TELKOM
NIKA
, Vol.14, No
.1, March 2
0
1
6
, pp. 144~1
5
5
ISSN: 1693-6
930,
accredited
A
by DIKTI, De
cree No: 58/DIK
T
I/Kep/2013
DOI
:
10.12928/TELKOMNIKA.v14i1.2538
144
Re
cei
v
ed Au
gust 28, 20
15
; Revi
sed
No
vem
ber 2
5
, 2015; Accepte
d
De
cem
ber
13, 2015
Co-simulation and Experiment Research on a Novel
Erection Mech
anism
Feng Jiang
t
ao*, Huan
g Xianxiang, G
a
o Qinhe, G
uo Xiaosong
, Yao Xiaoguang
Xi
’a
n Hig
h-tech
Researc
h
Institute,
T
ongxi
n
Roa
d
No. 2, Hong
qi
n
g
T
o
w
n
, Baqi
a
o
District, Xi’
a
n
,
Shaan
xi, Ch
in
a
*Corres
p
o
ndi
n
g
author, em
ail
:
fengjt29
108
2
217
@12
6
.com
A
b
st
r
a
ct
Th
e
e
r
e
c
tio
n
m
e
ch
an
i
s
m
wi
th
mo
vab
l
e b
a
c
k h
i
ng
ed
b
e
a
r
in
g
i
s
a
no
ve
l
e
r
e
c
ti
on
m
e
cha
n
i
sm
and
the form
of its mov
i
ng
proce
ss is complic
a
t
ed. T
he nov
el
erection
mec
han
is
m ne
eds
to be extens
i
v
ely
tested to prov
e its valu
e a
n
d
to ens
ure it
w
o
rks properl
y
. Kine
matic
a
nalysis w
a
s
a
ccomplis
he
d a
n
d
m
a
them
atic
al m
o
del of t
he hydraulic system
was
ac
quir
e
d.
Fu
z
z
y
ada
ptive PID c
ontrol
wa
s adopted for the
erectio
n
mech
anis
m
taki
ng a
d
vanta
ge of
fuzz
y
co
ntro
l
and
PID control. T
he n
o
vel
erecti
on
mec
han
is
m
w
a
s
valid
ated
by vir
t
ual prot
otype t
e
chn
o
lo
gy rea
l
i
z
e
d
by
co-s
i
m
u
l
atio
n metho
d
. T
he mech
anic
a
l, hydra
u
l
i
c an
d
control
mode
l
s
w
e
re respe
c
tively esta
bli
s
hed
in ADA
M
S, AMESim and S
i
mul
i
nk
. Experi
m
e
n
t w
a
s
compl
e
ted
on
a pl
atform. T
h
e resu
lts of si
mu
lati
on
a
nd
exper
iment i
n
d
i
cated th
at the
nove
l
erecti
o
n
mec
h
a
n
is
m co
uld
move b
a
se
d on
desi
g
n
e
d
sche
m
e
an
d t
he co
ntrol
effect of fu
zz
y
a
d
a
p
tive PID co
ntr
o
l
w
a
s excelle
nt. T
he nove
l
erec
tion mech
anis
m
ha
d gre
a
t practical va
lue.
Ke
y
w
ords
: Erection
mec
h
a
n
i
s
m, Co-si
m
ulat
ion, Virt
ua
l prot
otype tech
nol
o
g
y, F
u
zz
y c
ontr
o
l
Copy
right
©
2016 Un
ive
r
sita
s Ah
mad
Dah
l
an
. All rig
h
t
s r
ese
rved
.
1. Introduc
tion
The e
r
e
c
tion
mechani
sm
with mova
b
l
e ba
ck hing
ed be
ari
ng i
s
a
novel e
r
ectio
n
mech
ani
sm. Comp
ared wi
th
traditio
nal ere
c
tion
me
chani
sm it
add
s h
o
ri
zo
ntal
cylinder, th
eref
ore
back
hi
nge
d beari
ng ca
n move
in
h
o
ri
zontal dire
cti
on d
r
iven by
hydra
u
lic
cyl
i
nder.
The
n
o
vel
mechani
sm can
fulfill
the erection requirements
in strictured
space.
The erection mechani
s
m
need
s to be e
x
tensively tested to prove its
value an
d to ensure it works p
r
op
erly.
Experiment
s
norm
a
lly re
qu
ire exp
e
n
s
ive
equip
m
ent
a
nd lon
g
p
r
o
c
e
s
s of pa
rt de
sign an
d
system inte
gration. Com
p
rehen
sive si
m
u
lation
s are
often ne
ce
ssary and
helpf
ul. The erecti
on
mech
ani
sm
contai
ns me
cha
n
ical an
d
hydra
u
lic
system, there
f
ore al
one
software
can
not
compl
e
tely achieve its fea
t
ure. The dynamic b
ehavi
o
r of hydrauli
c
system i
s
highly non
-lin
ear
due
to
th
e phen
omen
a su
ch as non
linearitie
s of
valve cylind
e
r
com
b
inati
on, frictio
n
, fluid
comp
re
ssibilit
y and asso
cia
t
ed stiffness, whi
c
h cau
s
e
difficulties in t
he co
ntrol of su
ch sy
stem
s.
