TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol.12, No.6, Jun
e
201
4, pp. 4125 ~ 4
1
3
3
DOI: 10.115
9
1
/telkomni
ka.
v
12i6.508
7
4125
Re
cei
v
ed
No
vem
ber 1
0
, 2013; Re
vi
sed
De
cem
ber 3
1
,
2013; Accep
t
ed Jan
uary 1
4
, 2014
Adaptive Fuzzy
Sliding Mode Control for Hydraulic
Servo System of Parallel Robot
Fenglan Jia*, Li Hou, Yongqiao Wei, Yunxia You, Lili Yan
Schoo
l of Man
u
facturin
g Scie
nce an
d Eng
i
n
eeri
ng, Sich
ua
n Univ
ersit
y
,
NO.24, South Section 1, Yi
hu
an Ro
ad,
Ch
en
gdu, Ch
in
a, 61
006
5, +
86-81
2-
337
00
0
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: jfl198
9scu@
163.com
A
b
st
r
a
ct
The hydra
u
lic
servo contro
l system,
w
h
ich i
s
the imp
o
rtant
comp
on
ent
of para
lle
l robot, i
s
a hig
h
order, nonlinear, dynam
i
c
asymm
e
try, par
ameter unc
ertain
system
,
which serious
ly
affect dy
namic
perfor
m
a
n
ce
o
f
robot, so
it i
s
very
diffic
u
lt
to ga
in
go
od
perfor
m
a
n
ce w
i
th traditi
on
al c
ontrol. B
a
sed
on
slidi
ng
mo
de
varia
b
le struct
ure contro
l method co
mb
in
e
d
w
i
th fu
zz
y
c
ontrol a
nd ad
aptive co
ntrol,
an
ada
ptive
fu
zz
y
slidi
n
g
mod
e
c
ontrol
al
gor
ith
m
w
a
s
pr
es
ent
ed
by tak
i
ng
th
e Stew
art p
l
atform serv
o co
ntro
l
system of para
llel ro
bot as the res
earc
h
obj
ect, and the un
certain ter
m
of the slidin
g mo
de contro
ller w
a
s
appr
oxi
m
ate
d
w
i
th ada
ptive f
u
zz
y
c
ontro
l
method. T
h
e si
mulati
on r
e
sults
show
ed th
at c
o
mpar
ed w
i
th t
h
e
conventional P
I
D control applied to
the same hy
draulic c
o
ntrol system
,
t
he
new controller has
a strong
robustn
ess, g
ood trac
eab
ilit
y w
i
th regard
to mo
del
unc
e
r
tainties, u
n
kn
ow
n externa
l
disturb
ances
a
n
d
chan
ges i
n
the oper
ation c
ond
itions, as w
e
ll a
s
much
better ada
ptive ch
ara
c
teristics. T
he simulati
on res
u
lt
s
also s
howed t
hat the adaptive fu
z
z
y
s
liding mode control system
c
an
solve the dynam
ic
asy
m
m
e
trical
perfor
m
a
n
ce a
nd po
or stabi
lit
y of parall
e
l ro
bot, and
ca
n gr
eatly i
m
pr
ove the rob
o
t contro
l precisi
on.
Ke
y
w
ords
:
parallel robot, Stewart platform
, hydraulic serv
o system
, adaptiv
e fu
z
z
y
sliding m
o
de control
Copy
right
©
2014 In
stitu
t
e o
f
Ad
van
ced
En
g
i
n
eerin
g and
Scien
ce. All
rig
h
t
s reser
ve
d
.
1. Introduc
tion
Seriou
s probl
ems like environment
al p
o
ll
ution, energy sho
r
t
age, a
r
e
incre
a
si
ng seriou
s,
therefo
r
e, en
ergy co
nserv
a
tion and e
m
issi
on r
edu
ction technol
ogy naturally
draw m
o
re
and
more
attentio
n. Ro
bots ca
n not
only
re
duce the
wa
ste of resources, b
u
t al
so
effectively re
duce
the ene
rgy lo
ss, an
d grad
ually have attracted
th
e a
ttention of the governmen
t and re
sea
r
ch
institutes. Sin
c
e
Hunt, a re
nowned Au
st
ralian i
n
stit
uti
onal p
r
ofe
s
so
r puts fo
rward the theo
ry of
usin
g 6 d
e
g
r
e
e
s of freed
om
Stewart
platform [1] (as
sh
own i
n
Fig
u
re
1) a
s
a
ro
bot
mechani
sm i
n
1978, the pa
rallel ro
bot, as a new
kind
of robo
t, has been attracting much att
ention of ma
ny
schola
r
s. Th
e
Stewart platf
o
rm is a
para
llel me
ch
ani
sm comp
osed
of six same
hydrauli
c
servo
system
s
[2]. Comp
ared wi
th
tradition
al seri
al
me
c
hanism, it is qualified
with
better
stiffness,
car
r
y
i
ng
cap
a
cit
y
,
locat
i
o
n
ac
cu
ra
cy
a
nd sm
aller
worki
ng
spa
c
e.
