Int
ern
at
i
onal
Journ
al of Ele
ctrical
an
d
Co
mput
er
En
gin
eeri
ng
(IJ
E
C
E)
Vo
l.
10
,
N
o.
1
,
Febr
uar
y
2020
, pp. 84
9~85
5
IS
S
N: 20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v10
i
1
.
pp849
-
855
849
Journ
al h
om
e
page
:
http:
//
ij
ece.i
aesc
or
e.c
om/i
nd
ex
.ph
p/IJ
ECE
On tra
ck
ing contr
ol pro
blem for
polysol
en
oid motor
model p
re
di
ctive app
roac
h
Nguy
e
n H
ong
Quang
1
,
N
guy
en Phung
Qu
ang
2
,
D
o
Tr
u
ng Hai
3
,
Ngu
yen Nh
u Hien
4
1,3,4
Depa
rtment
o
f
Autom
at
ion,
T
hai
Ngu
y
en
Uni
ver
sit
y
of
T
ec
hn
olog
y
,
Vi
et Nam
2
Instit
ute for
Control
Engi
ne
eri
n
g
and
Autom
a
ti
o
n,
Hanoi
Univer
sit
y
of
Sc
ie
n
ce
a
nd
Technol
og
y
,
Viet
Nam
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Ja
n
25
, 2
019
Re
vised
Ma
y
5
,
20
19
Accepte
d
Se
p
27
, 20
19
The
Pol
y
sol
enoid
Li
nea
r
Motor
(PLM)
have
been
play
ing
a
cru
c
ia
l
role
i
n
m
an
y
industrial
aspe
ct
s
due
to
i
ts
func
ti
ons,
in
which
a
strai
ght
m
oti
on
is
provide
d
direct
l
y
without
m
ediate
m
ec
han
ical
ac
tuators.
Recent
l
y
,
wi
th
seve
ral
comm
ons
on
m
at
hematic
m
o
del
,
som
e
cont
rol
m
et
hods
for
PLM
base
d
on
Rota
tional
Motor
have
bee
n
applied,
but
positi
on,
vel
oc
ity
an
d
cur
ren
t
constrai
nts
which
ar
e
i
m
porta
nt
in
r
eal
s
y
stems
have
b
ee
n
igno
red
.
In
thi
s
p
ape
r
,
p
ositi
on
tr
ac
k
ing
cont
ro
l
prob
lem
for
PLM
wa
s
conside
red
un
der
stat
e
-
independe
nt
disturb
a
nce
s
via
m
in
-
m
ax
m
odel
pre
dic
tive
cont
rol
.
The
proposed
c
ontrol
ler
forc
es
tra
ck
ing
positi
o
n
err
ors
conve
r
ge
to
sm
al
l
reg
ion
of
origi
n
and
sati
sfies
st
at
e
in
cl
uding
po
siti
on,
ve
loc
i
t
y
and
cur
ren
ts
constra
in
ts.
Furt
her
,
a
num
eri
cal
sim
ula
ti
on
wa
s
implemente
d
to
validate
the
p
erf
orm
anc
e
of
the proposed cont
roller
.
Ke
yw
or
d
s
:
Ca
scade c
on
t
rol
Con
ti
nu
ou
s
c
on
t
ro
l
set
m
o
del
pr
e
dicti
ve
c
on
t
ro
l
Mi
n
m
ax
m
od
e
l pr
e
dicti
ve
Perm
anen
t m
a
gn
et
li
ne
ar
synch
ron
ous m
otor
Po
ly
so
le
no
i
d
li
near m
oto
r
Copyright
©
202
0
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
:
Nguyen
Ho
ng
Qu
a
ng,
Dep
a
rtm
ent o
f Au
t
om
ation
,
Thai
Nguyen
Un
i
ver
sit
y o
f Te
ch
no
l
og
y,
666, 3/
2
St
reet,
Tich L
uong
w
ard, T
hai Ng
uyen
ci
ty
, V
ie
t
N
a
m
.
Em
a
il
: qu
an
g.n
gu
ye
nhong@t
nu
t.e
du.
vn
1.
INTROD
U
CTION
Linear
Mot
or
t
ran
sm
issi
on
syst
e
m
s
are
widely
app
li
ed
to
pr
ov
i
de
directe
d
strai
gh
t
m
otions
in
wh
ic
h,
m
echan
ic
al
act
uator
s
a
re
el
i
m
inate
d,
resul
ti
ng
in
bette
r
pe
rfor
m
ance
of
m
otion
s
yst
e
m
s.
Gen
er
al
ly
,
Po
ly
so
le
no
i
d
Linear
Mot
or
(P
LM)
has
a
du
ra
ble
str
uctu
re,
operat
ion
s
acco
r
ding
to
el
ect
ro
m
agn
et
ic
ph
e
nom
eno
n
with
pr
inci
ples
as
sho
wn
i
n
[
1
-
12
]
an
d
va
ri
ou
s
a
ppli
cat
io
ns
s
uch
as
CN
C
Lat
he
[
13
]
,
sli
din
g
door
[14
]
.
W
it
hout
the
nee
d
of
a
ny
gear
box
f
or
m
otion
transfo
rm
ation
,
the
PLM
syst
em
beco
m
es
se
ns
it
ive
du
e
to
exte
rn
al
im
pacts
su
ch
as
f
rict
ion
al
f
orce,
e
nd
–
e
ff
e
ct
,
cha
nged
loa
d
a
nd n
on
-
sine
of f
lu
x.
