotato
r
y,
U
n
i
vers
it
y Ib
n-Kh
ald
o
u
n
of Ti
aret,
A
l
g
e
ria
Ar
ticle
Info
A
B
S
T
RAC
T
A
r
t
i
c
l
e hi
sto
r
y:
Rec
e
i
ve
d Ma
r 1,
201
9
Rev
i
sed
J
u
n
8
,
2
019
A
c
ce
p
t
e
d
Ju
l
8
,
2019
T
o
i
mp
rov
e
t
h
e
p
erfo
rm
ance
o
f
perm
an
ent-mag
n
et
s
yn
chro
n
o
u
s
m
ot
o
r
(
PMSM
)
d
r
ives
p
owered
b
y
t
h
e
v
o
l
t
age
inv
e
rter
w
it
h
PW
M
c
on
trol
,
a
s
e
ns
orles
s
c
o
n
t
r
ol
s
ch
eme
based
on
a
M
od
el
R
ef
eren
c
e
A
dap
t
i
v
e
S
ystem
(M
RA
S)
a
f
uzzy
l
og
ic
c
on
troller
(
FLC
)
bas
e
d
in
w
it
h
f
u
zzy
s
up
ervis
o
r
s
t
ru
cture.
T
he
m
ajo
r
d
raw
b
acks
of
t
h
e
c
on
vent
io
nal
MRAS
,
name
l
y
ch
att
e
ring
p
heno
m
e
na,
hig
h
-o
rder
h
arm
oni
cs
a
n
d
ex
tern
al
n
oi
se,
a
re
d
i
s
c
us
sed.
T
h
e
se
d
rawb
acks af
fect
t
h
e
e
s
timat
ed
s
peed
ac
c
uracy
of
the
MR
AS
and
r
e
duce
t
he
c
ont
rol
rel
i
ability
o
f
t
h
e
system.
To
e
l
i
mi
nat
e
t
h
ese
d
r
awb
acks,
a
n
FL
C
i
s
de
sign
e
d
a
n
d
i
n
t
e
g
ra
te
d
i
n
to
t
he
MR
AS
t
o
a
d
j
u
s
t
t
h
e
o
b
s
e
rver
g
ai
n
t
o
r
ed
uce
th
e
chat
terin
g
i
n
cl
o
s
e
d
l
o
o
p
sp
ee
d
and
c
l
o
s
e
d
l
o
o
p
cu
rrent
/to
rqu
e
.
Co
m
p
arati
v
e
s
i
m
u
l
a
tio
n
s
u
s
i
ng
t
h
e
pro
p
o
s
ed
Fuzz
y-
MR
A
S
an
d the con
v
ent
i
o
n
al
MRAS
a
re
p
erf
o
rmed
t
o
v
a
lidat
e
t
h
e
e
ff
ecti
v
en
ess of
t
he
p
r
op
os
ed
FLC
s
tru
c
t
u
re.
P
e
rf
o
r
m
a
nce
si
m
u
l
a
tions
o
f
t
h
e
o
v
erall
pro
pos
ed
Fu
zz
y
-
MR
A
S
b
ased
s
en
sorl
ess
cont
rol
schem
e
a
re
p
erf
o
rmed
t
o
v
e
rif
y
t
he
robust
ness
a
nd
c
ontrol
r
el
i
a
bilit
y
of
t
he
s
ys
t
e
m.
T
he
r
esul
t
s
s
how
t
hat
t
h
e
p
r
op
os
ed
Fuzzy MRA
S
h
a
s
s
atisfactory
perf
orm
a
n
ces
w
i
t
h
reductio
n
o
f
t
o
t
al
h
a
rm
on
ic
d
ist
o
rtio
n
g
e
nerated
in t
he
p
hase
c
urrent
s.
Key
w
ords:
PM
SM
P
u
l
s
e Wi
dth
Mo
dul
a
t
i
o
n
PWM
S
e
nsorless
s
pe
e
d
c
on
tro
l
M
o
d
e
l
ref
e
rence
adap
ti
ve
MRAS
F
u
z
z
y
S
uperv
isor C
o
n
tr
ol
Copyr
i
g
h
t © 20
1
9
Ins
t
itu
t
e
of A
d
van
ced E
n
g
i
n
e
er
in
g a
nd Sci
e
nce.
All r
i
ghts reserv
ed.
C
o
rres
pon
d
i
n
g
Au
t
h
or:
Larbi M’hame
d
,
L2GEG
I-Labo
t
at
ory,
U
n
i
v
er
sity I
b
n
-K
hal
d
o
u
n
of
T
i
are
t
,
14
00
0,
A
lger
ia
.
Em
ail
:
larb
i
_m
h@
ya
ho
o.fr
1.
INTRODUCTI
O
N
In
nume
r
o
u
s
a
p
p
lica
t
io
ns
t
o
d
a
y,
P
e
r
m
a
nent
M
ag
ne
t
S
ync
hr
ono
us
M
o
tor
(
PM
S
M
)
is
t
he
p
ref
e
r
r
ed
A
C
d
ri
ve
c
ompa
red to
t
he
o
t
h
e
r
s
beca
use
o
f
i
ts
s
pec
i
a
l
f
e
a
t
ures
,
f
o
r
ins
t
a
n
ce
,
hi
gh
p
o
w
e
r
de
n
s
it
y,
h
i
g
h tor
que
to
c
urre
n
t
r
at
io
a
nd
hi
gh
e
f
fic
i
e
n
c
y
.
H
e
nce
,
PMSM
h
as
b
een
w
i
d
ely
u
s
ed
i
n
v
a
riou
s
fie
l
ds
a
nd
a
pp
l
i
ca
t
i
o
n
s
suc
h
a
s
m
a
nufa
c
tur
i
n
g
e
q
u
i
pm
ent,
p
r
oduc
tion
m
ach
in
es,
tra
n
sp
ort
a
tio
ns,
e
t
c
[1].
A
s
an
i
m
por
tan
t
a
p
p
lic
at
io
n
of
PMS
M
,
t
h
e
mo
t
i
on
c
o
n
t
r
ol
r
e
q
ui
re
s
no
t
on
ly
t
h
e
a
cc
u
r
ate
k
n
o
w
l
e
dg
e
o
f
rotor
pos
i
tio
n
for
f
i
e
l
d
or
i
e
n
t
a
t
i
on
bu
t
als
o
t
he
i
n
f
orma
tio
n
of
r
otor
s
pe
e
d
f
or
c
lose
d-
loo
p
c
o
n
tro
l
;
thus,
pos
it
ion
tra
n
sd
u
c
ers
suc
h
a
s
o
p
t
i
ca
l
e
n
co
der
s
a
n
d
r
eso
l
vers
a
re
n
e
e
d
ed
t
o
be
i
nsta
ll
e
d
on
t
h
e
sh
aft
[2-3].
