Internati
o
nal
Journal of Ele
c
trical
and Computer
Engineering
(IJE
CE)
Vol
.
4
,
No
. 5, Oct
o
ber
2
0
1
4
,
pp
. 65
8~
66
7
I
S
SN
: 208
8-8
7
0
8
6
58
Jo
urn
a
l
h
o
me
pa
ge
: h
ttp
://iaesjo
u
r
na
l.com/
o
n
lin
e/ind
e
x.ph
p
/
IJECE
Robust Cont
rol
Strat
e
gy for Pn
eumatic Dri
v
e System vi
a
Enhanced Nonlinear PID Controller
Sy Najib Sy
S
a
lim
1
, M
F
Rahmat
2
, AAM Faudz
i
3
, Z
H
I
s
mail
3
, NH Su
nar
2
,
S
h
a
m
s
u
l
A
nua
r Sa
msu
d
i
n
4
1
Department of
Control, Instrumenta
tion
and Automation, Faculty of Elect.
Eng
,
U
n
iv T
e
knika
l Ma
la
y
s
ia
Melak
a
2
Department of
Control
and Mechatr
onic Eng
i
neering, Faculty
of
Elect.
Eng
,
Univ
ersiti
Tekno
logi
Mala
y
s
ia
3
Centr
e
for
Arti
fici
al In
tellig
ence &
Roboti
c
s (C
AIRO),
Universi
ti
Teknolog
i Mal
a
y
s
i
a
, Kual
a
Lu
m
pur, Malay
s
i
a
Article Info
A
B
STRAC
T
Article histo
r
y:
Received
J
u
n 22, 2014
Rev
i
sed
Au
g
24
, 20
14
Accepted
Sep 10, 2014
This paper pre
s
ents the pneu
m
atic position
i
ng sy
st
em
controlled
b
y
Enhanced Nonlinear PID (NPID) contro
lle
r.
The cha
r
ac
terist
ic of ra
t
e
varia
tion of the
nonline
a
r gain th
at ar
e read
il
y av
ail
a
ble in NP
ID controll
er i
s
utili
zed
to im
prove the p
e
rfor
m
ance of th
e c
ontrolle
r. A Sel
f-regula
tio
n
Nonlinear
Function (SNF) is used to
re
proce
ss the
e
rror signals with the
purpose of co
ntinuously
gen
e
rating
the
valu
es for the r
a
te variation.
Subsequently
, the
contro
ller has
successfully
b
een
implemented on
d
y
namically
changing loads and pressu
res
.
The com
p
aris
on with the othe
r
avai
labl
e m
e
tho
d
s
u
ch as
. NP
I
D
and conv
enti
onal P
I
D ar
e pe
rform
ed and
evalu
a
ted
.
The
effectiven
ess of this
method with Dead Zone C
o
mpensator
(DZC) has also been succes
sfull
y
d
e
m
onstrated and prov
en through
simulations and
expe
rime
nta
l
studie
s.
Keyword:
pne
um
at
i
c
posi
t
i
oni
n
g
sy
st
em
NPI
D
SNF
dead-z
one
com
p
ensation
robustness
Copyright ©
201
4 Institut
e
o
f
Ad
vanced
Engin
eer
ing and S
c
i
e
nce.
All rights re
se
rve
d
.
Co
rresp
ond
i
ng
Autho
r
:
M.F. Ra
hm
at,
Depa
rt
m
e
nt
of
C
ont
r
o
l
a
n
d
M
echat
r
oni
c
En
g
i
neeri
n
g,
Facu
lty of Electrical Eng
i
n
eerin
g
,
Un
iv
ersiti Tekn
o
l
o
g
i
Mal
a
ysia.
Em
a
il: fu
aad
@fk
e
.u
tm
.
m
y
1.
INTRODUCTION
Pne
u
m
a
ti
c act
uat
o
rs a
r
e
wi
d
e
l
y
used
i
n
i
n
dust
r
i
e
s t
o
pe
r
f
o
r
m
m
a
ny
t
a
sks s
u
c
h
as
pi
c
k
a
n
d
pl
ace
ap
p
lication
,
g
r
ip
p
i
ng
, clam
p
i
n
g
,
d
r
illin
g and
sp
raying
.
They are categorized
u
n
d
e
r fl
u
i
d
p
o
wer co
n
t
ro
l and
appl
y
t
h
e
pri
n
c
i
pl
es of
usi
n
g
com
p
ressed
ga
s as a sou
r
ce
o
f
p
o
we
r t
o
per
f
o
rm
a vari
et
y
of t
a
s
k
s. It
dea
l
s wi
t
h
m
echani
cal
pr
ope
rt
i
e
s o
f
ga
s
e
s an
d o
ffe
r s
e
veral
a
dva
nt
ag
es su
ch as si
m
p
le to
m
a
in
tain
, fast m
o
tio
n
,
low
co
st, h
i
g
h
p
o
wer to
weigh
t
ratio
, free
from o
v
e
rh
eatin
g an
d
reliab
l
e [1
]. Th
e ab
ility
to
o
p
e
rate at a h
i
gh
num
ber
of cy
c
l
es per
wo
rk
da
y
i
s
al
so one
of t
h
e ad
va
nt
ages o
f
t
h
i
s
dri
v
e.
Due to t
h
e
s
e advanta
g
es, it has
been prom
oted as an alternative to
hydra
u
lics and electric s
e
rvo m
o
tors i
n
m
a
ny autom
a
ted tasks.
In
spi
t
e of
th
ese adv
a
n
t
ages, pn
eu
m
a
tic
actu
a
to
rs are su
bj
ect to
non
lin
earities du
e to co
m
p
ressib
ilit
y o
f
air, h
i
gh
frictio
n
forces and
dea
d
ba
nd of the spool
m
ovem
e
n
t
in the valve
[2]. These
nonli
n
earities m
a
ke
an accurate position
diffic
u
lt to ac
h
i
eve, a
n
d
it re
q
u
ires a
n
a
p
pr
o
p
riate c
ont
rolle
r f
o
r
better pe
rf
orm
a
nce.
In
early 1
900
s, th
e use of th
i
s
actu
a
to
r was li
mited
to
a certain
app
licatio
n
du
e to
t
h
e
d
i
fficu
lty o
f
obtaini
ng a good
pe
rform
a
nce.
Thus,
resea
r
c
h
on this c
o
m
pone
nt is ra
rely perform
e
d for
decade
s
until there is
a d
e
m
a
n
d
to
b
e
ap
p
lied in
th
e au
t
o
m
a
tio
n
indu
stry circa 19
50
s [3
].
Research on
pn
eu
m
a
tic p
o
s
itio
n
i
ng
co
n
t
r
o
l
h
a
s
gro
w
n
si
g
n
i
f
i
cantly in
th
e 1990
s
w
h
en
m
a
n
y
con
t
ro
l techn
i
q
u
e
s
h
a
v
e
b
e
en
ex
am
in
ed
on
the
syste
m
as r
e
por
ted
i
n
[4-
6
].
A
lth
oug
h th
e
co
nv
en
tio
n
a
l
PID
con
t
ro
ller i
s
no
t su
itab
l
e
for th
e systems with
high nonlinearity, but it is
still popula
r
with the idea of m
odification as
a study conducted by [7-9]
.