Th
ese
eff
ect
s
encou
nter
both
in
the
l
ongitu
din
al
a
nd
in
th
e
trans
versal
di
recti
on,
w
hic
h
is
al
ong
with
s
at
ur
at
io
n
in
supp
li
e
d
vo
lt
age
, m
ake
good c
on
t
ro
l
pe
rfor
m
ance
f
rom
the li
near
dri
ve
a
dif
ficult
ta
sk
.
Ther
e
are
se
ve
ral
resea
rch
es
t
akin
g
int
o
acc
ount
the
posit
ion
c
ontr
ol
of
P
LM
in
prese
nc
e
of
e
xter
nal
disturba
nces.
The
a
uthor
s
in
[
15
]
pr
ese
nte
d
a
co
ntr
ol
desi
gn
m
et
ho
d
t
o
r
egu
la
te
velocit
y
base
d
on
PI
–
sel
f
-
tun
in
g
c
om
bin
ing
with
ap
propriat
e
est
i
m
a
ti
on
te
ch
nique
at
slow
vel
oc
it
y
zon
e,
but
if
load
is
c
ha
ng
e
d,
PI
–
sel
f
-
t
un
i
ng
con
t
ro
ll
er
will
be
not
ef
fici
ent.
I
n
order
t
o
ov
e
rc
om
e
chan
ge
d
loa
d,
m
od
el
ref
e
re
nce
c
on
t
ro
l
m
et
ho
d
base
d
on
Ly
ap
unov
sta
bili
ty
theor
y
was
em
plo
ye
d
i
n
[
16
]
.
Additi
on
al
l
y,
the
com
pen
sat
io
n
appr
oac
hes
we
re
propose
d
i
n
resea
rch
[
17
]
on
wh
ic
h,
t
he
fr
ic
ti
onal
f
orce
wer
e
e
sti
m
at
ed
by
L
ugrie
an
d
Stribec
k
f
rict
io
n
m
od
el
res
pec
ti
vely
.
In
[
18
]
,
the
ad
va
ntage
of
t
hat
the
sli
di
ng
m
od
e
co
ntr
ol
ap
plied
i
n
L
inea
r
Motor
is
that
real
posit
ion
va
lue
tracks
set
po
i
nt.
H
ow
e
ve
r,
the
disad
va
ntages
of
this
m
et
ho
d
are
fin
ding
sli
din
g
s
urface
an
d
c
hatte
rin
g.
I
n
t
he
view
of
non
li
nea
r
sys
tem
s,
the
stu
dy
in
[19
]
a
pply
li
near
iz
at
ion
m
et
hod
to
PLM
syst
em
bu
t
this
m
et
hod
is
rest
ri
ct
ed
by
uncer
ta
in
par
am
et
er
and
disturba
nces.
It
is
cl
ear
that
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
1
,
Febr
uar
y
2020
:
84
9
-
855
850
the
previ
ous
r
esearche
s
do
no
t
m
ention
posit
ion,
vel
ocity
and
c
urren
t
s
co
ns
trai
nts
a
s
well
as
im
p
act
of
exter
nal d
ist
urban
ce
which
is
i
m
po
rta
nt pr
operti
es
of the c
on
t
ro
l sy
ste
m
s.
2.
DYN
AM
I
C M
ODEL
Po
ly
so
le
no
i
d
li
near
m
oto
r
i
s
co
ns
tr
ucted
accor
ding
to
el
ect
r
om
agn
et
ic
inducti
on
a
s
show
n
in
Figure
1.
Figure
1.
Com
po
sit
io
n of
pol
ysole
noid
li
nea
r
m
oto
r
Let
u
s c
onsi
de
r
a d
y
nam
ic
m
o
del of
PLM i
n
[2
0
-
24
]:
=
−
+
(
2
)
+
,
=
−
−
(
2
)
−
(
2
)
+
,
(1)
=
2
(
+
(
−
)
)
−
1
,
=
.
Wh
e
re
,
,
,
sta
nd
for
cu
rr
e
nt,
veloc
it
y
and
posit
io
n
res
pecti
vely
.
In
a
ddit
ion
,
t
he
const
ant
pa
ra
m
et
er
of
PLM
inclu
des:
,
,
,
,
,
,
as
resist
a
nce,
inducto
r,
pole
pair
,
po
l
e
ste
p,
flu
x
a
nd
m
ass
of
ro
t
or.
The
in
put
volt
age
is
pr
ese
nt
ed
as
,
an
d
is
un
m
easur
e
d
ex
te
r
nal
f
or
ce
.
Suppose
t
hat,
can
be
cl
assify
into
finite
set
as
=
{
1
,
2
,
.
.
.
,
}
.
It
is
w
or
th
to
note
that,
the
m
od
el
(1)
is
sim
il
ar
t
o
per
m
anent
m
agn
et
r
otati
on
sync
hron
i
zat
ion
m
oto
r
m
od
el
.
He
nce,
le
t
us
assum
e
that,
and
ha
s
the
a
ppr
ox
im
ate
sim
i
la
r
val
ue
s
a
nd
the
te
rm
(
−
)
can
be
ig
nore
d
i
n
the
thir
d
e
qu
at
io
n
of
(
1)
t
hat
le
ads
to
li
near
relat
ion
s
hip be
tween c
urre
nt
and posit
io
n
-
ve
locit
y.
3.
CONTR
OL D
ESIGN
In
this
paper,
le
t
us
se
par
at
e
dynam
ic
m
od
el
(1)
i
nto
c
urren
t
s
ubsyst
em
an
d
posit
ion
su
bsy
ste
m
.
As
a
forem
entio
ne
d,
posit
io
n
subsyst
em
ca
n
be
c
onside
re
d
as
a
li
nea
r
ti
m
e
inv
aria
nt
s
yst
e
m
un
de
r
e
xter
nal
disturba
nces
a
nd
the
n,
th
e
posit
ion
c
on
tr
ol
le
r
is
design
e
d
with
the
ai
d
of
existe
d
m
in
-
m
ax
m
od
el
predict
ive
con
t
ro
l
the
or
y
[
25
]
.