Therefore, it
can
red
u
ce the
static e
r
rors (owin
g
to its hi
gh rigi
dity) an
d the dy
nami
c
erro
r (o
win
g
to its low m
o
ment of ine
r
tia)
whe
n
appli
e
d
to the robot.
Becau
s
e of t
he viscos
ity
of the oil, the fr
iction bet
ween the
cylinder
and pi
ston, n
online
a
r
se
rvo valve flow
and time
-v
arying
system
para
m
eters, Stewart
pl
atform
often use
s
the symmet
r
i
c
al valve co
ntrolling a
s
y
mmetrical cy
linder [3
-4]. Stewart platf
o
rm
hydrauli
c
co
ntrol syste
m
tends to be high or
der system wit
h
highly non
linear, pa
ra
meter
uncertaintie
s
,
whi
c
h
serio
u
s
ly affect th
e
dynamic
pe
rf
orma
nce of t
he Stewart
pl
atform hyd
r
a
u
lic
control
syste
m
. And th
e
nonlin
ear an
d pa
ram
e
te
r un
ce
rtaintie
s m
a
ke the
system
dyna
mic
cha
r
a
c
teri
stics compl
e
x, as it is
difficu
lt to
establi
s
h accu
rate
mathemati
c
al
model
and
the
traditional
co
ntrol algo
rith
m is difficult to perfo
rm sati
sfacto
ry co
ntrolling effect [5-6].
It is hard to
impleme
n
t the traje
c
tory t
r
ac
kin
g
control of the pa
rallel ro
bot st
eadily,
pre
c
isely and
quickly. Literature [7] sta
t
ed that
the traje
c
tory tra
c
king p
e
rfo
r
m
ance of slidi
n
g
mode control wa
s better th
an that of the fuzzy cont
rol.
Sliding mode
variable st
ru
cture
cont
rol for
the sy
stem
para
m
eter p
e
rturbation a
nd
external
disturban
ce
s has st
rong
robu
stne
ss,
so it
provides a good sol
u
tion to comp
li
cated nonlinear system
contro
lling probl
em [8]. Fuzzy control
is
set
up
ba
sed o
n
fu
zzy reasonin
g
, an
d it i
s
u
nne
cessary
to e
s
t
ablish a
p
r
e
c
i
s
e
mathem
atical
model. It i
s
a
goo
d
way to
solve
un
ce
rt
ain
system
s,
whi
c
h
ha
s b
e
en p
r
e
s
ente
d
and
ap
plied
[9].
The
resea
r
ch
obje
c
t of ad
a
p
tive cont
rol t
heory
ca
n al
so be
aimed
at
uncertai
n
sy
stem
s, whi
c
h
is
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 6, June 20
14: 4125 – 4
133
4126
a kin
d
of non
linear
co
ntrol
method. So
we
c
an d
e
si
g
n
an ad
aptive fuzzy slidin
g mode
co
ntrol
system
with
more superi
or performance through usi
ng the re
asoning abilities of
fuzzy
cont
rol,
the learni
ng a
b
ility of adaptive control a
n
d
the
rapidity of sliding mo
de co
ntrol [10
-
11].
F
i
g
u
r
e
1
.
S
t
ew
art
P
l
at
f
o
r
m
In this pape
r, the adaptive fuzzy slidin
g mode
control algorith
m
on the ba
sis of a
daptive
control algo
ri
thm combi
n
i
ng fuzzy co
ntrol and
sli
d
ing mod
e
control was p
r
opo
se
d for
the
hydrauli
c
se
rvo system,
and the
un
certaintie
s
of
the sli
d
ing
mode
controller
wa
s fu
zzy
approximated
by usi
ng the
adaptive fu
zzy co
ntro
l m
e
thod
[12
-
1
4
].
