TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol.12, No.4, April 201
4, pp. 2753 ~ 2
7
6
1
DOI: http://dx.doi.org/10.11591/telkomni
ka.v12i4.4277
2753
Re
cei
v
ed Au
gust 12, 20
13
; Revi
sed O
c
t
ober 2
8
, 201
3; Acce
pted
No
vem
ber 1
6
,
2013
Design of Electric Load Measuring and Controling
System for Steering Sevro
Pan Wen
t
ao,
Peng Zhaoq
in*, Caochun
Li Bo
Schoo
l of Auto
mation Sci
enc
e and El
ectrica
l
Engi
neer
in
g, Beih
ang U
i
nv
e
r
sit
y
NO.37 Xue
y
u
a
n
Roa
d
, Hai
d
ia
n District, Chin
a, 010-8
2
3
381
92
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: pengz
ha
oqi
n
@
bu
aa.ed
u.cn
A
b
st
r
a
ct
An autom
a
tic
loading si
mulation and
testi
ng
system of steering servo wa
s designed in
this
paper
.
Design of hardware and s
o
ftware of
the testing system
were
illustrated, and
the
Load testing principle
and
test strategy w
e
re a
n
a
l
y
z
e
d
.
T
o
rque s
e
rvo
motor
w
a
s use
d
to si
mulate
the l
o
a
d
in
g a
d
d
ed to th
e ste
e
r
i
n
g
servo. T
he t
o
rq
ue
ma
gn
itud
e
w
a
s control
l
ed
prop
ortion
al
to
the a
ngu
lar
dis
p
lac
e
ment
of the ste
e
rin
g
ser
v
o
to simulate
ela
s
tic load. F
u
zzy PID control
alg
o
ri
th
m w
a
s used to
mak
e
torque s
e
rvo
motor fo
llow
the
change of
angular
di
s
p
lac
e
ment quickly
and accur
a
tely. The effectiveness
of
the autom
a
t
i
c testing syst
e
m
w
a
s proved by
the exper
i
m
ent
al resu
lts.
Ke
y
w
ords
: aut
omatic test, loading sim
u
lation, fuz
z
y PID
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
In the field of
flight control
system, loa
d
si
mulato
r i
s
o
ne of the mai
n
devices
whi
c
h
can
simulate
the
loadin
g
ad
de
d to the
actu
ator. Th
e lo
ad si
mulato
r
can i
m
itate t
he ae
ro
dyna
mic
load
s of the
actuato
r
suffe
red by the
aircraft d
u
ri
n
g
the flight by a
dding l
oadin
g
to the actu
ators
s
y
s
t
em of the mis
s
iles
or aircrafts
.
Re
cent years, many institut
es do researche
s
on the l
oadin
g
simul
a
tor [1-3]. Th
e main
loadin
g
mo
de
of the
loa
d
simulators
are
ele
c
tro
-
hyd
r
aulic loa
d
ing
and
ele
c
tro
l
oadin
g
. Elect
r
o
-
hydrauli
c
lo
a
d
ing m
e
thod
has th
e a
d
va
ntage of
high
pre
c
isi
on,
wid
e
freq
uen
cy b
and, big
torq
ue,
etc. However, there is
still some di
sadvantages, such
as o
il leakage, pollution, and inconveni
ent
maintena
nce. Electro l
oadi
ng metho
d
h
a
s the
advant
age of n
o
he
avy oil sou
r
ce, light polluti
on,
easy to maintain, high re
spon
se sp
eed,
simple st
ruct
ure, high reli
ability. With the develop
m
ent
of the the
o
ry
of d
r
iver an
d control, th
e ele
c
tro lo
a
d
si
mulato
r n
o
w
ca
n
reali
z
e
bigg
er torque
loadin
g
, high
preci
s
io
n, and wide freq
uen
cy band.
Electro loa
d
i
ng simul
a
tor
is the best choice
for the lowe
r l
oadin
g
motor [4, 5]. Torque motor is
sui
t
able for torq
ue loadi
ng m
easurin
g syst
em
with low
sp
ee
d, high torqu
e
and a
c
cu
rat
e
torqu
e
outp
u
t. Servo box coul
d be red
u
ce
d wh
en te
st
system u
s
ed
torque moto
r as the load, so the
erro
r that was cau
s
ed by mech
anical backla
s
h
coul
d be avoi
ded.
Torq
ue
outpu
t wa
s lin
ear
with the
an
gu
lar di
spl
a
cem
ent of the
ste
e
ring
servo
o
u
tput for
simulatin
g
th
e real
state.