O
n
the
oth
er
ha
nd,
the
c
ro
ss
-
c
urren
t
co
m
pen
sat
ion
m
e
thod
betwee
n
is
us
ed
to
tra
nsfo
rm
first
of
two
e
quat
ion
i
n
(
6)
in
to
li
near
f
or
m
.
More
ov
e
r,
the
current
co
ntr
ol
le
r
is
based
C
CS
m
od
el
pr
e
di
ct
ive
con
t
ro
l
w
hich ca
n
s
olv
e
cur
re
nt constrai
nt
prob
le
m
.
3.1.
C
ontr
ol of curre
nt
subs
ys
te
m
By
ap
plyi
ng
de
couplin
g
c
on
t
r
ol law a
s foll
ow:
=
−
(
2
)
+
1
,
=
(
2
)
+
(
2
)
+
2
.
(
1
)
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
On tracki
ng control
prob
le
m
f
or
po
ly
s
olen
oi
d mo
t
or
model
pr
e
dicti
ve appr
oa
c
h
(
Ng
uye
n Hon
g Qua
ng
)
851
The
c
urre
nt subsyste
m
is transf
orm
ed
to li
ne
ar s
yst
em
:
=
−
+
1
,
=
−
+
2
(
2
)
To
desig
n
CC
S
-
MPC
in
cu
r
ren
t
cl
os
ed
lo
op,
from
(3
),
by
us
i
ng
sim
ple
m
ann
er
,
le
t
us
obta
in
the d
isc
rete t
im
e p
re
dicti
ve m
od
el
as f
oll
ows:
(
+
+
1
)
=
(
+
)
+
̄
(
+
)
,
∀
=
0
,
1
,
.
.
.
,
−
1
(
3
)
Wh
e
re
is
a
predict
ion
horiz
on,
(
+
+
1
)
=
[
(
+
+
1
)
,
(
+
+
1
)
]
is
(
i
+1)
th
.
Estim
at
ed
current
vector
on
dq
-
c
oor
din
at
e
a
nd
(
)
=
(
)
=
[
(
)
,
(
)
]
at
sa
m
ple
tim
e
k.
The
pr
e
dicti
ve
co
ntr
ol
in
pu
t
vecto
r
i
s
presente
d
as
̄
(
+
)
=
[
̄
1
(
+
)
,
̄
2
(
+
1
)
]
,
̄
1
(
)
=
1
(
)
,
̄
2
(
)
=
2
(
)
. W
it
h
sta
nd for sam
pling
ti
m
e, the
sta
te
a
nd in
pu
t m
at
rix
,
are:
=
[
−
0
0
−
]
,
=
[
1
1
]
(
4
)
The
c
ontr
ol
ta
sk
is
to
fin
d
the
s
eq
ue
nce
of
c
on
t
ro
l
in
pu
t
vect
or
̄
(
)
,
̄
(
+
1
)
,
…
.
.
,
̄
(
+
−
1
)
, which
m
ini
m
i
ze the
fo
ll
owin
g
c
os
t f
unct
ion:
=
∑
(
−
(
+
+
1
)
)
1
(
−
(
+
|
)
)
=
1
(6)
Wh
e
re
1
=
[
0
;
01
]
is
po
si
ti
ve
m
at
rix,
is
ref
e
ren
ce
i
nput
vecto
r
fro
m
po
sit
ion
c
ontr
oller
w
hich
is
pr
ese
nt
i
n
t
he
nex
t
sect
io
n,
r
epr
ese
nt
pro
portion
al
rati
o
be
tween
d
-
cu
rren
t
error
|
−
|
an
d
q
-
curren
t
error
|
−
|
.
I
n
order
to
sim
plify
the
optim
iz
ation
(
6),
le
t
us
as
sum
e
that
the
refe
ren
ce
i
nput
ve
ct
or
is
const
ant
in
the
predict
io
n
hor
iz
on
ti
m
e.
The
assum
ption
is
com
m
on
in
pract
ic
al
exp
e
ri
m
ent
du
e
to
th
e
fact
that
current
lo
op
tra
ns
it
io
n
r
esp
on
se
is
co
nsi
der
a
bly
faster
than
it
s
posit
ion
lo
op.
Mo
r
eov
e
r,
the
pr
e
dicti
ve
con
t
ro
l
in
pu
t
norm
al
l
y
subje
ct
ed
t
o
t
he
l
inear
co
ns
trai
nt
̄
(
+
)
<
.
The
c
urre
nt
c
on
t
ro
ll
e
r
fr
e
qu
e
ntly
ope
rate
in
sm
all
sam
ple
,
the
reby
the
on
e
ste
p
horizo
n
=
1
w
as
ta
ke
i
nto
a
ccount
for
the
co
ntr
ol
de
sign.
A
nd
the
n,
by
sel
ect
ing
op
ti
m
al
var
ia
ble
̄
(
)
=
(
)
,
the
m
ini
m
i
zat
ion
(
6)
can
be
rewrit
te
n
as:
̄
(
)
=
̄
(
)
(
)
̄
(
)
+
2
(
(
)
−
)
2
̄
(
)
+
.
.
̄
(
)
<
,
∀
=
1
,
2
,
.
.
.
,
−
1
(
5
)
3.2.
C
on
tr
ol
of p
osi
tion
s
ubs
yste
m
The
dy
nam
ic
of
posit
ion
su
bsy
ste
m
is
sign
ific
a
ntly
slow
e
r
than
c
urren
t
s
ub
syst
e
m
s.
