Int
ern
at
i
onal
Journ
al of
P
ower E
le
ctr
on
i
cs a
n
d
Drive
S
ystem
s
(
IJ
PEDS
)
Vo
l.
12
,
No.
1
,
M
a
r 202
1
, p
p.
10
~
19
IS
S
N:
20
88
-
8694
,
DOI: 10
.11
591/
ij
peds
.
v12.i
1
.
pp10
-
19
10
Journ
al
h
om
e
page
:
http:
//
ij
pe
ds
.i
aescore.c
om
On
fini
te
-
time
ou
tput fe
edback sli
ding mo
de cont
rol of an
elastic
multi
-
mo
t
or syst
em
Pha
m
Tu
an
T
ha
n
h
1
, Tr
an
X
ua
n
Ti
nh
2
,
D
ao P
huong
Na
m
3
, Da
o
Sy L
ua
t
4
, Ng
uyen
Ho
n
g Qu
ang
5
1,2
Le
Quy
Don
Univer
sity
of Technology,
Hano
i,
Viet
na
m
3
Hanoi
Univer
si
t
y
of
Sci
ence and
Technol
ogy
,
Ha
noi,
Vi
et
n
am
4
Dong Nai
Te
ch
nology
Univer
si
t
y,
Bi
ên
Hòa
,
Vi
e
tna
m
5
Tha
i
Nguyen
U
nive
rsity
of Technology,
Tha
i
N
guyen,
Vi
et
n
am
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
J
ul
28
, 2
0
20
Re
vised
Jan
15
, 2021
Accepte
d
Fe
b
15
, 20
2
1
In
thi
s
pape
r
,
th
e
tracki
ng
con
tr
ol
sche
m
e
is
pre
sente
d
using
th
e
fra
me
work
of
finite
-
ti
m
e
sl
idi
ng
mode
con
trol
(SM
C)
la
w
and
high
-
g
ai
n
observe
r
for
disturbe
d/un
ce
rt
ai
n
mu
lt
i
-
mot
or
drivi
ng
sys
tems
under
the
conside
ra
ti
on
mul
ti
-
ou
tput
sys
te
ms.
The
con
ver
gence
ti
m
e
of
slidi
ng
mod
e
con
trol
is
esti
mated
in
co
nnec
t
ion
with
linear
m
at
r
ix
inequali
t
ie
s
(LMIs)
.
The
input
stat
e
stabilit
y
(
ISS
)
of
proposed
con
tro
ller
w
as
analyz
ed
by
Lya
punov
stabi
lity
the
ory
.
Final
ly,
th
e
exte
nsive
simu
la
t
ion
result
s
are
giv
e
n
to
v
alidate
the
adva
nt
age
s o
f
proposed contr
ol
design
.
Ke
yw
or
d
s
:
Finit
e
-
ti
me con
trol
Linear
matri
x i
nequali
ti
es
M
ulti
-
m
otor s
yst
ems
Ou
t
pu
t
fee
db
ac
k
c
on
t
ro
l
Sli
din
g m
od
e
c
on
t
ro
l
This
is an
open
acc
ess arti
cl
e
un
der
the
CC
BY
-
SA
l
ic
ense
.
Corres
pond
in
g
Aut
h
or
:
Ngu
yen Ho
ng
Qu
a
ng,
Thai
Ngu
yen
Un
i
ver
sit
y o
f Te
ch
no
l
ogy,
666, 3/
2
St
reet,
Tich L
uong
W
ard, T
hai Ng
uyen
Ci
ty
-
T
hai
Ngu
yen Pro
vinc
e, V
ie
tna
m
Emai
l:
qu
a
ng.
nguye
nhong@t
nu
t.e
du.
vn
1.
INTROD
U
CTION
M
ulti
-
m
otor
dri
ve
s
ys
te
ms
ha
ve
been
em
pl
oy
e
d
i
n
s
ys
te
ms
m
ovin
g
pa
per,
meta
l,
ma
te
rial
bein
g
qu
it
e
popula
r
in
man
uf
a
ct
ur
i
ng
s
ys
te
ms
an
d
researc
he
d
by
man
y
aut
hor
s
in
the
rece
nt
ti
mes.
The
c
ontr
ol
method
util
iz
e
d
arti
fici
al
ne
ur
al
net
wor
k
(
ANN)
te
c
hn
i
que
has
bee
n
pres
ente
d
by
B
ou
c
hib
a
et
al
.,
[1].
Howe
ver,
the
disad
va
ntage
is
to
in
vestigat
e
the
ap
pro
pr
ia
t
e
net
works
with
as
so
ci
at
ed
le
arn
i
ng
r
ules
i
n
con
t
ro
l
desig
n.
Furthe
r
more,
t
he
ef
fec
ti
ven
ess
of
tra
ckin
g
pro
blem
or
the
sta
bili
zat
ion
of
t
he
ca
scade
s
ys
te
m
a
re
not
sti
ll
con
side
re
d
unde
r
the
in
f
luences
of
us
i
ng
ne
ural
net
w
ork
a
p
proach.
Dominiq
ue
K
ni
tt
el
,
et
al
.
,
pro
po
s
ed
man
y
li
near
c
on
t
ro
ll
ers
unde
r
the
c
on
si
derat
ion
of
the
a
ppr
oximat
e
model
of
m
ulti
-
mo
to
r
sy
ste
ms
withou
t
el
ast
ic
,
fr
ic
ti
on
as
a
li
near
s
yst
em
to
de
sig
n
the
c
ontr
oller
base
d
on
the
tr
ansf
e
r
f
un
ct
io
n
te
ch
nique
[2
]
,
[
3].
The
f
rame
wor
k
of
the
cl
assic
al
PI
c
ontroll
er
a
nd
H
in
fin
it
y
to
el
imi
nat
e
disturba
nce
was
pro
posed
in
th
e
work
of
[2
]
,
[
3].
In
the
el
as
ti
c
mu
lt
i
-
m
otor
dr
i
ve
sy
ste
m
s,
it
is
necess
ary
to
est
imat
e
the
belt
te
nsi
on
to
est
ablish
t
he
a
s
so
ci
at
ed
sta
te
f
eedb
ac
k
co
ntr
ol
le
r.
H
owe
ver
,
the d
iffic
ulti
es o
f
the
c
on
tr
ol d
esi
gn
li
e
in
th
e
fact
that
meas
ur
e
m
ent
of
this
belt
te
ns
io
n
by
us
i
ng
se
nsors
.
T
he
s
l
i
d
i
n
g
m
o
d
e
c
o
n
t
r
o
l
(
S
M
C
)
t
e
c
h
n
i
q
u
e
b
a
s
e
d
s
t
a
t
e
f
e
e
d
b
a
c
k
c
o
n
t
r
o
l
e
n
a
b
l
e
s
t
o
e
l
i
m
i
n
a
t
e
i
n
f
l
u
e
n
c
e
o
f
d
i
s
t
u
r
b
a
n
c
e
s
a
n
d
u
n
k
n
o
w
n
p
a
r
a
m
e
t
e
r
s
w
a
s
p
r
o
p
o
s
e
d
i
n
[1
]
,
[
4].
In
[
4]
,
the
sli
ding
m
ode
c
ontr
oller
wa
s
c
ombine
d
with
disturba
nce
observ
e
r
as
well
as
it
erati
ve
le
arn
i
ng
con
t
ro
l
(
ILC)
,
feedfo
rw
a
r
d
con
t
ro
ll
er.