H
owe
v
er,
these
se
n
s
ors
a
r
e
expe
n
s
i
v
e
an
d
v
e
ry
s
e
n
siti
v
e
t
o
en
vi
ronmen
t
a
l
co
nst
r
ai
nt
s
su
ch
a
s
vib
r
a
t
i
o
n
a
n
d
t
e
m
p
e
r
a
t
u
r
e
[
4
]
.
I
n
o
r
d
e
r
t
o
o
v
e
r
c
o
m
e
the
s
e
pro
b
l
em
s,
i
nste
ad
o
f
us
i
ng
p
o
s
iti
on
se
n
s
ors,
a
s
ensor
l
e
s
s
c
o
n
t
r
ol
m
et
h
o
d
has bee
n
d
e
v
e
l
o
p
e
d
f
or
c
o
n
tr
ol
of
t
he
m
ot
or.
T
h
e
ba
s
i
c
pr
inc
i
ple
o
f
s
e
n
s
o
rle
ss
co
n
t
ro
l
is
t
o
d
ed
uce
t
h
e
ro
t
o
r
spee
d
a
n
d
p
o
sit
i
on
us
i
n
g
v
a
r
i
ous
in
form
atio
n
a
n
d
me
ans,
i
nc
lu
din
g
d
ire
c
t
c
al
cula
t
i
o
n
,
pa
ra
m
e
ter
ide
n
ti
fic
a
ti
o
n
,
co
nd
iti
o
n
e
stim
a
tio
n,
i
nd
i
r
ec
t
m
e
a
s
u
r
ing
a
n
d
so
o
n.
T
he
s
t
a
tor
c
u
r
r
ent
s
a
nd
v
o
l
t
a
ges
are
ge
ner
al
ly
u
s
e
d
t
o
c
a
l
cu
la
te
t
he
i
nform
a
ti
on
o
f
spee
d a
nd rot
o
r posi
tio
n
[5-6]
.
The
M
o
del
Re
fere
nce
A
d
a
p
ti
ve
S
yste
m
m
a
kes
use
o
f
t
he
r
e
d
un
da
nc
y
o
f
tw
o
m
a
c
h
ine
mode
l
s
o
f
di
ffere
n
t
s
t
ruc
t
ure
s
t
ha
t
e
s
tim
ate
t
h
e
sam
e
s
t
a
te
v
a
r
ia
b
l
e
(rot
or
s
pe
e
d
)
of
d
i
f
fere
nt
s
e
t
o
f
in
pu
t
v
a
riab
le
s.
T
he
e
s
tim
ator
t
ha
t
doe
s
n
o
t
i
n
v
o
l
ve
t
he
q
uan
t
it
y
to
b
e
e
s
t
i
ma
te
d
i
s
c
h
o
se
n
a
s
t
he
r
e
f
er
ence
m
o
d
e
l
,
a
n
d
the
ot
h
e
r
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
SS
N
:
2088-
86
94
S
e
ns
o
r
l
e
ss c
ont
ro
l
of
PMS
M
wi
t
h
fu
z
z
y
mo
de
l
r
e
f
e
ren
c
e
ada
pti
v
e
sy
st
e
m
(La
r
b
i
M’ha
med
)
1
773
e
s
tim
ator
m
ay
b
e
re
garde
d
a
s
t
h
e
ad
j
u
stab
le
m
ode
l
.
T
he
e
r
r
o
r
be
t
w
een
t
he
e
st
i
m
a
t
ed
q
uan
t
it
ie
s
obt
ai
ne
d
b
y
t
h
e
t
w
o
mo
d
e
l
s
i
s p
r
o
p
o
r
ti
on
al t
o
t
h
e
a
ngul
a
r
d
is
pl
ac
e
m
e
n
t
bet
w
een
t
he
tw
o
e
st
im
ate
d
f
l
u
x
vec
t
ors [6-7]
.
H
o
w
e
ve
r,
t
he
c
onve
nti
ona
l
MRAS
f
o
r
PM
SM
s
en
sorle
s
s
c
o
n
t
r
o
l
su
ffer
s
cha
tte
rin
g
p
rob
l
em
[
8-
9]
.
Th
e
p
r
esen
c
e
o
f
e
x
t
e
rn
al
d
istu
rb
an
ce
s
a
n
d
p
a
ra
me
t
e
rs
v
a
r
i
a
t
i
o
n
s
i
n
t
h
e
mo
to
r
li
mi
ts
t
h
e
d
y
n
a
m
i
c
pe
rform
a
nc
es of t
h
e
t
r
ad
i
t
i
o
n
a
l
vect
or
c
on
tr
ol m
eth
o
d
,
w
h
en c
on
ve
nt
io
na
l
PI
r
e
gulat
ors a
r
e
used.
S
o
t
he
a
im
o
f
the
prese
n
t
st
ud
y
is
t
o
d
e
ve
lo
p
a
sim
p
le
c
on
tro
l
s
t
ra
te
gy,
w
h
i
ch
e
xp
loi
t
s
t
h
e
a
dva
n
t
age
s
o
f
t
h
e
v
e
c
t
or
c
on
t
r
ol
s
t
r
ate
g
y
of
t
he
PM
SM
a
nd
o
v
e
rco
m
es
t
he
l
i
m
it
ati
o
ns
o
f
PI
r
e
g
ul
a
t
o
r
i
n
t
h
e
c
o
n
v
e
n
t
i
ona
l
M
R
A
S
obse
v
e
r
a
nd
in
c
urre
n
t
l
oo
ps.
S
o
,
the
c
o
mbi
n
a
t
i
on
o
f
c
l
a
ssic
a
l
PI
regu
la
tor
an
d
adap
t
i
ve
fuzz
y
s
u
perv
is
or
m
ake
s
i
t
p
o
ssi
bl
e
t
o
i
nc
rea
s
e
t
h
e
pre
c
is
ion
o
f
t
he
m
at
hem
a
ti
ca
l
al
g
o
r
it
hm
i
n
the
cl
assica
l
co
nt
ro
ll
e
r
s
wi
t
h
f
l
e
xi
bil
i
ty
and
si
m
pl
i
c
it
y
of
f
u
z
z
y
li
ngu
i
s
t
i
c
f
or
mal
i
sm
[
14].