Thi
s
co
n
t
ro
ller is wi
d
e
ly ap
p
lied
i
n
in
du
stries co
m
p
ared
to
the
other techniques
due t
o
its good characteristics
,
low
co
st an
d easy to
im
p
l
e
m
en
t as well as m
a
tu
re in
th
eoreti
cal analysis [10-12]. T
h
e a
dva
nc
ed c
ont
rol st
rategies
suc
h
as ada
p
t
i
v
e co
nt
r
o
l
,
f
u
z
z
y
l
ogi
c co
nt
r
o
l
,
ne
u
r
al
net
w
or
k a
nd
ot
he
rs
were a
g
gressi
vel
y
i
nvest
i
g
at
ed a
n
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE Vo
l. 4
,
N
o
. 5
,
O
c
tob
e
r
20
14
:
658
–
6
67
6
59
applied on t
h
e early of
2
000s. However, in the last
decade m
a
ny re
searchers found t
h
at the techniques that
in
teg
r
ate with
PID con
t
ro
ller
are m
o
re p
r
actical. It referred to
th
e in
creasi
n
g
n
u
m
b
e
r of
p
u
b
licatio
n
s
written
by
[1
3-
2
1
]
am
on
gst
ot
hers
. F
o
r e
x
am
pl
e, du
e t
o
t
h
e dra
w
b
ack o
f
ada
p
t
i
v
e cont
r
o
l
l
e
rs t
h
at
are not
fast
eno
u
g
h
to
fo
llow th
e
p
a
ram
e
ter v
a
ri
atio
n
,
[1
6
]
h
a
v
e
p
r
op
o
s
ed
Mu
lti-
m
o
d
e
l co
n
t
ro
ller b
a
sed on
sev
e
ral fi
xed
PD-
co
n
t
ro
llers. This tech
n
i
qu
e is
p
r
op
o
s
ed
for t
h
e
p
o
s
ition
con
t
ro
l of a
p
n
e
umatic cy
lin
d
e
r
u
n
d
e
r v
a
riab
le
lo
ad
s.
Fiv
e
PD con
t
ro
llers were tuned
b
a
sed
on
the esti
m
a
ted
mo
d
e
l fo
r th
e fi
ve fix
e
d
lo
ad
that h
a
s b
een
id
en
tified
earlier. Exp
e
ri
men
t
al resu
lts
sh
ow
t
h
at th
is
tech
n
i
qu
e
sign
ifican
tly i
m
p
r
ov
ed th
e ab
ility o
f
produ
cing
a good
perform
a
nce even under va
riable loa
d
c
o
ndi
tions.
Th
is p
a
p
e
r d
e
als with
th
e i
n
v
e
stig
atio
n
on
th
e
robu
stness of th
e
p
n
eu
m
a
t
i
c actu
a
to
rs
wh
ich
cont
rol
l
e
d
by
t
h
e no
vel
Sel
f
-re
gul
at
i
o
n N
o
nl
i
n
ear
P
I
D
(
S
NPI
D
)
c
ont
rol
l
er
t
h
at
ha
d b
een p
ubl
i
s
he
d
i
n
t
h
e
pre
v
i
o
us
wo
rk
[2
2]
. The sy
st
em
i
s
exam
i
n
ed base
d
on t
h
e
vari
at
i
o
n o
f
l
o
ad a
n
d p
r
ess
u
re i
n
i
n
c
r
easi
n
g an
d
decreasi
n
g val
u
es.
The
DZC
was a
dde
d t
o
t
h
e re
al syst
em
,
and the
conse
que
nce t
o
the
s
y
ste
m
was obs
e
rve
d
.
Th
e exp
e
rim
e
n
t
s were p
e
rfo
r
med
to
co
nfirm
th
e cap
ab
ility o
f
t
h
is co
n
t
ro
ller. Th
e co
mp
ariso
n
s wit
h
t
h
e
o
t
h
e
r
exi
s
t
i
n
g
m
e
t
hods i
n
cl
u
d
i
n
g
PID
an
d
NP
I
D
c
ont
r
o
l
l
e
r a
r
e p
e
r
f
o
r
m
e
d base
d o
n
t
r
an
si
ent
an
d st
ea
dy
-st
a
t
e
per
f
o
r
m
a
nce.
The
rest
o
f
t
h
i
s
pape
r i
s
or
gani
ze
d as
f
o
l
l
o
ws:
In
sect
i
o
n
2
,
re
searc
h
m
e
t
hod
i
s
de
scri
be
d
startin
g
with
math
e
m
atica
l
m
o
d
e
llin
g
an
d fo
llowed
b
y
th
e d
e
si
g
n
o
f
th
e con
t
ro
ller. Th
e sim
u
late
d
and
expe
ri
m
e
nt
al
r
e
sul
t
s
usi
n
g
M
A
TLAB
/
S
I
M
U
LI
NK a
r
e
descri
bed i
n
s
ect
i
on 3
.
Fi
n
a
l
l
y
, sect
i
on 4
cont
ai
ns
som
e
concl
u
di
ng
rem
a
rks.
2.
R
E
SEARC
H M
ETHOD
2.
1 Sys
t
em
Model
Th
e system
u
n
d
e
r con
s
id
er
atio
n is sh
own
in Fig
.
1
.
I
t
co
nsi
s
t
s
of
5/
3
pr
op
ort
i
onal
di
rect
i
onal
cont
r
o
l
val
v
e, d
o
u
b
l
e
-act
i
ng wi
t
h
do
ubl
e ro
d cy
l
i
nder,
di
spl
ace
m
e
nt
t
r
ansducer
,
pressu
re sens
ors,
dat
a
acqui
si
t
i
on
s
y
s
t
em
,
P
C
a
n
d
m
a
s
s
p
a
y
l
o
a
d
.
The t
r
a
n
s
f
er
fu
nct
i
on
of t
h
e sy
st
em
i
s
obt
ai
ned usi
ng S
y
st
em
Ident
i
f
i
cat
i
o
n
.
Fo
r
th
is pur
pose, 2
000
d
a
ta p
o
i
n
t
s r
e
pr
esen
tin
g
th
e inpu
t an
d
ou
tpu
t
sig
n
a
l of
th
e open
lo
op
system w
e
r
e
collected wit
h
a sam
p
ling time of 0.01
second.
A state s
p
ac
e m
odel
as sh
o
w
n
i
n
(1
) a
n
d (
2
) i
s
use
d
a
s
a
m
odel
structure
of the
syste
m
.
)
(
)
(
)
(
)
(
t
Ke
t
Bu
t
Ax
Ts
t
x
(1
)
)
(
)
(
)
(
)
(
t
e
t
Du
t
Cx
t
y
(2
)
where,
A
R
nxn
,
B
R
nxm
,
C
R
1xn
and
D
R
1xm
are th
e
matr
ic
es o
f
th
e syst
e
m
. W
h
ile
x(t)
R
n
,
y(t)
R, u
(
t)
R
m
,
and
K
R
n×m
re
present the state-vector, m
eas
ured output, m
easured i
n
put
s
i
gnal
and
n
o
ise, respectively.
T
h
e
est
i
m
a
ti
on o
f
t
h
e val
u
es
of
t
h
e param
e
t
e
rs i
s
perf
orm
e
d usi
n
g t
h
e P
r
e
d
i
c
t
i
o
n
-
Er
ro
r
M
i
nim
i
zati
on
(PEM
)
t
echni
q
u
e
wi
t
h
i
n
M
A
T
L
AB
. Thr
o
ug
h t
h
i
s
m
e
t
hod, t
h
e
param
e
t
e
rs are cal
cul
a
t
e
d by
m
i
nim
i
zi
ng a cost
fu
nct
i
o
n of
t
h
e pre
d
i
c
t
i
o
n
er
ro
rs,
(
t
) gi
vi
ng;
N
t
N
N
t
N
Z
V
1
2
)
(
2
1
)
,
(
(3
)
where
Z
N
and
N
d
e
no
tes th
e
set o
f
d
a
ta an
d n
u
m
b
e
r of
d
a
ta sa
m
p
les, resp
ectiv
ely. Fo
r
ex
a
m
p
l
e, th
e in
pu
t-
o
u
t
pu
t d
a
ta o
v
e
r a ti
m
e
in
terv
al o
f
1
i
s
represen
t
e
d by
:
)]
(
),
(
.......