Virtual p
r
otot
ype technol
o
g
y provide
s
an effe
ctive approa
ch
for investigatin
g the
novel
ere
c
tion m
e
chani
sm. On
e
widely
used
simul
a
tion
software
is Si
mulink,
whi
c
h is
a dia
g
ram
prog
ram
m
ing
metho
d
a
n
d
provide
s
a
g
eneral e
n
viro
nment. ADA
M
S is a
po
werful m
e
chani
cal
desi
gn and simulation software, whi
c
h provide
s
t
ool
s to build mech
ani
cal structures a
nd 3
D
visual
simula
tion. AMESim offers
a
grap
hical mo
deling a
p
p
r
o
a
ch
and m
a
ny libra
ries
of
comp
one
nts
particula
rly in
hydrauli
c
s. Co-sim
ulation
method h
a
s been
extensively adopted
in
resea
r
ch and
simulation of
hydrauli
c
ste
e
l-belt overwi
nd buffer dev
ice [1], braki
n
g perfo
rman
ce
of a vehicle [2], a two-axis trackin
g
sy
stem [3],
four-whe
el-d
rive
hybrid ele
c
tric vehicle [4], a
turbojet fuel
system [5], parallel hyb
r
id l
oade
r [6
], an electri
c
b
u
s
with motori
ze
d whe
e
ls [7], the
comp
osite A
BS control
o
f
vehicle
s
[8], and so
on.
The co-sim
ulation platfo
rm exploit
s
the
advantages of
different software.
The si
mulation re
sult
s illustrate that
virtual
protot
ype
techn
o
logy h
a
s a
go
od a
p
p
licatio
n in m
any scop
es
a
nd is an
effective approa
ch
for inve
stigat
ing
behavio
rs
of com
p
lex sy
stem
s. It ca
n also save
time and
cost. Fu
zzy l
ogic
co
ntrol
is an
alternative a
ppro
a
ch to a
c
hieve d
e
si
re
d goal
s.
Che
n
C. Y. [9] p
r
opo
se
d an i
n
tegrate
d
fuzzy
controlle
r to achi
eve a synch
r
on
ou
s p
o
sitioni
ng
obj
ective for a dual-cylind
e
r
electro-hydra
u
lic
liftin
g
s
y
s
t
em w
i
th u
nba
la
nc
e
d
lo
ad
in
gs
, s
y
ste
m
un
ce
rtaint
ies
and
di
sturban
ce
s.
The
experim
ental results sho
w
ed that
the controlle
r ca
n effectively ac
hieve the obj
ective of posi
t
ion
synchro
n
ization. Li L [10]
propo
sed
a fuzzy ad
a
p
tive sliding
mode control
sch
eme wh
ich
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Co-sim
ulation
and Expe
rim
ent Re
sea
r
ch
on a No
vel Erectio
n
Me
ch
anism
(Fen
g Jian
gtao)
145
combi
ned
fu
zzy co
ntrol with sliding
mode co
ntro
l
to a
c
hieve
nonlin
ear co
ntrol of t
he
erctin
g
mech
ani
sm.
Chia
ng M.
H.
[11] combin
ed fu
zzy lo
gi
c a
nd
slidin
g
mode
control to devel
op
an
electro-hydra
u
lic vel
o
city
contro
lle
r th
at
wa
s robu
st t
o
extern
al
lo
ad di
sturban
ces. Z
heng
J.
M.
[12] used fu
zzy
co
ntrol t
o
an
ele
c
tro
-
hydrauli
c
po
sition
se
rvo
and p
r
ovid
ed
good
tra
c
ki
ng
perfo
rman
ce
even
subj
ect
ed to
externa
l
distu
r
ba
nce
s
. Otto
Ce
rm
an [13] i
n
trod
uce
d
a
meth
od
for desi
gn of a fuzzy slidi
n
g mode co
ntroller for
ele
c
t
r
o-hydra
u
lic
servo mechani
sm. The re
su
lts
from above rese
arch sho
w
ed that fuzzy contro
l ha
s bee
n su
ccessfully use
d
in the comp
lex
pro
c
e
ss
with good p
e
rfo
r
m
ance. It has a short ri
se tim
e
and a sm
all
oversh
oot.
We introdu
ce
d co
-sim
ulati
on metho
d
to simulate the
ere
c
tion me
chani
sm. The
ere
c
tion
system
wa
s modele
d
in ADAMS, AMESim and Ma
tl
ab/Simulin
k. The advanta
ges of ADAM
S in
mech
ani
cal system, AMESim in hydraulic syst
e
m
a
nd Simulink i
n
advanced
controlle
r de
sign
were exploite
d and
co
mbi
ned to offe
r
an integ
r
ated
simul
a
tion f
o
r the
ere
c
ti
on me
ch
anism.
Fuzzy ada
ptive PID cont
ro
l wa
s ado
pte
d
to co
ntro
l th
e novel e
r
e
c
tion me
cha
n
si
m. It was abl
e of
adaptatio
n
to
paramete
r
chang
es and
deal with
n
o
n
linea
r dyn
a
m
ic b
ehavio
r asso
ciated
with
hydrauli
c
m
o
tion syste
m
. The aim
o
f
this re
se
arch i
s
to inv
e
stigate th
e
novel e
r
e
c
tion
mech
ani
sm a
nd prove its v
a
lue.
2. Compositi
on of the
No
v
e
l Erection Mecha
n
ism
The e
r
e
c
tion
mech
ani
sm
with movabl
e
back hin
ged
beari
ng i
s
mainly com
p
ose
d
of
ere
c
tion arm, lock devi
c
e,
rail, slider, e
r
ectio
n
and h
o
rizontal cyli
nders, as
sh
own in Figu
re 1.
Erectio
n
a
r
m
is u
s
ed to
su
pport
and
ere
c
t load f
r
om
hori
z
ontal
sta
t
e to vertical
state o
r
ba
ck to
flat. Lock
dev
ice i
s
ap
plied
to fixing and
limiting load
again
s
t vertical and l
a
teral
movement o
n
ere
c
tion arm. Erection cyl
i
nder p
u
she
s
load and e
r
ectio
n
arm
rotating ro
un
d back hing
ed
beari
ng.
