TELKOM
NIKA
, Vol.13, No
.2, June 20
15
, pp. 539 ~ 5
4
6
ISSN: 1693-6
930,
accredited
A
by DIKTI, De
cree No: 58/DIK
T
I/Kep/2013
DOI
:
10.12928/TELKOMNIKA.v13i2.1258
539
Re
cei
v
ed
De
cem
ber 1, 20
14; Re
vised
Ma
rch 22, 20
15; Accepted
April 12, 201
5
Design and Simulation of Small Space Parallel Parking
Fuzzy Controller
Qiulin Sheng*
1
,
Jie Min
2
, Xing
Zhang
3
, Zheng
w
e
n
Z
h
ang
4
, Yi Li
5
, Guang
y
a Liu
6
Coll
eg
e of Elec
trical an
d Elect
r
onic En
gin
eer
i
ng, Hub
e
i U
n
iv
ersit
y
of T
e
chn
o
lo
g
y
, W
u
h
an,
430
06
8, Hube
i Provinc
e
, Chin
a
*Corres
p
o
ndi
n
g
author, em
ail
:
sql103@
qq.c
o
m
1
, mj9048
20
557
@16
3
.com
2
,
64537
151
9
@
qq.com
3
,
669
03
736
4@q
q
.com
4
, 1029
5
199
40@
qq.co
m
5
, w
h
lt
xz@a
li
yu
n.com
6
A
b
st
r
a
ct
Based
on the
nonlinearity and time-v
ariation of autom
atic
parking pat
h tracking control system
,
w
e
use fu
zz
y
control th
eor
ie
s and
metho
d
s
to expl
ore
th
e
control r
u
l
e
s to i
m
pr
ove fu
zzy control
l
ers
a
n
d
desi
gn a
n
aut
omob
ile ste
e
ri
ng contr
o
ll
er. T
hen w
e
bu
i
l
d
the si
mu
lati
on exp
e
ri
ment
platfor
m
of an
auto
m
o
b
il
e i
n
Simuli
nk to s
i
mu
late t
he r
e
versin
g
settin
g
s
of par
all
e
l
p
a
rkin
g. T
h
is p
aper
ad
opts t
he
Mamda
n
i co
nt
rol rul
e
s; the
me
mbers
h
ip f
unctio
n
is
th
e
Gauss functi
on. T
h
is
pap
e
r
verifies th
e f
u
zz
y
control
l
er'
s
kin
e
matic mod
e
l
and th
e adv
an
tages of fu
zz
y
control rul
e
s. Simulati
on res
u
lts show
that the
desi
gn of
the control
l
er all
o
w
s
t
he auto
m
obil
e
to st
op
i
n
to the
park
i
n
g
spac
e s
m
all
e
r tha
n
the
sp
ace
obtai
ne
d by
pl
ann
ing
pat
h, and
auto
m
atic
parki
ng
bec
o
m
es
poss
i
bl
e i
n
the
parki
ng
plot. T
he c
ont
rol
system is ch
aracteri
z
e
d
by sma
ll trackin
g
er
ror, fast respon
se and h
i
g
h
reli
abil
i
ty.