We carrie
d on
a simul
a
tion
resea
r
ch a
b
out the
sin
g
l
e
chan
nel v
a
lve cont
rol
unsymm
e
trical hydrauli
c
cylinde
r of t
h
e
hydrauli
c
po
sition se
rvo system of the
Stewart
plat
form of paral
lel rob
o
t. Co
mpared with
the
conve
n
tional
PID control, adaptive fu
zzy sli
d
ing
mode
control
method not
only has b
e
tter
robu
stne
ss a
nd tra
c
king
a
b
ility, but also ha
s
a go
o
d
ad
aptive a
b
ility about t
he vari
ability of
para
m
eter an
d extern
al di
sturb
a
n
c
e. It
meets
th
e
re
quire
ment
s o
f
dynam
ic an
d ste
ady
state
index, and h
a
s go
od traj
e
c
tory tra
c
king
pre
c
isi
on an
d the ability to sup
p
ress crosslin
k
cou
p
l
i
ng
load.
2. Mathema
t
i
cal Model
The rese
arch
ed hydrauli
c
servo
sy
st
e
m
wa
s co
nsi
s
t
ed of
se
rvo valve, hydrauli
c
moto
r
and lo
ad
and
so fo
rth. Th
e
schem
atic
di
agra
m
of valv
e-controlled
h
y
drauli
c
moto
r was sho
w
e
d
in Figure 2, which h
a
s
com
posite lo
ad
wi
th the
mass, dampin
g
an
d the sp
ring, an
d the syste
m
'
s
external lea
k
age was ig
no
red [15].
Figure 2. Sch
e
matic Di
ag
ram of Valve-controlle
d Hyd
r
auli
c
Motor
In orde
r to establi
s
h ma
thematical m
odel
of the hydrauli
c
sy
stem, the following
ass
u
mptions
were made [2]:
1) Th
e four
throttling wi
n
dows of
serv
o va
lve we
re
matchin
g
a
nd symm
etri
cal, the
sup
p
ly oil pre
s
sure
wa
s co
nstant; the ret
u
rn oil p
r
e
s
su
re wa
s
zero.
2) Conn
ectin
g
pipelin
e wa
s short a
nd t
h
ick, fr
iction l
o
ss in pip
e
lin
e, the impa
ct of fluid
mass, and pi
peline’
s dyna
mics
we
re ign
o
red.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Adaptive F
u
zzy Slidin
g Mo
de Co
ntrol for Hyd
r
auli
c
Se
rvo S
ystem
of Parallel… (F
engla
n
Jia
)
4127
3) The flo
w
in
g state of the internal a
nd e
x
ternal lea
k
a
ge wa
s lamin
a
r flow.
4) The p
r
e
s
sure of liquid chambe
r in the hy
drauli
c
m
o
tor wa
s eq
u
a
l. Bulk mod
u
lus an
d
oil temperature were co
nst
ant.
5) Wh
en the
hydrauli
c
cyl
i
nder pi
ston
rod wa
s protracting, the servo valve core was
shifting rig
h
t, namely of
0
v
x
.
Und
e
r the
a
bove a
s
sum
p
tions,
we
can g
e
t the
d
y
namic equ
a
t
ion of the
h
y
drauli
c
sy
st
em.
2
2
4
Lq
v
c
L
mm
L
Lm
t
m
L
e
mm
s
mL
m
m
L
QK
x
K
P
dV
dP
QD
C
P
dt
dt
dd
TD
P
J
B
G
T
dt
dt
(1)
Whe
r
e
L
Q
is load flow of th
e valve,
v
x
is spool di
spla
ce
ment,
c
K
is flow-p
re
s
s
ur
e
coeffici
ent of
valve,
L
P
is load pressure,
q
K
is flow en
han
ceme
nt of val
v
e on
steady
wo
rki
ng-
point,
m
D
is th
eoreti
c
al di
splacement of
hydrauli
c
m
o
tor,
m
is an
gular
displa
cement of
hydrauli
c
mo
tor,
tm
C
is total leakag
e co
efficient of hydrauli
c
motor,
m
V
is total volume of
con
n
e
c
ted pi
pes, Hyd
r
auli
c
motor an
d valve chamb
e
r,
e
is effectiv
e bulk mod
u
l
u
s of workin
g
oil,
J
is the total inertia (th
e
inertia bet
we
en sp
ool of
Hydra
u
lic m
o
tor and lo
ad
(exclu
ding oil
)
trans
fer into t
he motor shaf
t),
m
B
is viscou
s dampin
g
coef
fi
cient of load
and moto
r,
G
is t
o
r
s
i
o
n
spri
ng st
if
f
n
e
ss of
loa
d
,
l
T
is external load torque
loaded on the motor shaft.