),........
2
(
),
2
(
),
1
(
),
1
(
[
N
y
N
u
y
u
y
u
Z
N
(4
)
The dat
a
i
s
used fo
r est
i
m
a
t
i
n
g a m
odel
.
The di
fference bet
w
een t
h
e o
b
ser
v
ed o
u
t
put
an
d
predi
c
t
i
ng o
u
t
put
i
s
kn
own
as t
h
e r
e
si
dual
or p
r
edi
c
t
i
on error
w
h
i
c
h i
s
gi
ven
by
)
(
ˆ
)
(
)
(
t
y
t
y
t
(5
)
where
)
(
t
y
and
)
(
ˆ
t
y
represent
obse
r
v
e
d ou
put
and
p
r
edi
c
t
e
d out
put
, respect
i
v
el
y
.
In ge
neral
,
t
h
e out
put
y(t
)
can be re
prese
n
ted as;
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Rob
u
s
t C
o
n
t
ro
l
S
t
ra
teg
y
f
o
r Pn
euma
tic Drive S
y
stem via En
han
ced Non
lin
ear PID Con
t
ro
ller
(
M
.F. R
a
hm
at)
66
0
)
(
)
(
)
(
)
(
)
(
t
e
q
H
t
u
q
G
q
t
y
k
n
(6
)
Th
e estim
ated
p
a
ram
e
ter is ob
tain
ed b
y
m
i
n
i
m
i
z
i
n
g
(3) as fo
llo
ws;
N
N
N
N
N
Z
V
Z
,
min
arg
ˆ
ˆ
(7
)
The
following
equation is
the
discrete state-s
p
ace e
qua
ti
on
obtaine
d t
h
rough this i
d
entifi
cation
process
.
0
5
.
0
0
0
0
2
10
544
.
8
350
.
1
.846
2
1
A
0
0
125
.
0
B
2
2
-2
10
967
.
3
10
765
.
3
10
3.494
-
C
;
]
0
[
D
The co
nt
i
n
uo
u
s
t
r
ansfe
r
f
u
nct
i
on ca
n be o
b
t
a
i
n
ed
usi
n
g t
h
e
Zero
Or
der
H
o
l
d
(
Z
O
H
) c
o
n
v
ersi
on m
e
t
h
o
d
wi
t
h
sam
p
lin
g
ti
m
e
,
Ts =
0
.
01
s. Th
is conv
ersi
on m
e
th
o
d
g
e
n
e
rates th
e co
n
tinu
o
u
s
tim
e in
pu
t sign
al
b
y
ho
ld
ing
each sam
p
le va
lue consta
nt over
one
sam
p
le peri
od.
Fi
gu
re
1.
Ex
pe
ri
m
e
nt
al
Set
up
Fi
gu
re 2.
System
with
SNPID con
t
ro
ller
a.
Co
ntr
o
ller De
sign
In
g
e
n
e
ral,
th
e tran
sfer fu
n
c
tion
o
f
PID con
t
ro
ller
with
n
o
i
se filter is
g
i
v
e
n b
y
:
1
1
1
)
(
)
(
s
N
T
s
T
s
T
K
s
E
s
U
d
d
i
p
PID
(8
)
whe
r
e
K
p
is th
e p
r
op
ortion
a
l gain
and
T
i
is t
h
e in
teg
r
al tim
e.
Bo
th
p
a
ram
e
ters can
b
e
tun
e
d to
im
p
r
o
v
e
t
h
e rise
tim
e
and eliminate the steady
state error,
resp
ectiv
ely. Mean
wh
ile, th
e d
e
riv
a
tiv
e ti
m
e
,
T
d
can give the
effect
o
f
i
n
creasi
n
g
t
h
e stab
ility o
f
th
e system
b
y
i
m
p
r
o
v
i
n
g
t
h
e tran
sien
t resp
on
se. In
con
t
ro
l syste
m
d
e
sig
n
,
stab
ility is th
e first criteri
o
n
th
at n
e
ed
s t
o
b
e
con
s
id
ered. In o
r
d
e
r to m
a
in
t
a
in
th
e stab
ility o
f
t
h
e system
, th
e
co
nd
itio
ns as
written
i
n
(9)
m
u
st b
e
co
m
p
lied
.
1
BT
j
L
(9
)
whe
r
e
is a m
a
gni
t
ude
of
t
h
e
open
l
o
o
p
sy
st
em
.
The speed of t
h
e respo
n
se i
s
one of t
h
e cri
t
eri
ons t
h
at
need t
o
be consi
d
ered t
o
obt
ai
n t
h
e opt
im
al
perf
orm
a
nces. It
l
eads t
o
co
n
s
i
d
eri
ng t
h
e
ba
ndwi
d
t
h
f
r
eq
u
e
ncy
of
t
h
e sy
st
em
. In gene
r
a
l
,
t
h
e speed
o
f
t
h
e
respo
n
se i
s
i
n
creased wi
t
h
res
p
ect
t
o
t
h
e i
n
creasi
ng of
ban
d
wi
dt
h.
Ho
we
ver, i
t
i
nvol
ves
a t
r
ade-of
f be
t
w
een
speed and r
o
b
u
st
ness of t
h
e respo
n
se, and
hi
gh ba
ndwi
d
t
h
m
a
k
e
s th
e s
y
ste
m
sen
s
it
iv
e
to
th
e n
o
i
se. T
h
u
s
, in
orde
r t
o
pr
ovi
d
e
a g
o
o
d
res
u
l
t
s
i
n
a
wi
de ra
n
g
e o
f
perf
orm
a
nce i
n
cl
udi
n
g
s
t
abi
l
i
t
y
, speed and
ro
bust
n
ess
,
t
h
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE Vo
l. 4
,
N
o
. 5
,
O
c
tob
e
r
20
14
:
658
–
6
67
6
61
desi
gn sho
u
l
d
corresp
on
d t
o
vari
ous cri
t
e
ri
a i
n
cl
udi
ng ga
i
n
m
a
rgi
n
, phase
m
a
rgi
n
, gai
n
crossover f
r
eq
uency
an
d
m
a
x
i
m
u
m sen
s
itiv
it
y. In
o
r
d
e
r to
en
sure th
e o
p
tim
u
m
p
e
rfo
rm
an
ce to
b
e
acq
u
i
red
,
sev
e
ral sim
u
lat
i
o
n
s
based on
di
ffer
e
nt
Gai
n
M
a
rgi
n
(GM
)
and P
h
ase M
a
rgi
n
(PM
)
were con
duct
e
d as depi
ct
ed i
n
Tabl
e 1
.
B
a
sed
on t
h
ese resul
t
s
, t
h
e
opt
im
u
m
val
u
e
of
GM
and PM
are
15
.7
dB
an
d
71.
9
at f
r
e
q
u
e
n
c
y
1
.
38
0 H
z
an
d
0
.
286
Hz, respectively. These values provid
e an appropriate trad
e-off
between
speed perf
orm
a
nce and ro
bu
st
ness.
Accor
d
i
ng t
o
[
23]
, i
n
pract
i
c
e for wel
l
-
t
uned
sy
stem
t
h
e value of GM
and PM
shoul
d be bet
w
een 6 dB
t
o
20
dB
and
35
to
8
0
, resp
ectiv
ely. Th
e
m
a
x
i
mu
m
p
eak
fo
r the sen
s
itiv
it
y fu
n
c
tio
n
is less th
an
6d
B.
Tabl
e 1.