Hen
ce,
in
the
c
on
t
rol
desig
n
of
po
sit
io
n,
we
assum
ed
that
the
desi
red
curre
nt
eq
ua
ls
to
act
ual
current
.
By
set
ti
ng
=
+
(
2
)
−
1
̈
,
the
la
st
tw
o
e
qu
at
ion
of
(
1
)
ca
n
be
rewrit
te
n
as
disc
rete
sta
te
sp
ace
m
od
el
o
f
tra
cki
ng er
rors
a
s:
+
1
=
+
+
,
=
.
(
6
)
Wh
e
re
:
=
−
,
=
−
,
=
[
,
]
,
=
(
)
,
=
(
)
=
[
1
0
1
]
,
=
[
2
2
2
(
)
−
1
]
,
=
[
2
2
−
−
1
]
,
=
[
1
,
0
]
,
The
c
on
t
ro
l
in
pu
t
is
desig
ne
d
to
f
or
ce
er
ror
vecto
r
converg
e
to
sm
al
l
reg
ion
c
entere
d
at
or
i
gin
.
More
ov
e
r,
,
sat
i
sfies the
foll
owin
g
c
onstrai
nt
s:
∈
,
∈
and
z
→
0
∈
,
(
7
)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
1
,
Febr
uar
y
2020
:
84
9
-
855
852
wh
e
re
0
≜
{
(
1
,
2
)
∈
ℝ
2
|
1
2
0
0
0
0
}
0
∈
≜
{
(
1
,
2
)
∈
ℝ
2
|
1
2
}
≜
(
∈
ℝ
.
)
To
ac
hieve
co
ntr
ol
ob
j
ect
ive
(
9
),
m
in
-
m
ax
m
od
el
pr
e
dicti
ve
co
ntr
ol
propose
d
in
[
6
]
was
a
ppli
ed
.
The
op
e
rati
on
of
t
he
posit
ion
co
ntro
ll
er
is
dev
i
ded
into
two
m
od
e
s
:
an
“i
nn
e
r”
a
nd
a
n
“
ou
te
r
”
co
nt
ro
ll
er
.
The
in
ner
co
nt
ro
ll
er
is
act
ive
d
w
hen
the
sta
te
is
in
the
r
ob
us
t
c
ontrol
in
va
riant
set
0
,
an
d
it
s
r
ole
is
to
ke
ep
the
error
sta
te
in
0
un
de
r
exte
rnal
disturbance
.
The
in
ner
c
ontrolle
r
is
li
near
feedback
=
an
d
it
is
i
m
po
rtant
in
the
c
on
st
ru
c
ti
on
of
t
he
con
t
ro
l
r
obus
t
in
var
ia
nt
se
t
0
w
hich
is
s
el
ect
ed
based
on
(
+
)
=
0
,
is
po
sit
iv
e
integer
nu
m
ber
.
To
be
sp
eci
fic,
le
t
us
sel
ect
0
=
∑
(
+
1
)
wh
e
re,
is distu
rb
a
nce set.
The
oute
r
c
ontrolle
r
w
orks
w
hen
the
er
ror
s
ta
te
is
ou
tsi
de
the
in
var
ia
nt
s
et
0
an
d
ste
e
rs
t
he
syst
em
sta
te
to
the
inv
aria
nt
set
.
For
the
ou
te
r
co
nt
ro
ll
er,
we
us
e
m
in
–
m
ax
m
od
el
pr
e
dicti
ve
con
t
ro
l
,
a
nd
c
on
si
der
a fix
e
d h
or
iz
on
for
m
ulati
on
.
I
n
this
secti
on, t
he
sel
ect
ed
qua
dr
at
ic
c
os
t f
un
ct
ion
as:
(
̱
,
̱
,
̱
)
=
∑
(
+
+
+
+
+
)
−
1
=
0
,
≥
0
,
>
0
.
(
8
)
Wh
e
re
̱
=
[
,
+
1
,
.
.
.
,
+
−
1
]
,
̱
=
[
,
+
1
,
.
.
.
,
+
−
1
]
,
̱
=
[
+
1
,
+
2
,
.
.
.
,
+
−
1
]
.
Her
ei
n,
se
quen
ce
vect
or
̱
is
ch
os
e
n
to
m
ini
mize
cost
functi
on
(
10)
i
n
t
he
w
or
st
ca
se
w
her
e
,
̱
is
m
axi
m
u
m
po
i
nt of
(10
).
The follo
wing
al
gorithm
su
m
m
arize t
he
ope
rati
on of
posit
ion co
ntr
oller.
Algorithm
1
:
(
Po
sit
io
n
Co
ntr
oller)
Data
:
If
∈
0
,
set
=
.
Ot
herwise,
fi
nd
the
so
luti
on
of
(
11
)
a
nd
set
to
th
e
fi
rst
c
ontrol
i
n
t
he
opti
m
a
l
seq
uen
ce
̱
.
4.
NUMER
IC
A
L SIM
ULATI
ON
In
t
his
sect
io
n,
we
sim
ulate
t
he
po
sit
io
n
tra
ckin
g
of
whol
e
syst
e
m
un
de
r
sta
te
,
in
pu
t
c
on
st
raint
a
nd
exter
nal
dist
urban
ce
in
f
ollo
wing
ta
ble.
W
e
us
e
the
sam
e
cu
rr
e
nt
c
on
t
ro
ll
er
an
d
t
wo
dif
fer
e
nt
pr
e
di
ct
ion
horizo
ns
of
posit
ion
c
on
tr
olle
r
to
com
par
e
qual
it
y
of
eac
h
con
t
ro
ll
er
.