It
can
be
see
n
t
hat,
the
cl
assi
cal
nonlinear
con
t
ro
l
la
w
ha
s
bee
n
inv
est
igate
d
i
n
rob
otic
syst
em
s
s
uch
as
em
plo
yi
ng
the
fu
zz
y
-
c
ontrol
in
in
ver
te
d
pe
ndulum
s
ys
te
ms
[
5],
r
obust
adap
ti
ve
co
ntr
ol
sche
mes
i
n
bilat
eral
te
le
operators
[
6],
wheel
ed
m
ob
il
e
r
obotic
syst
ems
[7],
t
racto
r
tra
il
er
[8
].
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
P
ow Elec
& Dri S
ys
t
IS
S
N: 20
88
-
8
694
On fi
nite
-
ti
me ou
t
pu
t f
ee
dbac
k sli
ding
mode
con
tr
ol
of an
el
as
ti
c mult
i
-
m
oto
r syste
m (P
ham
T
uan T
hanh
)
11
Additi
on
al
l
y,
t
he
o
ptimal
c
ontrol
a
nd
opti
miza
ti
on
pro
blem
are
al
so
c
onsidere
d
in
rece
nt
ti
me
via
t
he
w
ork
i
n
[9
]
-
[12
].
T
he
s
li
din
g
mode
ba
sed
c
on
t
ro
l
has
bee
n
paid
muc
h
at
te
ntio
n
in
recent
years
be
cause
it
is
a
wi
dely
releva
nt
co
ntr
ol
method
ology
for
un
ce
rtai
n/
disturbe
d
s
ys
te
ms.
S
M
C
as
a
rob
us
t
co
ntr
ol
scheme
f
oc
use
s
on
reducin
g
t
he
di
sadv
a
ntage
of
exter
nal
distu
rb
a
nces
a
s
m
uc
h
as
possible
base
d
on
t
he
de
sign
of
a
ppropr
ia
te
sli
din
g
s
urfac
e
as
well
as
impleme
nting
the
e
quivale
nt
co
ntr
oller
[
13]
-
[
21].
An
a
dap
ti
ve
sc
hem
e
wa
s
pr
opos
e
d
i
n
[
22]
with
ou
t
a
ny
knowle
dge
of
the
bound
on
the
disturba
nc
e
an
d
thei
r
de
rivati
ves.
T
he
high
order
S
M
C
de
sign
wa
s
in
ves
ti
gated
in
c
on
t
ro
l
sy
ste
m
of
unic
ycle
unde
r
t
he
c
onside
rati
on
of
un
ce
rtai
nt
ie
s
of
matc
hed
a
nd
unmatc
he
d
te
rm
[
23].
Additi
onal
ly,
the
act
uator
sat
urat
io
n
w
as
al
so
c
on
si
de
red
in
sli
ding
mode
con
t
ro
l
desi
gn
f
or
s
pace
c
raft
sy
ste
m
s
[24].
Th
e
pro
blem
of
in
vestigat
in
g
sta
te
obse
rvers
for
syst
em
s
is
a
n
importa
nt
dire
ct
ion
in
the
co
ntr
ol
li
te
ratur
e.
I
n
[
25]
,
high
-
gain
obser
ve
r
(HGO)
has
be
e
n
in
vestigat
e
d
sinc
e
1980
s
by
the
work
of
Pete
rsen
a
nd
H
ollot
on
H
in
fini
ty
co
ntro
l
bas
ed
on
the
a
dj
us
ti
ng
of
ap
propriat
e
coeffic
ie
nt
to
s
at
isfy
the
li
nea
r
matri
x
in
eq
ua
li
ti
es
(LM
Is
).
A
uthors
in
[
25]
desc
ribe
d
t
he
de
velo
pm
e
nt
from
the
tra
diti
on
al
exam
ple
to
di
fferentia
l
obse
rvabil
it
y
us
i
ng
T
aylo
r
e
xp
a
ns
i
on
with
c
onside
rin
g
the
ti
me
i
nter
val
[
−
,
]
.
It
sug
gested
that
the
H
G
O
sho
uld
be
desig
ne
d
by
interme
diate
va
riables
obta
in
ed
f
rom
the
diff
e
re
ntial
ob
serv
a
bili
ty
work
[
25].
M
ore
ov
e
r,
in
rece
nt
ti
me,
t
he
HGO
desig
n
f
or
a
cl
ass
of
m
ulti
-
in
pu
t
mu
lt
i
-
outp
ut
(
M
I
MO)
unif
ormly
obse
rv
a
bl
e
sy
ste
ms
were
al
so
co
ns
i
de
red
a
nd
the
e
xt
ensio
ns
of
previo
us
idea
wer
e
im
plemente
d
f
or
unce
rtai
n
no
nlinear
s
ys
te
m
s
with
sa
mp
le
d
outp
uts
[15
].
The
work
i
n
[26
]
mentio
ned
tw
o
pr
ob
le
ms
,
in
cl
ud
in
g
sta
te
est
imat
ion
via
Neural
Net
w
orks
a
nd
bac
kst
epp
i
ng
te
c
hniqu
e
in
deali
ng w
it
h i
nput sat
ur
at
io
n.
I
t i
s wo
rth n
oting
t
hat the
outpu
t
fee
db
ac
k
c
on
t
ro
l sc
heme
bein
g
the
fram
ewor
k
of
sta
te
fee
dba
ck
c
ontrol
an
d
obse
rv
e
r
was
desig
ne
d
us
in
g
dy
na
mic
gain
an
d
exte
nd
e
d
sta
te
obser
ver
[27].
The
fi
xed
-
ti
m
e
SM
C
has
be
en
me
ntio
ned
unde
r
the
c
onside
rati
on
of
model
sepa
ra
ti
on
f
or
D
ual
-
M
ot
or
Dr
i
ving
sy
ste
ms
but
the
se
par
at
io
n
te
ch
ni
qu
e
prob
le
m
was
el
imi
nate
d
[
28
]
,
[
29]
.
Howe
ver,
mos
t
of
the
pr
e
vious
w
ork
i
n
m
ulti
-
m
otor
dr
i
ve
co
ntr
ol
s
ys
te
ms
we
re
no
t
onl
y
menti
on
e
d
to
fi
nite
t
ime
c
onverge
nc
e
in
SM
C
te
c
hn
i
que,
but
al
so
not
care
d
a
bout
i
nf
l
uen
ce
of
el
a
sti
c
and
fr
ic
ti
on.
Furthe
rm
or
e
,
obse
rv
e
r
desi
gn
f
or
te
ns
io
n
was
only
c
onside
red
as
li
near
a
ppr
ox
imat
e
model
a
s
well
as
th
e
separa
ti
on
pr
inciple
has
not
been
mentio
ned.
I
n
t
his
pa
per,
we
c
on
si
der
t
he
L
M
I
base
d
fi
nite
-
ti
me
SM
C
for
mu
lt
i
-
m
oto
r
s
ys
te
ms
in
pres
ence
of
el
ast
ic
,
fr
ic
ti
on
as
well
a
s
t
he
high
gain
obse
rv
e
r
te
c
hn
i
que
is
deter
mine
d
in
our
work.
It
is
w
or
t
h
noti
ng
tha
t
the
S
M
C
base
d
sta
te
fee
db
ac
k
c
ontr
ol
sc
he
me
gu
a
ra
ntees
the
el
imi
natio
n
of
disa
dv
a
nta
ge
of
disturba
nc
e
an
d
un
ce
rtai
nties.
Ther
e
f
or
e,
co
nsi
der
at
io
n
of
usi
ng
high
-
gai
n
ob
s
er
ver
is
i
nvest
igate
d
to
c
omp
ute
the
te
nsi
on
i
n
this
sy
ste
m
a
nd
co
mb
i
ne
wit
h
the
sta
te
fee
dback
c
ontr
ol
scheme
to
i
mp
le
ment
the
ou
t
pu
t
feedback
c
on
trolle
r
gu
a
ra
nteed
the
separ
at
io
n
pr
i
nciple.