Per
f
o
r
m
a
nce of t
h
e S
y
st
e
m
with
resp
ect to
GM and
PM
Gm
Pm
tr (s)
ts (s)
Number of
oscillation
Robustness criterion
27.
200
85.
000
4.
140
7.
560
-
19.
700
78.
200
1.
490
2.
820
-
15.
700
71.
300
0.
748
1.
460
-
13.
700
66.
400
0.
518
0.
881
<
1
cycle
11.
900
61.
000
0.
390
1.
170
<
1
cycle
9.
470
51.
700
0.
287
1.
490
<2
cy
cle
2.
340
13.
500
0.
143
5.
360
6
cy
cle
×
2.3. De
terminati
on of
Nonli
n
ear Gain
The no
nl
i
n
ea
r gai
n
,
k
x
(e)
w
h
i
c
h
bou
nd
ed
in
th
e sector
max
0
e
k
e
k
x
as indicated in (10) is use
d
t
o
i
n
c
r
eases
t
h
e pe
rf
orm
a
nce
o
f
t
h
e sy
st
em
i
n
t
e
rm
s of s
p
eed.
Thi
s
gai
n
re
prese
n
t
s
t
h
e
co
nt
i
n
uo
us
dy
nam
i
c
n
o
n
lin
ear fun
c
t
i
o
n
.
Th
is fun
c
tio
n is th
en
co
mb
in
ed
in cascad
e
with
PID con
t
ro
ller.
2
exp
exp
e
e
e
k
i
i
x
(1
0)
whe
r
e:
max
max
max
e
e
e
sign
e
e
e
e
e
α
i
and
e
max
re
p
r
esent
t
h
e
rat
e
vari
at
i
o
n
of
n
o
n
l
i
n
ear
gai
n
a
n
d
ra
nge
o
f
va
ri
at
i
on,
res
p
ect
i
v
el
y
.
The
val
u
e
of
no
nl
i
n
ea
r gai
n
k
x
(e
)
is au
to
matically
v
a
ried
d
e
p
e
nd
s on
the v
a
lu
e
o
f
α
i
t
h
at
i
s
on
-l
i
n
e
gene
rat
e
d
usi
n
g (
1
1
)
.
Fi
gu
re
2 s
h
ows
t
h
e
bl
oc
k
di
ag
ram
of t
h
e
sy
st
em
wi
t
h
S
N
P
I
D c
ont
rol
l
e
r.
1
)
(
)
(
s
ds
d
s
e
s
i
(1
1)
In
ord
e
r to ensu
re t
h
e stab
ilit
y o
f
t
h
e system, th
e m
a
x
i
m
u
m v
a
lu
e of
n
o
n
lin
ear
g
a
i
n
k
(
e
ma
x
)
shou
ld
be
d
e
term
in
ed
in
ad
v
a
n
ce. Th
is
is p
e
rform
e
d
via Pop
o
v
st
ab
ility criterio
n
.
Th
e
p
r
o
cedur
e of th
is criterio
n
b
a
sed
on
seco
n
d
o
r
d
e
r sy
st
em
has
di
scuss
e
d
by
o
t
her re
searc
h
er
i
n
[
24]
. Si
nce
t
h
e pl
a
n
t
i
s
re
prese
n
t
e
d a
s
a
t
h
i
r
d
o
r
d
e
r system
, t
h
e u
s
e
o
f
a M
A
TLAB fun
c
ti
o
n
is m
o
re p
r
actical to
g
e
t th
e Po
pov
p
l
o
t
. Fig
u
re 3
illu
strate th
e
Pop
o
v
p
l
o
t
of t
h
e tested system
. It is p
o
ssi
b
l
e to
co
nstruct a straigh
t
lin
e
with
a po
sitiv
e
slo
p
e
p
a
ssing
thro
ugh
the intersect point between Popov and the
real ax
is wh
ere t
h
e Popo
v
p
l
o
t
is en
tirely to
th
e rig
h
t
of th
is lin
e. It
can be see
n
that the Popov plot of
G
(j
)
is
crossi
ng the
real axis at th
e poi
nt
(
-
0
.
2
3
4
,
j0
). T
h
e m
a
xim
u
m
val
u
e o
f
t
h
e
no
nl
i
n
ear
gai
n
ca
n be obt
ai
ned
usi
n
g (1
2
)
. Th
eref
ore
,
t
h
e
al
l
o
wa
bl
e ran
g
e of
n
o
n
l
i
n
ear
ga
i
n
k
(
e
)
is (0,
4.
274).
j
G
e
k
1
max
(1
2)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Rob
u
s
t C
o
n
t
ro
l
S
t
ra
teg
y
f
o
r Pn
euma
tic Drive S
y
stem via En
han
ced Non
lin
ear PID Con
t
ro
ller
(
M
.F. R
a
hm
at)
66
2
Figure
3
.
P
o
p
o
v
pl
ot
Fi
gu
re
4.
R
e
l
a
t
i
ons
hi
p
bet
w
ee
n
an
d
Su
bseq
ue
nt
l
y
, t
h
e desi
g
n
pa
r
a
m
e
t
e
rs are det
e
rm
i
n
ed by
i
d
ent
i
f
y
i
ng t
h
e r
e
l
a
t
i
onshi
ps be
t
w
een
and
i
n
o
r
der t
o
pr
o
duce t
h
e m
a
xi
m
u
m
val
u
e of
rat
e
va
ri
at
i
on (
α
i
)
with expone
ntial decay. It perform
e
d using
Particle Swarm
Op
ti
m
i
za
tio
n
(PSO) techniq
u
e
.
Details on
t
h
i
s
t
ech
ni
q
u
e
were
e
xpl
ai
ned
i
n
[
2
2]
. T
a
bl
e
2
indicates the re
sults of
and
t
h
r
o
u
g
h
t
h
i
s
opt
i
m
i
zati
on t
echni
que
. The
r
e
l
a
t
i
onshi
ps be
t
w
een
and
can
be
pl
ot
t
e
d
as s
h
o
w
n
i
n
Fi
g
u
re
4
.
T
hus
, t
h
e e
q
uat
i
o
n
as e
x
pr
essed i
n
(1
3
)
c
a
n t
h
en
be
ap
p
l
i
e
d t
o
det
e
rm
ine t
h
e
val
u
e o
f
and
.
519
.
0
(1
3)
Tabl
e 2. Param
e
t
e
r
det
e
rm
i
n
ati
on
v
i
a Particle Swarm
Op
timizatio
n
Pa
ra
m
e
ter
Opt
1
Opt
2
Opt
3
Opt
4
Opt
5
167.
90
2
141.
20
2
158.
90
4
144.
50
2
129.
51
0
324.
41
1
267.
51
3
305.
21
1
285.
30
1
248.
53
1
:
0.
518
0.
528
0.
521
0.
506
0.
521
The rat
e
va
ri
at
i
on (
α
i
) is d
e
sig
n
e
d
to
provid
e
a certain
v
a
lu
e of non
lin
ear
g
a
in
at th
e b
e
g
i
nn
ing
for th
e
p
u
rp
o
s
e t
o
o
v
erco
m
e
th
e static frictio
n
.
This rate v
a
riatio
n
is th
en
d
ecreasin
g
startin
g fro
m
th
is v
a
lu
e
and
endi
ng at 0
where the steady
state response i
s
achieve
d.
Fo
r b
e
tter in
terpretatio
n
,
it can
be elab
o
r
ated
throug
h
th
e fo
llowing
deriv
a
tion
.