Ho
rizontal
cylin
d
e
r p
u
lls ba
ck hi
nge
bea
ring movin
g
along
the
rai
l
, reali
z
ing
lo
ad
moving in ho
rizo
ntal dire
ct
ion. Two ho
ri
zontal
cylind
e
rs a
r
e sym
m
etrically arrange
d to ensure
stability.
Figure 1. Co
mpositio
n of t
he novel erection mech
ani
sm
Comp
ared wi
th
traditio
nal ere
c
tion
me
chani
sm,
the
ere
c
tion
me
chani
sm with movable
back hin
ged
beari
ng a
d
d
s
rail, slid
er a
nd ho
riz
ontal
cylinde
r. It adopts
ere
c
tio
n
and h
o
ri
zo
ntal
cylinde
rs to realize erectio
n
pr
ocess. Erection
cylind
e
r
is
used
to
a
l
ter amplitu
d
e
and
hori
z
o
n
tal
cylinde
r to transfer h
o
ri
zont
al position. T
he nov
el ere
c
tion mech
ani
sm expan
ds the moving form
of erectio
n
m
e
ch
ani
sm.
The erectio
n
pro
c
e
ss
can
be divided int
o
three
stage
s:
(1) Ea
rly erection stage. Horizontal cylin
der is
lo
cked.
Erection cyli
nder p
u
she
s
the load
and erectio
n
arm rot
a
ting round b
a
ck hi
nged b
eari
ng
to a certain a
ngle.
(2) Co
ope
rat
i
on sta
ge.
Horizontal
cyli
nder
sta
r
ts t
o
move. Ere
c
tion a
nd
ho
rizo
ntal
cylinde
rs m
o
ve togethe
r. The load
and e
r
ectio
n
arm
rotate rou
nd b
a
ck hing
ed b
earin
g a
s
well
a
s
move in hori
z
ontal dire
ctio
n.
(3) V
e
rtical a
d
justme
nt sta
ge.
Ho
ri
zonta
l
cylinde
r
ce
ase
s
whe
n
e
r
ectio
n
an
gle
attains
about 80
°~85
°. The load i
s
ere
c
ted to vert
ical state by ere
c
tion cylin
der alo
ne.
Hydra
u
lic p
r
i
n
cipl
e of the
ere
c
tion sy
stem
is
sho
w
n in Fig
u
re
2. Hydrauli
c
system
inclu
d
e
s
hyd
r
aulic
pum
p, overflow valve, bidire
ctio
n
a
l bala
n
ce v
a
lve, hydra
u
li
c lo
ck, hydra
u
lic
cylinde
r,
electro-hyd
r
auli
c
prop
ortio
nal valv
e and other comp
one
nts. From el
ectro
-
hyd
r
auli
c
prop
ortio
nal
7 and 8 which cont
rol the
spee
ds of
h
y
drauli
c
cylin
ders, pre
s
su
re oil flow in the
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 14, No. 1, March 2
016 : 144 – 1
5
5
146
forwa
r
d
ch
am
bers of e
r
e
c
ti
on cylin
de
r 1
8
and
ho
ri
zo
ntal cylind
e
r
15 an
d 16.
T
h
is a
c
tuali
z
e
s
the
load rotatin
g
round b
a
ck hi
nged b
eari
ng
whi
c
h move
s in hori
z
ontal
dire
ction.
1-Motor; 2-Pu
mp ; 3-Check v
a
lve; 4-T
h
rottle valve; 5-
Press
u
re ga
ge; 6-R
e
lief valv
e; 7,8-Electro-h
y
d
r
au
l
i
c
prop
ortion
al va
lve; 9-Bil
a
teral
pilot-c
ontrol
l
ed
va
lve; 10-B
ilat
e
ral b
a
la
nce va
lve; 11,12,
13,1
4
-Pressur
e
sensor; 15,1
6
-
H
orizo
n
tal c
y
l
i
n
der; 17-D
i
sp
l
a
c
e
ment sens
or; 18-Erecti
on c
y
l
i
nd
er
Figure 2. Hyd
r
auli
c
pri
n
cipl
e of the ere
c
tion syste
m
3. Mathema
t
i
cal Model of
the Erec
tion
Mecha
n
ism
3.1. Kinematic Analy
s
is o
f
the Me
chan
ism
In orde
r to
acq
u
ire
kin
e
m
atic featu
r
e
of
the novel ere
c
tion
mech
ani
sm,
kinem
atic
analysi
s
i
s
a
c
compli
she
d
firstly. In e
r
e
c
tion
process the lo
ad
an
d e
r
e
c
tion
arm rotate
s aro
und
back hin
ged
bearin
g and
also move
s in horizonta
l
directio
n. Kinematic m
o
del is sho
w
n
in
Figure 3.
Figure 3. Kinematic mo
del
of the mecha
n
ism
In plane co
o
r
dinate
syste
m
oxy
,
P
1
is the origi
n
of coordinate
system. Suppo
se
s
c
o
or
d
i
na
te
o
f
P
3
is (
x
2
(
t
),
y
2
(
t
)) and
coo
r
dinate of
P
4
is (
x
1
(
t
),
y
1
(
t
)). The followin
g
equation
s
can
be acquired b
a
se
d on ge
o
m
etric
relatio
n
shi
p
.