A
n
a
da
p
t
i
v
e
fuzz
y
s
uper
v
i
s
o
r
c
on
tro
ller,
h
a
v
e
bee
n
p
r
o
p
o
se
d
for
t
h
e
PM
S
M
s
ens
o
rless
control
in
c
l
o
s
e
d
l
o
op est
i
ma
t
e
spee
d
a
nd
clo
s
ed
l
o
op
c
u
rre
nt/t
orq
u
e.
2.
M
O
D
E
L
I
N
G
O
F
TH
E
S
Y
ST
EM
The
s
y
n
c
hron
o
u
s
pe
rma
n
en
ts
m
agne
ts
m
a
c
h
i
ne
c
a
n
b
e
ela
bor
ate
d
b
y
c
arr
y
in
g
o
u
t
a
m
ode
lin
g
w
ith
in
the
me
an
in
g of
P
ark.
T
he ma
c
hi
ne
m
ode
l
in
t
he tur
ni
n
g
dys
pha
sic
r
e
f
e
r
en
ce
(
d
-
q
)
i
s w
r
itt
en [8-9]
:
I
I
ω
⎣
⎢
⎢
⎢
⎢
⎡
I
ωI
ωI
I
ω
L
L
I
I
ϕ
I
ω
⎦
⎥
⎥
⎥
⎥
⎤
⎣
⎢
⎢
⎡
0
0
00
⎦
⎥
⎥
⎤
V
V
1
The
elec
trom
agne
tic
tor
que
i
s
give
n b
y
:
(2)
3.
M
R
AS SENSORL
E
S
S SPEED CONTROL
F
i
gure
1
s
how
s
the
M
R
A
S
b
ase
d
s
pee
d
e
s
tim
atio
n
sc
hem
e
.
It
u
ses
t
h
e
out
put
s
of
t
wo
m
o
d
e
l
s
:
on
e
in
de
pe
n
d
e
n
t
o
f
r
ot
or
s
peed
(
Re
fe
re
nce
M
o
del)
a
n
d
t
he
o
t
h
er
d
epe
nde
n
t
on
ro
t
o
r
spe
e
d
(
A
d
j
u
s
t
a
b
le
M
o
d
e
l
),
to
f
or
m
a
n
e
rr
or
s
ign
a
l.
A
PI
c
on
tro
lle
r
is
u
se
d
i
n
t
he
a
d
a
pta
t
i
o
n
m
echa
n
i
s
m
for
co
nverg
e
n
c
e
i
n
t
h
e
sy
s
t
em [
11-1
2
-
13
].
F
i
gure
1.
S
t
r
ucture
M
RA
S
for
the
est
i
m
a
te
s
peed
The
sta
t
e
spac
e
d-
q axis sta
t
o
r
curr
ents of
PMSM
d
es
ig
n
e
d
a
s
re
fere
nce
mode
l is g
i
v
en
by:
0
0
0
(3)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N:
2088-8694
I
nt J P
ow
Ele
c &
Dri S
y
st
V
ol
.
10,
No. 4, D
ec 2019
:
1772 –
1
780
1
774
The
sta
t
e
spac
e
d-
q axis sta
t
o
r
curr
ents of
PMSM
d
es
ig
n
e
d
a
s
a
d
jus
t
a
b
le
m
odel i
s
gi
v
e
n
b
y
:
0
0
0
(
4
)
A
f
t
e
r
de
ve
lop
i
n
g
a
d
j
usta
ble
a
n
d
refere
nc
e
m
ode
ls,
t
h
e
a
d
a
p
ta
t
i
o
n
mec
h
a
n
i
s
m
w
i
ll
b
e
bui
lt
fo
r
MRAS met
h
od.
Th
e
ad
ap
t
a
t
i
on
me
c
h
ani
s
m
is
d
e
s
ig
n
e
d
in
a
w
ay
t
o
g
e
n
e
rat
e
t
h
e
v
a
l
ue
o
f
est
i
ma
t
e
d
s
p
ee
d
use
d
s
o
to
m
i
ni
m
i
ze
t
he
e
rror
betw
e
e
n
t
h
e
es
t
i
m
a
te
d
an
d
refere
nce
d-q
a
x
i
s
s
ta
tor
cur
r
en
ts.
T
h
e
e
rror
be
tw
ee
n
the
e
s
tim
ated
a
n
d
r
efe
r
enc
e
d-
q
axi
s stat
or
c
urr
e
nt
s
ar
e de
fine
d a
s
:
(
5
)
The
sta
t
e
cur
r
ent
s
e
rror
com
pone
n
t
is
:
(
6
)
The
sta
t
e
er
ror mode
l
o
f
t
he
P
M
S
M i
n
the
d
-
q
syn
chr
o
n
o
u
s
r
efe
r
e
n
c
e
f
r
am
e i
s
g
iv
e
n
a
s fl
o
w
:
ε
A
ε
W
(
7
)
Where
is the
e
r
r
o
r state
vec
t
o
r
.
is t
he ou
t
pu
t ve
c
t
or of t
h
e fee
d
ba
ck
b
l
o
ck.
The
sys
t
e
m
i
s
asympt
o
tica
l
l
y
h
y
p
e
r
s
t
a
bl
e
w
h
e
n
t
he
c
ou
nter-r
eac
t
i
o
n
b
l
o
ck
r
ea
cts
to
t
he
P
opo
v i
n
eq
ua
li
ty
[
1
1
-12-
13]
:
ε
W
dt
γ
in
w
h
i
ch
t
0
(8)
F
i
n
a
l
l
y,
w
e ca
n c
oncl
u
de th
a
t
the
ob
serve
d
ro
t
or
s
pee
d
s
at
i
s
f
i
e
rs the
f
ol
low
i
ng
ada
p
tat
i
o
n l
a
w
s
:
(
9
)
A
k
I
ε
I
ε
(10)
(11)
Where
a
nd
k
ω
ar
e
t
h
e
P
I
spee
d
ob
s
er
ver
contro
l
l
e
r
The
ro
tor
e
s
ti
m
a
ted
s
p
ee
d
is
g
e
n
era
t
ed
f
ro
m
the
a
d
ap
ta
tion
m
e
ch
an
ism
us
in
g
the
er
ro
r
be
t
w
e
e
n
t
h
e
e
s
tim
ated
a
n
d
r
efe
r
enc
e
c
urre
nts
obta
i
ne
d b
y
the
mode
l as
f
ol
lo
ws:
(
12)
F
i
nal
l
y
,
t
h
e
e
s
t
i
m
a
ted r
o
t
o
r
p
o
s
i
t
i
on is o
b
t
a
i
ned
b
y
i
nte
g
rat
i
n
g t
h
e
est
i
m
a
t
e
d rot
o
r
spe
e
d.