After Lapla
c
e
transfo
rmatio
n for (1
), we can obtain:
2
()
(
)
()
()
()
(
)
()
4
()
()
()
(
)
()
Lq
v
c
L
m
Lm
m
t
m
L
L
e
sm
L
m
m
m
m
L
Qs
K
x
s
K
P
s
V
Q
s
D
s
s
C
Ps
Ps
TD
P
s
J
s
sB
s
s
G
s
T
s
(2)
From (2), we c
an obtain the trans
f
e
r func
tion:
2
32
2
2
hq
m
m
vh
h
h
KD
x
ss
s
(3)
Whe
r
e
h
is natural fre
que
ncy of hydraulic system
,
h
is hydrauli
c
dam
ping ratio.
)
1
2
(
2
2
s
s
D
k
h
h
h
m
q
Figure 3. Con
t
rol Structu
r
e
of Hydrauli
c
Servo Syste
m
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 6, June 20
14: 4125 – 4
133
4128
The structu
r
e
of hydraulic
servo
system
is
sho
w
n in F
i
gure 3. F
r
om
the figure we
kno
w
that the syste
m
is co
nsi
s
te
d of electro
-
h
y
drauli
c
se
rv
o valve, servo amplifier, h
y
drauli
c
moto
r,
sen
s
o
r
s a
nd
so f
o
rt
h.
I
n
orde
r t
o
si
mplif
y
t
he se
rv
o sy
st
e
m
,
sup
p
o
s
ed th
a
t
the system
is only ine
r
tia
load
s,
while the
ela
s
tic loa
d
, viscou
s loa
d
were 0, a
nd the dynami
c
cha
r
a
c
teri
st
ic
s of
se
rv
o v
a
lv
e,
amplifiers an
d other
ele
c
tri
c
al
comp
one
nts are ign
o
red. The transfer functio
n
of
serv
o
sy
st
e
m
is
simplified as:
2
22
2
h
hq
m
m
vh
h
KD
xs
s
(4)
By (4), the status eq
uation
can be
writte
n as:
m
x
Ax
B
yC
x
(5)
Whe
r
e:
12
1
2
,,
TT
xx
x
x
x
x
2
2
01
,0
/
,
2
T
qh
m
hh
h
AB
k
D
10
C
3. Design of
Con
t
roller
Con
s
id
er the
followin
g
SISO uncertainly
system:
12
2
(,
)
(
,
)
(
)
(,
)
xx
x
fx
t
g
x
t
u
t
d
x
t
yx
(6)
Whe
r
e
12
1
1
T
x
xx
xx
is
t
h
e
sy
st
e
m
’s
st
a
t
e v
e
ct
o
r
,
ut
is t
he
control in
put
of the sy
ste
m
,
y
is the
ou
tput,
,
dx
t
is a
bo
unde
d di
sturban
ce,
,,
,
f
xt
g
x
t
are the
unkno
wn n
o
n
linea
r fun
c
ti
on, an
d
,0
,
(
,
)
,
gx
t
d
x
t
D
D
is the up
per bo
und fu
nctio
n
.
A
ssu
ming
ˆ
(,
)
(
,
)
(,
)
f
xt
f
x
t
f
x
t
,
ˆ
(,
)
f
xt
is
the e
s
timated
value of
(,
)
f
xt
,
(,
)
f
xt
is t
he
uncertainty of model.
The tra
cki
ng
error vecto
r
is defined:
12
1
1
[,
]
[
,
]
[
,
]
TT
T
T
dd
ee
e
x
x
x
x
e
e
(7)
The nonli
nea
r equation i
s
o
b
tained:
12
2
(,
)
(
,
)
(
)
(,
)
(
)
dd
d
d
ee
e
f
ex
t
g
ex
t
u
t
d
ex
t
x
t
(8)
The slidi
ng m
ode switching
function is
se
lected a
s
:
()
st
e
(9)
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Adaptive F
u
zzy Slidin
g Mo
de Co
ntrol for Hyd
r
auli
c
Se
rvo S
ystem
of Parallel… (F
engla
n
Jia
)
4129
Whe
r
e
1
1
1
(0
)
,
t
he co
rrel
a
t
i
on
coef
f
i
ci
ent
sat
i
sf
ie
s t
h
e Hu
r
w
it
z
polynomial, n
a
mely
()
()
0
st
st
. So we obtain the sliding mod
e
surface equ
ation as:
11
2
()
0
st
e
e
(10
)
Namely
,
11
()
(
,
)
(
,
)
()
(
,
)
(
)
0
dd
d
d
s
t
e
f
ex
t
g
ex
t
u
t
d
ex
t
x
t
(11
)
So the sliding
control la
w is desig
ned a
s
:
11
11
1
()
[
(
,
)
(
,
)
(
)
]
(,
)
1
[(
,
)
(
,
)
(
)
]
(,
)
dd
d
s
w
d
ds
w
u
t
e
f
ex
t
d
ex
t
x
t
u
ge
x
t
ef
x
t
d
x
tx
t
u
gx
t
(12
)
Whe
r
e
sg
n(
)
,
0
sw
uk
s
k
,
1,
0
sg
n(
)
0
,
0
1,
0
s
ss
s
Then we ca
n obtain:
()
s
g
n
(
)
sw
st
u
k
s
(13
)
Then,
()
()
()
0
st
st
d
t
s
k
s
(14
)
Bu
t in
th
e ac
tu
a
l
s
e
r
v
o
s
yste
m
,
(,
)
,
(,
)
,
(
)
f
xt
g
x
t
d
t
are un
kn
own,
th
e cont
rol
la
w
()
ut
can
not b
e
i
m
pleme
n
ted
in practi
cal
appli
c
at
ion
s
.