22
22
2
0
()
()
(
)
t
ei
x
ty
t
l
v
d
(1)
22
12
1
2
[
(
)
(
)]
[
(
)
(
)]
em
x
tx
t
y
t
y
t
l
(2)
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Co-sim
ulation
and Expe
rim
ent Re
sea
r
ch
on a No
vel Erectio
n
Me
ch
anism
(Fen
g Jian
gtao)
147
21
12
()
(
)
(
)
arct
an
()
(
)
yt
y
t
t
x
tx
t
(3)
Whe
r
e
13
ei
P
Pl
is the initial len
g
th of ere
c
tion cylind
e
r
and
34
em
P
Pl
repre
s
ents th
e
distan
ce b
e
tween two ju
ncti
ons.
θ
(
t
) refers to the ere
c
ti
on angl
e.
v
1
(
t
) and
v
2
(
t
) a
r
e
the spee
ds o
f
two hydra
u
lic
cylinde
r pisto
n
rod
s
.
Kinematic b
a
l
ance equ
atio
ns of the l
oad
and ere
c
tion
arm are as fol
l
ows:
1
()
c
o
s
(
)
(
)
(
)
(
)
(
)
pf
t
w
F
tt
F
t
F
t
F
t
M
x
t
(4)
11
()
s
i
n
(
)
(
)
(
)
po
F
t
t
N
t
Mg
My
t
(5)
(
)
()
()
()
()
pp
G
w
P
F
tl
t
M
g
l
t
M
t
J
t
(6)
Whe
r
e
F
p
(
t
) i
s
thru
st force
of erectio
n
cylinder.
γ
(
t
) repre
s
e
n
ts the
angle bet
we
en thru
st
force
an
d po
sitive
x
axi
s
.
F
f
(
t
) i
s
fri
c
tion force
bet
wee
n
rail
an
d slid
er.
F
t
(
t
) is pull force of
hori
z
ontal cyl
i
nder.
F
w
(
t
) is calculated
wind load.
M
repre
s
e
n
ts m
a
ss of the lo
ad and
ere
c
ti
on
arm.
N
o
1
(
t
) is
s
u
pport force of rail to s
l
ider.
l
p
(
t
) is th
ru
st force a
r
m of
ere
c
tion
cylin
der to p
o
int
P
4
.
l
G
(
t
) is loa
d
gravity arm t
o
poi
nt
P
4
.
M
w
(
t
) is the
moment
of calcul
ated
win
d
loa
d
.
J
P
is the
moment of in
ertia of the load and e
r
e
c
tion arm to poi
nt
P
4
.
l
p
(
t
) and
l
G
(
t
) are d
e
fined a
s
:
22
11
1
22
1
22
()
()
()
()
s
i
n
(
)
a
r
c
t
a
n
()
()
()
em
p
lx
t
y
t
yt
lt
t
x
t
xt
y
t
(7)
40
()
c
o
s
[
(
)
]
GG
lt
P
P
t
(8)
Wind lo
ad is
given by:
()
wi
i
F
tq
S
(9)
Whe
r
e
S
i
is cal
c
ulate
d
wi
ndward are
a
and
q
i
rep
r
e
s
ents wind
p
r
essure
corre
s
po
ndin
g
cal
c
ulate
d
wi
ndward a
r
e
a
.
3.2. Mathem
atical Model
of the
H
y
draulic Sy
stem
Servo valve
and
hydra
u
li
c
cylinde
r a
r
e two
im
po
rt
ant pa
rts i
n
the ele
c
tro
-
h
y
drauli
c
actuato
r
syst
em
a
s
sh
own
in
Fi
gu
re 4. The cylin
d
e
r
ports a
r
e con
necte
d to a propo
rtion
a
l valve
,
and
pisto
n
m
o
tion i
s
obtai
ned
by mo
dul
ating the
oil
flow
i
n
to and
out
of
th
e cylinder cham
be
rs.
P
s
is the hy
drauli
c
sup
p
l
y
pressu
re
a
nd
P
0
is
the return press
u
re.
x
v
i
s
the spool val
v
e
displ
a
cement
.
q
1
and
q
2
are fluid flow from and to cy
linder.
P
1
an
d
P
2
are the fluid pre
s
sure
in
the forwa
r
d a
nd return cyli
nder chamb
e
r
s
of the
actu
ator, respe
c
tively.
A
1
is the
pisto
n
side
a
r
ea
and
A
2
is the rod sid
e
are
a
. When diffe
ren
c
e
s
betwe
en
P
1
and
P
2
exist, the hydrauli
c
cylin
d
e
r
extends o
r
co
mpre
sse
s
.
The flow th
ro
ugh a
re
stri
ction is
gene
rall
y turbulent a
n
d
pro
p
o
r
tiona
l to the sq
uared ro
ot
of the pre
s
sure d
r
op. Eq
uation
s
of flow thr
ough t
he valve orif
ice
s
co
me from orifice flow
equatio
ns ex
plaine
d in the
following e
q
u
a
tions.
11
2
()
dv
s
g
qC
w
x
p
p
(10
)
22
0
2
()
dv
g
qC
w
x
p
p
(11
)
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 14, No. 1, March 2
016 : 144 – 1
5
5
148
Figure 4. Sch
e
matic dia
g
ra
m of the hydraulic
system
Whe
r
e
C
d
i
s
d
i
scharge coef
ficient,
w
is th
e spo
o
l are
a
gradi
ent and
ρ
is the fluid
den
sity.
In ord
e
r to o
b
tain a
math
ematical
de
scriptio
n of i
n
stantan
eou
s
pre
s
sure in
si
der th
e
forwa
r
d
and
return
cylinde
r cham
be
rs, t
he fluid flo
w
balan
ce
equa
tions a
r
e
con
s
ide
r
ed fo
r t
w
o
control volum
e
s at ea
ch si
de of the cylinder.
The d
e
rivative of the lo
ad
pre
s
sure is gi
ven by the tot
a
l load
flow t
h
rou
gh th
e a
c
tuator. It
divided by th
e fluid
cap
a
ci
tance
and
al
so takin
g
le
akage a
nd
com
p
re
ssi
bility into co
nsi
d
e
r
ati
on.