T
he
sta
bili
zat
ion
of
c
ascade
s
ys
te
m
is
sat
isfie
d
by
the
outp
ut
fee
db
ac
k
con
t
ro
l al
gorithm a
nd s
how
n by the
or
et
ic
al
analysis,
sim
ulati
on
s.
2.
DYN
AM
I
C M
ODEL O
F
A ROBOT
M
A
NIP
ULATO
R
AND
PR
OBL
EM ST
ATEM
ENT
As
t
he
w
ork
descr
i
bed
in
[
30],
the
m
ode
l
of
a
m
ulti
-
m
otor
s
ys
te
ms
with
fr
ic
ti
on,
back
la
s
h
an
d
el
ast
ic
can
be
r
epr
ese
nted
as
(
1)
:
1
1
1
2
2
2
1
1
1
1
1
1
1
1
1
21
1
2
2
2
2
2
2
2
2
2
12
2
12
12
1
1
2
2
12
12
21
12
1
1
2
2
21
12
12
1
.
(
)
(
)
(
)
1
.
(
)
.
(
)
(
)
1
(
1
)
.
1
(
1
)
.
rL
rL
L
c
b
L
L
L
c
b
L
L
LL
LL
k
f
k
g
T
r
F
K
k
f
k
g
T
r
F
K
F
C
r
r
F
Cl
F
C
r
r
F
Cl
yF
=
−
=
−
=
+
−
+
=
+
−
−
=
−
+
=
−
+
=
(1)
In
w
hic
h,
the
par
a
mete
rs
of
t
his
s
ys
te
m
are
sh
ow
n
i
n
Tabl
e
1
.
I
n
orde
r
t
o
co
ns
i
der
the
model
(1)
i
n
the stat
e sp
a
ce
r
e
p
r
e
s
e
n
t
a
t
i
o
n
,
t
h
e
s
t
a
t
e
v
a
r
i
a
b
l
e
s
a
n
d
c
o
n
t
r
o
l
i
n
p
u
t
s
a
r
e
g
i
v
e
n
t
o
o
b
t
a
i
n
t
h
e
s
t
a
t
e
s
p
a
c
e
m
o
d
e
l
(2):
1
2
3
4
5
6
1
2
1
2
2
1
1
2
T
T
LL
x
x
x
x
x
x
x
F
F
=
=
;
1
2
1
2
TT
rr
u
u
u
==
;
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8
694
In
t J
P
ow
Ele
c
&
D
ri
S
ys
t,
V
ol
.
12
, N
o.
1
,
Ma
rch
20
21
:
10
–
19
12
1
1
3
2
2
4
3
1
1
1
1
1
1
1
1
5
1
4
2
2
2
2
2
2
2
2
6
2
5
12
1
3
2
4
5
12
6
12
1
3
2
4
6
12
;
1
.
(
)
.
(
)
(
)
1
.
(
)
.
(
)
(
)
1
(
1
)
.
1
(
1
)
.
c
b
L
L
c
b
L
L
x
u
x
x
u
x
x
k
f
x
k
g
x
T
r
x
K
x
k
f
x
k
g
x
T
r
x
K
x
C
r
x
r
x
x
Cl
x
C
r
x
r
x
x
Cl
=
−
=
−
=
+
−
+
=
+
−
−
=
−
+
=
−
+
(2)
Table
1.
D
yn
a
mic pa
rameter
Para
m
aters
Exp
lain
atio
n
J
1
, J
2
, J
L1
, J
L2
Moto
rs’s I
n
er
tia
m
o
m
en
t,
Load
s’s
Inertia
m
o
m
en
t
(kg
m
2
)
,
,
∅
The to
rqu
e of
Moto
r,
L
o
ad
(
Nm
),
rot
o
r’
s Flu
x
(
W
b
)
Ro
to
r’
s Self
-
in
d
u
ctio
n
(
H)
,
,
,
,
Ro
ller
‘s Rad
iu
s, v
elo
city
r
atio
,
roto
r’
s electric
ang
le vel
o
city
,
stato
r’
s an
g
l
e
v
elo
city
,
b
elt tens
io
n
ω
1
, ω
2
,
ω
r1
, ω
r2
The ang
le velo
city
of m
o
to
r
,
lo
ad
c
1
, c
2
, b
1
, b
2
The coeffi
cien
ts o
f
Stif
fnes
s an
d
fr
icti
o
n
∆
1
,
∆
2
The er
rors o
f
an
g
le
sp
eeds
un
d
er
th
e influ
en
ce of back
la
sh
,
elastic
Re
mark 1
:
Un
li
ke
the
de
scriptio
n
was
est
ablishe
d
i
n
[
1
]
-
[
4],
this
w
ork
c
onside
rs
mu
lt
i
-
m
otor
s
yst
ems
in
prese
nce
of
nonli
near
pro
per
t
y,
bac
klas
h,
fr
ic
ti
on,
el
ast
ic
ph
e
nome
non.
T
he
refor
e
,
the
trans
fer
-
f
un
ct
i
on
base
d
a
ppr
oa
ch
in
[
1
]
-
[
4]
has
no
t
bee
n
m
entione
d
i
n
thi
s
pa
per
due
to
these
chall
enges.
The
mai
n
co
ntro
l
obje
ct
ive
to
fin
d
the
velo
ci
ti
es
12
,
rr
ob
ta
ini
ng
the
trac
king
of
sta
te
va
riab
le
s
vecto
r
=
[
∆
1
,
∆
2
,
1
,
2
,
21
,
12
]
.
Furth
erm
or
e,
beca
use
it
is
ha
r
d
t
o
e
sta
blish
the
se
ns
ors
in
thi
s
mu
lt
i
-
m
otor
s
ys
te
m,
the
co
ntr
oller
needs
to
be
ad
de
d
more
t
he
sta
te
obse
rv
e
r
obta
ining
the
se
pa
rati
on
pr
i
nciple.
Fo
r
the
outp
ut
feedbac
k
c
on
t
r
ol d
e
sig
n,
t
he
a
ssu
m
ptio
ns
a
re
intr
oduce
d
as
fo
ll
ows:
Assum
ption
1.
The
diamet
er
of
eac
h
m
ot
orcycle
is
ne
glig
ible
com
par
e
d
to
le
ngth
of
t
he
conve
yor
bel
t
of
a
mu
lt
i
-
m
otor s
yst
em.
Assum
ption
2.
T
he
fr
ic
ti
on
a
nd
sli
p
c
oeffic
ie
nts
of
the
c
onve
yor
belt
of
a
m
ulti
-
mo
t
or
s
ys
te
m
are
c
onsta
nt
and
the
y
we
re
no
t
de
pende
d
on
loa
ds
,
diam
et
er
of
eac
h
m
otorcycle
is
ne
gligible
c
ompa
red
t
o
le
ng
th
of t
he
c
onve
yor belt
of
a m
ulti
-
mo
t
or
sy
ste
m.
3.
FINITE
-
TI
M
E OUTP
UT F
EE
DBACK S
LIDIN
G
MO
DE CONT
ROL
D
ESIG
N
3.1.