From
(1
1
)
, let;
1
)
(
,
s
s
G
C
onsi
d
ere
d
i
m
pul
se
res
p
onse
rep
r
ese
n
t
s
t
h
e
err
o
r
si
g
n
al
, t
h
us;
)
(
exp
1
)
(
)
(
,
1
,
t
s
G
L
t
g
t
(1
4)
Perf
o
r
m
differ
e
ntial of
(
1
4
)
;
)
(
exp
)
(
)
(
2
,
t
t
g
dt
d
t
t
(1
5)
Based
o
n
th
e i
n
itial v
a
lu
e t
h
eo
rem
;
2
2
0
exp
lim
)
(
lim
t
t
t
t
(1
6)
B
a
sed
on
t
h
e
fi
nal
val
u
e t
h
eo
r
e
m
;
-0
.
4
-0
.
3
-0
.2
-0
.
1
0
-0
.6
-0
.4
-0
.2
0
0.
2
R
e
[(
j
w
)]
wI
m
[
G
(
j
w
)
]
-0
.
2
4
-0
.23
5
-0
.
2
3
-1
0
-5
0
x 1
0
-3
-0
.
2
3
4
240
260
280
300
320
340
120
130
140
150
160
170
Value o
f
Value o
f
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE Vo
l. 4
,
N
o
. 5
,
O
c
tob
e
r
20
14
:
658
–
6
67
6
63
0
exp
lim
)
(
lim
2
t
t
t
t
(1
7)
2.
4. De
ad-z
o
n
e
C
o
mpen
sa
ti
on
In
pract
i
ce, t
h
e wi
dt
h o
f
dead
-zone i
s
un
kn
o
w
n. T
hus
, t
h
e
DZC
i
s
em
pl
oy
ed t
o
of
fset
t
h
e del
e
t
e
ri
ous
effects of dead-zone. A simila
r co
m
p
ensator as in [22] is
used t
o
overcom
e t
h
i
s
probl
em
. It
is im
pl
em
en
t
e
d by
usi
ng
t
h
e fol
l
o
wi
ng rul
e
s:
0
e
DZC
d
U
U
then
e
e
if
P
DZC
d
u
U
then
U
And
e
e
if
0
n
DZC
d
u
U
then
U
And
e
e
if
0
where
U
e
0
,
u
p
and
u
n
are inpu
t co
m
p
en
sat
i
o
n
b
a
sed
o
n
erro
r,
p
o
s
itiv
e d
ead
-zon
e co
m
p
e
n
satio
n
an
d
neg
a
tiv
e
dead-zone com
p
ensa
tion, respectively.
B
a
sed on t
h
ese co
ndi
t
i
ons, t
h
ere
i
s
no force i
m
posed t
o
t
h
e pay
l
oad
wh
en
th
e
o
u
t
pu
t o
f
th
e
DZC
is represented by
U
e
0
. Fo
r t
h
e o
t
h
e
r co
nd
itio
n
s
in
wh
ich th
e p
o
s
itio
n
error is
exceeded, ed is in either positive or ne
gative direction, the output of the
c
ontroller is added to the
dead zone
com
p
ensator
u
p
and
u
n
, respect
ively.
3.
RESULTS
A
N
D
DI
SC
US
S
I
ON
The pe
rf
orm
a
nce of t
h
e p
n
e
u
m
a
t
i
c
posi
t
i
oni
ng sy
st
em
cont
rol
l
e
d
by
SNP
I
D was exam
i
n
ed usi
ng t
h
e
d
i
fferen
t
step
i
n
pu
t an
d tested
to th
e
v
a
ri
o
u
s of lo
ad
a
n
d
p
r
essu
re.
The
di
ffe
rence
with t
h
e
nom
inal load
a
n
d
p
r
essure w
e
re tested
to
illu
strate th
e robustn
ess of th
is
co
n
t
ro
ller.
The p
e
rform
a
n
ce o
f
th
is tech
n
i
q
u
e
is
com
p
ared t
o
t
h
e
ot
her
t
ech
n
i
ques
nam
e
l
y
con
v
e
n
t
i
onal
PID
an
d
NP
I
D
co
nt
r
o
l
l
e
r.
The
param
e
t
e
rs of
t
h
e
pr
o
pose
d
co
nt
r
o
l
l
e
r i
n
cl
u
d
i
n
g
SNF an
d ot
he
r param
e
t
e
rs are t
a
bul
at
ed i
n
Tabl
e 3. F
i
g
u
r
e
5
d
e
m
o
n
s
t
r
a
t
e
s
t
h
e
sim
u
l
a
t
e
d resu
l
t
of t
h
e o
u
t
p
ut
resp
onse
o
b
t
a
i
n
ed fr
om
the sy
st
em
cont
rol
l
e
d by
SN
PID,
NPI
D
a
n
d PI
D
controller. The
result indicates that
these
controllers are able to follow
the input with different position and
di
rect
i
on. Tho
u
gh, i
t
can be seen t
h
at
t
h
e SN
PID of
fer
faster response with lower
steady-state
error com
p
ared
t
o
t
h
e ot
her
m
e
t
hods. The st
eady
-
st
at
e error for t
h
e sy
st
em
wi
t
h
NPID i
s
cl
ose
m
i
m
i
cs
t
h
e resul
t
obt
ai
n
e
d by
t
h
e sy
st
em
wi
th SNPI
D cont
r
o
l
l
e
r. However
,
i
t
provi
des t
h
e sl
ower respo
n
se co
m
p
ared t
o
t
h
e ot
hers.
For a
sy
st
em
wi
t
h
PID co
nt
rol
l
e
r, t
h
e per
f
o
r
m
a
nce i
s
do
ggere
l c
o
m
p
ared to other due to
the
presence of ove
r
shoot
t
h
at
can reduce t
h
e sy
st
em
robust
n
ess. In
ord
e
r t
o
val
i
d
at
e the perf
orm
a
nce of t
h
e SNPI
D cont
r
o
l
l
e
r, t
h
e resul
t
from
t
h
e sim
u
lat
i
on i
s
com
p
ared t
o
t
h
e
resul
t
obt
ai
ned f
r
o
m
t
h
e real
-t
im
e sy
st
em
. As can be see
n
f
r
om
Fi
gure
6, t
h
e
resp
o
n
se
o
b
t
a
i
n
ed
ba
se
d
on
ex
pe
ri
m
e
n
t
al is qu
ite si
milar with
th
e
si
m
u
latio
n
.
Table
3.
Param
e
ters o
f
t
h
e c
o
n
t
roller
Contr
o
l str
a
tegies
Contr
o
l Param
e
ter
s
Parameter Abbreviation
Value
PID
Pr
opor
tional Gain
K
p
2.
099
Integral Gain
T
i
104.
60
3
Derivative Ga
in
Filter
T
d
N
0.
035
12.
207
SN-
F
unction
Para
m
1
129.
51
0
Para
m
2
248.
53
1
Var
i
ation of E
r
r
o
r
e
ma
x
1.
350
Dead-zone co
m
p
e
n
sator
Contr
o
l value in the r
a
nge of desir
e
d
e
ss
u
e0
0.
010
+ve dead-zone com
p
ensation
u
p
0.
500
-ve dead-zone co
mpensation
u
n
-
0
.
650
Desired e
ss
e
d
0.
005
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Rob
u
s
t C
o
n
t
ro
l
S
t
ra
teg
y
f
o
r Pn
euma
tic Drive S
y
stem via En
han
ced Non
lin
ear PID Con
t
ro
ller
(
M
.F. R
a
hm
at)
66
4
Fi
gu
re
5.
Si
m
u
l
a
t
i
on res
u
l
t
f
o
r
di
ffe
rent
c
o
nt
rol
l
e
r
Fig
u
re 6
.