We can obtai
n the flow co
ntinuity equat
ions:
11
1
11
1
2
()
ec
ic
e
dV
V
d
p
qC
p
C
p
p
dt
d
t
(12
)
01
1
1
11
1
2
()
ec
i
c
i
c
e
VA
y
dp
dy
qA
C
C
p
C
p
dt
d
t
()
(13
)
22
2
22
1
2
()
ec
i
c
e
dV
V
d
p
qC
p
C
p
p
dt
dt
(14)
02
2
2
22
2
1
()
e
c
ic
ic
e
VA
y
dp
dy
qA
C
C
p
C
p
dt
d
t
()
(15)
Whe
r
e
V
1
an
d
V
2
are the total fluid volumes in the two cylinde
r ch
ambe
rs,
V
01
and
V
02
are
the
origi
nal total flui
d volume
s o
f
the tw
o
cylinder
cham
b
e
rs (i
ncl
udin
g
the vol
u
m
e
of
pipelines
and initial cylinder
chambers),
A
1
y
an
d
A
2
y
rep
r
e
s
ent t
he flo
w
rates as a fu
nctio
n
of
volume
cha
n
ge du
e to
pist
on motio
n
,
β
e
is the
bul
k m
odulu
s
of
hyd
r
auli
c
fluid.
C
ic
and
C
ec
a
r
e t
he
internal a
nd e
x
ternal lea
k
a
ge co
efficient
.
The dyn
a
mic equatio
n of
the hy
draulic system
with
mass
m
is
d
e
s
c
r
ib
e
d
b
y
Se
c
ond
Ne
wton La
w:
2
11
2
2
2
dy
mA
p
A
p
F
dt
(16)
Whe
r
e
m
is the equival
ent
mass,
y
is t
he pisto
n
di
splacement, a
n
d
F
is the friction force a
nd
external force
on the cylind
e
r.
3.3. Contr
o
l Sy
stem Desi
gn
PID co
ntrol h
a
s b
een
wid
e
ly use
d
in i
ndu
stry
be
ca
use
of its
si
mplicity. Ho
wever, PID
control i
s
no
t suitable
fo
r sy
stem
with a la
rge
a
m
ount of la
g, paramete
r
variation
s
a
n
d
uncertainty in
the model. F
u
zzy co
ntrol
has fo
und m
a
ny appli
c
ation
s
in a va
riety of fields,. It has
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Co-sim
ulation
and Expe
rim
ent Re
sea
r
ch
on a No
vel Erectio
n
Me
ch
anism
(Fen
g Jian
gtao)
149
the advanta
g
e
that it doe
s not re
quire a
n
accu
rate
m
a
thematical model of
the
system [1
4, 1
5
].
Hen
c
e, fuzzy
adaptive PID control wa
s
develop
ed
to
utilize the a
d
v
antage
s of b
o
th PID co
ntro
l
and fuzzy lo
gic
control. Figure
5 re
pre
s
ent
s the
st
ructure of fuzzy adaptiv
e
PID cont
rolle
r. By
usin
g the referen
c
e
sign
al
and the feedback sig
nal
that come
s from the sen
s
ors, the controller
transmits the
control si
gn
als that a
r
e
cal
c
ul
ate
d
by the co
ntrol algo
rithm
to operate
the
prop
ortio
nal valve. The aim
of control a
c
t
i
on is to mini
mize the tra
c
king e
r
ror.
Figure 5. The
stru
cture of
fuzzy adaptive
PID controll
e
r
The
purpo
se
of fu
zzy l
ogi
c
cont
rolle
r i
s
to
adju
s
t t
h
ree
pa
ram
e
ters of PI
D
controlle
r.
Fuzzy logi
c
controlle
r
con
s
ists of t
w
o in
put va
ria
b
le
s
and th
ree
out
put varia
b
le
s. The in
puts a
r
e
the error
e
a
nd the
cha
n
ge in e
r
ror
ec
, whi
c
h
are
blurred
and
exported to
fuzzy infe
re
nce
module
with
fuzzy rule
s. The outp
u
ts
are
ᇞ
K
P
,
ᇞ
K
I
and
ᇞ
K
D
, wh
ich a
r
e
re
spe
c
tively added
to
initial variable
K
P
0
,
K
I
0
and
K
D
0
[16, 17].
Fuzzy control
l
er
con
s
ist
s
o
f
input fuzzifi
c
at
ion, fu
zzy
control rule
s, fuzzy i
n
fere
nce
an
d
output defu
z
zificatio
n
. It
works a
c
cording to a
se
t of linguistic rule
s and
calcul
ates o
u
tput
variable
s
. In
p
u
t value
s
go
throug
h fu
zzi
f
ication i
n
terf
ace
an
d a
r
e
conve
r
ted
to
fuzzy li
ngui
sti
c
values. T
hen,
fuzzy control
rule
s
are
u
s
ed to i
n
fer th
e outp
u
t vari
able
s
. Finally
, defuzzification
method
conv
erts the fuzzy output values into
si
gnals to be
sent out. The inf
e
rence
process
con
s
i
s
ts of
a
set of rule
s d
r
iven by the li
ngui
stic
valu
e
s
of the
error
and
the erro
r in
ch
ang
e
[18].
Table
1
sho
w
s the fu
zzy control
rule
s t
o
dete
r
mine
the p
r
op
er
co
ntrol a
c
tion
s t
hat are fea
s
i
b
le
for the cu
rren
t condition.