Fini
te
-
time
t
r
ackin
g
sli
din
g mod
e c
on
tr
ol
design
Accor
ding to
the
m
o
d
e
l
(
3
)
a
n
d
A
s
s
u
m
p
t
i
o
n
1
,
2
,
t
h
e
t
r
a
c
k
i
n
g
e
r
r
o
r
m
o
d
e
l
c
a
n
b
e
o
b
t
a
i
n
e
d
a
s
(3)
:
2
2
5
1
1
2
1
5
1
2
2
6
2
2
5
2
1
2
1
5
1
2
2
6
3
1
5
4
2
6
5
1
1
3
1
1
3
1
1
2
1
6
2
2
4
2
2
4
2
2
1
2
1
.
(
)
.
(
)
(
)
1
.
(
)
.
(
)
(
)
c
b
L
L
c
b
L
L
r
x
x
x
C
r
x
C
r
x
l
r
x
x
x
C
r
x
C
r
x
l
x
u
x
x
u
x
x
k
f
x
k
g
x
T
r
x
K
x
k
f
x
k
g
x
T
r
x
K
=
−
−
=
−
−
=−
=−
=
+
−
+
=
+
−
−
(3)
It can
b
e
r
e
wr
it
te
n
by:
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
P
ow Elec
& Dri S
ys
t
IS
S
N: 20
88
-
8
694
On fi
nite
-
ti
me ou
t
pu
t f
ee
dbac
k sli
ding
mode
con
tr
ol
of an
el
as
ti
c mult
i
-
m
oto
r syste
m (P
ham
T
uan T
hanh
)
13
12
1
12
2
11
12
1
12
2
22
33
44
1
1
55
66
2
2
0
0
0
0
.
.
0000
0
0
0
0
.
.
0000
0
0
0
0
1
0
1
0
0
0
0
0
0
0
0
1
.
0
1
0
0
0
0
0
0
0
0000
0000
0
0
0
0
0
L
L
C
r
C
r
xx
C
r
C
r
xx
xx
xx
r
K
xx
xx
r
K
−
−
−
−
=+
−
(
)
1
2
.,
0
0
u
u
d
x
t
+
(4)
wh
e
re:
1
12
T
z
x
x
=
;
2
3456
T
z
x
x
x
x
=
;
11
00
00
A
=
;
1
2
1
1
2
2
12
1
2
1
1
2
2
00
00
C
r
C
r
A
C
r
C
r
−
=
−
1
21
1
2
2
00
00
0
0
L
L
r
A
K
r
K
−
=
;
22
0
0
1
0
0
0
0
1
0
0
0
0
0
0
0
0
A
−
−
=
2
1
0
0
0
0
1
0
0
0000
0000
B
=
;
1
0000
0000
B
=
(
)
1
1
3
1
1
3
1
2
2
1
2
2
4
2
2
4
2
2
0
0
1
.
(
)
.
(
)
,
1
.
(
)
.
(
)
c
b
L
L
c
b
L
L
k
f
x
k
g
x
T
d
z
t
K
k
f
x
k
g
x
T
K
+−
=
+−
therefo
re
, t
he
t
rack
i
ng er
ror mo
del (4)
can
be rep
rese
nted as:
(
)
1
2
1
1
2
2
1
1
2
1
2
1
1
2
2
1
2
1
2
22
2
1
2
2
00
00
00
00
00
0
0
1
0
00
0
0
0
1
0
,
0
0
0
0
0
0
0
0
0
L
L
C
r
C
r
z
z
z
C
r
C
r
r
z
z
z
B
u
d
z
t
K
r
K
−
=+
−
−
−
−
=
+
+
+
(5)
Con
si
der the sl
iding va
riable
(
)
(
)
12
1
T
s
e
e
A
P
e
=
−
(6)
wh
e
re
z
r
e
f
ez
=−
an
d
(
)
1
11
ee
−
=−
,
HI
=
,
the
fi
nite t
ime sl
iding m
od
e
c
on
t
ro
l l
a
w
is
propose
d
as
d
e
sc
ribe
d
in
the
f
ollow
in
g The
or
e
m 1.
The
or
em
1
: C
onside
r
the
s
ys
te
m (5) wit
h t
he
d
ist
urba
nce
(
2
(
,
)
)
be
ing
boun
ded
by
2
d
and the
S
MC
is
giv
e
n
as
;
(
)
=
(
)
−
2
−
1
(
)
(7)
in
w
hich
(
)
(
)
1
1
2
1
2
T
d
n
de
u
t
B
A
z
A
P
z
dt
−
=
−
−
−
,
(
)
(
)
(
)
(
)
(
)
(
)
(
)
2
s
g
n
s
g
n
s
u
t
d
s
e
s
e
s
e
s
e
=
+
+
,
;
ar
e
the
posit
ive
co
ns
ta
nt
numb
e
r
s.
The
pro
pos
ed
S
M
C
ena
bl
es
us
to
obta
in
the
finite
-
ti
me
Inp
ut
Stat
e
Stabil
it
y
(I
S
S)
sta
bili
ty.
Pr
oo
f
: The
Proof is
div
ide
d
i
nto t
w
o
ste
ps
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8
694
In
t J
P
ow
Ele
c
&
D
ri
S
ys
t,
V
ol
.
12
, N
o.
1
,
Ma
rch
20
21
:
10
–
19
14
In Step
1,
we p
rove t
hat the cl
os
e
d
s
ys
te
m
re
aches t
o
the
sli
ding s
urface i
n fin
it
e ti
me:
Ba
sed on t
he L
yapu
nov
ca
ndidate
fu
nction u
sing t
he
sli
di
ng v
a
riable as:
(
)
(
)
(
)
(
)
2
1
2
T
V
s
e
s
e
s
e
=
(8)
ta
kin
g
the ti
me
d
e
rivati
ve o
f
t
his L
yapu
nov f
un
ct
io
n
al
ong t
he
s
ys
te
ms
(5),
w
e
obta
in:
(
)
(
)
(
)
(
)
(
)
(
(
)
(
)
)
(
)
(
)
(
)
1
1
2
1
2
2
1
12
21
22
2
2
1
12
2
2
,
=
s
,
T
T
T
dd
TT
d
dg
e
V
s
e
s
e
s
e
s
e
H
z
A
P
A
z
A
z
B
u
H
z
z
d
z
t
dt
dg
e
e
A
z
A
P
B
u
z
d
z
t
dt
=
=
−
+
+
+
−
−
+
−
+
−
+
(9)
su
bst
it
uting
(
7) into (
9)
giv
e
s:
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
2
2
2
s
g
n
sg
n
,
TT
TT
V
s
e
s
e
s
e
s
e
s
e
s
e
s
e
d
s
e
s
e
d
z
t
=−
−+
(10)
accor
ding
t
o
(
)
(
)
2
2
x
,
ma
d
d
z
t
, we
hav
e:
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
2
21
2
m
i
n
m
i
n
2
2
2
2
V
s
e
s
e
s
e
V
s
e
V
s
e
+
−
−
=
−
−
(11)
Choose
(
)
(
)
(
)
(
)
1
/
2
2
m
in
2
m
in
2
2
0
,
2
0
,
1
/
2
1
+
=
=
=
+
, acc
ordin
g
to
the i
ne
qu
al
it
y (
11),
the close
d
s
ys
t
em r
eac
hes
t
o
s
li
din
g
surface
in finit
e ti
me g
i
ven as
(12
)
:
(
)
(
)
(
)
(
)
2
1
2
2
0
2
2
2
2
2
1
ln
1
r
V
s
e
t
t
−
+
=
−
(12)
wh
e
re
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
2
0
0
0
1
/
2
T
V
s
e
t
s
e
t
s
e
t
=
(13)
In
Step
2,
on
c
e
the
cl
ose
d
s
ys
te
m
traj
ect
ory
reac
hes
t
he
sli
ding
s
urfac
e,
the
e
rro
rs
c
onve
rg
e
to
at
tract
ion
reg
i
on
in
fi
nite
ti
me.