T
he
pa
ram
et
er
of
po
ly
s
olen
oid
li
near
m
oto
r
and
c
on
trolle
r
s
how
in
Table
1
.
As
c
an
be
seen
in
Figure
2
,
bo
t
h
posit
ion
a
nd
velocit
y
error
sta
y
in
sta
te
con
st
rain
ts
reg
i
on
an
d
co
nv
e
r
ges
to
sm
all
ball
centere
d
at
or
igi
n.
T
he
c
urre
nt
is
sat
isfyi
ng
input
const
raint
unde
r
ti
m
e v
aryi
ng
exter
nal force
s
.
Table
1
.
T
he
param
et
er o
f
po
l
ysole
noid li
nea
r
m
oto
r
a
nd contr
oller
Para
m
eter
Valu
e
Co
n
troller para
m
e
t
er
Valu
e
Po
le pair
1
0
0
0
.2 (
m
m
)
Po
le step
2
0
(
m
m
)
0
0
1
.5 (
m
/s)
Ro
to
r
m
ass
0
.17
(
k
g
)
10
Ph
ase coil
Resis
ta
n
ce
1
0
.3 (
)
K
[
-
300
-
5]
d
-
ax
is in
d
u
ctan
ce
1
.4
(
m
H
)
U
[
-
5
0
,
5
0
]
q
-
ax
is in
d
u
ctan
ce
1
.4
(
m
H
)
Flu
x
0
.03
5
(
W
b
)
Figure
2. The
perform
ance o
f
prop
os
ed
cont
ro
ll
er
(
c
on
ti
nu
e
)
k
z
k
u
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
On tracki
ng control
prob
le
m
f
or
po
ly
s
olen
oi
d mo
t
or
model
pr
e
dicti
ve appr
oa
c
h
(
Ng
uye
n Hon
g Qua
ng
)
853
Figure
2. The
perform
ance o
f
prop
os
ed
cont
ro
ll
er
5.
CONCL
US
I
O
N
We
al
so
il
lustr
at
ed
the
im
pact
of
pr
e
dicti
on
horizo
n
on
perform
ance
of
the
syst
em
.
If
is
sm
a
ll
,
the
dyna
m
ic
cl
os
ed
loo
p
syst
em
is
fast,
set
tl
ing
tim
e
is
s
m
al
l
but
inp
ut
co
ns
trai
nts
m
a
y
no
t
hold.
We
hav
e
t
o
c
hoos
e
N
is
suffici
e
ntly
la
rg
e
to
sat
isf
y
co
ns
trai
nts
.
The
propose
d
two
casca
de
lo
op
s
base
d
on
MPC
with
s
uf
ic
ie
nt
ly
s
m
al
l
pr
ed
ic
ti
on
ho
rizo
n
(
)
of
the
cu
r
ren
t
l
oop
t
o
reduce
cal
culat
ion
l
oad
of
m
ic
ro
process
ors
du
e
to
sm
all
sam
pling
tim
e
of
the
cu
rr
e
nt
loop.
Whe
n
consi
der
i
ng
th
e
cu
rr
e
nt
re
spon
s
e
is
ideal
, th
e
syst
em
p
erf
orm
ance is total
ly
dep
e
nd
i
ng
on m
in
-
m
ax
MPC
of t
he ou
te
r
lo
op.
ACKN
OWLE
DGE
M
ENTS
This
researc
h
was
sup
porte
d
by
Re
searc
h
F
oundat
io
n
f
unde
d
by
T
hai
N
guye
n
Un
i
ver
sit
y
of
Tech
no
l
og
y.
REFERE
NCE
S
[1]
Ng.
Ph.
Quang
,
J.
A.
Di
tt
ri
ch,
V
ec
tor
Con
trol
of
Thr
ee
-
Phase
AC
Mac
hin
es
-
S
yst
em
Dev
el
opment
in
the
Practice
,
2nd
Edition,
Spr
inge
r
B
erl
i
n
Hei
del
ber
g
,
2008
.
[2]
Danie
l
Aus
der
au
,
"
Pol
y
so
le
noid
-
Li
n
ea
r
ant
ri
eb
m
it
g
enut
e
te
m
Sta
t
or
,"
Zur
ic
h
.
PhD
The
ss
is
,
2004.
[3]
Danh
Hu
y
Ng
u
y
en
,
et
al
.
,
"
Nonline
ar
Con
t
rol
of
an
A
ct
i
ve
Magne
t
ic
B
ea
ring
wi
th
O
utput
Contr
ai
nt
,
"
In
te
r
na
ti
on
al J
ou
r
nal
of Elect
ric
al an
d
Co
m
pu
te
r
Engine
er
ing
(IJ
ECE)
,
vol.
8
(
5
)
,
pp
.
366
6
-
3677,
2018
.
[4]
H.
Chen,
K.
Lia
ng,
R.
Nie,
X.
L
iu,
"Three
-
Dim
ensiona
l
Elec
tro
m
agne
ti
c
Anal
ysis
of
Tubul
ar
Perm
ane
nt
Magn
et
Li
ne
ar
L
aunc
h
er
",
Applied
Super
conduc
t
iv
i
ty IE
E
E
Tr
ansacti
ons
on
,
vol
.
28
,
no
.
3
,
pp
.
1
-
8
,
2018
.
[5]
Hos
sein
Torkam
an,
Aghil
Ghahe
ri
,
Ali
Ke
y
h
ani
,
"A
xi
al
flux
sw
it
che
d
re
luct
anc
e
m
a
chi
nes:
a
comprehe
nsi
ve
rev
ie
w of
design
and topol
ogi
es",
Elec
tri
c Powe
r
Appl
ic
a
ti
ons I
ET
,
vol
.