Table 1. Fu
zzy control rule
s
NB
NM
NS
ZO
PS PM PB
NB
PB/NB/NS
PB/NB/PS
PM
/NM
/
PB
PM
/NM
/
PB
PS/NS/PB
ZO
/
Z
O
/
P
M
ZO
/
Z
O
/
NS
NM PB/NB/NS
PB/NB/PS
PM
/NM
/
PB
PS/NS/PM
PS/NS/PM
ZO
/
Z
O
/
PS
NS/ZO
/
ZO
NS PM
/NB/
ZO
PM
/NM
/
PS
PM
/NS/P
M
PS/NS/PM
ZO
/
Z
O
/
PS
NS/PS/PS
NS/PS/ZO
ZO
PM
/NM
/
ZO PM
/NM
/
PS
PS/NS/PS
ZO
/
Z
O
/
PS
NS/PS/PS
NM
/P
M
/
PS
NM
/P
M
/
ZO
PS PS/NM
/
Z
O
PS/NS/ZO
ZO
/
Z
O
/
ZO
NS/PS/ZO
NS/PS/ZO
NM
/P
M
/
ZO
NM
/PB/
ZO
PM
PS/ZO/NB
ZO/
Z
O/PS
NS/PS/NS
NM
/PS/
NS
NM
/P
M
/
NS
NM
/PB/
NS
NB/PB/NB
PB
ZO/
Z
O/
NB
ZO/
Z
O/
N
M
NM
/PS/
N
M
NM
/P
M
/
N
M
NM
/P
M
/
NS
NB/PB/NS
NB/PB/NB
The in
puts an
d output
s a
r
e
defined
by u
s
ing
seve
n ve
rbal val
u
e
s
such
a
s
:
NB
-n
egative big,
NM
-
negative me
dium,
NS
-ne
gative sm
all,
ZO
-z
er
o,
PS
-positive small,
PM
-po
s
itive medium,
PB
-
P
K
I
K
D
K
de
dt
0
0
0
PP
P
II
I
DD
D
K
KK
K
KK
K
KK
ec
K
e
K
P
G
I
G
D
G
ec
e
ᇞ
K
P
ᇞ
K
I
ᇞ
K
D
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 14, No. 1, March 2
016 : 144 – 1
5
5
150
positive
bi
g. The cont
rol
l
a
ws con
s
ist of
49 cont
rol
rule
s define
d
a
s
’if
e
is
NB
and
ec
is
NB
,
then
ᇞ
K
P
,
ᇞ
K
I
and
ᇞ
K
D
is
PB
,
NB
a
nd
NS
’. The fu
zzy rul
e
s have
b
een
de
sign
ed
and
adj
uste
d
based on th
e
simulatio
n
a
nd expe
rt kn
owle
dge of
h
y
drauli
c
cylin
der featu
r
e a
nd PID control
algorith
m
. It is rea
s
on
able
to present t
hese lin
g
u
isti
c te
rms by triangul
ar-sh
a
p
e
mem
bershi
p
function
s
whi
c
h
are
u
s
ed
to determine
the d
egr
ee
of input
s in
the fu
zzifi
cati
on p
r
o
c
e
ss.
The
deci
s
io
n-m
a
ki
ng outp
u
t ca
n be
obtain
e
d
usi
ng
a ma
x-min fuzzy i
n
feren
c
e
whe
r
e the
output
is
cal
c
ulate
d
by the Cente
r
of Area metho
d
.
The
stru
ctu
r
e
of
control
system i
s
sho
w
n i
n
Fi
gure
6. Th
e
sen
s
ors a
c
q
u
ire
ere
c
tion
angle
and
di
spla
cem
ent
of hori
z
o
n
tal
cylinde
r
pi
ston ro
d. The
error si
gnal
is o
b
taine
d
by
comp
ari
ng acquire
d sign
al
with de
sire
d
sig
nal.
Th
e control sign
al
is obtain
ed by
cal
c
ul
atio
n
of
fuzzy ad
aptive PID cont
ro
ller. The
con
t
rol ele
c
tr
ical
current i
s
o
b
tained th
rou
gh propo
rtion
a
l
magnifie
r
. Th
e ele
c
tri
c
al
current in
put
electro-
hydra
u
lic
propo
rtio
nal valve
to
adju
s
t op
eni
ng
whi
c
h adju
s
ts the flow and
hydrauli
c
cylinder pi
ston rod spe
ed, therefo
r
e me
ch
anism
can m
o
ve
based on exp
e
ctation.
Figure 6. Structure
of control
system
Erectio
n
process ge
nerally use
s
the
u
n
iform
a
c
cel
e
ration and deceleration planni
ng
method. Du
e
to the accel
e
ration
cu
rve
is not c
onti
nuou
s, there
is flexible impact in e
r
e
c
ting
pro
c
e
ss. Acceleratio
n
sel
e
ction is q
u
ite con
s
e
r
vative and ere
c
tion time is long. We ad
opt
comp
osite
si
n
e
functio
n
to
plan e
r
e
c
tion
angle to
solve the p
r
obl
e
m
.
θ
0
is i
n
itial
value of erecting
angle
an
d
θ
1
is final
val
ue. Ere
c
ting
time i
s
T
,
τ
=
t
/
T
.
θ
(
t
) is dete
r
mine
d
by the foll
o
w
ing
expre
ssi
on
s:
10
()
(
)
(
)
ts
(17
)
si
n(
4
)
1
0
44
8
24
1
7
(
)
9
c
o
s
(
)
/
4
43
6
8
8
4s
i
n
(
4
2
)
7
1
44
8
k
k
s
k
(18
)
Con
s
tant
2
44
k
.