Co
ns
ide
r
t
he
cl
os
ed
syst
em
in
the
sli
ding
s
urface
(
)
(
)
12
1
0
T
s
e
e
A
P
e
=
−
=
.
Choose
an
d
k
are
c
on
sta
nt
c
oeffici
ents,
i
n
wh
ic
h
(
t
)
;
f
e
k
.
The
pa
rameters
0
,
1
0
và
2
0
an
d
the
matri
ces
0
,
0
,
0
,
0
X
Q
W
G
an
d
Y
ar
e
ch
os
en
to
s
at
isfy
the
f
ollow
i
ng
Linea
r
M
at
rix
Ine
qu
al
it
ie
s (
LMIs):
1
1
1
1
1
2
1
2
0
T
T
T
A
X
X
A
A
Y
Y
A
Q
X
XW
+
−
−
+
−
(14)
1
2
1
2
2
0
T
A
A
X
XG
−
(15)
1
2
2
0
IQ
−
(16)
2
2
2
0
IW
−
(17)
2
12
2
0
k
−
(18)
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
P
ow Elec
& Dri S
ys
t
IS
S
N: 20
88
-
8
694
On fi
nite
-
ti
me ou
t
pu
t f
ee
dbac
k sli
ding
mode
con
tr
ol
of an
el
as
ti
c mult
i
-
m
oto
r syste
m (P
ham
T
uan T
hanh
)
15
The
n
the
s
ys
t
em
(
5)
is
sta
bili
zed
i
n
fi
nite
ti
me
wit
h
the
at
tract
ion
re
gion
(
)
1
e
t
k
.
T
he
Pro
of
is
simi
la
rly
im
plemented
as a
bo
ve
ste
p wit
h
th
e co
rr
es
pondin
g
L
ya
punov F
unct
ion
1
1
1
1
1
(
)
,
T
V
e
e
P
e
P
X
−
==
3.2.
High
-
gain
ob
s
erver
c
ontrol
design
of mul
t
i
-
mo
to
r
syste
ms
As
descr
i
be
d
i
n
[
25]
,
it
is
ha
rd
to
fin
d
th
e
obser
ve
r
f
or
mu
lt
i
-
ou
t
pu
t
sy
ste
ms
beca
use
the
data
colle
ct
ion
nee
ds
t
o
be
im
plemented
in
the
su
f
fici
ently
sm
al
l
interval
as
well
as
em
ployin
g
t
he
ta
ylor
series
appr
ox
imat
io
n.
Th
ere
fore,
the
hi
gh
-
ga
in
obs
erv
e
r
was
pr
opos
e
d
by
[15]
e
nab
li
ng
us
t
o
de
al
with
m
ulti
-
ou
t
put
sy
ste
ms
. Consi
der the cla
ss
of
non
li
nea
r u
nif
ormly
obser
va
ble s
ys
te
ms:
(
)
1
,
z
f
u
z
y
C
z
z
=
+
==
(
19
)
wh
e
re
t
he
sta
te
va
riables
n
zR
wit
h
12
1
;
1
,
2
,
3
,
.
.
.
.
,
;
.
.
.
;
k
n
n
k
qk
k
z
R
k
q
n
n
n
n
n
=
=
=
,
the
in
pu
t
uU
is
a
com
pact set
of
R
m
,
the
ou
t
pu
t
var
ia
bles
1
n
yR
.
(
)
1
1
2
1
12
1
1
1
2
3
2
2
1
1
(
u
,
z
,
z
)
0
(
u
,
z
,
z
,
z
)
;
,
;
0
;
,
0
,
.
.
.
,
0
(
u
,
z
)
0
q
n
n
n
n
n
q
q
q
q
f
z
f
z
z
f
u
z
C
I
f
z
−
−
=
=
=
=
1
1
2
1
,
0
,
.
.
.
,
0
q
n
n
n
n
n
CI
=
This
a
bove
s
yst
em
al
so
sat
isfi
es
se
ver
al
ass
umpti
ons
as
des
cribe
d
i
n
[15
].
It
is
hard
to
fi
nd
directl
y
the
Hi
gh
-
G
ai
n
ob
s
er
ver
f
or
m
od
el
(19
),
s
o
t
hat
it
can
be
tr
ansf
e
rr
e
d
i
nto
the
f
ollo
wing
f
orm
to
ea
sie
r
de
sign
base
d on the tr
ansfo
rmati
on a
s foll
ow
s:
1
2
1
1
1
12
1
(
.)
1
2
3
2
00
2
(
.)
1
1
(
u
,
z
,
z
)
(
u
,
z
,
z
,
z
)
:
,
(
z
)
(
u
,
z
)
k
k
f
nq
n
z
q
f
q
z
i
z
f
f
z
f
+
−
−
=
→
=
=
;
We
im
ply
0
1
(
u,
)
z
A
yC
=
+
+
==
and
c
onti
nue
t
o
f
in
d
the inter
mediat
e obser
ve
r
[
15
]:
1
1
1
1
1
1
ˆ
ˆ
ˆ
ˆ
(
u
,
)
(
u
,
)
TT
A
S
C
S
C
−
−
−
−
−
−
=
+
−
−
.
Be
ca
us
e
1
0
AA
−
=
và
0
CC
=
then
em
ployin
g
0
=
, we
ob
ta
in
the
dyn
a
mic eq
uat
ion
of
ob
se
r
ver erro
rs:
1
1
1
0
ˆˆ
(
A
S
C
C)
(
(
u
,
)
)
(
u
,
)
TT
C
z
−−
=
−
+
−
−
.
I
n
order
to
analyze
the
ob
serv
e
r
er
rors,
sel
ect
ing
the
cand
i
date
f
unct
ion
1
()
T
VS
=
,
ta
ke
the
de
rivati
ve
of
this
functi
on
al
ong
the
sy
ste
m
traj
ect
ory of
t
r
ansfo
rme
d
s
ys
t
ems:
21
00
1
1
0
m
in
ˆ
(
u
,
(
)
)
2
(
S
)
(
s
)
c
VV
z
−
+
+
+
wh
e
re
x
m
in
(
s
)
(
S
)
(
s
)
ma
=
base
d
on the
w
ork
i
n
[
15]
, we
imply
that:
(
)
2
1
1
1
1
1
q
c
V
c
V
V
−
−
−
+
(
20
)
the
fact t
hat is
the foll
owin
g
s
el
ect
ion
s:
1
1
1
1
11
(
I
,
I
,
.
.
.
,
I
)
q
n
n
n
b
l
o
c
k
d
i
a
g
−
=
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8
694
In
t J
P
ow
Ele
c
&
D
ri
S
ys
t,
V
ol
.
12
, N
o.
1
,
Ma
rch
20
21
:
10
–
19
16
ˆ
(
z
)
+
is t
he
le
ft i
nve
rse of
b
l
ock d
i
agonal mat
rix
ˆ
(
z
)
with
1
12
1
1
21
1
ˆ
ˆˆ
(
u
,
z
)
(
u
,
z
,
z
)
ˆ
(
z
)
b
l
o
c
k
d
ia
g
,
,
.
.
.