13
,
no
.
3
,
p
p.
310
-
321
,
201
9.
[6]
Hao
Chen,
Yi
m
in
g
Zha
n,
Rui
Nie,
Shu
y
an
Z
hao,
"M
ultiobjecti
v
e
Optimiz
at
i
on
Design
of
T
ubula
r
Perm
anen
t
Magne
t Li
ne
ar L
aunc
her
"
,
P
lasma Sc
i
ence
IE
EE
Tr
ansacti
ons on
,
vol. 47, no. 5, p
p.
2486
-
2492
,
2
019.
[7]
Jordi
Garc
ia
-
A
m
orós,
"Line
ar
H
y
brid
Re
luc
t
an
ce
Motor
with
High
Densit
y
F
orc
e",
Ene
rgi
es
,
vol.
11,
pp
.
280
5,
2018.
[8]
B.
Tomczuk,
G.
Schrode
r,
A.
W
ai
ndok,
"F
ini
t
e
El
ement
Anal
y
sis
of
the
Magne
tic
Fiel
d
and
El
e
ct
rom
ec
h
ani
c
a
l
Para
m
et
ers
Calc
ula
ti
on
for
a
Slo
tt
ed
Perm
ane
nt
Magne
t
Tubular
Li
ne
ar
Motor",
IEE
E
Tr
ans
on
ma
gnetic,
vo
l.
43
,
no.
7
,
pp
.
3229
-
3236,
2007
.
[9]
I.
I.
Abdal
la,
T
.
Ibra
him
and
N.
M.
Nor,
"A
nal
y
sis
of
Tubular
Li
ne
ar
Motors
for
Diffe
ren
t
Sha
pes
of
Magne
ts,
"
in
IE
EE A
c
ce
ss
,
vol.
6
,
pp
.
10297
-
10310,
2018
.
d
oi:
10
.
1109/AC
CESS
.
2017.
277
5863
[10]
H.
R.
Esfa
han
ian,
S.
Hasan
za
d
e
h
and
M.
He
y
da
ri,
"P
erf
orm
ance
Anal
y
s
is
and
F
o
rce
Com
ponents
Im
prove
m
ent
of
an
Tubular
Li
ne
ar
Induc
ti
on
Mot
or
Us
ed
in
a
No
vel
Magn
et
i
c
Tr
ai
n
b
y
3
-
D
FEM,"
Elec
tric
al
Eng
ine
ering
(
ICEE
)
,
Iranian
Confe
re
nce
on
,
Mashhad
,
2018
,
pp
.
1368
-
1372.
doi
:
10
.
1
109/ICE
E
.
2018.
8472417
[11]
X.
Z.
Huang
,
J.
Li
,
Q
.
Ta
n
,
C
.
M
.
Zha
ng
and
L
.
Li,
"D
esign
Pri
nci
pl
es
of
a
Pha
se
-
Shift
Modula
r
Slotl
ess
Tubular
Perm
ane
nt
Mag
net
Li
n
ea
r
S
y
nc
hronous
Motor
W
it
h
Thre
e
Se
ctional
Prim
ari
es
and
Anal
y
s
is
of
Its
Dete
nt
Force
,
"
in
IE
EE
Tr
ansacti
ons
on
Industrial
El
ectronics
,
vol.
65
,
no.
12,
pp.
9346
-
9355,
Dec
.
2018.
d
oi:
10.
1109/T
I
E.
20
18.
2813992
[12]
V.
Ti
unov,
"P
ra
ct
i
ca
l
Appli
catio
n
and
Methods
of
Cal
culat
ion
for
Li
ne
ar
Induc
t
ion
Motors,"
20
18
Inte
rnationa
l
Mult
i
-
Conf
ere
nc
e
on
Industrial
Engi
ne
ering
and
Mode
rn
Technol
ogie
s
(
FarEastCon)
,
Vladi
vostok
,
2018,
pp.
1
-
6.
doi:
10
.
1109/Far
Ea
stCon.2018
.
8
602929
[13]
Yang
Ze
qing
,
Li
u
Li
bi
ng,
W
a
ngzuoj
i
e,
Chen
Yingshu,
Xiao
Quan
y
ang
,
"
Stat
i
c
and
D
y
n
a
m
ic
Chara
cteri
s
ti
c
Sim
ula
ti
on
of
Feed
S
y
stem
Drive
n
b
y
L
ine
ar
Motor
in
High
Speed
Com
pute
r
Num
eri
cal
Control
Lathe
,
"
TEL
KOMNIKA
(Tele
communic
a
t
ion,
Comput
ing,
El
e
ct
ronics
and
Control)
,
vol.
11
(7)
,
pp
.
3673
-
36
83
,
2013
.
[14]
A
y
m
en
Lachhe
b
,
Jalel
Khedir
i
,
Li
lia
E
l
Am
rao
ui,
"
Perform
ances
Anal
y
sis
of
a
Li
n
ea
r
Motor
for
Slidi
ng
Door
Applic
a
ti
on,
"
I
nte
rnational
J
ournal
of
Po
wer
El
e
c
troni
c
s
and
Dr
iv
e
Syste
m
(
IJPEDS)
,
vol.
8(
3)
,
pp.
1139
-
1146
,
2017
.
[15]
Jul
-
Ki
Seok,
Jong
-
Kun
Le
e,
D
ong
-
Choon
Le
e
,
"
Sensorless
Sp
ee
d
Control
of
Nons
al
ie
nt
Per
m
ane
nt
Magne
t
S
y
nchr
onous
Motor
Us
ing
Rotor
-
Pos
it
ion
-
Track
ing
PI
Control
ler,
"
IEEE
Tr
ansa
ct
ions
on
Indust
rialE
lectroni
cs
,
vol
.