So the torqu
e
mo
tor m
u
st be able to
follow the
si
gnal of an
gu
lar
displ
a
cement
quickly. In orde
r to imp
r
ove
the co
ntrolling perf
o
rma
n
ce,
torque signal wa
s
controlled
with fuzzy PID
controlle
r. Fu
zzy
PID
cont
rolle
r
con
s
i
s
ts of
fuzzy
co
ntrolle
r a
nd
PID
controlle
r, which
contai
ns both
the accuracy of PIDcontroll
er
a
nd re
al-time
feature of fu
zzy
controlle
r. Its good robu
stn
e
ss ca
n cate
r to need of no
nlinea
r syste
m
[6, 7].
2. Structu
r
e
of the Sy
stem
The a
u
tomati
c ele
c
tro loa
d
i
ng si
mulatio
n
and
mea
s
u
r
ing
system
d
e
sig
ned i
n
thi
s
p
ape
r
is to reali
z
e t
o
rqu
e
loa
d
in
g for
some
ki
nds
of stee
ri
ng sen
s
ors,
and
wa
s u
s
e
d
to finish
so
me
static an
d dynamic pe
rformance autom
at
ic testing fo
r the stee
ring
servo
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 4, April 2014: 2753 – 2
761
2754
2.1. Mechani
cal Struc
t
ure
The me
cha
n
i
c
al platfo
rm
wa
s co
nsi
s
te
d of
base, tested
stee
rin
g
motor, co
upling,
encode
r, be
a
r
ing
seat, torque
se
nsor, l
oadin
g
mo
to
r and
all
kin
d
s
of
fixtures.
The m
e
chani
cal
stru
cture of
the exp
e
rim
e
nt pla
tform
was sh
own as
Figu
re
1.
T
h
e teste
d
stee
ring
servo
is
the
work obj
ect
of the testing platform, a
nd the
testin
g system te
sted the stat
ic and dyn
a
m
ic
perfo
rman
ce
of the steeri
n
g motor auto
m
atically
. Th
e requi
re
d torque was a
d
d
ed by the loa
d
ing
motor, the
va
lue of l
oadi
n
g
torque
vari
ed
with
the
chang
e of th
e
input volta
g
e
of the
loa
d
i
ng
motor which resultin
g in co
ntinuou
s an
d smooth lo
adi
ng.
2.2. Structu
r
e of Mea
s
ure
m
ent and
Co
ntrol Sy
stem
The p
e
rfo
r
m
ance of th
e
measurement
and
control
syste
m
d
e
ci
ded th
e a
ccurate
of
steeri
ng serv
o measuri
n
g
system. Th
e efficien
cy, quality, and co
st of hard
w
are sh
ould
be
c
o
ns
ide
r
ed
.
Figure 1. The
Mecha
n
ical
Structu
r
e of the System
Becau
s
e the
system n
eed
control the lo
ading
moto
r
and the teste
d
actuato
r
be
side
s the
data analy
z
in
g and p
r
o
c
e
s
sing. Th
e structure of
up
and lo
we
r co
mputers was adopted i
n
this
system. Th
e
up co
mput
er which ru
n
in wind
ow
s operating
system mainl
y
to finish d
a
ta
acq
u
isitio
n, data saving,
data analy
z
in
g, data di
spl
a
ying, etc, a
nd the lower comp
uter was
mainly to take as loading
controller. According to
th
e req
u
irem
en
ts of the load
ing in the sta
t
ic
and
dynami
c
perfo
rma
n
ce
testing,
DSP
tech
nique
was
ado
pted i
n
the lo
we
r
compute
r
, whi
c
h
wa
s co
nsi
s
te
d of DSP minimum sy
ste
m
, A/D,
D/A, RS485 com
m
unication
module, sen
s
or
sign
al tra
n
sm
itter, power m
odule, a
u
xilia
ry circ
uit
s
. Th
e co
mpo
s
itio
n diag
ram
of
the sy
stem
was
s
h
ow
n
in
F
i
gu
r
e
2
.
Figure 2. Overall Structu
r
e
Di
ag
ram of M
easurin
g System
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
De
sign of Ele
c
tri
c
Load M
e
asu
r
ing a
nd
Contro
ling S
y
stem
for Steering Se
vro
(Pan We
ntao)
2755
Torq
ue motor and steeri
n
g
servo we
re b
o
th fixed on the mechani
cal platform, and we
re
con
n
e
c
ted int
o
a whol
e through the
cou
p
ling
s
, torque
sen
s
o
r
and
angul
ar di
spl
a
cem
ent sen
s
or.
Industri
a
l co
mputer a
nd
DSP boa
rd o
u
tput excita
tion sig
nal of steering
se
rvo
and torq
ue
motor,
and
acqui
sition b
oards co
llected
outp
u
t sig
nal
of
se
nso
r
s an
d th
e feed
ba
ck signal
of ste
e
ring
serv
o.