(13)
4.
PM
SM
F
U
Z
ZY SUPERVIS
O
R
CO
N
T
ROL
Th
is
s
ystem
i
s
c
ompose
d
o
f
tw
o
parts
:
a
b
loc
k
PI
t
y
pe
c
or
rec
t
io
n
a
n
d
t
h
e
bloc
k
fo
r
t
h
e
fuzz
y
sup
e
r
v
isor
c
on
t
r
ol [14]
.
The con
t
r
o
l
syste
m
b
loc
k
dia
gra
m
sug
g
e
s
t
e
d
is pre
s
en
ted i
n
F
ig
u
r
e
2.
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
SS
N
:
2088-
86
94
S
e
ns
o
r
l
e
ss c
ont
ro
l
of
PMS
M
wi
t
h
fu
z
z
y
mo
de
l
r
e
f
e
ren
c
e
ada
pti
v
e
sy
st
e
m
(La
r
b
i
M’ha
med
)
1
775
F
i
gur
e 2.
Bl
o
c
k
dia
gr
am
of
the
prop
ose
d
co
n
t
ro
l sys
t
em
.
4.
1.
C
l
a
ssic
a
l PI
c
orr
ect
ion
The
P
I
r
egula
t
or
i
s
use
d
t
o
pr
od
uc
e
a
c
o
n
t
ro
l
re
sp
ons
e
t
h
a
t
i
s
a
fu
nct
i
on
of
t
he
e
rr
or
m
a
gnit
ude.
T
he
P
r
oportional
t
er
m
(P)
of
t
he
r
eg
ulator
i
s
for
m
ed
by
m
u
ltiplying
t
he
e
rro
r
sig
n
a
l
by
a
P
ga
in
.
A
s
t
he
e
r
r
o
r
si
gna
l
bec
o
m
e
s
large
r
,
the
P
ter
m
o
f
t
h
e
re
g
u
l
a
tor
be
com
e
s
l
a
rg
er
t
o
pro
v
i
d
e
m
o
r
e
c
or
rec
t
i
o
n.
I
n
shor
t,
P
i
s
a
m
a
gni
fy
f
u
n
c
t
i
on.
Th
e
Int
e
g
r
a
l
t
erm
(I
)
o
f
t
h
e
r
eg
ul
at
o
r
i
s
u
s
ed
t
o
eli
m
i
n
a
t
e
sma
ll
steady
errors.
The
(I
)
ter
m
c
alculates
a
co
n
t
i
nuo
us
r
unn
in
g
t
o
tal
o
f
t
he
e
rror
s
i
g
n
a
l
.
T
her
e
fore
,
a
sm
a
ll
ste
a
d
y
s
ta
te
e
rror
acc
umula
t
es
i
n
t
o
a
la
rge
erro
r
v
a
lu
e
o
v
e
r
t
i
m
e.
T
hi
s
acc
u
m
ul
a
t
e
d
e
rror
sign
al
i
s
m
u
ltipl
ie
d
by
a
n
I
ga
in
f
ac
t
o
r
and
bec
o
m
e
s
t
h
e
(
I
)
ou
t
p
u
t
term
of the
P
I re
g
u
la
to
r.
4.
2.
Fuzzy Logi
c
control
F
u
z
z
y
lo
g
i
c is able
t
o
u
se
h
u
m
a
n
r
eason
s not
i
n term
s
o
f
d
i
s
cre
t
e
s
y
m
bol
s
and
nu
mb
ers,
b
ut
i
n
t
e
rms
of
f
uz
z
y
s
e
t
s
.
The
se
t
er
m
s
a
re
qu
i
te
f
le
x
i
ble
w
i
t
h
r
e
s
pe
c
t
t
o
t
he
de
f
i
n
i
t
i
o
n and
val
u
es. Th
e
Ba
s
ic
co
n
f
i
g
u
ra
t
i
o
n
of a
F
uz
zy Log
ic
C
on
tro
l
l
e
r
(F
LC)
consi
s
t
s
o
f t
h
e
fo
l
l
ow
i
n
g
com
po
ne
nt
s
[1
4-1
5
]:
a.
Fuzzif
i
er,
b.
Fuzzy
rule bas
e,
c.
In
fe
re
nc
e
en
gine
d.
De
fu
zzifie
r
Th
e
Fu
zzi
fi
e
r
c
h
a
ng
e
s
t
h
e
i
np
ut
(
cri
s
p
sig
n
a
l
s)
i
nt
o
fu
z
z
y
v
a
l
u
e
s
.
Th
e
fu
z
z
y
r
ul
e
ba
se
c
on
sist
s
of
ba
sic
da
t
a
a
n
d
lin
g
u
i
s
t
i
c
ru
le
s.
T
he
e
ng
ine
i
s
t
he
b
ra
in
o
f
a
f
uzz
y
c
on
trolle
r
w
h
i
c
h
a
b
ili
ty
t
o
simu
la
t
e
t
he
hum
an
d
ec
is
i
on
base
d
on fina
ll
y,
t
he se
c
o
n
d
tra
nsform
atio
n
con
v
er
t
s
v
a
l
ue
s in
to t
he
r
ea
l
val
u
es.
A
f
u
zz
y
co
nt
rol
l
e
r
is
a
s
p
e
c
i
al
f
u
z
zy
s
y
s
t
e
m
t
h
at
c
an
b
e
u
s
e
d
a
s
a
co
ntro
l
l
e
r
c
om
po
ne
nt
i
n
a
c
l
o
s
e
d
lo
o
p
s
yst
e
m
.
T
he
i
n
t
e
g
ra
t
i
o
n
o
f
a
f
u
zzy
s
ys
te
m
in
to
a
c
l
o
se
d
lo
o
p
i
s
sh
ow
n.
S
pec
i
a
l
e
mp
ha
sis
is
p
u
t
o
n
t
o
t
h
e
tra
n
sfer
b
eha
v
i
o
r
o
f
f
uz
zy
c
on
tro
l
l
e
rs,
whic
h
ar
e
anal
yze
d
u
s
i
n
g
d
i
ffere
n
t
c
onf
ig
ur
at
i
o
ns
o
f
sta
ndar
d
Mem
b
er
shi
p
f
u
n
ct
i
ons
[
1
4
-1
5].