,
ˆˆ
q
k
n
k
i
f
f
I
zz
−
+
=
=
wh
e
re
S
is t
he
po
sit
ive
d
e
finit
e so
l
ution o
f
th
e algeb
raic
Lya
puov
e
qu
at
ion
(
21)
:
0
TT
S
A
S
S
A
C
C
+
+
−
=
(
21
)
come
back
to
th
e model
(19), t
he
e
qu
i
valent
obser
ve
r
ca
n be
determi
ned as
(22)
:
11
ˆ
(
z
)
ˆ
ˆ
ˆ
ˆ
ˆ
(
)
f
(
u
,
z
)
(
z
)
(
z
z
)
T
z
S
C
C
z
+
+
−
−
=
=
−
−
(22
)
al
tho
ug
h
im
ple
mentin
g
th
e
observ
e
r
i
nterme
diate
ly, we sti
ll
obtai
n
the
dire
ct
resu
lt
(23)
:
1
1
2
1
2
2
11
1
1
1
(
u
,
z
)
ˆ
(
z
)
(
u
,
z
)
q
n
q
T
q
qq
k
q
k
i
CI
f
C
x
SC
f
C
z
+
+
−
−
+
−
+
=
=
(
23
)
wh
e
re
1
1
2
(
i,
j)
(
1
)
i
j
j
i
j
n
S
C
I
+−
+−
=−
with
!
;1
i,
j
q
!
(
j
i)
!
i
j
j
C
i
=
−
in
orde
r
t
o
fi
nd
co
ntr
ol
de
sign
a
ppr
opria
te
HGO
for
mu
lt
i
-
m
otor
s
ys
te
ms
(1),
w
e
nee
d
to
i
m
plement
al
te
rn
at
el
y:
Con
si
der the
f
ollow
i
ng
m
ulti
-
m
oto
r
s
ys
te
ms
:
12
2
1
1
2
2
1
3
3
12
1
21
2
22
3
12
3
1
()
1
(
)
K
(
)
(
)
(
T
.
)
1
(
1
)
.
L
C
L
TC
x
u
x
T
x
J
f
x
u
x
f
x
r
x
K
x
C
r
x
r
x
x
Cl
yx
=−
=
+
−
−
+
=
−
+
=
w
he
re
31
1
1
1
2
1
1
2
1
2
2
2
1
2
3
1
2
2
2
2
2
3
2
1
2
;
;
r
r
x
x
x
F
x
x
x
x
x
x
F
=
=
=
=
=
=
accor
ding
t
o
th
e res
ult (
22),
w
e obtai
n
th
e e
quivale
nt
obse
rver:
1
2
3
3
2
2
1
1
2
2
1
3
3
3
3
3
12
1
21
2
22
3
3
3
12
3
1
ˆ
ˆ
ˆ
(
u
x
)
3
(
x
)
1
ˆ
ˆ
ˆ
ˆ
ˆ
ˆ
(
)
K
(
)
(
)
(
T
.
)
(
)
1
ˆ
ˆ
ˆ
ˆ
ˆ
(
1
)
(
)
.
y
L
C
L
TC
L
xx
T
x
J
f
x
u
x
f
x
r
x
x
x
KT
x
C
r
x
r
x
x
rJ
x
x
Cl
x
=
−
−
−
=
+
−
−
+
+
−
=
−
+
+
−
=
(24)
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
P
ow Elec
& Dri S
ys
t
IS
S
N: 20
88
-
8
694
On fi
nite
-
ti
me ou
t
pu
t f
ee
dbac
k sli
ding
mode
con
tr
ol
of an
el
as
ti
c mult
i
-
m
oto
r syste
m (P
ham
T
uan T
hanh
)
17
4.
OFFLINE
SI
MU
L
ATIO
N RESULTS
To
cl
ea
rly
val
idate
the
e
ff
ic
acy
of
the
pro
po
s
ed
outp
ut
f
eedb
ac
k
co
ntr
ol
sc
heme,
a
mu
lt
i
-
m
otor
dr
i
vi
ng
s
ys
te
m
is
est
a
blishe
d
in
Fig
ure
1,
wh
ic
h
sho
ws
t
he
physi
cal
m
eanin
g
of
pa
ra
mete
rs
a
s
well
as
the
com
plete
d
c
on
trol
s
ys
te
m
.
T
he
sim
ulati
on
resu
lt
s
i
n
Fig
ures
2
an
d
3
de
scribe
the
be
ha
vior
of
te
n
si
on
a
nd
velocit
ie
s
wh
e
n
t
he
l
oad
is
c
hange
d
from
T
L1
=
100
Nm
to
T
L2
=
50
N
m.
It
sh
oul
d
be
note
d
t
hat
the
c
omparis
on
betwee
n
the
respo
ns
e
us
in
g
propose
d
s
olu
ti
on
a
nd
the
cl
assic
al
a
ppr
oach
es
has
bee
n
in
vestigat
ed.
Fu
rt
hermo
re
,
the
sim
ulati
on
resu
lt
s
i
n
Fi
gure
4
s
how
the
respo
ns
e
of
propose
d
high
ga
in
ob
se
r
ver
in
Mult
i
-
M
ot
or syst
ems
.
Figure
1. The
s
tructu
r
e
of m
ulti
-
mo
to
r
c
ontro
l system
(a)
(b)
Figure
2.
(a
)
T
he
respo
ns
e
of
te
ns
io
n
P
ID
wi
th the l
oad
var
i
at
ion
(b)
T
he r
esp
on
se
of te
nsi
on
LMI
-
S
M
C
with
the loa
d variat
ion
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8
694
In
t J
P
ow
Ele
c
&
D
ri
S
ys
t,
V
ol
.
12
, N
o.
1
,
Ma
rch
20
21
:
10
–
19
18
(a)
(b)
Figure
3.
(a
)
T
he
respo
ns
e
of
sp
ee
d
P
ID
wit
h
the
loa
d vari
at
ion
(b)
T
he r
esp
on
se
of s
pe
ed
L
MI
-
S
M
C
with
the
load va
riat
ion
(a)
(b)
Fig
ure
4. The
re
spon
se
of
obs
erv
e
r,
(
a)
tensi
on obse
rv
e
r,
(
b) s
peed O
bs
e
rver,
with
gear g
ap
=
0.1
5
ra
d
(8,59
0
)
5.
CONCL
US
I
O
N
To
deal
with
the
mu
lt
i
-
m
otor
sy
ste
ms
i
n
pr
e
sence
of
nonli
near
pr
op
e
rty,
back
la
s
h,
el
ast
ic
,
the
finite
ti
me
sli
din
g
m
od
e
c
ontrolle
r
com
bin
in
g
with
hi
gh
gain
observ
e
r
is
de
vel
op
e
d
in
this
w
ork.
T
he
co
nve
rg
e
nce
ti
me
of
S
M
C
is
cl
early
est
imat
ed
by
LMI
te
chn
i
qu
e
a
nd
th
e
ob
se
r
ver
is
de
al
t
with
mu
lt
i
-
outp
ut
sy
ste
m
s.
The
pro
po
se
d
c
ontr
oller is
validat
ed via si
m
ulati
on
s
in
compa
rison w
it
h
e
xisti
ng so
l
utio
n.
ACKN
OWLE
DGE
MENTS
This
r
e
s
e
a
r
c
h
w
a
s
s
u
p
p
o
r
t
e
d
b
y
R
e
s
e
a
r
c
h
F
o
u
n
d
a
t
i
o
n
f
u
n
d
e
d
b
y
T
h
a
i
N
g
u
y
e
n
U
n
i
v
e
r
s
i
t
y
of T
echnolo
gy.
REFERE
NCE
S
[1]
Bousmaha
Bouc
hiba
,
Isma
il
Kh
al
il
Bouss
erh
an
e,
Mohammed
Kari
m
Fellah,
Haz
za
b
,
“Ne
ur
a
l
Ne
twork
Slid
i
ng
Mode
Contro
l
f
or
Multi
-
Ma
chine
We
b
W
inding
Sys
te
m
,
”
R
e
vue
Roumaine
des
Sc
ie
nc
es
Te
chni
ques
-
S
eri
e
Él
e
ct
rote
chni
qu
e
e
t Éne
rgét
ique
,
vol. 62,
no.