2.3. Structu
r
e of the So
ftw
a
r
e
Sy
stem
The fun
c
tion
of the soft
ware was to
reali
z
e o
n
li
ne integ
r
ated
manag
emen
t for the
s
y
s
t
em,
perf
o
rmance tes
t, s
e
c
u
rity, protec
tion
an
d
monito
ring
of stee
ring
servo m
e
a
s
uri
ng
system. Mod
u
lar
stru
cture
was
adopte
d
in the softwa
r
e sy
stem. Each m
odula
r
wa
s inde
pen
dent
of ea
ch oth
e
r
. The
software
of mea
s
uring
sy
stem
mainly in
clu
ded t
w
o p
a
rt
s: ho
st
comp
uter
prog
ram a
nd
lowe
r co
mput
er program.
1) Ho
st com
p
uter pro
g
ram:
The
fun
c
tion
s
of
this
pa
rt
i
n
clu
de pa
ram
e
ter setting
s, status
displ
a
y, data displ
a
y and p
r
ocessin
g
, an
d se
rial
com
m
unication
s. The ho
st co
mputer
pro
g
ram
wa
s prog
ram
m
ed in L
abv
iew e
n
viron
m
ent wi
th hi
gher
re
adabil
i
ty and com
patibility. The
function of th
e prog
ram
wa
s sh
own in Figure 3 .
p
k
i
k
d
k
Figure 3. Software Fun
c
tio
n
s of System
Figure 4. The
Control Blo
c
k of System
2) Lo
we
r co
mputer p
r
o
g
ram: The pro
g
ram m
a
inly
controlled t
he torq
ue m
o
tor an
d
steeri
ng
serv
o by DSP b
oard. T
he p
r
ogra
m
wa
s
prog
ram
m
ed
with C p
r
og
ram lan
gua
g
e
in
CCStudi
ov3.3. The co
nt
rol
block diag
ra
m was
sho
w
n
in Figure 4.
DSP boa
rd receive
d
orde
r from the
ho
st co
m
pute
r
throu
gh
seri
al
comm
uni
cation, and
output the
control si
gn
al to steeri
ng se
rvo.
Accordi
ng to
the signal
steeri
ng a
ngula
r
displ
a
cement
signal
and t
o
rqu
e
sen
s
or, the DSP bo
ard o
u
tput the cont
rol si
g
nal to the torque
motor. Thu
s
the mea
s
u
r
ing
system coul
d
load ela
s
tic l
oad to stee
rin
g
servo.
3. Contr
o
l Algorithm
In the mea
s
uring
sy
stem
, the co
ntrol
sign
al of torque m
o
tor m
u
st follo
w th
e output
sign
al
of
th
e angul
ar displ
a
cem
ent of
the stee
ri
ng
servo. So th
e
system
mu
st
respon
se
qui
ckly
to the chan
g
e
of inp
u
t si
g
nal. Tradition
al PID
contro
l method
was widely
used
in the in
du
stri
al
control proce
ss
cu
rre
ntly. The pa
ram
e
ter of PI
D co
n
t
rol wa
s defin
ed by engin
e
e
ring
experi
e
nce,
Ho
wever, It wa
s difficult
and time-co
n
s
umin
g to ge
t perfect pa
ra
meter. The
speed of resp
onse
coul
d not me
et the requi
re
ments if
the measuri
ng sy
stem was
con
t
rolled ju
st by PID control.
Fuzzy contro
ller me
asure
d
co
ntroll
ed
obje
c
t, and
converted
the
mea
s
uri
ng
signal to
fuzzy
amou
nt that de
scrib
ed by
huma
n
lang
uage
through
the fu
zzy interfa
c
e
conve
r
ter. T
h
en
the fuzzy
con
t
rol output fu
zzy am
ount b
y
the ru
les b
a
se
d on hu
m
an lang
uag
e. Accu
rate o
u
tput
coul
d be
got
by defuzzifica
tion. Fuzzy control
h
a
s bet
ter ro
bu
stne
ss an
d ad
apta
b
ility and was fi
t
to the mea
s
u
r
ing
system [
8
].The pa
ram
e
ters
p
k
,
i
k
and
d
k
we
re the
ke
rnel
of PID co
ntro
l. Self-
tuning fuzz
y
PID
contr
o
ller w
a
s
c
o
mposed by PI
D
c
o
ntr
o
l and fuzzy c
ont
r
o
l. The par
a
meters of
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 4, April 2014: 2753 – 2
761
2756
PID
control
were co
rre
ct
ed
o
n
line
by fuzzy co
nt
rol.
Fu
zzy PI
D
control
not
onl
y maintaine
d
the
feature
s
of PID co
ntrol, but
also ha
d fl
exibility, better control a
c
cura
cy and qui
ckl
y
r
e
spon
se.