Ke
y
w
ords
: Pa
ralle
l Parki
ng, Fu
zz
y
Co
ntrol,
Matlab Si
mulat
i
on
1. Introduc
tion
The pu
rpo
s
e
of this study is to improve t
he co
ntrol scheme of intell
igent automo
b
iles
so
that in reversi
ng the sma
r
t automobil
e
can be more
close to the ideal reversin
g trajecto
ry and
it
can
stop
into
a sm
all pa
rki
ng spa
c
e. Thi
s
technol
ogy
can
ea
se traffic p
r
e
s
sure
a
nd en
han
ce t
h
e
safety perfo
rmance of vehicle
s
.Autom
atic parkin
g
techn
o
logy is currently on
e of the most
popul
ar sm
art
technolo
g
y, whi
c
h ha
s attracte
d the
attention of man
y
resea
r
che
r
s. There are two
the mo
st
cla
s
sic ap
proa
ch
es [1]
-
[2]: (1
)
By pat
h
pl
an
ning method,
whi
c
h previously wa
s
give
n
a
geomet
ric
pa
th,
com
b
ined
with
th
e car's dynami
c
model and
p
a
rki
ng re
stri
ctions gen
erat
ion
control st
rate
gies,thi
s
is
a
resea
r
ch pa
th trac
king
control of the
vehicle
ba
sed on th
e visual
environ
ment t
hat ha
s
bee
n
more m
a
ture
[3]-[4]. (2
) B
a
se
d o
n
skill
acq
u
isitio
n m
e
thod
s,whi
c
h
by
fuzzy logi
c or
neural networks [5
], learnin
g
more
skill
e
d
peopl
e parking tech
nolo
g
y
transferred
to
the autom
atic pa
rking
co
ntrolle
r. Thi
s
method
is
n
o
t desi
gne
d
referen
c
e p
a
th, whi
c
h
co
n
t
rol
strategi
es wa
s impl
emente
d
to control
the po
siti
on
a
nd o
r
ientatio
n
angle
a
c
cord
ing to the
ca
r in
the
pa
rki
ng. This pap
er studie
s
fuzzy
lo
gic cont
ro
l m
e
thod. On
the
basi
s
of
prev
ious
studi
es,
we
will desi
gn
a f
u
zzy controller that
wa
s
opt
imized its fuzzy control rul
e
s
to control
t
he car so
that
it
can a
c
hi
eve a more id
eal
control pe
rformance,
incl
udi
ng the contro
l pre
c
isio
n an
d the minimu
m
size of parkin
g
spa
c
e
s
an
d
parki
ng time.
2. Kinematic Model of Car Parallel Parking
2.1. Dete
ctio
n of Car Par
k
ing Spaces
Before the
car a
u
tomatically stoppi
ng
into
the pa
rking
sp
aces,
it must be
detecte
d
arou
nd the
p
a
rki
ng
spa
c
e
s
. The im
age
sen
s
o
r
or
ul
traso
n
ic
se
nsor is i
n
stall
e
d
arou
nd
car t
o
detect the p
a
rki
ng spa
c
e
s
. For exam
ple, t
he image se
nsor p
hotograph
ed
sce
nery aro
und,
combi
ned
wit
h
an
algo
rith
m, the car
ca
n identif
y pa
rking
sp
aces.
The i
rre
gula
r
spa
c
e
s
tuni
n
g
a
s
a rule rectan
gular
spa
c
e
s
in Figure 1. If t
he parki
ng
spa
c
e i
s
larg
e enou
gh, the parking
spa
c
e
s
will be the parking
space of sm
art car.T
ables
and Figures are
presented center, as shown
bel
o
w
and cite
d in the manu
scri
p
t
.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No. 2, June 20
15 : 539 – 54
6
540
2.2. The Esta
blishing Rev
e
rsing Mode
l of Car Esta
blished
2.2.1. The Ac
kerman angl
e of motor tw
o
w
h
eels i
n
actu
a
l mo
v
e
ment
In orde
r to si
mplify the reverse mod
e
l o
f
the car, the
car
side
-sli
p
won't be
con
s
ide
r
ed
and the
parki
ng spee
d is
very low. Car whe
e
ls in
th
e rotation foll
ow the
prin
ci
ple of Acke
rman
[13] angle a
s
Figure 2.
Two corne
r
s of car
two wh
eel
s computation e
quation
s
are sho
w
n in the
followin
g
:
ta
n
ta
n
in
in
ou
t
ou
t
in
o
u
t
L
R
L
R
RB
R
(1)
Acco
rdi
ng to equatio
n (1
), followin
g
equ
ation will be
came abo
ut
in
an
d
out
:
ta
n
+
=
ta
n
in
in
ou
t
i
n
RB
R
(2)
There i
s
a
re
lationship
of
formula
(2) o
f
the two
wh
eels rotation
al an
gle i
n
th
e a
c
tual
movement.