The
c
o
mbi
n
a
t
i
on
of
c
l
a
ssic
a
l
c
on
tro
l
a
nd
t
h
e
fuzz
y
s
u
p
e
r
v
is
or
c
o
n
t
r
o
l
a
r
e
b
a
s
e
d
o
n
t
h
e
s
p
e
e
d
e
r
r
o
r
a
n
d
its
v
ar
ia
ti
o
n
,
t
h
e
s
y
s
t
em
p
erforma
n
ce
s
r
e
spec
t
i
ve
ly.
W
e
u
se
t
he
f
i
r
st
a
d
v
a
n
t
a
ge
i
n
tr
ans
i
en
t
st
a
t
e
a
n
d
t
h
e
sec
o
n
d
adv
a
nta
g
e
i
n
t
h
e
s
t
e
a
d
y
st
at
e.
T
h
e
p
ro
po
si
ti
on
b
ases
o
f
fuzz
y
s
upe
r
v
i
s
or
r
ule
s
t
o
gener
a
te
t
he
w
eig
h
t
s
to be
a
p
pli
e
d
to
e
a
c
h r
e
gu
lat
o
r ga
i
n
s
K
P
, K
i
.
The
use
d
h
eur
i
st
i
c
laws
are
:
Incr
ease
K
p
i
f
t
he
s
ystem
res
p
onses
f
a
r
from
the
referen
c
e
to
i
ncrease
the
c
o
n
v
e
rg
en
ce
s
p
e
ed
a
nd
de
cre
a
se
K
i
if t
he sys
tem
r
e
sponse
is ne
a
r
from
the
r
e
f
ere
n
ce
to a
n
tic
i
p
ate
t
he
o
ve
rs
ho
o
t
.
Thi
s
s
t
r
a
t
eg
y
of
i
mp
ro
vi
n
g
t
he
c
on
t
r
o
l
l
a
w
i
n
t
r
a
n
si
t
o
ry
s
t
a
t
e
hel
p
s
us
t
o
ob
ta
in
b
e
tter
re
sp
onse
s
t
ha
n
us
in
g
on
l
y
the
PI
regula
t
or
a
lo
ne [1
4
-1
5].
P
e
rform
a
nc
e
suc
h
a
s
re
s
p
o
n
s
e
t
i
m
e,
over
s
h
o
o
t
a
t
star
t
u
p,
overs
h
o
o
t
f
or
L
oa
d
a
ppl
ica
t
i
on,
t
he
L
oad
r
e
ject
io
n t
i
me
a
nd t
h
e c
h
a
t
terin
g
r
educ
i
ng a
r
e
the e
s
se
n
t
i
a
l
s
e
lem
e
nts of
t
h
i
s
stra
te
g
y
.
The
in
pu
ts
o
f
the
F
u
zz
y
s
u
p
e
rvis
or
PI
con
t
rol
l
er
a
re
t
he
e
rr
or
(
e
)
,
the
err
o
r
c
h
an
ge
(
de/
d
t
)
a
nd
the
sa
tura
tio
n
(
,
T
he
F
i
s
u
sed
in
t
h
i
s
stu
dy
i
s
a
s
ug
u
no
ty
pe.
The
cho
s
en
m
em
ber
ship
f
unc
t
i
o
n
for the
i
npu
t
v
a
riab
les
has tria
ng
u
l
ar
shapes
as show
n
in
F
igur
e 3
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN:
2
0
8
8
-
86
94
In
t
J
Po
w
Elec
& D
r
i
Sy
st Vo
l
.
1
0
,
N
o
.
4
, D
ec
2
0
1
9
:
1
772
–
1
780
1
776
F
i
gur
e
3.
I
npu
t
&
out
put
m
e
m
be
r
s
h
i
p
f
u
nc
ti
o
n
.
The
r
u
le
base
e
xpr
esse
d b
y
lin
gu
is
tic
ter
ms i
s
show
n i
n
T
a
b
le
1
.
No
te tha
t t
h
e distri
b
u
t
i
o
n
of rule
s in
tab
l
e
is
s
ym
me
t
r
i
c
.
A
f
t
e
r
“
d
é
f
uzz
i
f
i
c
a
t
i
o
n”
,
t
h
e
r
o
bu
st
ness
a
n
a
l
y
s
i
s
is
qua
nti
f
ie
d
by
t
he
f
unc
t
i
o
n
u
=
f
(
error
,
d
–
error
,
V_sat
)
w
h
i
c
h
c
o
r
r
e
spo
n
d
s
t
o
a
s
ur
f
ace
p
r
e
sente
d
i
n
F
i
gur
e
4.
Table
1
. F
u
z
zy
Rul
es Used
N
/
N
Z
N
/
P
PG
P
/
N
P
G
P
/
P
Z
Z
/
/
P
N
/
/
Z
P
N
/
Z
/
P
N
N
N
Z
PG
P
P
Z
P
G
F
i
gur
e
4.
F
uzzy
f
u
n
c
t
i
o
n
(
u
)
Vs
at
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
SS
N
:
2088-
86
94
S
e
ns
o
r
l
e
ss c
ont
ro
l
of
PMS
M
wi
t
h
fu
z
z
y
mo
de
l
r
e
f
e
ren
c
e
ada
pti
v
e
sy
st
e
m
(La
r
b
i
M’ha
med
)
1
777
F
i
gure
5.
Bloc
k
dia
gra
m
o
f se
nsor
l
e
ss
v
ec
t
o
r
contro
l of
PM
S
M
5.
S
I
MULAT
I
ON RESUL
TS
AND
D
I
S
CU
SSIONS
To
d
emon
st
rat
e
t
h
e
p
erfo
rma
n
ce
o
f
t
h
e
p
r
opo
s
e
d
con
t
rol
sch
e
me
(
s
e
e
F
i
g
u
r
e
5
)
,
a
s
e
t
o
f
s
i
m
u
l
a
t
i
o
n
s
i
s
c
a
r
ri
ed
o
ut
o
n
a
P
M
SM
m
o
d
e
l
by
u
si
ng
SI
MU
L
I
NK/
M
AT
L
A
B
.
The
par
a
m
e
ters
o
f
the
te
ste
d
PM
SM
are
gi
ve
n
i
n
T
ab
le
2
.