Espe
cially
when th
e di
st
urba
nce item
d
is
large
r
, the switch gain
k
of controller will be also
increased, which wi
ll ca
use chattering. In order
to redu
ce th
e
chatteri
ng, u
s
ing the fuzzy
system to a
p
p
roximate the
control la
w
()
ut
is prop
osed
in this pa
pe
r, namely
ˆˆ
ˆ
,,
f
gh
ap
proximate
s
,,
s
w
f
gu
. In the fuzzy
system of
approa
chin
g
equivalent co
ntrol,
the
()
s
t
and
()
s
t
are
sel
e
cte
d
as i
nput va
riable
s
, then
the output
()
ut
of
f
u
zzy
co
nt
rol
is
co
nt
rolle
d
by
f
u
z
z
y
r
u
l
e
s
of
f
u
zz
y
contr
o
ller
.
The fuzz
y r
u
les
ar
e
given in
th
e
following form [5]:
()
12
:(
)
(
)
(
)
j
jj
j
R
I
F
st
i
s
A
a
n
d
st
i
s
A
T
H
E
N
u
t
i
s
B
Based o
n
the
adopting of the sin
g
le valu
e fuzzifi
cation
, the produ
ct inferen
c
e e
ngi
ne
and the cente
r
avera
ge def
uzzifier, the o
u
tput of fuzzy
system is giv
en as:
1
1
1
1
()
()
()
n
m
ij
ii
j
i
n
m
j
ii
j
i
yA
x
yx
Ax
(15
)
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 6, June 20
14: 4125 – 4
133
4130
Whe
r
e
()
j
ii
A
x
is the
membershi
p
function
of
i
x
.
)
x
(
i
s intro
duced vecto
r
, namely
()
)
T
yx
x
=
(
,
where
1
=[
y
]
,
mT
y
1
)
[
()
()
]
.
m
x
xx
(
1
1
1
()
)
()
n
j
ii
i
n
m
j
ii
j
i
Ax
x
A
x
(
(16
)
A
ssu
ming
ˆ
ˆ
ˆ
(|
)
,
(|
)
,
(
|
)
f
gh
fx
g
x
h
s
are sele
cted a
s
above fuzzy logic
system, then:
ˆ
ˆ
ˆ
(|
)
(
)
,
(|
)
(
)
,
(
|
)
(
)
TT
T
ff
g
g
h
h
f
xx
g
x
x
h
s
s
(17
)
The
ˆˆ
ˆ
,,
f
gh
are repl
a
c
ed by
,,
f
gh
, so t
he control la
w turns i
n
to t
he fuzzy ve
ctor of
()
x
and
()
s
.
The
T
f
,
T
g
,
T
h
are
ch
ang
ed
based
on t
he
cha
nge
of ad
aptive la
w. Wh
ere
ˆ
(|
)
s
g
n
(
)
,
0
h
hs
k
s
k
D
k
k
,
.
The ada
ptive law is d
e
si
gn
ed as:
1
2
3
()
()(
)
()
f
g
h
rs
x
rs
x
u
t
rs
s
(18
)
The optimal p
a
ram
e
ter is d
e
fined a
s
:
ˆ
arg
m
i
n
[s
up
|
(
|
)
(
,
)
|
]
ˆ
ar
g
m
i
n
[
s
u
p
|
(
|
)
(
,
)
|
]
ˆ
ar
g
m
i
n
[
s
u
p
|
(
|
)
|
]
ff
n
gg
n
n
hh
ff
xR
gg
xR
hh
s
w
xR
f
x
f
xt
g
x
g
xt
hs
u
(
1
9
)
Whe
r
e
f
,
g
,
h
are
the set of
f
,
g
,
h
.