But in orde
r t
o
simplify the
model, a unif
i
ed Acke
rman
angle
will be
defined a
s
,
which
is midpoint rotational an
gl
e of car fro
n
t axle following
Fig.3.
This
angle
will be u
s
e
d
for the math
ematical m
o
d
e
l behin
d
. Th
e angle
of the two
whe
e
ls A
c
kerman an
gle h
a
s a
ce
rtain relation
ship
wi
th the
.
The e
quation
s
a
r
e
sho
w
n in th
e
following:
ta
n
=
2
in
L
R
B
RR
(3)
Acco
rdi
ng to front equ
ation
s
(1
), (2), (
3
);
ta
n
ta
n
2
=
ta
n
ta
n
2
in
out
LB
LB
(4)
2.2.2. Car re
v
e
rsing mod
e
l under the
simplified model
Two
wh
eel
s o
f
car were
co
ntrolled
with
two moto
rs a
nd two contro
l chip
s
re
sp
ectively in
the pap
er, a
nd wheel
s ro
tational angl
e
were dete
r
mined by the
Acke
rma
n
a
ngle. In build
ing
model, Acke
rman an
gle
wa
s re
pre
s
e
n
ted with
i
n
the pape
r.
Automatic p
a
rallel p
a
rkin
g
algorith
m
i
s
b
a
se
d o
n
kine
matic m
odel
of the
ca
r.
In building
kine
matic
m
odel
of
the ca
r,
first
of
all mod
e
l pa
rameters
nee
d to be
dete
r
mined. F
o
r
e
x
ample, the
coordi
nate
of the
cente
r
of t
he
car i
s
（、
x
y
), the symb
ol of
represents
speed of ca
r, the symb
ol of
l
rep
r
e
s
ent
s sp
eed of left
whe
e
l of ca
r and the
symbol of
r
rep
r
e
s
ents spee
d o
f
right whe
e
l of car. Th
e
r
e
pr
es
e
n
t
s
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
De
sign a
nd Sim
u
lation of Sm
all Space Parallel Pa
rki
n
g Fuzzy Cont
rolle
r (Qiuli
n Sheng
)
541
angle
betwee
n
X-axis and
cent
ral axi
s
of ca
r. The
repre
s
e
n
ts
sli
p
of car th
at is the
rate of
cha
nge
of
. Every time the travel trajec
to
r
y
o
f
r
e
ar
has
impo
r
t
an
t re
la
tio
n
s
h
ip
w
i
th
d
i
s
t
an
ce
o
f
whe
e
ls
and
axle, and A
c
kerman
angl
e of the
cent
er of fro
n
t ax
le. The
spe
e
d
of ca
r h
a
ve no
influenc
e
on t
he travel trajec
tory [2]-[3].
Kinematic
e
q
uation of ca
r are sho
w
n in
the followin
g
:
cos
(
)
si
n
(
)
ii
i
d
t
xi
xi
i
i
yi
yi
i
i
（
+1
）
=
（）
+
（）
（
+1
）
=
（）
+
（
+1
）（
+1
）
（
+1
）
=
（）
+
（
+1
）（
+1
）
(5
)
Speed eq
uati
on of ca
r left and rig
h
t wh
e
e
ls are sh
own in the following:
()
()
()
2
()
()
2
rl
rl
ii
i
ii
i
=
（）
=
(6)
In each
sam
p
le peri
od, the ca
r ca
n o
b
tain the rot
a
tional sp
ee
d of the left
and rig
h
t
whe
e
ls by the
above equ
ations.
Figure 1. Det
e
ction a
n
det
ermin
a
tion th
e car
parking spa
c
es
Figure 2. The
model of Ackerma
n
angl
e
Figure 3. Ackerma
n
angl
e of car
Figure 4. The
reversing mo
del of car
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No. 2, June 20
15 : 539 – 54
6
542
3. The Desig
n
of
Controll
er of Car
Fuzzy
co
ntrol
l
er is th
e core of Fuzzy co
ntrol sy
stem,
and the d
e
si
g
n
of Fuzzy co
ntrolle
r
dire
ctly affects the accuracy of
final control out
come. The de
sign o
f
the controll
er sh
ould ap
ply
mode
rate p
r
i
n
cipl
es th
at repre
s
e
n
t not
only hi
gh a
ccura
cy, but al
so
simplified
fuzzy infe
ren
c
e
operation rel
a
tively. Because
compl
e
x control syst
e
m
will lead to delayed re
spon
se of system
that can have an influence on real-tim
e cont
rol of
car that will decrease ac
curacy of control in
turn [8].