In
t
h
i
s
se
c
t
i
o
n,
t
he
s
pe
e
d
s
e
tti
n
g
i
s
trea
te
d
w
ith
a
d
a
pt
iv
e
fu
z
z
y
co
nt
ro
ll
e
r
a
sso
c
ia
t
e
d
wi
th
Mo
de
l
Refe
ren
c
e
A
d
ap
ti
ve
S
ystem
pow
er
ed
by
the
vo
lta
ge
i
nve
rter
w
i
t
h
PWM
c
o
n
t
r
ol
.
The
s
p
ee
d
r
e
fe
renc
e
tra
j
e
c
t
or
y
i
s
g
ive
n
b
y t
h
e
fo
l
l
ow
ing
be
nc
hm
ark:
(
0,
+
10
0,
-10
0,
0)
rad/s.
Tab
l
e
2.
P
a
r
ameter
s
of t
he
m
ot
or
C
o
m
pone
nts
Va
lue
s
U
nits
C
n
Ω
n
R
s
L
s
P
J
F
5
1000
1.
67
1.
45
3
0.
17
3.
10
0.
013
Nm
tr
/
min
Ω
mH
- - -
-
-
Wb
Kg
.
m
²
Nm
s/
r
a
d
F
i
gure
6
a
n
d
(Zo
o
m
of
S
tart
up
a
n
d
L
oa
d
a
p
p
l
i
c
at
i
o
n)
i
l
l
u
s
trate
the
s
i
m
u
lat
i
on
r
esu
lts
b
y
the
m
ode
l
r
e
fe
renc
e ada
p
t
i
ve
s
ystem
(
M
RAS
) w
i
t
h
fuzz
y
s
u
pperv
is
or has
s
u
p
er
ior
ity
an
d
g
ives
t
he
bes
t per
for
m
a
n
ce an
d
r
obust
n
ess
re
la
t
i
ve
t
o
the
M
o
de
l
Re
fe
renc
e
A
d
ap
ti
ve
S
ys
t
e
m
of
s
en
sorle
ss
vec
t
or
c
ontr
o
l
i
n
t
e
r
m
s
o
f
low
spee
d
be
hav
i
o
r
,
spe
e
d
reve
rsi
o
n
a
nd
loa
d
r
e
j
ecti
o
n
as
s
h
o
w
n
i
n
Ta
b
l
e
3,
c
ompe
nsa
t
e
co
ns
idera
b
ly
t
h
e
di
st
urba
nce
s
c
ause
d
b
y
t
he
l
oa
d
varia
t
i
o
ns
a
nd
mi
nimi
ze
t
he
t
or
q
u
e
r
i
ppl
es
(
c
u
rre
nt
)
devel
o
ped
by
t
h
e
motor
(
s
e
e
F
igure
7,8
and 9)
.
Ta
ble
3.
S
ummary
o
f pr
o
p
ose
d
con
tro
l
sim
u
l
a
t
i
o
n
p
e
rfor
ma
nce
C
ont
r
o
ll
e
r
D
d
(%)
T
r
(s)
T
m
(s
)
E
s
(%)
D
p
(%
)
T
p
(s)
M
R
AS
1
.
258
0
.
0
4
0.
04
0
1
3
0.
017
F
u
zz
y
-
MR
A
S
0.
8
0
.
04
0
.
0
4
0
4.
8
0.
013
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN:
2
0
8
8
-
86
94
In
t
J
Po
w
Elec
& D
r
i
Sy
st Vo
l
.
1
0
,
N
o
.
4
, D
ec
2
0
1
9
:
1
772
–
1
780
1
778
D
d
(%)
D
p
(%)
T
r
(
s)
T
m
(s
)
E
s
(
%
)
T
p
(
s)
V
sd
,
V
sq
I
sd
, I
sq
I
sd
*
I
sq
*
Ω
∗
R
s
L
sd
, L
sq
J
F
P
C
r
The o
v
e
r
sh
oo
t
at star
t
up
Th
e
ov
ersho
o
t
f
o
r
l
o
ad
app
l
i
cat
i
on
The r
e
sponse time
Th
e
r
ise
t
i
me
The sta
t
ic
e
rr
or
The l
o
ad
r
eje
c
t
i
o
n
t
i
m
e
S
t
ator
w
i
n
d
i
n
g
d,
q
a
xis
vo
lt
a
g
e
r
e
spec
t
i
v
e
ly
S
t
a
t
or
w
ind
i
ng
d,
q
a
xis
cur
r
e
nt
r
e
s
pe
c
t
ive
l
y
R
e
fer
e
nce
sta
t
or
w
ind
i
ng
d,
q
a
xis
cur
r
ent
r
e
spec
ti
ve
ly
The
e
l
ec
tr
ic
r
ot
or
s
pe
ed
R
e
f
e
r
e
n
c
e ro
tor
s
p
eed
E
s
ti
ma
t
e
d
ro
t
o
r
s
p
e
e
d
Rot
o
r
pos
i
tio
n
Perman
en
t Mag
n
e
t
Flux
S
t
a
t
o
r
p
h
a
s
e
r
e
s
i
s
t
a
n
c
e
The
sta
t
or
i
nd
u
c
t
a
nce
s
o
f
the
axi
s
d
,
q
Inertia
o
f turn
i
ng pa
r
t
s
V
iscous
f
r
i
c
t
ion
coe
f
f
i
c
i
e
n
t
Poles pa
i
r
s num
b
er
Loa
d
torque
F
i
gur
e
6.
T
he
r
ota
t
io
na
l
spee
d
(r
eal,
estima
t
ed,
r
e
f
e
r
e
nce
)
i
n
the
vec
t
or
c
o
n
tr
o
l
b
ased
on
M
R
A
S
F
i
gur
e
7
:
S
imulat
i
on
r
e
su
lts:
m
e
a
s
ur
ed
,
measu
r
ed
,
m
easur
ed
pha
se
s
tat
o
r
cur
r
e
nt
(
V
e
ctor
c
ontr
o
l
base
d
cla
s
s
i
ca
l
contr
o
l)
.
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
SS
N
:
2088-
86
94
S
e
ns
o
r
l
e
ss c
ont
ro
l
of
PMS
M
wi
t
h
fu
z
z
y
mo
de
l
r
e
f
e
ren
c
e
ada
pti
v
e
sy
st
e
m
(La
r
b
i
M’ha
med
)
1
779
F
i
gure
8. S
imulat
i
on re
su
lts:
m
easur
ed
,
me
as
u
r
e
d
, me
a
sured
pha
se
s
t
a
t
o
r c
u
rrent
(
V
e
c
t
or
c
o
n
tr
ol
b
ase
d
ada
p
tive
fuzz
y c
ontro
l)
F
i
gure
9.