Hori
zo
ntal cylinders pull
back hi
nge
d bea
ri
ng
moving alo
n
g
hori
z
o
n
tal
dire
ction,
compl
e
ting
erection
process tog
e
t
her
wit
h
erectio
n
cylinder. Re
qui
rement of
ho
ri
zontal
cylin
de
r
is moving
sm
oothly. Displa
ceme
nt, velocity, accelera
tion and im
p
a
ct curve
s
chang
e sm
oot
hly
and
avoid
m
u
tation. We
adopt
polyno
m
ial inte
rpol
a
t
ion meth
od
to plan
e
r
e
c
tion a
ngle.
S
0
is
initial value
of
ho
rizontal
cy
linder
di
spla
cement
and
S
1
is final value. Movement time is
T
.
S
(
t
) is
given by:
10
()
(
)
(
)
St
S
S
s
(19
)
76
5
4
(
)
20
70
84
35
s
(20
)
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Co-sim
ulation
and Expe
rim
ent Re
sea
r
ch
on a No
vel Erectio
n
Me
ch
anism
(Fen
g Jian
gtao)
151
4. Co-simula
t
ion of th
e Erection M
e
c
h
anism
Mech
ani
cal
model
is e
s
ta
blish
ed i
n
Pro/E and
ADA
M
S. Hydrauli
c
m
odel
is e
s
tablish
ed
in AMESim.
Control model is es
ta
blish
ed in Simulin
k. The mo
del
s esta
blished
in AMESim and
ADAMS are
transmitted t
o
Simulin
k
which
is th
e m
a
in
simulatio
n
environme
n
t. AMESim and
ADAMS are
assi
stant si
mulation env
ironm
ent. Informatio
n is
excha
nge
d throu
gh software
interface.
4.1. Mechani
cal model
ADAMS is
used to a
nalyze
virtual me
ch
anical sy
stem
and it a
dopt
s Lag
rang
e e
q
uation
s
to establi
s
h
kinematic m
o
d
e
l. ADAMS/View is
a
protot
yping modul
e
that allows u
s
ers to buil
d
a
compl
e
x mechani
cal sy
ste
m
. Use
r
s
ca
n define vari
ous p
r
op
ertie
s
su
ch a
s
: material, den
sit
y
,
stiffness an
d
so o
n
. Th
ree
-
dime
nsi
onal
model
ca
p
abi
lity of ADAMS is limited,
however, it can
excha
nge file
s with othe
r advan
ced CA
D softwar
e to
ensu
r
e data
con
s
i
s
ten
c
y. Mech
ani
sm/Pro
adopt
s
seam
less
conn
ecti
on
with Pro/
E and t
r
an
smits mo
del t
o
ADAMS/View to
co
ndu
ct
comp
re
hen
si
ve kinemati
c
analysi
s
. Mechani
cal mod
e
l
of the erecti
on me
chani
sm establi
s
he
d in
ADAMS is sh
own in Figu
re 7. Three
-
di
mensi
onal m
odel wa
s e
s
tablished in Pro/E, and it
wa
s
transmitted
t
o
ADAMS/View usi
ng M
e
ch
ani
sm/Pro
. The model
is compo
s
e
d
of seven rigid
bodie
s
which
are lin
ked
with others by re
volu
te pair, prismatic p
a
ir a
nd fixed pair.
Figure 7. Mechani
cal mod
e
l in ADAMS
As the m
ode
l in ADAMS
must inte
ra
ct with the
co
ntrol mo
del i
n
Simulin
k, it sho
u
l
d
define
suitab
le “state va
ri
able
s
”
whi
c
h
are u
s
e
d
to exch
ang
e
informatio
n. The inp
u
t st
ate
variable
s
are
pisto
n
rod
speed
si
gnal
s of hyd
r
aul
i
c
cylind
e
rs. T
he o
u
tput
st
ate vari
able
s
are
ere
c
tion an
gl
e, piston ro
d displ
a
cement
of horiz
o
n
tal cylinde
r, and
load
s
of hydraulic
cylinde
r.
4.2. H
y
draulic Model
AMESim use
s
gra
phi
cal
model ap
pro
a
ch that ea
ch comp
one
nt contain
s
a set of
equatio
ns a
n
d
is lin
ked
with others through a
syst
e
m
of port
s
. Firstly sel
e
ct h
y
drauli
c
pum
p,
relief valve, bilateral b
a
la
nce valve, h
y
drauli
c
lock,
hydrauli
c
cy
linder, an
d e
l
ectro
-
hyd
r
aul
ic
prop
ortio
nal v
a
lve from
hyd
r
auli
c
lib
ra
ry,
then e
s
tabli
s
h mo
del i
n
a
c
cordan
ce
wit
h
the
se
que
n
c
e
of sket
c
h mo
de, sub
-
mo
de
l mode, para
m
eter mo
d
e
, and ru
n mod
e
. Hydrauli
c
model e
s
tabli
s
he
d
in AMESim is sho
w
n in
Figure 8. Th
e co
-si
m
ulati
on interfa
c
e
of AMESim and Simulin
k is
achi
eved by
con
n
e
c
ting th
e men
u
i
c
on
in AMESim to the S-fu
ncti
on in Sim
u
lin
k. The
mod
e
l
in
AMESim is compiled in S-f
unctio
n
whi
c
h
can be u
s
e
d
in Simulink.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 14, No. 1, March 2
016 : 144 – 1
5
5
152
Figure 8. Hyd
r
auli
c
mod
e
l in AMESim
4.3. Co-simu
l
ation Re
sults and An
aly
s
is
Simulink blo
c
k of the ere
c
tion mechani
sm is sh
o
w
n in
Figure 9. It is mainly comp
ose
d
of
four pa
rts: pl
aned
curve, f
u
zzy adaptiv
e PID c
ontroll
er, hydra
u
lic
model, and m
e
ch
ani
cal mo
del.