3.1. The Blo
ck Diagr
a
m of Fuzz
y
Controller
The blu
r
re
prese
n
ts that the pre
c
i
s
e di
gital variable
chan
ge
s into fuzzy vari
a
b
le. The
pro
c
e
s
s of fu
zzy i
n
ference
prese
n
ts that
acco
rdin
g to
the fuzzy
rule
ba
se, a
nd th
e data
b
a
s
e
h
a
s
been
esta
blished, the fu
zzy inputsa
re p
r
ocesse
d to
gene
rate the
corre
s
p
ondin
g
co
ntrol in
p
u
t
s
and co
ntrol strategy.
Anti-blur
process
rep
r
e
s
ent
s th
at the fu
zzy
outputs chan
ge into
preci
s
e
nume
r
ic vari
a
b
les, which were u
s
ed to control outp
u
t obje
c
t.
3.2. Block Di
agram of the
Fuzzy
Contr
o
ller
The main diff
eren
ce
s bet
ween fuzzy co
ntrol
sy
stem and o
r
dina
ry comp
uter n
u
m
eri
c
al
control
syste
m
is the
u
s
e
of a fu
zzy co
ntrolle
r. Th
e
perfo
rman
ce
of a fu
zzy co
ntrolle
r i
s
d
e
cided
these fa
cto
r
s that are
the
stru
ct
ure of t
he fuzzy
cont
rolle
r, synthet
ic rea
s
oni
ng
algorith
m
s
an
d
fuzzy control rule
s.
The ba
sic blo
ck
di
a
g
ram of
the fuzzy co
ntrol
as Fig
u
re 6 shown, the fuzzy
controlle
r is
the its co
re
,and fuzzy controlle
r
cont
rol rul
e
s i
s
impleme
n
ted
by a comput
er
prog
ram. Im
plementatio
n
pro
c
e
s
s
of
fuzzy
co
ntrol
algo
rithm i
s
de
scribe
d a
s
follo
ws:first
l
y,
comp
uter obt
ain the
p
r
e
c
i
s
e val
u
e
of
controlle
d va
ri
able
s
by
the
interrupt
sam
p
ling, the
n
th
is
value com
pares with a giv
en variabl
e to obtain a
differen
c
e of si
gnal E, as an
input; seco
n
d
ly,
the a
c
curate
differen
c
e
of sign
al E chan
ge into
blur,
a
nd u
s
ing
app
ropriate
lang
u
age
rep
r
e
s
en
ts
the differe
nce
E, then ba
se
d on th
e
synthesi
s
of
rule
-based
rea
s
o
n
ing, co
mbin
e
d
with
he fu
zzy
r
e
lation
R
obtains
fuzzy c
ontr
o
l volume. c
o
mputation equations
ar
e s
h
ow
n in the
following:
ue
R
.
3.3. The fuzzy
Controller
Design
Bec
a
us
e
the
pro
c
e
ss of re
versin
g is very complex, we ne
ed to d
e
sig
n
a com
p
lex fuzzy
controlle
r so t
hat the car
can be
cont
roll
ed preci
s
el
y.
The len
g
th of
the vehicle i
s
defin
ed in 4
m
,
and
th
e width
is 1.6m;
the
cente
r
po
sition
coo
r
dinat
e
of the vehi
cle
is
set
(x, y),
and th
e o
r
din
a
te
rep
r
e
s
ent
s th
e dista
n
ce fro
m
the
car to the pa
rking
sp
ace. F
u
zzy
controlle
r h
a
s
three i
nput
s
x,
y, and
in th
e p
r
ocess of
reve
rsi
ng; a
nd a
outp
u
t
equal
to the
rotation
angl
e of
whe
e
ls
approximatel
y. This th
re
e-dimen
s
ion
a
l f
u
zzy contro
ll
er to
control
have a
total
of 18
rul
e
s.