T
he Lyapu
no
v function is selecte
d
as:
2
12
3
11
1
1
()
2
TT
T
f
fg
g
h
h
Vs
rrr
(20
)
Whe
r
e
12
3
,,
rr
r
are the
positive co
nstant,
,
f
ff
,
g
gg
.
hh
h
Th
en:
12
3
12
3
11
1
11
()
()(
)
1
ˆ
()
[
(
)
(
|
)
]
TT
T
ff
g
g
h
h
TT
T
T
f
ff
g
g
g
TT
hh
h
h
Vs
s
rr
r
sx
s
x
u
t
rr
s
ss
d
t
h
s
s
w
r
(21
)
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Adaptive F
u
zzy Slidin
g Mo
de Co
ntrol for Hyd
r
auli
c
Se
rvo S
ystem
of Parallel… (F
engla
n
Jia
)
4131
Whe
r
e
ˆ
ˆ
(,
)
(
|
)
(
(
,
)
(
|
)
)
fg
wf
x
t
f
x
g
x
t
g
x
u
i
s th
e ap
proxim
ate erro
r. Be
cause of
ˆ
(|
)
s
g
n
(
)
h
hs
s
, we c
an obtain as
:
12
12
3
3
12
12
3
3
11
[
(
)
]
[
()(
)
]
1
[(
)
]
(
)
(
)
11
[
(
)
]
[
()(
)
]
1
[(
)
]
TT
f
fg
g
T
hh
TT
ff
g
g
T
hh
Vr
s
x
r
s
x
u
t
rr
rs
s
s
d
t
s
w
D
s
r
rs
x
r
s
x
u
t
rr
rs
s
s
w
s
r
(22
)
Based
on
the
ada
ptive la
w, we
ca
n al
so
obtain
Vs
w
s
. Acc
o
rding to the princ
i
ple
of universal fuzzy app
rox
i
mation, the
w
may be very infinitesimal,
so we can obtain
0
V
.
Namely it is p
r
oved that the
control
sy
ste
m
is stable in
the sen
s
e of
Lyapun
ov.
4
.
Simulation and discus
s
ion
In orde
r to verify the effectivene
ss of
the cont
rolle
r and get the
accurate si
mulation
result, MATLAB is
us
ed to mak
e
s
i
mulation for t
he hydrauli
c
servo
system. Acco
rding to loo
k
i
ng
up the releva
nt para
m
eter of hydrauli
c
servo
sy
ste
m
, the transfe
r function
of hydrauli
c
se
rvo
system i
s
given as [5, 16]:
2
891572.069544
()
81.978
18667.756
9
Gs
ss
(23
)
The statu
s
eq
uation is give
n as:
11
22
01
0
18667.
7569
81.
978
1
m
xx
xx
(24
)
Whe
r
e th
e i
n
itial state v
a
lue of
syst
em is
00
, the adaptive p
a
rameter i
s
se
lected
as
12
3
0.5
,
0.1
,
1
rr
r
, the output
si
gnal i
s
given
as
0.2
c
os(
)
d
x
t
, the switch
ing fun
c
tion i
s
given as
11
2
1
()
,
5
st
e
e
.
The slidi
ng
mode contro
l with the followin
g
five kind
s of membe
r
ship
function
approximate
s
the
()
ut
, namely:
2
2
2
2
2
(
)
e
x
p
[
((
/
6
)
/
(
/
1
2
))
]
(
)
e
x
p[
(
(
/
12)
/
(
/
1
2)
)
]
(
)
e
x
p[
(
/
(
/
12
)
)
]
(
)
e
x
p[
(
(
/
1
2
)
/
(
/
1
2)
)
]
(
)
e
x
p[
(
(
/
6
)
/
(
/
12
)
)
]
PM
PS
ZO
NS
NM
ss
ss
ss
ss
ss
(25
)
The memb
ership fun
c
tion
of switching f
unctio
n
is defi
ned a
s
:
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 6, June 20
14: 4125 – 4
133
4132
2
1
()
1
e
x
p
(5
(
3
))
()
e
x
p
(
)
1
()
1e
x
p
(
5
(
3
)
PM
ZO
NM
s
s
ss
s
s
(26
)
Figure 4. Position Tra
cki
ng
using PID
C
ontroller
Figure 5. Position Tra
cki
ng
using Ada
p
tive
Fuzzy Sliding
Mode Co
ntro
ller
Figure 6. Position Tra
cki
ng
Error u
s
in
g PID
C
ontroller
Figure 7. Position Tra
cki
ng
Error u
s
in
g
Adaptive Fuzzy Sliding Mo
de Co
ntrolle
r
Figure 8.
,
f
xt
and Estimated
,
f
xt
Figure
9.