S
im
ulat
ion
resu
l
t
s:
z
oom
of
me
asure
d
a
nd m
easure
d
6.
CONCLUSION
A
robus
t
c
o
nt
r
o
l
se
ns
orless
o
f
a
p
e
r
ma
nent
m
a
gnet
sy
nc
hr
on
o
u
s
m
ot
or
i
s
pre
s
ente
d.
T
h
e
s
im
ul
a
tio
n
r
e
s
u
l
t
s
o
b
t
a
i
n
e
d
i
n
t
h
i
s
w
o
r
k
c
o
n
f
i
r
m
i
t
s
f
e
a
s
i
b
i
l
i
t
y
a
n
d
v
a
l
i
d
a
t
e
exc
e
l
len
t
dyna
mic
perfor
m
a
n
ce.
T
he
r
esults
show
a
g
o
o
d
e
st
ima
tio
n
u
nde
r
di
ffe
re
nt
ope
r
a
ti
n
g
c
o
n
d
iti
o
n
s
a
n
d
l
ow
s
e
n
si
ti
vit
y
t
o
ex
ter
n
a
l
d
istur
b
a
n
ce
s
t
h
e
y
al
lo
wed
u
s
t
o
g
e
t
ri
d
of
e
sp
ec
i
a
lly
m
e
c
h
a
n
i
cal
s
p
e
e
d
s
en
so
r
o
r
p
o
s
i
t
i
o
n
,
w
h
ic
h
is
e
x
p
e
n
si
ve
a
nd
f
r
a
gi
l
e
.
C
onc
l
ude
d
aga
i
n
s
t
t
h
e
Mo
de
l
Re
fere
nce
A
d
a
p
ti
ve
S
ystem
is
s
imple
t
o
i
m
p
l
ement
,
d
o
n
't
t
ak
e
i
n
to
a
cc
oun
t
t
h
e
m
e
a
s
u
r
em
ent
no
ise
or
t
he
e
nv
i
r
o
n
me
nt.
It
n
ot
r
eq
uire
a
l
on
g
c
a
l
cu
la
tio
n
ti
me
,
an
d
h
a
s
a
goo
d
dyn
a
m
i
c
r
e
spo
n
se
s
peed
a
nd
go
o
d
d
ist
u
rba
n
ce
r
e
jec
t
i
on,
it
s
how
a
r
esp
o
n
se
t
i
m
e
a
n
d
e
f
fi
ci
en
t
rob
u
stn
e
ss.
A
c
c
o
rd
i
n
g
to
t
he
s
im
ul
a
t
ion
re
su
l
t
s
t
h
e
m
odel
refere
nc
e
ada
p
t
i
ve
s
ys
t
e
m
(
MRAS
)
w
i
t
h
f
u
z
z
y
s
u
p
p
e
r
v
i
s
o
r
h
a
s
s
u
p
e
r
i
o
r
i
t
y
a
nd
gi
ves
t
h
e
best
p
er
form
ance
a
nd
rob
u
s
t
ne
ss
re
l
a
tiv
e
to
t
he
M
od
el
R
e
f
ere
n
ce
A
dap
t
i
v
e
Syst
em
o
f
se
n
s
orle
ss vec
t
o
r co
n
t
rol
i
n
t
e
r
m
s
of l
o
w
s
p
e
e
d
beha
v
i
o
r
,
sp
e
ed
reve
r
s
i
o
n
a
nd
loa
d
re
jec
t
i
o
n
w
i
th
r
e
duc
t
i
on
o
f
to
ta
l ha
rm
o
n
ic
di
s
tort
io
n ge
ne
rate
d i
n
the
p
h
a
se
c
urre
nts.
REFERE
NC
E
S
[1]
A
.
Asri,
A
.
Dah
m
an, A
.
M
a
lek
Sai
d
in
a,
S
. Iqb
a
l,
M
.
Ab
dul Razak
,
"A new sp
eed
s
ens
o
rles
s
field
orient
ed c
o
n
t
r
o
l
le
r
for
PM
SM
b
a
s
e
d
on
MR
AS
a
n
d
PS
O,"
J. E
l
ect
r
i
c
a
l
Syst
ems
,
v
ol.
1
2
(
3),
p
p
.
5
6
5
-573,
2
0
1
6
.
[2]
H
.
L
i
u
a
n
d
S
.
L
i
.
,
"
S
p
e
e
d
c
on
tro
l
f
or
P
MS
M
s
e
rv
o
s
y
s
t
e
m
u
sing
p
r
edi
c
t
i
v
e
f
unctio
n
a
l
c
o
nt
r
o
l
a
n
d
e
x
te
n
d
e
d
s
ta
t
e
obs
e
rver,"
IEE
E
Tr
ans
Ind E
l
ectr
o
n
, vo
l
.
59
(2
), p
p. 11
7
1
-1
18
3,
2
01
2.
[3]
B.
A
lecsa,
e
t
al
.
,
"
S
i
m
uli
n
k
m
o
d
e
l
i
n
g
a
nd
d
es
ign
of
a
n
eff
i
c
i
en
t
hard
ware-
c
on
st
r
a
i
n
e
d
FP
GA-ba
s
e
d
P
MSM
spe
e
d
co
nt
roll
er,"
IEEE T
r
ans
In
dInf
o
rma
t
,
vo
l.
8(3
),
p
p
.
5
44
-56
2
, 2
01
2.
[4]
D.
J
ia
ng
,
e
t
al
.
,
"
A
s
li
d
i
n
g
m
o
d
e
o
b
s
e
r
v
e
r
for
PMSM
sp
e
e
d
a
n
d
r
otor
p
o
s
it
i
o
n
c
o
nsideri
n
g
sa
liency,"
in
Powe
r
E
l
ectr
onics Sp
ecia
l
ists C
onfer
ence,
PES
C
20
08.
IEE
E
, pp
.
8
09
-8
14
, 2
00
8.
Is
d
Is
q
(
A
)
Is
d
Is
q
(
A
)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N:
2088-8694
I
nt J P
ow
Ele
c &
Dri S
y
st
V
ol
.
10,
No. 4, D
ec 2019
:
1772 –
1
780
1
780
[5]
C.
J
ia
nb
o,
e
t
a
l
.
,
"
A
n
impro
v
e
d
slidin
g
mo
de
o
bse
r
ve
r
for
p
o
sit
ion
sens
orl
e
ss
vecto
r
c
on
trol
d
ri
v
e
o
f
P
M
SM,
"
i
n
P
o
wer Electro
nics a
n
d
M
o
ti
on
Cont
rol Co
n
f
eren
c
e
,
IEEE 6th Intern
ati
o
n
a
l
,
pp
.