1,
p
p.
109
-
113
,
201
7.
[2]
Haka
n
Ko
c,
Do
mi
nique
Kni
tt
e
l,
Mich
el
de
Ma
th
el
in
,
and
Gabri
el
Abba
,
“Mode
l
in
g
and
Robust
Co
ntrol
of
Wi
nd
ing
Sys
te
ms f
or
E
la
s
ti
c
We
bs
,
”
IEEE
Tr
ansacti
on
on
Control
Syst
e
ms
Technol
ogy
,
vo
l. 10,
no.
2,
pp.
19
7
-
208,
2002
.
[3]
Domini
que
Knit
te
l
,
Edou
ard
La
r
oche
,
Dani
el
Gi
gan,
and
Hak
an
Koç,
“T
ension
C
ontrol
for
W
inding
Sys
te
ms
W
ith
Two
-
Degre
es
-
of
-
Free
dom
Hinf
Control
le
rs
,
”
IE
EE
Tr
ansacti
ons
on
Industry
Ap
pli
cation
,
vol
.
3
9,
no
.
1,
pp
.
11
3
-
120,
2003
.
[4]
Jie
-
Shiou
Lu
,
M
ing
-
Yang
Cheng
,
Ke
-
Han
Su,
M
i
-
Chi
Tsai
,
“
Wi
r
e
t
ension
cont
rol
of
an
au
tom
a
tic
mot
or
winding
ma
ch
ine
an
it
er
a
ti
ve
learni
ng
sli
ding
mod
e
con
tr
ol
appr
o
ac
h
,
”
R
oboti
cs
and
Co
mputer
-
Inte
grat
ed
Manufa
ct
urin
g
,
vol.
50
,
pp
.
50
-
6
2,
2018
.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
P
ow Elec
& Dri S
ys
t
IS
S
N: 20
88
-
8
694
On fi
nite
-
ti
me ou
t
pu
t f
ee
dbac
k sli
ding
mode
con
tr
ol
of an
el
as
ti
c mult
i
-
m
oto
r syste
m (P
ham
T
uan T
hanh
)
19
[5]
Mohamm
ed
A
.
A.
Al
-
Mekhla
f
i,
Herm
an
Wa
hid
,
Azia
n
Abd
Azi
z
,
"A
dapt
iv
e
Neu
ro
-
Fuzzy
Contro
l
Approac
h
for
a
Single
Inv
erted
Pendulum
Sys
tem
,"
In
te
rnationa
l
Journal
of
E
lec
tric
al
and
Comp
ute
r
Eng
ineering
(IJ
ECE
),
vol
.
8,
no.
5
,
pp
.
3657
-
3665
,
2018
.
[6]
Li
u
Y.
C
.
,
Dao
N.,
Zh
ao
K.
Y.
,
“On
Robust
Control
of
Nonlin
ea
r
Te
l
eope
r
at
ors
u
nder
Dynamic
U
nce
rt
ai
nt
ie
s
with
Vari
able
T
im
e
Dela
ys
and
wi
th
o
ut
Relati
v
e
Vel
oci
ty
,
”
I
EE
E
Tr
ansacti
ons
on
In
dustrial
Informatic
s,
v
o
l.
16
,
no
.
2,
pp.
1272
-
1280
,
2019
.
[7]
Nguyena
,
T
.
,
Hoang,
T.,
Pham,
M.,
&
Dao
,
N
.
,
“A
G
aussian
wave
let
n
et
wor
k
-
base
d
robust
ada
pt
ive
tracki
n
g
cont
roller
for
a
whee
le
d
mobi
l
e
robot
with
unkn
own
whee
l
slips,
”
Inte
r
nat
ional
J
ournal
of
Control
,
vol
.
92,
no.
11
,
pp.
2681
-
2692
,
2019.
[8]
Binh
,
N.
T.,
Tu
ng,
N.
A.
,
Nam
,
D.
P.,
Quang
,
N.
H.
,
“An
ada
p
ti
ve
ba
ckste
ppin
g
tr
aj
e
ct
ory
tr
acking
cont
rol
of
a
tra
c
tor
tra
i
le
r
wh
ee
l
ed
mobi
l
e
rob
ot,
”
I
nt
ernati
ona
l
Journal
of
Con
t
rol,
Aut
omat
ion
and
S
yste
ms
,
vo
l.
17,
no.
2
,
pp
.
465
-
473,
2019
.
[9]
Dw
i
Prihant
o
,
I
r
awa
n
Dw
i
Wahyono,
Suw
asono
and
Andrew
Nafa
lski,
"V
irt
u
al
La
bor
at
ory
f
or
Li
n
e
Fol
lowe
r
Robot
Compe
t
ition
,"
Inte
rnat
io
nal
Journal
of
El
e
ct
rica
l
and
Computer
Engi
n
ee
ring
(IJ
ECE
)
,
vol.
7,
no
.
4,
p
p.
2253
-
2260,
201
7.
[10]
Phuong
Nam
Dao,
Hong
Quang
Nguyen,
Phuon
g
N
am
Tra
n,
Ha
i
Yen
Tra
n
Thi
,
“Ada
pti
v
e
Dyna
mi
c
Progr
am
ing
base
d
Optimal
C
ontrol
for
a
Rob
ot
Mani
pul
at
or
,
”
Inte
rnationa
l
Jo
urnal
of
Powe
r
El
e
ct
ronics
and
El
e
ct
rica
l
Dr
iv
e
s
(IJ
PE
DS)
,
vol
.
1
1,
no
.
3
,
pp
.
112
3
-
1131,
2020
.
[11]
Nam,
Dao
Phuong,
et
a
l.
,
“R
obust
contro
l
f
or
buck
conv
er
te
r
base
d
on
o
pti
mization
,”
20
17
Int
ernati
ona
l
Confe
renc
e
on
S
yste
m Scien
ce a
nd
Engi
n
ee
ring
(
ICSSE)
,
Ho
Chi M
inh
Cit
y
,
2017
,
pp
.
704
-
707
.
[12]
Quang
N.
H.,
et
al
.
,
“
Mul
ti
pa
r
am
e
tri
c
mode
l
p
red
ictive
cont
ro
l
base
d
on
la
gu
e
rre
model
for
pe
rma
nen
t
ma
gn
et
li
ne
ar
synchrono
us
mot
ors,”
Int
e
rnational
Journal
of
Elec
tric
al
and
Computer
E
ngine
ering
(IJ
E
CE),
vo
l.
9,
no.
2,
pp.
1067
-
1077
,
2019.
[13]
El
-
Sous
y,
“Ada
pti
ve
Dyna
mic
Slidi
ng
Mode
Control
Sys
te
m
using
Rec
urr
e
nt
RBF
N
for
High
-
Perform
an
ce
Induc
ti
on
Motor
Servo
Driv
e,
”
I
EE
E
Tr
ansacti
o
ns on
Industrial Inform
ati
cs
,
vo
l
.
9,
no.
4,
pp.
192
2
-
1936,
2013
.
[14]
Zi
ti
ng
Ch
en,
Zh
i
jun
Li
,
Phil
ip
Khan,
Akme
li
awa
ti
,
Bha
tt
i
,
Khan,
“Robust
St
abi
liz
at
ion
of
Underac
tua
t
ed
non
li
ne
ar
Sys
te
ms:
A fa
st
t
erm
in
al
Sl
idi
ng
Mode
Approac
h
,
”
ISA
Tr
ansacti
o
ns
,
vol. 66, pp. 2
41
-
248,
201
7
.