Validatio
n
resu
lt for SNPID
3.1. Robus
tne
ss
Tests
Robustness can be defined a
s
the ability of
a contro
l sys
t
em
to be insensitive
to the variation of the
pl
ant
param
e
t
e
rs. I
n
o
r
de
r t
o
t
e
st
t
h
e sy
st
em
rob
u
st
ness, s
o
m
e
i
nvest
i
g
ati
ons are
perf
orm
e
d t
o
t
h
e sy
st
em
.The
ab
ili
ty o
f
th
e
SNPID co
n
t
ro
l
l
er to
co
m
p
en
sate
th
e sy
s
t
e
m
when there are change
s occur in the load and
pressure is exa
m
ined. The perform
ance
is a
n
alyzed fo
r both
co
n
d
itio
n
s
in
th
e case o
f
th
e lo
ad
/p
ressure is
i
n
creasi
ng or
decreasi
ng. Th
e
m
e
asurem
ent
of t
h
e perfo
rm
ance i
s
based on t
h
e di
st
ance of 200
mm
.
C
o
m
p
ari
s
on
wi
t
h
t
h
e ot
he
r m
e
t
hods are
perf
orm
e
d as a perf
orm
a
nce benchm
ark. The det
a
i
l
s
of t
h
e
perform
a
nce based on the
experim
e
nts for
all cases are
t
a
bul
at
ed i
n
Ta
b
l
es 4 t
h
r
o
u
g
h
7. Th
e r
e
su
lt in
d
i
cates
that the S
N
P
I
D+DZC a
n
d
N
P
ID+
D
ZC c
o
n
t
roller are
m
o
re
r
obu
s
t
th
an
P
I
D
+
D
Z
C. I
t
ca
n
b
e
s
e
en
th
a
t
,
wh
e
n
t
h
e
m
ovi
n
g
m
a
ss i
s
decrease
d
fr
om
8.4k
g
t
o
3.
1 k
g
, t
h
e
ove
rs
ho
ot
f
o
r
t
h
e PI
D co
nt
r
o
l
l
e
r i
s
si
gni
fi
cant
l
y
increase
d
. It
be
com
e
s
m
o
re aggra
v
ated if t
h
e
mass is incr
eased
f
r
o
m
8
.
4kg
to
13
.5
kg
and
u
lti
m
a
tel
y
affected
th
e stab
ility o
f
th
e syste
m
. T
h
e sam
e
si
tu
atio
n
o
ccurs wh
en
th
e pressu
re is d
ecreased an
d
in
creased fro
m
0.
6 M
P
a t
o
0
.
4
5
M
p
a a
nd
0.
75 M
P
a
,
res
p
ect
i
v
el
y
.
M
eanw
h
i
l
e
, t
h
e s
y
st
em
wi
t
h
SNPI
D c
ont
rol
l
er has
succee
ded t
o
a
c
hieve
better perform
a
nce. T
h
e consistency
of t
h
e pe
rform
a
nce for all cas
es indicates that this
co
n
t
ro
ller is less sen
s
itiv
e to
t
h
e ch
ang
e
s
o
f
l
o
ad
and
pr
essure. Th
e ov
erall an
alysis o
n
t
h
ese find
ing
is p
l
o
tted
i
n
Fi
g
u
re
7.
Th
e num
bers
of
1
,
2,
3 a
n
d
4
o
n
t
h
e x
-
axes
re
p
r
esent
t
h
e
ex
pe
ri
m
e
nt
s of t
a
bl
e 1, t
a
bl
e
2, t
a
bl
e 3
and
table 4, res
p
ectively.
Tabl
e 4. Perf
or
m
a
nce of t
h
e s
y
st
em
for M
=
3
.
1 k
g
wi
t
h
nom
i
n
al
load M
=
8.
4 k
g
Per
f
orm
a
nce
Co
n
t
ro
ller
SNPID+
DZC
NPID+
D
ZC
PID+D
Z
C
Settling Ti
m
e
(t
s
)
0.
659
1.
524
1.
123
Rise Ti
m
e
(t
r
) 0.
314
1.
268
0.
317
Over
shoot
(%OS)
0.
000
0.
000
7.
973
Steady-stat
e
erro
r (e
ss
)
0.
043
0.
112
0.
267
Tabl
e 5.
Per
f
o
r
m
a
nce of t
h
e s
y
st
em
for M
=
1
3
.5
kg
wi
t
h
nom
i
n
al
load M
=
8.
4 k
g
Per
f
orm
a
nce
Co
n
t
ro
ller
SNPID+
DZC
NPID+
D
ZC
PID+D
Z
C
Settling Ti
m
e
(t
s
)
0.
679
1.
803
1.
403
Rise Ti
m
e
(t
r
) 0.
241
1.
455
0.
306
Over
shoot
(%OS)
0.
000
3.
375
20.
869
Steady-stat
e
erro
r (e
ss
)
0.
046
0.
118
0.
269
5
10
15
20
-
10
0
0
10
0
20
0
15
16
17
-1
0
0
-5
0
0
50
10
0
15
0
13
.
8
9
13
.
8
9
2
13
.
8
9
4
-1
0
0
-9
9
.
9
9
-9
9
.
9
8
In
p
u
t
SN
P
I
D
PI
D
NP
I
D
15
20
25
30
-2
0
0
-1
0
0
0
10
0
20
0
Ti
m
e
(
s
)
D
i
sp
l
a
ce
me
n
t
(
mm)
33
.
5
34
34
.
5
14
8
15
0
15
15
.
5
16
16
.
5
-1
0
0
-50
0
50
10
0
15
0
S
i
m
u
la
ti
o
n
Ex
p
e
r
i
m
e
nt
In
p
u
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE Vo
l. 4
,
N
o
. 5
,
O
c
tob
e
r
20
14
:
658
–
6
67
6
65
Tabl
e 6. Perf
or
m
a
nce of t
h
e s
y
st
em
when Ps
i
s
reduced
t
o
0.
45
M
P
a
Per
f
orm
a
nce
Co
n
t
ro
ller
SNPID+
DZC
NPID+
D
ZC
PID+D
Z
C
Settling Ti
m
e
(t
s
)
0.
797
1.
612
1.
115
Rise Ti
m
e
(t
r
) 0.
336
1.
173
0.
334
Over
shoot
(%OS)
0.
000
0.
000
8.
081
Steady-stat
e
erro
r (e
ss
)
0.
0162
0.
021
0.
196
Table 7.
P
e
rformance of the
s
y
s
t
e
m
when Ps
is
increas
ed
to 0.75 MPa
Per
f
orm
a
nce
Co
n
t
ro
ller
SNPID+
DZC
NPID+
D
ZC
PID+D
Z
C
Settling Ti
m
e
(t
s
)
0.
699
1.
276
1.
321
Rise Ti
m
e
(t
r
) 0.
340
0.
636
0.
325
Over
shoot
(%OS)
0.
000
3.
304
19.
992
Steady-stat
e
erro
r (e
ss
)
0.
019
0.
025
0.
367
(a)
(b
)
(c)
(d
)
Fi
gu
re
7.
R
o
bu
st
ness a
n
al
y
s
i
s
base
d
on
dec
r
e
a
si
ng
an
d i
n
cre
a
si
ng
o
f
l
o
a
d
4.
CO
NCL
USI
O
N
In
th
is
p
a
p
e
r,
a rob
u
s
tn
ess
of th
e SNPID co
n
t
ro
ller is presen
ted
.
In
itial
l
y
, th
e p
e
rfo
r
man
ces o
f
th
e
syste
m
with
th
is co
n
t
ro
ller are ex
a
m
in
ed
th
ro
ug
h
sim
u
la
tio
n
s
. Exp
e
rim
e
n
t
s to
th
e real p
l
an
t were p
e
rformed
for
val
i
d
at
i
on purp
o
ses and
fo
un
d onl
y
sl
i
ght
di
st
i
n
cti
ons
b
e
tween
th
e
m
in
th
e tran
sien
t p
a
rt. Su
b
s
equ
e
n
tly, th
e
robustness
of t
h
e syste
m
was investig
ated chiefly by decreasing and increa
sing the load.