As sho
w
n i
n
Figu
re 4, th
e erro
r
e
and t
he e
r
ror
rate
ec
were the input of the fuzz
y
control,
p
K
,
i
K
and
d
K
we
re the
out
put of fuzzy control. Th
e fu
zzy a
m
ou
nt
E
and
EC
of
e
and
ec
could b
e
obtaine
d from
Equation (1
)
.
2
2
L
H
L
H
e
e
e
e
e
e
m
K
(1)
Whe
r
e
H
e
,
L
e
are
the maximum and mini
mum of
e
,
m
is the maximum
of fuzzy
amount.
e
K
is the fuzzy amount
of
e
. The a
ccu
rate
output
p
k
,
i
k
and
d
k
could b
e
obtaine
d from
Equation (2
).
2
2
L
H
L
H
u
u
U
l
u
u
u
(2)
Whe
r
e
u
is the accurate o
u
t
put of fuzzy control,
H
u
an
d
L
u
are the m
a
ximum and
minimum of o
u
tput,
l
is the
maximum of the fuzzy amo
unt, and
U
is fuzzy output o
f
the fuzzy
control.
The me
asuri
ng sy
stem a
s
sume
d the fu
zzy
sub
s
et
s
of
E
,
EC
and
U
were
both {
NB
,
NM
,
NS
,
ZO
,
PS
,
PM
,
PB
}
,
w
h
er
e
NB
is n
egative
b
i
g,
NM
is neg
ative
middl
e,
NS
is
negative sma
ll,
ZO
i
s
ze
ro, a
nd
PS
is positive small,
PM
is po
sitive middle,
PB
is positive
big.
The fuzzy co
ntrol choo
se
s Gau
ssia
n
fu
nction
2
2
i
i
i
b
a
x
A
e
x
u
as di
st
ribution fu
ncti
on
.
Whe
r
e
i
a
is th
e ce
ntral of f
unctio
n
an
d
i
b
is the
width o
f
function. Th
e ce
ntral of
{
NB
,
NM
,
NS
,
ZO
,
PS
,
PM
,
PB
} is
{-3
,
-2
,
-1,0,1,2,3}
.
The
sha
r
p
of the di
strib
u
tion fun
c
tio
n
wa
s
s
h
ow
n
in
F
i
gu
r
e
5
.
Figure 5. The
Membershi
p
Distri
bution F
unctio
n
The error
e
a
nd e
rro
r
rate
ec
were th
e in
pu
t sign
al
of fuzzy an
d PID
control. T
h
e
fu
zzy
amount
E
and
EC
could b
e
get
by fuzzy con
t
rol. The m
e
a
s
uri
ng
syste
m
set the initi
a
l values
0
p
k
,
0
i
k
and
0
d
k
of the
PID co
ntrol.
The th
ree fu
zzy
amou
nts
p
K
,
i
K
and
d
K
coul
d b
e
obtaine
d by fuzzy inferen
c
e table, whi
c
h
was
sho
w
n i
n
Table 1.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
De
sign of Ele
c
tri
c
Load M
e
asu
r
ing a
nd
Contro
ling S
y
stem
for Steering Se
vro
(Pan We
ntao)
2757
Table 1.
p
K
/
i
K
/
d
K
Fuzzy Inferen
c
e
Table
E
C
U
E
NB
N
M
NS
Z
O
PS
P
M
PB
NB
P
B
/
NB
/P
S
P
B
/
NB
/NS
P
M
/NM/NB
P
M/NM/NB
P
S
/
NS
/NB
ZO/ZO /N
M
ZO/ZO /NS
NM
P
B
/
NB
/P
S
P
B
/
NB
/NS
P
M
/NM/
NB
P
S
/
NS
/NM
P
S
/
NS
/NM
ZO/ZO
/NS
NS
/ZO /ZO
NS
P
M/NB
/ZO
P
M
/NM/NS
P
M/NS
/N
M
P
S
/
NS
/NM
ZO/ZO /NS
NS
/P
S
/N
S
NS
/P
S
/ZO
ZO
P
M/NM /ZO
P
M
/NM/NS
P
S
/
NS
/NS
ZO/ZO /NS
NS
/P
S
/NS
NM/P
M/NS
N
M
/
P
M
/
Z
O
P
S
P
S
/
NM /ZO
P
S
/
NS
/ZO
ZO/ZO /Z
O
NS
/P
S
/ZO
NS
/P
S
/ZO
NM/P
M/
ZO
N
M/P
B
/ZO
P
M
P
S
/
ZO /P
B
ZO/ZO /NS
NS
/P
S
/P
S
NM/P
M/P
S
N
M/P
M
/P
S
NM/P
B
/P
S
N
B
/
P
B
/P
B
P
B
Z
O/ZO /P
B
ZO/ZO /P
M
NM/P
S
/P
M
NM/P
M/P
M
N
M/P
M
/P
S
NB
/P
B
/P
S
NB
/P
B
/P
B
The a
c
cu
rat
e
outp
u
t
p
k
,
i
k
and
d
k
be o
b
tai
ned th
rou
g
h
defu
zzifi
cati
on. The
para
m
eters
p
k
,
i
k
and
d
k
were co
rrecte
d by the Equation (3)
as sho
w
n.