53
(
2
)
,
pp
.
3
99
-
405,
2006
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
1
,
Febr
uar
y
2020
:
84
9
-
855
854
[16]
Yuan
-
Rui
Ch
en,
Jie
W
u,
No
ber
t
Cheung
,
"
L
y
apunov
'
s
Sta
bil
ity
The
or
y
-
Based
Model
R
efe
ren
ce
Adap
tive
Control
for
Per
m
ane
nt
Magne
t
Li
ne
ar
Motor
Drive
s
,"
Proc
of
Powe
r
El
e
ctr
onic
s
Syste
ms
and
Appl
i
cat
ion
,
pp.
260
–
266
,
2
004
.
[17]
C.
Huang,
Li
-
Chen
Fu
,
"
Adap
ti
ve
B
ac
kst
eppi
n
g
Speed
/Pos
it
io
n
Control
wi
th
Frict
ion
Com
pe
nsati
on
for
Li
n
e
ar
Induc
ti
on
Motor
,"
Prod.
of
the 4
1st I
EEE
Conf
ere
nce on
De
ci
sio
n
and
Control
,
US
A,
pp.
474
–
479
,
2002
[18]
G.
Ta
p
ia
,
A.
T
a
pia
,
"
Slid
ing
-
Mode
Control
fo
r
Li
ne
ar
Perm
an
ent
-
Magn
et
m
o
tor
Pos
it
ion
Tr
acking
,"
Proc
ee
di
ng
of
th
e
I
FA
C
Wor
ld
Congress
,
200
7
.
[19]
Quang
H.
Nguy
en,
et.
al
.
,
"
Design
an
Exa
c
t
Linea
ri
za
t
ion
Control
ler
for
Perma
nent
Stim
ulatio
n
Sy
n
chr
onou
s
Li
ne
ar
Motor
Pol
y
sol
enoi
d
,
"
SSRG
Int
ernati
onal
Journa
l
of
E
lectric
a
l
and
Elec
tronic
s
Engi
n
ee
rin
g
,
vol.
4
(1)
,
2017
.
[20]
Quang.
N.
H,
et
al.
,
"
Flat
n
ess
B
ase
d
Control
Struct
ure
for
Pol
y
s
ole
noid
Perm
an
ent
Stim
ula
ti
on
Li
ne
ar
Motors,
"
SSRG
Inte
rnat
io
nal
Journal
of
Elec
tri
cal
and
E
lectronic
s
Engi
ne
e
ring
,
vol
.
3(12)
,
pp.
31
-
37
,
2016
.
[21]
Quang
N.
H.,
e
t
al.
,
"
Multi
Pa
ramet
ric
Progra
m
m
ing
base
d
Model
Predic
t
iv
e
Control
for
tr
ac
king
Contro
l
of
Pol
y
soleno
id Li
nea
r
Motor
,
"
Sp
e
ci
al
issue
on
Me
asur
eme
nt,
Cont
rol an
d
Au
tomat
ion,
vo
l. 19, pp.
31
-
37,
2017
.
[22]
Ngu
y
en
Hong
Quang,
e
t
al
.
,
“
Min
Max
Mode
l
Predictive
Con
trol
for
Pol
y
sol
e
noid
Li
n
ea
r
Mot
or
,”
In
te
rnation
al
Journal
of
Power
El
e
ct
ronics
an
d
Dr
iv
e
S
yste
m (
IJP
EDS)
,
vol
.
9
(
3
)
,
pp
.
1666
-
167
5,
2018
.
[23]
Ngu
y
en
Hong
Quang.
,
et
al.
,
"
Multi
par
ametr
ic
m
odel
pre
dic
t
iv
e
cont
rol
base
d
on
la
guer
re
m
odel
for
per
m
ane
n
t
m
agne
t
l
inear
s
y
nchr
onous
m
otors
,
"
Inte
rnat
io
nal
Journal
of
El
e
ct
rica
l
and
Computer
Enginee
ring
(
IJE
C
E
)
,
vol.
9
(
2
)
,
pp
.
10
67
-
1077,
2019
.
[24]
Dao
Phuong
Na
m
,
et
al
.
,
"
Multi
Para
m
et
ric
Prog
ramm
ing
and
Exa
ct
L
ineari
z
at
ion
base
d
Model
Predictive
Contro
l
of
a
Perm
ane
n
t
Magne
t
Li
ne
ar
S
y
nchr
onous
Mot
or,
"
Inte
rnat
iona
l
Confe
ren
ce
on
Syste
m
Sc
ie
nc
e
and
Engi
ne
ering
(
ICSSE
)
,
pp.
743
-
747,
2017
.
[25]
P.
O.
M.
Scokaert,
D.
Q.
Ma
y
ne,
"
Min
-
Max
Feedba
ck
Model
Pre
dic
ti
v
e
Co
ntrol
f
or
Constrai
ned
Li
ne
ar
S
y
stems
,
"
IEE
E
Tr
ansacti
o
ns On A
utomatic
Control
,
vol
.
43
(
8
)
,
1998
.
BIOGR
AP
H
I
ES
OF
A
UTH
ORS
Hong
Quan
g
Ngu
y
en
recei
ve
d
the
B.
S
degr
ee
in
el
e
ct
r
ic
a
l
engi
ne
eri
ng
fr
om
Tha
i
Ngu
y
en
Univer
sit
y
of
tec
hnolog
y
(TNUT
),
Vie
tna
m
,
200
7,
th
e
Mast
er’
s
degr
ee
in
cont
ro
l
eng
ine
er
ing
an
d
aut
om
at
ion
from
Hanoi
Univer
sit
y
of
Sci
enc
e
and
Te
chnol
og
y
(H
US
T),
Viet
Nam
,
2012
and
Ph.D
fr
om
Tha
i
Ngu
y
en
Univer
sit
y
of
te
chnol
og
y
(TN
UT),
Viet
nam,
2
019.