The
variable
s
me
mbershi
p
function are sh
o
w
n in the Fig
u
re 7, Figu
re
8, Figure 9, F
i
gure 1
0
. In this
pape
r, Mamd
ani co
ntrol rul
e
s were u
s
e
d
,
and membe
r
shi
p
functio
n
s
use Gau
s
si
an functio
n
[9].
The
co
ntrol rules of fuzzy controlle
r we
re sho
w
n in T
able 1:
On the pa
rt of the fuzzy rul
e
s of pa
rallel parking will
b
e
explained a
s
follows:
If x is S, a
nd
y is S, it
mea
n
s th
at the
car
ha
s b
een
reversed
into t
he p
a
rking.
Whe
n
is N, it mean
s that angle o
f
dire
ction of the vehicl
e at this time
is a negative valu
e, the directio
n
of the vehicle
need to be st
raighte
ned
ra
pidly.
If x is S, and y is S, it means that the
ca
r has b
een re
verse
d
into the parking. When
is
Z, it mean
s t
hat angl
e of
dire
cti
on
of th
e vehicl
e at t
h
is time
is
ze
ro, then
reve
rsing
the p
r
o
c
ess
of the car e
n
d
s
.
If x is B,
y is B,and
is P,
it means that
the car ha
s
rea
c
he
d the outsid
e
of parkin
g
spa
c
e
s
, an
d
ready to i
n
to
the pa
rki
ng
spa
c
e
s
; t
hen
the directio
n
of the a
ngle
of the
car i
s
positive, and i
t
began to turn the steeri
n
g
wheel
s to
the right makin
g
the car ba
ck to the parking
spa
c
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
De
sign a
nd Sim
u
lation of Sm
all Space Parallel Pa
rki
n
g Fuzzy Cont
rolle
r (Qiuli
n Sheng
)
543
Figure 5. The
block diag
ra
m of Fuzzy
c
ontroller
Figure 6. Block di
agram of
the fuzzy co
ntrolle
r
Figure 7. Membershi
p
fun
c
tion of x coo
r
dinate
of the car
Figure 8. Membershi
p
fun
c
tion of y coo
r
dinate
of the car
Figure 9. Membershi
p
fun
c
tion of the di
rectio
n
angle
Figure 10. Membe
r
ship fu
nction fu
zzy
controlle
r out
put
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No. 2, June 20
15 : 539 – 54
6
544
Table 1. The
control rul
e
s
of fuzzy co
ntroller
x
/
y
S
B
VB
=N
S
B
VB
PB
PM
PB
PB
PB
PM
=Z
S
B
VB
Z
Z
Z
PB
PB
Z
=P
S
B
VB
NB
NM
Z
Z
PM
NB
4. Simulink
Simulation Module Stru
cture
s
The length of
the vehicle is set to 4m, the wi
dth is set to 1.6m, a
nd the vehicl
e spee
d is
set to 5m/
s
[1
0] in sim
u
lati
on; there a
r
e
three fu
z
z
y
c
ontrolle
r in
put
u (1
), u
(2
), Contr
o
lle
r, w
h
ich
are x co
ordi
n
a
te, y coordi
n
a
te, and the dire
ction
of the vehicle respectively; a fuzzy cont
roll
er
input re
pre
s
e
n
ts ch
ang
ed rate of the an
gle direct
ion
of the vehicle
.