d
d
d
i
i
i
p
p
p
k
k
k
k
k
k
k
k
k
0
0
0
(
3
)
The m
e
a
s
u
r
i
ng
system
h
ad a
c
cu
rate
output a
n
d
q
u
ickly respon
se
and
a
c
hie
v
ed the
desi
r
ed
cont
rol effect throu
gh fuzzy PID control.
4. Testing Pr
inciple
The mea
s
u
r
in
g system
wa
s sho
w
n in Fi
g
u
re 6. Ze
ro p
o
sition
cha
r
a
c
teri
stic, line
a
r
ity and
transfe
r coef
ficient, frequ
ency ch
aract
e
risti
c
and maximum torque of sreering sevro was
measured through the me
asu
r
ing
syst
e
m
that was b
e
desi
gne
d to verity the performa
n
ce of the
system. Th
e
prin
ciple
s
a
n
d
method
s t
hat we
re
u
s
e
d
to mea
s
u
r
e the pa
ram
e
ters
of sree
ring
s
e
vro
were as
follows
.
Figure 6. The
Measu
r
in
g System
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 4, April 2014: 2753 – 2
761
2758
1) When th
e
input si
gnal o
f
steerin
g servo
wa
s zero.
The feed
ba
ck voltage
sig
nal an
d
angul
ar di
spl
a
cem
ent sig
n
a
l of steerin
g
servo
were
both ze
ro, an
d the feedba
ck voltag
e si
gnal
wa
s
symmet
r
ic ab
out th
e o
r
igin
with
the i
nput
si
gnal.
Ho
wev
e
r, the
r
e
wa
s al
way
s
so
me
deviation
s which
wa
s
ca
use
d
by el
ectrical, m
agne
tic, mechani
cal, fabri
c
atio
n, assem
b
ly
and
other facto
r
s betwe
en ele
c
trical zero po
sition an
d me
cha
n
ical ze
ro
position. The
feedba
ck
sig
nal
of stee
ring
se
rvo was
alwa
ys not
zero when the
i
nput
sign
al was
ze
ro, an
d the
r
e
wa
s a
no-li
ne
ar
regio
n
ne
ar t
he ori
g
in. Th
e purpo
se of
zero po
sitio
n
testing
wa
s to mea
s
u
r
e the deviati
ons
betwe
en ele
c
trical zero po
sition an
d me
cha
n
ical ze
ro
position.
The tri
angul
a
r
wave
contro
l sign
al
wa
s treated
a
s
th
e i
nput
signal
of
stee
ring
serv
o, and
the data acquisitio
n
boa
rd coll
ecte
d
the f
eedba
ck voltag
e signal of the
steerin
g se
rvo
simultan
eou
sl
y. The volta
ge valu
e of
feedba
ck
wa
s
cal
c
ulate
d
and
record
ed by th
e
h
o
st
comp
uter p
r
o
g
ram.
2) T
r
an
sfer
coefficient a
n
d
linearity a
r
e t
w
o p
a
ra
meters that reflect t
he de
gre
e
of
electri
c
linear
of ele
c
tri
c
ste
e
rin
g
se
rvo. The
input an
d ou
tput of stee
ri
ng sevro
we
re p
r
op
ortion
al
relation
shi
p
theoretically. Ho
wever, the
r
e wa
s al
wa
ys varying d
egre
e
s of no
nlinea
rity due
to
system
atic error an
d ran
d
o
m
erro
r. Therefore, linea
rity and transfe
r coeffici
ent chara
c
te
risti
c
s of
steeri
ng
se
rvo nee
ded
to
be an
alyze
d
i
n
practi
cal
e
n
ginee
ring. In
the re
al expe
rimental, ta
ki
ng
m
voltages
i
x
in the rang
e -1
0
V
to +10V wa
s treate
d
as
t
he input si
gn
al of steerin
g servo, an
d
measured th
e corre
s
po
nd
ing a
ngul
ar d
i
spla
cem
ent
sign
al
i
y
.
The
best fitting
straight lin
e
wa
s
found by th
e
prog
ram,
and
all of the
out
put si
gnal
sh
ould b
e
a
s
cl
ose
a
s
po
ssib
le to the li
ne.
The
essen
c
e of fitting straight
line was
cal
c
ulate
d
the best sl
ope a
nd intercept
estimate. Fitting
method com
m
only use
d
commonly lea
s
t squ
a
re m
e
thod fitting parameters [9].