,
gx
t
and Estimated
,
gx
t
With the sam
e
variation of paramete
r
a
nd ex
ternal lo
ad, the simul
a
tions were made by
usin
g the ad
aptive fuzzy
slidin
g mode
control
a
n
d
traditional PID control. The re
sults
were
sho
w
e
d
from
Figure 4 to
Figure 9. Co
mpared
with
the co
nventio
nal PID
control algo
rithm,
the
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Adaptive F
u
zzy Slidin
g Mo
de Co
ntrol for Hyd
r
auli
c
Se
rvo S
ystem
of Parallel… (F
engla
n
Jia
)
4133
system with a
daptive fuzzy slidin
g mode
control
fast completed the
conve
r
ge
nce in 1.5 se
con
d
s
,
thus e
n
sure
d
the better tracking
effect, whi
c
h
fully i
n
ca
rnate
s
its highe
r resp
onse spee
d
and
better control
pre
c
isi
on. We can al
so
find t
hat the adaptive fuzzy slidin
g mode
control
l
er
effectively weaken the
chattering
ph
enome
non,
whi
c
h ma
de
the external
disturban
ce
and
para
m
eter
pe
rturb
a
tion of t
he servo
syst
em pr
edi
cted
and comp
en
sated effectively.
Moreove
r
,
the traditional
PID control
method exist
ed obviou
s
o
scill
ation ph
e
nomen
a in th
e trackin
g
ph
ase,
whi
c
h m
a
y result in
stabili
ty and o
scilla
tion phe
nom
ena eve
n
ca
usin
g reso
na
nce
and
dam
age
the ro
bot d
u
r
ing
wo
rking
time. Thu
s
the ad
aptive
fuzzy sli
d
in
g mod
e
cont
rolle
r
wa
s m
o
re
suitabl
e to co
ntrol the pa
ral
l
el robot.
5. Conclusio
n
Aiming at t
h
e
highly
nonli
n
ear, l
oad
sen
s
itivit
y, para
m
eter
un
ce
rtainty and
tim
e
-varyin
g
load
cou
p
ling
interferen
ce
of cha
nnel
s i
n
the St
ewa
r
t
platform of
hydrauli
c
con
t
rol syst
em, an
adaptive co
ntrol algo
rithm combi
n
ing fu
zzy control
wi
th sliding mo
de cont
rol is
prop
osed for
the
hydrauli
c
se
rvo
system. Comp
ared with
the
co
nventional PID control alg
o
r
ithm, the ab
ove
prop
osed m
e
thod
can
g
uara
n
tee the
better robu
st tra
cki
ng
perfo
rman
ce
and p
a
ram
e
ter
pertu
rbatio
n of the hydrau
lic se
rvo syst
em wit
hout n
eedin
g
an accurate mod
e
l. Moreover, the
external
dist
urba
nce i
s
compen
sate
d
effectively.
The
simulatio
n
results
also sho
w
that
the
probl
em of chattering i
s
in
hibited obvio
usly, and
the
ability of anti-interfe
r
en
ce,
anti-pa
ram
e
ters
pertu
rbatio
n, stability and
control qualit
y of the sy
stem are imp
r
oved effectively. At
the same
time, the sch
eme is
simple
and suita
b
le
for engin
e
e
r
in
g appli
c
ation.
Referen
ces
[1]
KH Hunt. Structural
Kin
e
mati
cs of in-Parall
e
l-Autom
a
ted
Rob
o
t Arm.
Tr
ans. ASME J.
Mechanis
m
s
,
T
r
ansmissio
n
s,
and Auto
matio
n
in Des
i
gn
. 1
9
83; 105: 7
05-7
12.
[2]
T
ang Rui, W
a
n
g
Sha
o
ji
an
g, Hou Li. D
o
u
b
le
Adaptiv
e F
u
zz
y Sl
idi
ng Mo
de
Control for
H
y
drau
lic Serv
o
S
y
stem of Par
a
lle
l Mach
ine.
T
r
ansactio
n
s o
f
the Chines
e Society for Ag
ricultura
l
Mach
iner
. 20
12;
43(1
0
): 229-
23
4. (In Chines
e)
[3]
Liu T
ao, Li
u Qi
ngh
e, Jia
ng J
i
hai. Ad
aptiv
e
F
u
zz
y
S
lid
ing
Mode
Co
ntrol f
o
r H
y
dr
ostatic
T
r
ansmission
Sy
s
t
e
m
.