[15]
Clé
me
n
t
Tré
a
ng
le
,
Mondher
Far
za
,
Mohammed
M.
Saad,
“Fi
lt
er
ed
high
gai
n
ob
serve
r
for
a
cla
ss
of
unce
rt
ai
n
nonli
ne
ar
sys
te
m
s with
samp
le
d
o
utput
s,”
Aut
omat
ic
a,
vol
.
101
,
pp
.
197
-
206,
2019.
[16]
Dong
Y.,
You
L.
,
B
ing
X.
,
W
ene
L
.
,
“Robust
fini
t
e
-
time
adaptive
cont
rol
algorithm
for
sa
telli
t
e
fast
at
t
it
ud
e
ma
neuv
er,”
Journal
of
the F
rank
l
in
Instit
u
te
,
vo
l.
357,
no
.
16
,
pp
.
11558
-
11583,
2
020
.
[17]
Mofid
O.,
Mob
aye
n
S.
,
“Ada
p
ti
ve
slid
ing
mo
de
cont
ro
l
for
f
ini
te
-
ti
m
e
stab
i
li
ty
of
qu
ad
-
rot
or
UA
Vs
with
par
ametr
i
c
un
ce
r
ta
inties,
”
IS
A
tra
nsacti
ons
,
vo
l.
7
2,
pp
.
1
-
14
,
201
8.
[18]
Sun
T.
,
Cheng
L.
,
W
ang
W
.
,
Pan
Y.
,
“
Se
mi
glo
bal
expone
nt
ia
l
cont
rol
of
Eu
le
r
-
La
gra
ng
e
sys
tems
using
a
sl
iding
-
mode
d
isturba
nc
e
observe
r,
”
Au
t
o
matic
a,
vol
.
11
2,
p
.
108677
,
20
20.
[19]
Yang
Xiaowe
i
,
J
ia
nyong
Yao
,
W
enxi
ang
D
eng
,
“
Output
fe
edba
ck
ada
pt
ive
sup
er
-
t
wisting
slidi
ng
mode
con
trol
of
hydra
ulic
sys
te
ms
with
distu
rba
nce
co
mpe
nsati
on
,”
ISA
transactions
,
2020.
htt
ps://
do
i.
org/1
0.
1016/j.isa
tr
a.
2
020.
09.
014
[20]
Fu
C.
,
Tian
Y.,
Huang
H.
,
Zh
a
ng
L
.
,
Peng
C
.
,
“Fin
it
e
-
ti
m
e
tr
a
je
c
tory
tracki
ng
con
trol
for
a
1
2
-
rotor
un
ma
nn
ed
ae
ri
al
v
ehicle wi
th
input sat
ur
atio
n,
”
IS
A trans
act
i
ons
,
vol.
81,
pp.
52
-
62,
2018
.
[21]
Xiao
L.,
Zhu
Y.,
“Sli
d
ing
m
ode
output
f
ee
d
bac
k
con
trol
b
a
sed
on
tra
ck
ing
err
or
observe
r
with
disturba
n
ce
esti
mator,
”
IS
A t
rans
act
ions
,
vol
.
53,
no.
4
,
pp.
10
61
-
1072,
2014
.
[22]
Christofe
r
Edwa
rds,
Yuri
Shtessel,
“Ad
apt
iv
e
C
onti
nuous
Highe
r
Order
Slidi
ng
Mode
Control,”
Aut
omati
ca
,
vol
.
65,
pp
.
183
-
190
,
2016.
[23]
Maria
Thomas,
Kama
l
,
Bandyop
adhya
y,
Va
chhani,
“Continuous
Higher
Ord
er
Sli
ding
Mode
Con
t
rol
for
a
class
of
Unce
rtain
MIM
O nonl
inear
Sys
t
em
s:
an
ISS
App
roa
ch,”
Europea
n
Journal
o
f
Con
trol
,
vo
l. 41, pp.
1
-
7,
2018
.
[24]
Zhe
ng
Zhu,
Xi
a
Fu,
“Ada
pt
ive
Slidi
ng
Mode
C
ontrol
for
At
ti
tu
de
Sta
bil
i
zation
with
Ac
tua
tor
S
at
ura
ti
on,
”
IEEE
Tr
ansacti
ons on Indus
trial
E
le
c
tronic
s
,
vo
l. 58, n
o.
10
,
pp
.
4898
-
4907,
2011
.
[25]
Hass
an
K.
Khal
il
,
La
ur
ent
Pral
y
,
“High
-
g
ai
n
o
bserve
rs
in
nonl
ine
ar
f
ee
db
ac
k
cont
rol
,
”
In
te
rn
ati
onal
Journal
of
Robust
and
Non
l
ine
ar Control
,
v
ol.
24
,
no
.
6
,
pp
.
993
-
1015,
2013
.
[26]
Ma
L.,
Zong
G.
,
Zh
ao
X.
,
Huo
X.,
“Observ
ed
-
b
ase
d
ad
apt
iv
e
fi
nit
e
-
t
ime
tr
ac
kin
g
cont
rol
for
a
c
la
ss
of
nonstri
c
t
fee
dba
ck
nonli
n
ea
r
sys
te
ms
wit
h
input
satur
at
io
n,
”
Journal
o
f
t
he
Franklin
Inst
it
ute
,
vol
.
357,
no.
16,
pp
.
1151
8
-
11544,
2019
.
[27]
Huang
Y.
,
Li
u
Y.,
“A
com
pa
ct
design
sche
me
of
ada
p
ti
ve
output
-
fee
db
ack
con
trol
for
u
nce
rt
ai
n
nonli
n
e
ar
sys
te
ms,”
In
te
rn
a
ti
onal
Journal of
Control
,
vol
.
92,
no
.
2
,
pp
.
26
1
-
269,
2019
.
[28]
Ze
ng
T
.
,
Ren
X
.
,
Zh
ang
Y.,
“Fixe
d
-
time
Slid
in
g
Mode
Contro
l
and
High
-
g
ai
n
Nonline
ar
it
y
Co
mpe
nsati
on
fo
r
Multi
-
mot
or
Dri
ving
Sys
te
m
,
”
I
EE
E
Tr
ansacti
o
ns on
Industrial Inform
ati
cs
,
vo
l. 16, no. 6, pp. 40
90
-
4098,
2020
.
[29]
Inli
n
Wa
ng
,
Xu
em
e
i
Ren
,
and
Qiang
Chen
,
“
Casca
de
Optima
l
Control
for
T
rac
king
and
Sy
nchr
onizati
onof
a
Multi
mot
or
Dr
i
ving
Sys
te
m,”
I
EE
E
Tr
ansacti
o
ns
on
Control
Syste
ms
Tecnol
og
y
,
vol
.
27
,
no.
3
,
pp.
1376
-
1384
,
2019.
[30]
Nam
D.
P., Th
an
h
P.
T.,
T
i
nh
T
.
X.,
Dat T.
T.,
Va
n
V.
M.,
“High
-
Gain
Obs
erv
er
B
ase
d
Output
Fe
e
dbac
k
Contro
ll
er
for
a
Two
-
Moto
r
Drive
Sys
te
m
:
A
Separ
ation
P
rinc
iple
Approa
ch,
”
AE
TA
201
7
-
Re
c
ent
Adv
an
ce
s
in
El
e
ct
ri
ca
l
Engi
ne
ering
and
Relate
d
Sc
ie
nc
e
s: Theory
and
A
ppli
cation
,
vol
4
65,
pp
.
840
-
859
,
2018.
Evaluation Warning : The document was created with Spire.PDF for Python.