Measuri
ng da
ta
point
s
i
i
y
x
,
cou
l
d be
fitted wi
th the lin
e
b
kx
y
. The relation
shi
p
betwe
en in
pu
t sign
al an
d f
eedb
ack
sign
al could
be
d
e
scrib
ed
by the fitting line.
The
pri
n
cipl
e
of
least squ
a
res
method wa
s to
make
the sum of
ab
sol
u
te
value
s
of
deviation
s
m
i
i
i
i
b
kx
y
1
or
m
i
i
i
i
b
kx
y
1
2
as small as possible.
The value of
k
and
b
coul
d b
e
estimated
by the equati
on (4
), whe
r
e
the slope of the
fitting line
k
is the tran
sfer
co
efficient of steerin
g se
rvo.
n
i
i
n
i
i
i
m
i
i
m
i
i
i
x
x
y
y
x
x
x
m
x
y
x
m
y
x
b
x
b
y
k
1
2
1
1
2
2
1
(4)
The m
a
ximu
m di
stan
ce
b
e
twee
n me
asuring
poi
nts
a
nd fitting lin
e
is
max
L
. The lin
eari
t
y
of steerin
g se
rvo coul
d be
obtai
ne
d thro
ugh the Equa
tion (5).
%
100
max
FS
L
Y
L
(5)
W
h
er
e
L
is the
linea
rity of
steerin
g
se
rvo, and
FS
Y
is full-scale
an
gula
r
displ
a
cement
output value
of the steerin
g servo.
3) F
r
eq
uen
cy
ch
ara
c
teri
sti
c
wa
s u
s
ed
to de
scrib
e
th
e mathe
m
atical mod
e
l of
steering
serv
o
sy
st
em
.
Frequ
en
cy
cha
r
a
c
t
e
ri
st
ic
ref
l
e
c
t
ed
the
inhe
rent
cha
r
acte
ri
stics of
stee
ring
se
rvo
and
wa
s ma
inly use
d
to
mea
s
ure th
e amplitu
de-f
r
equ
en
cy ch
ara
c
teri
stic,
pha
se-f
req
u
e
n
cy
cha
r
a
c
teri
stic and pa
ss b
a
nd.
In the frequ
e
n
cy characte
ristic te
sting,
t
he stee
ring
servo
wa
s
c
ontrolle
d by
0.1Hz to
10Hz sin
u
soi
dal sign
al. The data acqu
isiti
on boa
rd
colle
cted the
feedback vo
ltage sign
al and
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
De
sign of Ele
c
tri
c
Load M
e
asu
r
ing a
nd
Contro
ling S
y
stem
for Steering Se
vro
(Pan We
ntao)
2759
angul
ar di
spl
a
cem
ent
sign
al of stee
rin
g
servo.
T
he cu
rves of
amplit
ude-f
r
eq
uen
cy
cha
r
a
c
teri
stic
and ph
ase-freque
ncy characteri
stic
we
re dra
w
n a
c
co
rding to the
collecte
d
data.
The pa
ss b
a
n
d
and ph
ase lag wa
s mo
st
important p
a
ram
e
ters of
freque
ncy chara
c
te
risti
c
. Low p
a
ss b
and
woul
d limit the steeri
ng servo resp
on
se
spe
ed,
and hi
gh pa
ss ba
nd
would affect
ed by the noise.
The ph
ase la
g is the p
hase differen
c
e
betwe
en t
he
pha
se of inp
u
t
and pha
se
of corre
s
po
n
d
ing
output sig
nal
[10, 11].
4) In
the m
a
ximum torqu
e
characte
ri
stic te
sting, the
output
of torque m
o
tor in
cre
a
sed
with the incre
a
sin
g
of angu
lar displa
cem
ent. The data
acqui
sition b
oard
colle
cte
d
the maximu
m
torque u
n
til the angul
ar di
splacement of steeri
ng serv
o no long
er in
cre
a
sed.
4. The results
The stee
ring
se
rvo wa
s controlled
by
four
peri
od’
s 2V tri
ang
u
l
ar
sign
al; the data
acq
u
isitio
n board
colle
cte
d
the voltage feedba
ck
sign
al. The result of the zero p
o
siti
on
cha
r
a
c
teri
stic tests was sh
own
in Fi
gu
re 7. T
he bl
ack
curve
was
the c
ontrol
si
gnal
of ste
e
ri
ng
servo, a
nd th
e red
wa
s the
voltage feed
back si
gnal.
The average
value of voltage feedb
ack
wa
s
0.0239V whe
n
the cont
rol sign
al wa
s ze
ro.