S-function
module
of pa
rkin
g
are
use
d
ha
ndling
real
-ti
m
e imag
e di
splay a
nd
o
b
s
tacl
e avoid
a
nce i
n
pa
rkin
g proce
s
s. truck
kinem
atic mo
dule
s
is
sub
r
outine mo
dul
e for the ma
t
hematical mo
del of the tru
c
k; other m
o
d
u
les
are
som
e
of t
he auxilia
ry module i
n
si
mulation. Ne
st
ing of fu
zzy
logic
co
ntroll
er : ba
se
d on
front
the fuzzy con
t
rol rul
e
s sp
e
c
ify the form
at FIS
files e
m
bedd
ed int
o
the fu
zzy
controlle
r in th
ree
s
t
eps
[9],[11]:
(1). Usin
g MATLAB comm
a
nd rea
d
fis in t
he MATLAB main win
d
o
w
, type: new file name
= rea
d
fis ('file-nam
e') Ente
r.
(2).
Using th
e mou
s
e
in t
he FIS e
d
itor: Ty
pe "fuzzy
file-na
m
e" i
n
the MAT
L
AB main
wind
ow, then
the file will be
s
ent to the MATLAB work-spa
ce.
(3).
Do
uble
-
cl
ick the
fuzzy
controlle
r of
being
ne
sted,
and
by m
odi
fying the p
a
ramete
r
name, it was changed to
"new file name". Afte
r nested
Control
l
er, we
will i
n
spect the fuzzy
controlle
r to
verify
available. Rig
h
t key-fuzzy
co
ntroll
er, a
nd
sel
e
ct "look un
der ma
sk",
if th
e
fuzzy controll
er sh
ows FIS,
it represent
s ne
sting su
cce
ss; if t
he
displ
a
y is sffis, it represe
n
ts
not nestin
g
succe
ss
whe
n
the fuzzy con
t
roller
sho
u
ld
be re
-ne
s
ted
until Show FI
S.
Figure 11. Simulink
simul
a
tion module
structu
r
e
s
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
De
sign a
nd Sim
u
lation of Sm
all Space Parallel Pa
rki
n
g Fuzzy Cont
rolle
r (Qiuli
n Sheng
)
545
5. Simulink
Simulation Results
1. In the
foll
owin
g three
Figure 1
2
, th
e dy
na
mic proce
s
se
s of
reversi
ng
pa
rking
were
simulate
d in
different
starti
ng p
o
ints tha
t
are
co
or
din
a
tes (10.5, 3.5),
(10.
5,4.0),
and
(1
0.5, 4.
5)
respe
c
tively. From the th
ree co
ordinat
es
We c
an find that they have differe
n
c
e y-coo
r
din
a
t
es.
Based
on Fi
gure
13, wh
en the pit
c
h
is the b
e
st
d
i
stan
ce, the l
ongitudi
nal di
stan
ce h
a
ve
no
influen
ce on
automatic p
a
rking b
u
t maki
ng the minim
u
m parkin
g
space larg
er [1
2].
2. On th
e b
a
si
s of the
b
e
st
starting
point, a
c
cord
ing to the
re
verse
path
g
eometry
con
s
trai
nt bet
wee
n
the car
and the p
a
rki
ng sp
ace
obt
ains a m
e
tho
d
of steeri
ng
control strate
gy,
who
s
e the
si
mulation resu
lts are
sho
w
n
in Figure 1
4
.
3. On th
e ba
sis of the
be
st sta
r
ting
po
int,
the sim
u
l
a
tion results
of the fu
zzy
control
algorith
m
wa
s sh
own in Figure 1
5
.
Figure 12. Th
e reversin
g proce
s
s
Figure 13. Th
e reversin
g proce
ss
Figure 14. The reve
rsi
n
g
process on t
he
Figure 15. The rev
e
rsi
ng p
r
o
c
e
s
s on the ba
si
s
basis of path plan
ning
of fuzzy Cont
rol
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No. 2, June 20
15 : 539 – 54
6
546
6. Simulation Anal
y
s
is
Acco
rdi
ng to the simulatio
n
results in Fi
gure
1
2
, it can be se
e that the samlle
r distan
ce
betwe
en the
car
and th
e
obsta
cle i
s
m
o
re
stand
ard
tr
ajecto
ry.so
the be
st vertical di
stan
ce
is
0.5m.On the basi
s
of the best sta
r
ting
point of
the reverse, the same 6m pa
rking, Figure 1
4
sho
w
s the pa
th-ba
s
ed
con
t
rol method cannot ma
ke
the ca
r safe stop into parki
ng spa
c
e
s
, b
u
t
Figure
15 ba
sed on
the
fu
zzy cont
rol si
mulation re
su
lt
s sh
ow that
cars
ca
n be
safe stop
into t
h
e
parking spa
c
e.