Moreove
r, the
effect
caused by
va
ri
at
i
on of
pressu
res t
o
t
h
e sy
st
em
perform
ance is also exa
m
in
ed. The
syste
m
with SNP
I
D s
hows
a
superior perform
a
nce in ter
m
s of
accuracy,
speed and robustness com
p
ar
ed to anothe
r m
e
thod that were
ex
a
m
in
ed
in
th
is research
. Overall, i
t
p
r
o
v
i
des a lo
wer stea
d
y
state
erro
r an
d
is ab
le to
main
tain
th
e res
p
on
se
wi
t
hout
o
v
ersh
oot
.
ACKNOWLE
DGE
M
ENTS
Thi
s
re
searc
h
i
s
su
pp
ort
e
d
by
M
i
ni
st
ry
of
Hi
g
h
e
r
E
ducat
i
o
n
(M
O
H
E)
Malaysia, Un
iv
ersiti
Tekn
o
l
o
g
i
Mal
a
ysia (UTM) an
d Un
iv
er
siti Tekn
ik
al Malaysia Melak
a
(UTe
M) t
h
rou
g
h
Research
Univ
ersity
G
r
an
t (
G
U
P
)
Tier
1
vo
te num
b
e
r
Q
.
J13
000
0.712
3.00H
36
. Au
thor
s ar
e g
r
atefu
l
to
the Min
i
str
y
, UTM an
d
UTeM
fo
r s
u
p
p
o
r
tin
g the
w
o
rk
.
REFERE
NC
ES
[1]
M Karpenko and N Sepehri. "Development
and experimental evaluation of a fi
xed
-
gain nonlin
ear control for a low-
cost pneumatic actuator".
I
E
E
Pr
oceed
ings
Contr
o
l T
h
eor
y
and
A
pplicat
ions
.
200
6; 153: 629-640.
[2]
SR Pandian, F
Takemura, Y Hay
a
kawa
, and
S Kawamura. "Pressure observer-
controller desig
n
for pneumatic
c
y
lind
e
r a
c
tu
ator
s".
IEEE/ASM
E
Transactions on
Mechatronics
.
2
002; 7: 490-499.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Rob
u
s
t C
o
n
t
ro
l
S
t
ra
teg
y
f
o
r Pn
euma
tic Drive S
y
stem via En
han
ced Non
lin
ear PID Con
t
ro
ller
(
M
.F. R
a
hm
at)
66
6
[3]
P
Beat
er.
Pneumatic Drives (
S
ystem Design, Mod
e
ling
and Contr
o
l)
. Verlag
Ber
l
in Heidelberg: Sp
ringer. 2007.
[4]
T Shen, K T
a
m
u
ra, N
Henm
i,
and T
Nakazaw
a. "Robust
m
odel fo
llowing
co
ntroller app
lied
to position
i
ng o
f
pneum
atic
con
t
r
o
l va
lve
with fr
i
c
tion".
In
I
E
EE I
n
ternational Co
nference on
Con
t
rol Applica
tions
. 1998: 512-516.
[5]
L Reznik
, O Ghanay
em, and A Bourmistrov. "PID plus fuzz
y
controller structu
r
es as
a design base for industrial
applications".
Engineering
Applications o
f
Art
ifi
c
i
al In
tel
ligen
ce
.
2000; 13: 419-4
30.
[6]
E Richer and Y Hurm
uzlu. "A High P
e
rform
ance
P
n
eum
a
tic F
o
rce Actuator S
y
s
t
e
m
. P
a
rt 2 - Nonl
inear Contro
lle
r
Design".
ASM
E
Journal of Dyna
mic Sy
stems Measurement and
Control
. 2001
; 1
22: 426-434.
[7]
R Vilanova. "IMC based Robust
PID design:
Tun
i
ng guidelin
es and automatic tun
i
ng".
Journal of Process Control
.
2008; 18: 61-70.
[8]
RR Sumar, AAR
Coelho,
and Ld
S Coelho.
"Com
putation
a
l in
te
lli
ge
nce
approach to PID controller
design using the
universal model"
.
Information
Sc
i
e
nces
. 2010; 180
: 3980-3991.
[9]
RB van Varseveld and GM Bone. "Accurate
position contro
l of a
pneumatic
act
u
a
tor using on/off solenoid v
a
lves"
.
IEEE/ASME Transactions on M
e
chatronics
. 1997
; 2: 195-204.
[10]
WD Chang an
d SP Shih. "PI
D
controller design of
nonlinear s
y
stems using an improved particle swarm
optim
izat
ion app
r
oach".
Communications
in Non
l
inear Science an
d Numerical Sim
u
lation
.
2010; 1
5
: 3632-3639.
[11]
M Rahimi Khoy
gani, U Islamic
Azad,
S
Hajigha
s
e
m
i
, U Is
lam
i
c Azad, D S
a
nae
i
,
and U Is
lam
i
c Azad. "Des
ignin
g
and Simulation for Vertical Moving Control of
UAV Sy
stem using PID LQR
and Fuzzy
logic".
Internation
a
l
Journal of Electrical and
Co
mputer
Eng
i
neer
ing
(
I
JECE)
.
2013; 3:
651-659.
[12]
G Qin, Y. Ma, X. Zhang
,
and M. Zhang, "Design
of
F
u
zz
y
Adapt
i
v
e P
I
D Tem
p
erature Control
l
er
Bas
e
d on F
P
G
A,"
TELKOMNIKA Indonesian Journ
a
l of
Electrical
Engineering,
vol. 11
, pp
. 6008-6
016, 2013
.
[13]
K Ahn and T Th
anh. "Nonlinear
PID control to improve the
cont
rol perform
anc
e
of the pneum
at
ic
artif
ici
a
l m
u
scle
manipulator
usin
g neural n
e
twork
"
.
Journal of
Mechanica
l Scien
c
e and Techno
logy
. 2005; 19: 106-
115.
[14]
C Jun
y
i
,
C
Bing
gang,
Z Xin
i
ng,
and W
Guangnu
i. "Fraction
a
l Pr
oportional
Int
e
gr
al Con
t
rol for
Pn
eum
a
tic Position
S
e
rvo S
y
s
t
em
s
"
. P
r
es
ented
a
t
th
e
IEEE/ASM
E In
ternational Conference
, MESA, 2
008.
[15]
S Cho. "Trajector
y
track
ing control of a pn
eumatic
X-Y table using neur
al ne
twork bas
e
d PID control".
International Jo
urnal of
Precisi
o
n
Engin
eering
a
nd Manufacturin
g
. 2009; 10: 37-
44.
[16]
M Taghizad
eh,
F Najafi, and A Ghaffari
. "Multim
odel PD-c
ontrol of a pneum
atic actu
a
tor under
variabl
e
loads".
Int J Ad
v Manu
f
Technol
. 2010; 4
8
: 655-662.
[17]
MF Ra
hma
t
,
SNS Sa
lim,
AAM Fa
udz
i,
ZH I
s
mail, SI Samsudin, NH Sunar
, et al.
"Non-lin
ear Modeling and
Cascade Con
t
rol of an Industrial Pneumatic Actu
ator S
y
s
t
em".
Australian Journa
l of Basic and
Applied S
c
ien
ces
.
2011; 5: 465-47
7.
[18]
MF Rahmat, SNS Salim, NH Su
nar, AAM Faud
zi, ZH Ismail, and K Huda
. "Ide
ntification and non-linear contro
l
strateg
y
for indu
strial
pneumatic
actu
a
tor".