189
8-1
902
,
2
00
9.
[6]
M
.
H
o
rch
,
A
.Boum
edi
e
ne,
L.Bagh
li,
"
M
R
A
S
-ba
s
ed
S
ens
o
rles
s
Sp
eed
I
nt
egral
Ba
ck
st
eppi
ng
C
on
trol
f
or
I
nd
uctio
n
M
ach
ine,
u
s
i
ng
a
F
l
u
x
B
a
c
k
s
t
e
ppi
ng
O
b
s
erver,
"
In
tern
atio
na
l
Jo
urn
a
l
of
Po
wer E
l
ectr
onics
an
d D
r
ive
Syst
em
(I
J
P
E
D
S
)
,
vo
l
.
8
,
n
o.
4,
p
p
.
16
50
-16
6
2
, 20
1
7
.
[7]
C.
S
ch
aud
e
r,
"
M
o
del
Referen
ce
Ad
apti
ve
S
ystem
w
i
th
S
i
m
p
l
e
S
e
nso
r
less
Fl
u
x
O
b
s
erver
f
o
r
Inducti
o
n
Mot
o
r
Dr
iv
e
,
"
IEEE
Intern
at
ion
a
l
Con
f
e
r
ence o
n
P
o
wer
El
ec
t
r
o
n
i
c
s
,
Dr
ives
and
E
n
erg
y
Sys
t
em
s,
p
p.1
-
6
,
Decem
ber1
6-1
9
,
20
12
, Beng
a
l
u
ru
,
In
d
i
a.
[8]
Bao
DY,
W
ang
H,
W
ang
XJ,
Z
han
g
CR,
"
S
e
ns
orl
e
s
s
s
peed
c
on
t
r
ol
b
a
se
d
o
n
t
he
improved
Q
-MRAS
me
thod
f
o
r
i
n
d
u
cti
o
n
m
o
t
o
r d
r
iv
es,
"
E
nerg
i
e
s,
11:23
5
,
2
0
18.
[9]
Zo
r
ga
n
i
Y
A,
K
ou
ba
a
Y,
B
ou
ss
a
k
M
.,
"
MRAS
s
ta
te
e
s
t
i
m
a
t
or
f
or
s
peed
s
enso
rles
s
IS
FOC
i
n
d
u
ct
ion
mo
to
r
d
r
ives
w
i
t
h
L
uenb
erger
l
o
ad
to
r
qu
e
est
i
m
a
ti
on
,
"
ISA Tr
ans
, 6
1
:
30
8-1
7
,
2
01
6.
[10]
S
u
man
.
M
,
Ch
andan
.
C
,
S
a
b
t
asach
i.
S
,
"
S
i
m
ula
t
io
n
st
udies
o
n
m
o
d
el
r
ef
erence
a
dapt
iv
e
co
ntro
ll
er
b
ased
s
p
eed
es
timatio
n
tech
ni
qu
e
f
o
r
t
h
e
vect
or
c
on
tro
lle
d
perm
anen
t
magn
et
s
y
n
chro
no
us
m
otor
d
riv
e
,
"
E
L
S
EV
IER,
vo
l
.
1
7,
20
0
9
.
[11]
Z
.
X
ingmin
g,
et a
l
.
,
"
W
id
e-S
p
eed-Ran
ge
S
en
so
rl
ess
Control
of
I
nt
erio
r
P
M
S
M
Based
o
n
MR
AS,"
Interna
tion
a
l
Co
nfer
ence, Elect
ri
ca
l
M
a
c
hin
e
s
and
System
s
(
I
CEM
S
)
,
p
p
.
8
04–
80
8,
201
0.
[12]
L
a
rbi.
M
,
Gh
e
r
ab
i
.Z,
Do
ud
ar
.
K
,
"
A
Rob
u
s
t
s
ens
o
rl
ess
con
t
rol
o
f
P
M
S
M
b
as
ed
o
n
s
l
i
d
i
n
g
mod
e
o
bs
erver
an
d
m
o
del
ref
e
rence
adapt
i
v
e
s
ys
tem,"
In
ter
nat
io
na
l Jou
r
na
l of
P
o
wer
E
l
ectr
onics a
nd Dr
ive Sys
t
em
(
I
JPE
D
S),
v
o
l
.
8
,
n
o
.
3,
p
p.
1
0
16-1
0
2
5
,
201
7.
[13]
D
.
S
o
l
a
n
k
i
a
n
d
P
.
R
S
h
a
r
m
a
,
"
S
e
n
s
o
r
l
e
s
s
S
p
e
e
d
C
o
n
t
r
o
l
o
f
P
M
S
M
U
sing
M
RAS
Me
tho
d
,
"
in
3r
d In
ter
n
a
t
iona
l
Co
nfer
ence o
n
Rol
e
of
En
gi
neer
s
a
s
E
n
trep
reneu
r
s
in
Curren
t
S
cena
r
io
, 20
1
4
.
[14]
M
.
L
arb
i
,
S.
H
as
s
a
in
e
,
S
.
M
o
reau,
B.
M
azari,
"
A
ro
bu
st
f
uzzy
s
u
pe
rv
is
or
c
ont
ro
l
l
e
r
a
ppl
ied
t
o
a
p
erm
a
n
e
nt
m
agn
e
t
s
y
n
c
hron
ou
s
m
o
to
r,
"
IR
EE
,
v
o
l.
6,
n0
.
5
,
20
11
.
[15]
M.
I
r
f
a
n
,
Ma
c
h
mu
d
E
f
fe
nd
y,
N
ur
A
li
f,
L
a
i
lis
S
,
Ilh
a
m
Pa
ka
y
a
,
Am
ru
l
Faru
q,
"
P
e
rf
orm
a
n
ce
co
m
p
ari
s
on
o
f
f
u
z
zy
lo
gic
a
n
d
p
r
op
ortion
a
l-in
te
gr
a
l
f
o
r
a
n
e
l
ec
tro
n
i
c
loa
d
c
on
troll
er,"
Inter
n
a
t
i
ona
l J
o
ur
na
l o
f
P
o
wer E
l
ect
ronics
an
d
D
r
i
v
e System
(IJP
E
D
S
)
,
v
ol.
8
,
no.
3
,
p
p
.
117
6-118
3,
2
01
7.
Evaluation Warning : The document was created with Spire.PDF for Python.