Planned
cu
rves are de
sire
d ere
c
tion an
gle and di
sp
l
a
cem
ent of horizontal
cyli
nder pi
ston
rod.
There are two fu
zzy a
d
aptive co
ntro
llers
whi
c
h
separately co
ntrol e
r
e
c
tion
and h
o
ri
zon
t
al
cylinde
rs. Hy
drauli
c
mo
del
is establi
s
h
e
d
in AMESim and co
mpile
d in S-functi
on that can
be
use
d
in Sim
u
link. Me
ch
ani
cal m
odel
is
establi
s
h
ed i
n
ADAMS
an
d tran
smitted
to Simulin
k
by
softwa
r
e i
n
terface. T
he in
p
u
t variabl
es
of fuzzy ad
a
p
tive co
ntroll
ers a
r
e th
e e
rro
rs of pl
an
ned
ere
c
tion a
ngl
e and
displa
cement of h
o
ri
zontal
cy
lind
e
r
pist
on rod
with actual
ere
c
tion a
ngle
a
n
d
displ
a
cement
in sim
u
lation
pro
c
e
s
s. The
output va
ri
ab
les a
r
e
co
ntro
l sign
als
of el
ectro
-
hyd
r
auli
c
prop
ortio
nal
valves. The
input vari
abl
es of
hy
dra
u
lic mo
del a
r
e co
ntrol
sig
nals
of ele
c
t
r
o-
hydrauli
c
pro
portion
al valv
e an
d hyd
r
a
u
lic
cyli
nde
r l
o
ads.
The
out
put vari
able
s
are pi
ston
rod
spe
ed
sig
nal
s. Th
e in
put
variabl
es of
me
chani
cal
model are
pi
ston rod
spe
ed sign
als.
T
he
output varia
b
l
e
s a
r
e e
r
e
c
tio
n
angl
e, displacem
ent
of h
o
rizontal
cylin
der
pi
ston
ro
d and
hydra
u
l
i
c
cylinde
r load
s.
Figure 9. Simulink bl
ock of the ere
c
tion system
Simulation re
sults of erecti
on angle a
r
e
sho
w
n in Fig
u
re 10. Fig
u
re 10(a
)
is de
sire
d an
d
actual
ere
c
ti
on an
gle in
simulatio
n
p
r
oce
s
s. Fi
gu
re 10(b) i
s
th
e error of de
sire
d an
d a
c
tual
ere
c
tion an
gl
e. Displ
a
cem
ent simulatio
n
re
sults
of h
o
rizontal cyli
nder
piston
rod are sh
own in
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Co-sim
ulation
and Expe
rim
ent Re
sea
r
ch
on a No
vel Erectio
n
Me
ch
anism
(Fen
g Jian
gtao)
153
Figure 11. Fi
gure 1
1
(a) is
desi
r
ed a
nd a
c
tual di
spla
ce
ment in simul
a
tion pro
c
e
ss. Figure 11
(b
) is
the
erro
r
of desi
r
ed and actual displa
ceme
nt.
By
the
sim
u
lation
re
sults we can obtain
t
hat
ere
c
tion
angl
e an
d di
spl
a
cement
of ho
ri
zontal
cy
lin
de
r pi
ston
ro
d
chang
e
smoot
hly. The
error of
ere
c
tion
angl
e is
co
ntrolle
d in 0.0
6
°.Th
e displa
ceme
nt error
of ho
rizo
ntal
cylin
der
pisto
n
ro
d is
controlled
in
0.014 m. T
h
e
co
ntrol
pre
c
i
s
ion
is
gr
eat
and
satisfie
s
the control
ta
rget of
ere
c
ti
on
pro
c
e
ss. Co
mpared ere
c
t
i
on
a
ngle wit
h
di
spla
cem
e
nt of ho
rizont
al cylin
der pi
ston
rod
we
can
acq
u
ire th
at ere
c
tion a
ngl
e doe
sn’t
ch
ange
gre
a
tly
at the sta
r
t a
nd en
d of ho
rizo
ntal cylin
der
movement. It
indicates that
the impact of
two hydrauli
c
cylinders is li
ttle.
Figure 10. Simulation e
r
e
c
tion angle
Figure 11. Simulation di
spl
a
cem
ent
of horizontal cylin
der pi
ston rod
5. Experimental Verifica
tion
We
de
sign
ed
an
expe
rime
ntal ap
paratu
s
t
hat
can
re
alize
de
sired
moveme
nt t
o
p
r
ove
the effectiven
ess of the no
vel erectio
n
mech
ani
sm. Mech
ani
cal constitution i
s
sho
w
n in Fig
u
re
12. Mea
s
ure
m
ent and
co
ntrol sy
st
em
of experime
n
t
al appa
ratus are e
s
tabli
s
hed u
s
ing virtual
instru
ment te
chn
o
logy. We
choo
se the
hard
w
a
r
e of
PXI-1044
ca
se and PXI-62
59 multifuncti
on
data acqui
sition ca
rd an
d software of La
bVIEW to pro
g
ram.
Figure 12. Mech
ani
cal con
s
titution of the platform
0
20
40
60
80
100
0
20
40
60
80
10
0
期望
角度
角度
实际
0
20
40
60
80
10
0
-0
.
1
-0
.
0
5
0
0.05
0.1
0
20
40
60
80
100
0
0.2
0.4
0.6
0.8
期
望位移
位移
实际
0
20
40
60
80
100
-0
.
0
3
-0
.
0
2
-0
.
0
1
0
0.
01
0.
02
0.
03
Evaluation Warning : The document was created with Spire.PDF for Python.