Figure 7. The
Result of Zero Position Ch
ara
c
teri
stic T
e
sts
Takin
g
some
voltages in t
he ra
nge
-10
V
to +10V was tre
a
ted a
s
the input si
gnal of
steeri
ng
se
rvo, and m
e
a
s
ured
the
corresp
ondi
ng a
ngula
r
di
spla
ceme
nt si
gna
l. The fitting line
wa
s foun
d th
rough l
e
a
s
t sq
uare
metho
d
. The
origi
nal
curve
an
d fitting line
were
sho
w
n i
n
Fig
u
re
8 as follo
ws:
Figure 8. The
Result of Linearit
y and Transfe
r Co
efficient Te
sts
Whe
r
e
the
re
d curve
was t
he o
r
igin
curv
e an
d th
e bl
a
c
k was the
fitting
strai
ght li
ne. Th
e
angul
ar di
spl
a
cem
ent outp
u
t was lin
ear
with the i
nput
of steering
servo. The line
a
rity was 2.5
8
%
and the tran
sf
er co
efficient
2.01
V
.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 4, April 2014: 2753 – 2
761
2760
Figure 9. The
Result of Fre
quen
cy Re
sp
onse Test
s
The stee
rin
g
servo
wa
s co
ntrolled by fro
m
0.
1Hz to 1
0
Hz sinu
soi
d
al con
s
e
c
utive signal.
The data acquisitio
n
boa
rd coll
ecte
d the vo
ltage feedb
ack sig
n
a
l and ang
ular displa
cem
ent
sign
al of stee
ring
se
rvo. Freque
ncy resp
onse of the steering
se
rv
o
wa
s sh
own i
n
figure
9. Th
e
width of p
a
ss band
wa
s g
r
eater th
an 1
0
H
z. T
he
st
ee
ring
se
rvo co
uld qui
ckly follow the
co
ntrol
sign
al wh
en t
he freq
uen
cy
of control si
gnal was l
e
ss than 1
0
Hz.
The ph
ase
angle diffe
re
nce
betwe
en inp
u
t
and output signal was a
b
o
u
t 52 ° at 10Hz.
Figure 10. Th
e Re
sult of Maximum Torq
ue Test
s
The maximum torque
max
T
sh
ould be
esti
mated thro
ug
h the engi
ne
ering
experi
e
nce.
The ratio bet
wee
n
the an
g
u
lar di
spl
a
ce
ment of
the steering
se
rvo
and the in
put
of torque m
o
to
r
'
k
was
set firstly, so that the output of
to
rque moto
r wa
s greate
r
than maxi
mum torqu
e
of
steeri
ng serv
o when the a
ngula
r
displ
a
ceme
nt of
steerin
g se
rvo wa
s gre
a
ter than 90% of full
scale. Th
e a
m
plitude
of steering
servo
co
ntrol
sig
n
al incre
a
sed
slo
w
ly. The
data a
c
qui
siti
on
board
colle
ct
ed the
signal
of torque
se
nso
r
and
an
g
u
lar di
spla
ce
ment
sign
al o
f
stee
ring
servo.
The cu
rve
s
of torque an
d a
ngula
r
displa
ceme
nt we
re
sho
w
n in Fig
u
re 10.
Whe
r
e th
e bl
ack
curve
wa
s the
an
gula
r
displa
ceme
n
t
of stee
ring
servo
an
d th
e re
d i
s
the si
gnal
of t
o
rqu
e
se
nsor. The fo
rward
maxi
mum
to
rque
wa
s 4
8
.41Nm
a
nd
re
verse
maxim
u
m
torque wa
s
4
1
.01Nm.
4. Conclusio
n
This p
ape
r
mainly de
sig
ned an
d research
e
d
the
prin
ciple, h
a
rd
wa
re
syst
em, and
softwa
r
e
syst
em and
control stra
tegy of
the stee
ring
servo me
asuri
ng sy
stem. T
he torq
ue m
o
tor
wa
s treated
as the load f
o
r sim
u
lating
the real stat
e
of steering
servo. The a
c
curacy of torque
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
De
sign of Ele
c
tri
c
Load M
e
asu
r
ing a
nd
Contro
ling S
y
stem
for Steering Se
vro
(Pan We
ntao)
2761
and th
e
spe
e
d
of the
me
a
s
uri
ng
syste
m
were
im
p
r
oved th
roug
h
fuzzy PID
co
ntrol m
e
thod.
The
virtual instru
ment tech
nol
ogy and a
u
to
mated testin
g
techniq
u
e
s
were combin
ed for
compl
e
ting
the desi
gn an
d developm
e
n
t of the
steering se
rvo me
asu
r
ing
syste
m
.