International
Journal
of th
e Ph
ysica
l
S
c
ien
ces
. 2012; 7: 2565 -
2579.
[19]
S Salim, MF Ra
hmat, AA Faudzi, and
Z Is
mail.
"Position contro
l of pneumatic
actuator using an
enhancement of
NPID controller
based on the
characterist
ic of
rate v
a
riation
nonlinear gain"
.
The Internatio
nal Journal of
Advanced Manu
facturing Techno
logy
. 2014: 1-15
.
[20]
K Osman, AA
Mohd Faudz
i,
MF Rahmat, an
d K Suzumori. "Sy
s
tem
Identif
ication and Emb
e
dded Controller
Design for Pneum
a
tic Actu
ator
w
ith Stiffness Charac
terist
ic".
M
a
themati
c
al Pro
b
lems in Engin
e
ering
. 2014; 201
4:
13.
[21]
SNS Sa
lim,
MF
Ra
hma
t
,
AAM Fa
udz
i,
NH Suna
r,
Z
H
Is
mail, and SI Samsudin. "Tr
acking
performance
an
d
disturbance r
e
jection of
pneum
a
tic
ac
t
u
a
t
or s
y
st
em".
In
Control C
onference (
A
SC
C)
, 2013 9th
Asian
. 2013: 1-6.
[22]
SNS
Salim
, MF Rahm
at, AAM Faudzi,
ZH Ismail
, and NH S
unar. "Position Co
ntrol of Pneumatic Actu
ator Usin
g
Self-Regul
ation
Nonlinear
PID".
Mathematical Problems
in Engin
eering
. 2014; 20
14: 12.
[23]
E George.
Obse
r
v
e
rs in Control Sy
ste
m
s
. Lon
d
an:
Academic Press. 2002.
[24]
H S
e
raji
. "A n
e
w clas
s
of
nonlin
ear P
I
D
controllers with ro
b
o
ti
c appli
cat
ions
".
J
Robot S
y
st
.
1998
; 15: 161-81.
BIOGRAP
HI
ES OF
AUTH
ORS
Sy
e
d
Najib bin Sy
e
d
Salim
receiv
e
d B.En
g. (Mecha
t
ronic
s
) from
Universiti Tekno
logi
Malay
s
ia in 199
8. He com
p
lete
d his M. Eng. (
E
lectri
cal) at Universiti T
e
knol
ogi Malay
s
ia in
2003. He is a senior lecturer at t
h
e Universiti
T
e
knikal Mal
a
y
s
ia
Melaka, Malay
s
i
a
. Curren
t
l
y
, h
e
is pursuing a PhD degree in
Ele
c
tri
cal
Engin
eering a
t
Unive
r
siti Tekno
logi
Mala
y
s
ia
. His
res
earchs
ac
tivi
t
y
inc
l
udes
contr
o
l
s
y
s
t
em
s
desig
n
, instrumentatio
n and
automatio
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE Vo
l. 4
,
N
o
. 5
,
O
c
tob
e
r
20
14
:
658
–
6
67
6
67
Mohd Fua’ad Rahmat
receiv
e
d
degr
ee in E
l
ec
tric
al Engin
e
ering at
Univ
er
siti
T
e
knologi
Malay
s
ia
in 198
9. He started
hi
s
Master
degree b
y
taught course
specialized in
C
ontrol S
y
stem
Engineering
an
d graduated
in
1993 at The Univer
sity
o
f
Sheffield,
UK. Subsequently
, h
e
pursued his PhD degree in
Electr
onic Instrumenta
tion Engin
eering
at the School of
Engineer
ing
,
Sheffield Hallam University
,
UK
and graduated in 1996. Currently
, h
e
is a Professor in the
Department of
Control and M
echatronics
Eng
i
n
eer
ing, Faculty
of
Electr
ical Engineering
,
Universiti T
e
kn
ologi Malay
s
i
a
Skudai Johor. Hi
s field of specialization
i
n
clude S
y
s
t
em
Identif
ication an
d Estimation, Si
gnal Processing, Process Tomog
r
aph
y
for Industrial Process,
Process Control
Instrumentation
,
Sensors and
Ac
t
u
ators
,
H
y
dr
auli
c and
P
n
eum
a
ti
c
S
y
s
t
em
.
Ahmad `Athif
Mohd Faudzi
was
born in 1982. He receiv
e
d
the B. Eng. an
d the M
.
Eng.
degrees from
Universiti
Tekno
l
ogi Mala
ysi
a
, M
a
la
ys
i
a
and th
e
Dr. Eng.
in S
y
st
em
Integrat
ion
from Okay
ama University
, Japan in 2004, 2006, and
2010 respectiv
ely
.
He is now attach
ed with
the Cen
t
re for
Artifi
c
ial Int
e
ll
igence and
Roboti
c
s (CAIRO), Universiti
Teknolog
i Malay
s
i
a
as a
Senior Lecturer
. He is mainly
engaged
in the research f
i
elds
of pneumatic
actu
a
tors, soft
actu
a
tors, robotics automation
an
d their
applicatio
ns.
Z
ool H Ismail
obtain
e
d his Ph.D
. degr
ee
in Electrical
Eng
i
neer
in
g from Heriot-Watt Univ
ersity
,
Edinburgh, Un
ited Kingdom in
2011. In 2005
and 2007,
he received h
i
s B.
En
g and M.
Eng
.
degrees f
r
om Universiti Teknologi Malay
s
ia, S
kudai, Johor, M
a
lay
s
ia,
respectively
. He was
appointed as
a Senior Lectur
er at Universi
ti Teknolog
i Malay
s
ia
in 2011
exactly
after
completing h
i
s
Ph.D. He is
a
member of Soci
ety
for Und
e
rwater
Te
chnolog
y, IEEE Ocean
ic
Engineering Society
and Asian
Control Associa
tion
.
Currently, he is involv
e
d in nonlinear
control s
y
st
em
and his
m
a
in research in
terest is
in the develop
m
ent of nonlinear and adaptiv
e
robust task-space contro
l meth
ods
for regulation and trajecto
r
y
tr
ack
ing con
t
rol of robotic
s
y
ste
m
s.
NH Sunar
is currentl
y
pursuing
a Ph.D. degree in
Electri
cal
Engin
eering
in Univer
siti T
e
knologi
Malay
s
ia. In 2008, she received h
i
s B. Eng degr
ees from
Universiti Teknol
og
i Malay
s
ia, Skudai,
Johor, Malay
s
ia. She was appointed as a Compone
nt Design Engineer
at In
tel
Microelectron
i
c
Malay
s
ia
in 200
8 exactly
after completing her
B
.
Eng fo
r 2
y
e
ar
s. Currently
, she is involv
e
d in
adapt
i
ve
control
s
y
s
t
em
and
her
m
a
in res
e
a
r
ch
in
teres
t
is
in
the
d
e
velopm
ent
of
a
d
aptiv
e con
t
rol
methods for reg
u
lation
and
tr
ajector
y
tracki
ng
co
ntrol of
pneumatic
actu
a
tor
s
y
stem.
Shams
ul Anuar
Shamsu
d
in
received h
i
s
P
h
.D. i
n
Engine
ering fr
om
the Univers
i
t
y
of D
a
yton i
n
Day
t
on
, Ohio,
USA in
May
2013. His research
interests include mechan
isms,
machine
component design, engin
eering
design methods, engineering gra
phics and auto
mation. He has
been with UTe
M
(and KUTKM
) for 13 year
s
.
He is
a s
e
nior lectur
er a
t
t
h
e F
acult
y o
f
M
echani
cal
Eng
i
neering
in
UTe
M
.
Evaluation Warning : The document was created with Spire.PDF for Python.