He
is
cur
re
nt
working
as
a
le
c
ture
r
at
Dep
a
rtment
of
Industria
l
Autom
at
ion
,
Facul
t
y
o
f
El
e
c
tri
c
al
Engi
n
ee
r
in
g,
Tha
i
Ngu
y
e
n
Univer
sit
y
of
Te
chno
log
y
(TN
UT).
His
Rese
arc
h
In
te
r
ests
i
ncl
ude
Elec
tr
i
c
a
l
Drive
S
y
stem
s,
Adapti
ve
D
y
n
a
m
ic
Programming
Control
,
Robust
Nonline
ar
Model
Predic
ti
v
e
Control
,
Moti
on
Control
,
Control S
y
stem
and it
s
Applic
a
ti
ons,
M
ec
ha
choni
cs
.
Ngu
y
en
Phu
ng
Quan
g
rec
ei
ved
his
Dipl.
-
Ing.
(Uni.
),
Dr.
-
Ing
.
a
nd
Dr.
-
Ing.
habil.
degr
e
es
from
TU
Dresden,
Ge
rm
an
y
in
1975
,
1991
and
1994
r
espe
ctively
.
Prio
r
to
his
ret
urn
to
Viet
nam,
he
ha
d
worked
in
Germ
an
y
industr
y
for
m
an
y
y
e
ars,
con
tri
bute
d
to
create
inv
ert
ers
R
EFU
402
V
ec
tov
ar,
RD500 (RE
FU
El
ektronik);
Sim
over
t
6SE42
,
Ma
ster
Drive
MC (
Siemens).
From
1996
to
1998,
h
e
serve
d
as
lectur
er
of
TU
Dresden
where
he
wa
s
conf
err
ed
as
Privat
doz
ent
in
1997.
He
joi
n
ed
Hanoi
Univer
sit
y
of
Scie
n
ce
an
d
Technol
og
y
i
n
1999,
as
lect
u
rer
u
p
to
now.
He
is
cur
r
ent
l
y
a
profe
ss
or
of
HU
ST
and
honor
ar
y
profe
ss
or
of
TU
Dresden.
He
wa
s
aut
hor/
co
-
aut
h
or
of
m
ore
tha
n
170
journa
l and confe
ren
ce
p
apers
;
8
books wit
h
thre
e among them
was writ
te
n
in
Germ
an
and
on
e
in
Engl
ish
ent
i
tled
“
Vec
tor
Cont
rol
of
T
hre
e
-
Ph
ase
AC
Mac
hin
es
–
S
y
stem
De
vel
opm
ent
in
th
e
Prac
tice”
publi
s
hed
b
y
Springe
r
in
2008,
and
2nd
edi
t
ion
in
June
2015.
.
His
Rese
arc
h
In
te
r
ests
ar
e
El
e
ct
ri
ca
l
Drive
S
y
stems
,
Motion
Control,
Robot
i
c
Control,
Ve
ct
o
r
Control
o
f
E
le
c
tri
c
al
Ma
chi
n
es,
W
ind
an
d
Sol
ar P
ower
S
y
stems
,
Digit
al Cont
ro
l S
y
stems
,
Model
i
ng
and
Sim
ulati
on.
Do
Trun
g
Hai
born
in
1974
,
re
ce
iv
ed
th
e
PhD
.
degr
ee
in
Auto
m
at
ion
from
Ha
Noi
Univer
sit
y
of
Te
chno
log
y
and
Scie
nc
e
in
200
8
and
working
in
Thai
Ngu
y
en
Univer
sit
y
of
Te
chno
lg
y
now
.
His re
sea
r
ch
in
terests i
nc
lude m
oti
on
cont
rol
,
m
o
der
n
con
trol th
e
or
y
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
On tracki
ng control
prob
le
m
f
or
po
ly
s
olen
oi
d mo
t
or
model
pr
e
dicti
ve appr
oa
c
h
(
Ng
uye
n Hon
g Qua
ng
)
855
Ngu
y
en
Nh
u
Hi
en
rec
ei
v
ed
the
B.
S.
degr
e
e
in
elec
tr
ic
a
l
engi
n
ee
r
ing
from
Bac
Tha
i
Univer
sit
y
of
Mec
hanics
and
El
e
ct
ri
cs
(form
er
name
of
Thai
Ngu
y
en
Uni
v
e
rsit
y
of
Techno
log
y
)
in
1976,
the
M.S
degr
ee
and
the
PhD
degr
ee
in
aut
om
atio
n
and
cont
rol
fr
om
Ha
Noi
Uni
ver
sit
y
of
Sci
ence
and
T
ec
hnolog
y
in
1997
and
2002
respe
c
ti
ve
l
y
.
He
is
cur
re
ntly
a
As
socia
t
e
Profess
or
wit
h
Autom
at
ion
Division,
Facu
lty
of
El
e
ct
ri
ca
l
Enginee
ring
,
Th
ai
N
gu
y
en
Univ
ersity
of
Te
chno
log
y.
His
cur
ren
t
rese
a
rch
int
er
ests
inc
l
ude
adva
nc
e
con
trol
of
m
ult
i
-
axi
s
drive
s
y
stem,
co
ntrol
for
m
ult
i
-
var
ia
b
le
p
roc
ess
es
and rene
wab
l
e
en
erg
y
s
y
stem
s.
Evaluation Warning : The document was created with Spire.PDF for Python.