Ackn
o
w
l
e
dg
ements
This research
is a gene
ral
proje
c
t (6
130
5131
) su
ppo
rted by Natura
l Science Fou
ndation
of Chin
a. Spe
c
ial a
p
p
r
e
c
iati
on
shoul
d b
e
given to
Re
se
arch
Centre o
f
Fluid Po
we
r
Tran
smi
ssi
on
and
Control, Beihang Univ
ersity
for
the
sa
ke of their
selfle
ss h
e
lp.
Referen
ces
[1]
Cui
Hua
y
a
n
g
. Electro-H
y
d
r
a
u
lic Actu
ator
Loa
d
Sim
u
lati
on C
ontro
l S
y
stem.
Xi
’an.
North
w
e
s
te
r
n
Pol
y
t
e
ch
nica
l Univers
i
t
y
. 2
0
0
6
.
[2]
W
ang Sh
ouK
u
n
, W
ang Ju
nz
hen
g. D
y
n
a
mi
c Loa
din
g
for
Missile Act
uat
ors.
T
r
ansactio
n
s of Bei
jin
g
Institute of T
e
chno
logy.
2
007;
27(3): 247-
25
0.
[3]
Kang
Ch
en, J
i
anh
ua W
a
ng,
Jie Y
an. E
x
per
iment
and
Stu
d
y
of E
l
ectric
Loa
din
g
S
i
mul
a
tor for
Lli
n
e
a
r
Rud
der.
IEEE Intelligent Cont
r
o
l and Autom
a
t
i
on
. W
C
ICA 7th W
o
rld Co
ngr
ess on 20
08; 4
798 -4
80
2.
[4]
Liu
Jia
n
x
i
ng. R
e
searc
h
on E
l
e
c
tric Lo
adi
ng
S
y
stem
of R
u
d
d
e
r Bas
e
d
on
D
SP. Harb
in.
Ha
rbin
Institut
e
of T
e
chnol
og
y.
2010.
[5]
Yao
x
i
ng S
han
g, Shu
a
i W
u
,
Z
ong
xi
a Ji
ao.
Study o
n
Ulti
m
ate
Perfor
m
ance
of Li
ght-
duty Electro-
Hydra
u
lic T
o
r
q
ue-L
oad
Si
mu
l
a
tor. Ro
botics,
Auto
mati
on
a
nd Mec
hatro
ni
cs
. IEEE Conf
erenc
e on
200
8: 387 -3
92
.
[6]
W
ang Sh
u
y
an,
Shi Yu, F
e
n
g
Z
hong
xu. A Method for
Contr
o
lli
ng
a Lo
adi
n
g
S
y
stem B
a
se
d on a F
u
z
z
y
PID Contro
ller.
Mecha
n
ica
l
S
c
ienc
e an
d T
e
chno
logy for A
e
rosp
ace En
gi
neer
ing
. 2
0
1
1
; 30(1): 1
66-
172.
[7]
Rub
aai
A, C
a
stro-Sitiriche
MJ, Ofoli A.
DSP-Ba
se
d Impl
em
en
ta
ti
o
n
o
f
Fu
z
z
y-PID Co
n
t
ro
ll
e
r
U
s
i
n
g
Genetic Opti
mi
z
a
t
i
o
n
for
Hig
h Perfor
mance
Motor Dr
ives.
IEEE Industr
y
A
p
p
licati
ons
Conf
erenc
e
.
200
7: 164
9 -16
56.
[8]
Hua
ng W
e
ih
u
a
. F
u
zz
y C
ont
rol S
y
stems
a
nd A
ppl
icati
o
n
s
. Beij
ing.
Ele
c
tronics In
dust
r
y
P
u
b
lish
i
n
g
Hous
e. 201
2.
[9]
W
ang
Xi
nh
e,
Che
ng S
h
izh
o
u
. T
he Metho
d
of L
east S
q
uares
Carv
e
F
i
tting.
Jour
na
l of Xi
nji
a
n
g
Vocatio
nal U
i
n
v
ersity
. 2004; 1
2
(2): 84-8
6
.
[10]
W
ang Ju
n. T
he Stu
d
y
of
F
r
eque
nc
y C
h
aracteristic M
e
asurem
ent
S
ystem for a K
i
nd of M
i
ssil
e
Actuator. T
a
iyu
an. North Un
iv
ersit
y
. 2
008.
[11]
Xi
ao
qi
Hu
an
g.
T
he Ampl
itud
e F
r
eque
ncy-R
e
spo
n
se Ch
ar
acteristic an
d Phase F
r
eq
ue
ncy-Res
pons
e
Char
acteristic in measur
in
g
Har
m
on
ic F
l
ui
d T
e
mp
eratur
e.
Proce
e
d
i
ng
s of th
e 7t
h
Asian
T
herm
o
ph
ysic
al Prop
e
r
ties Confer
enc
e. 2004.
Evaluation Warning : The document was created with Spire.PDF for Python.