T
r
ans
actions
of the
Chi
nese S
o
ci
ety for Agricu
l
t
ural Mac
h
in
er
y
. 2010;
41(1)
: 29-33. (I
n
Chin
ese)
[4]
Bai Ha
n, W
a
n
g
Qingj
iu,
Xu
Z
hen, et al. Multipl
e
sli
d
in
g mode ro
bus
t adaptiv
e con
t
rol for valv
e
control
l
ed as
ymmetric c
y
l
i
n
der s
y
st
em.
T
r
ansactio
n
s of the Chin
e
s
e Society fo
r Agricultur
a
l
Machi
nery
. 20
09; 40(1
0
): 193
-198. ( in Ch
in
ese)
[5] Liu
Ji
nku
n
.
Slid
ing
mo
de v
a
ria
b
le structur
e control w
i
th
matl
ab si
mu
lati
on
.
Beiji
ng: T
s
ingh
ua u
n
ivers
i
t
y
press. 200
5.
[6]
T
ong S, Li HX. F
u
zzy
a
d
a
p
ti
ve slidi
ng
mo
d
e
control for MIMO nonlin
e
a
r s
y
stems.
IEEE Trans FS.
200
3; 11(3): 35
4–6
0.
[7]
Guoqi
n Gao, Y
i
Re
n, Ha
i
y
a
n
Z
hou,
Z
h
imi
n
g
F
ang. Smo
o
th
slidi
n
g
mod
e
c
ontrol f
o
r traj
ec
tor
y
track
i
n
g
of gree
nho
use
spra
yin
g
mo
b
ile ro
bot.
T
E
LKOMNIKA Indones
ian J
our
n
a
l
of Electric
al
Engi
neer
ing
.
201
3; 11(2): 64
2-65
2.
[8]
Che
n
CS, Ch
en W
L
. Robus
t adaptive sl
id
i
ng mod
e
contr
o
l usi
ng fuzz
y model
ing for
an inv
e
rted-
pen
dul
um s
y
st
em.
IEEE Trans IE
. 1998; 45(
2): 297–
30
6.
[9] Yong
qia
o
W
e
i,
Li H
ou, Zhij
u
n
Sun, Feng
la
n Jia, Bo
Li.
Backstepp
in
g Adaptiv
e Fuzzy
Sch
e
me for
SCARA
GRB4
00 Ro
bot.
T
E
LKOMNIKA Indon
esia
n Jour
nal
of El
ectric
al Eng
i
n
eeri
n
g
.
2013; 11(
8):
422
9-42
37.
[10]
N Noro
ozi, M Roop
aei, MZ
Jahromi. Ad
a
p
tive
fuzz
y
s
l
i
d
in
g mod
e
co
ntrol schem
e for uncertai
n
s
y
stems.
Co
mmu
n
No
nli
n
e
a
r Sci Nu
mer Si
mulati
on.
20
09; 14: 397
8–
39
92
.
[11]
Shi YF
, CJ Z
h
ou, Z
Li
ao. F
u
z
z
y
S
lid
in
g mo
d
e
Co
ntrol f
o
r th
e S
w
i
ng Arm
U
s
ed i
n
a
F
ouri
e
r T
r
ansform
Spectrometer.
Information T
e
chno
logy J
ourn
a
l.
201
1; 10(4):
736-7
47.
[12]
T
Z
W
u
, YT
Ju
ang. Ad
aptiv
e fuzz
y
s
l
i
d
in
g-m
ode
c
ontrol
l
er of
uncerta
in no
nlin
ear
s
y
stem
s.
ISA Trans
.
200
8; 47(3): 27
9–2
85.
[13]
W
e
i Meng, Z
u
de Z
h
o
u
, Quan
Liu, Qings
on
g
Ai. A
Practical
F
u
zz
y
Ad
aptiv
e Contro
l Strateg
y
for Multi-
DOF
Paralle
l R
obot.
ISCCCA
.
2013.
[14]
Otto Cerman. Fuzzy
model referenc
e control
w
i
th adap
tation mechanism.
Expert System
s with
Appl
icatio
ns
. 2
013; 40(
13): 51
81-5
187.
[15] ZS
Wu.
Hydraulic co
ntrol sys
tem
. Beij
in
g: Hi
gher Ed
ucati
o
n
Press. 2008. (In Chi
nese)
[16]
KW
T
ong, X Z
han
g, Y Z
hang
, Z
Xie, RX Ca
o. Slidi
ng Mod
e
Varia
b
le Stru
cture Contro
l o
f
Permanen
t
Magn
et S
y
nchr
ono
us Mach
ine
Based o
n
a N
o
vel R
eac
hin
g
La
w
.
Jo
urna
l of
Procee
din
g
s o
f
the CSEE
.
200
8; 28: 102-
106.
Evaluation Warning : The document was created with Spire.PDF for Python.