7. Conclusio
n
This pa
pe
r studies the p
a
rki
ng situ
ation at
low sp
eed
s, analyze of several
step
s in
parall
e
l pa
rki
ng, simplify parking
ste
p
s of
the ve
hicle mo
del,
and ab
stract
parking p
r
o
c
e
ss
para
m
eters,a
nd
re
sea
r
ch
the theo
ry of
fuzzy
cont
ro
l. Fuzzy
cont
rol i
s
appli
e
d
to the
pa
rall
el
automatic
pa
rkin
g sy
stem.
The inp
u
t, output and th
e fuzzy co
ntrol rule
s a
r
e
desi
gne
d in the
pro
c
e
ss d
e
si
gn of parking
. we can d
r
a
w
that t
he co
ntrolle
r desi
g
ned in this pa
per can ba
si
cally
meet the p
a
r
kin
g
d
e
man
d
. The p
a
rki
ng spa
c
e
where
the
car can
be
parked
ha
s be
come
smalle
r. The
trace h
a
s h
i
gher
cont
rol
pre
c
isio
n, which i
s
bette
r than that o
f
path planni
ng
method.
Ho
wever, we
can
kn
ow from
the
simulati
o
n
re
sults that
the controll
er de
sign
is
a
n
importa
nt a
s
p
e
ct, but
sel
e
cting a
startin
g
point i
s
al
so
essential
to succe
s
s pa
rki
ng, so that o
n
ly
two pa
rt must coope
rate t
o
make ca
r i
n
to parki
ng
spaces
safely. But in the actual p
r
o
c
e
s
s of
reversin
g the
situatio
n wo
uld n
o
t be
so ide
a
listic,
i
n
orde
r to m
a
ke
it control
real
vehi
cle,
the
controlle
r nee
ds furth
e
r imp
r
oveme
n
t.
Ackn
o
w
l
e
dg
ements
The
re
sea
r
ch is fund
ed
by Wuha
n
Fund
ame
n
tal Applie
d
Re
sea
r
ch P
r
oje
c
t,
No.20
130
124
0101
0 84
5, BSQD Fu
nd
Re
sea
r
ch Project, No. BSQD12
023,
and
sci
en
ce
an
d
technology support program of
Hubei Province, No.2014BAA135
Referen
ces
[1]
Yana
n Z
,
Emmanu
el GC Jr
. Robust
auto
m
atic par
all
e
l
parki
ng i
n
tig
h
t
spaces vi
a f
u
zz
y lo
gic.
Rob
o
tics an
d Autono
mous Sy
stems.
20
05; 5
1
: 111-1
27.
[2]
Jian
g H, Guo KH, Z
hang JW
. Design of aut
omatic
par
all
e
l
parkin
g
steeri
ng contro
ller b
a
sed o
n
pat
h
pla
n
in
g.
Journ
a
l of Jili
n Univ
e
r
sity (Engin
eeri
ng an
d T
e
chn
o
l
ogy Ed
ition).
2
011; 41(
2): 293
-297.
[3]
Yueh
ai W
,
Ning C. Path Plan
nin
g
Optimizatio
n
for T
eaching an
d
Pla
y
back W
e
ldi
ng R
obot.
T
E
LKOMNIKA Indon
esi
an Jou
r
nal of Electric
al Eng
i
ne
eri
ng.
2013; 1
1
(2): 9
60-9
68.
[4]
Yan
X, W
u
Q,
Liu H. An Impr
oved R
o
b
o
t Path
Plann
in
g Alg
o
rithm Base
d o
n
Genetic Alg
o
rithm.
T
E
LKOMNIKA Indon
esi
an Jou
r
nal of Electric
al Eng
i
ne
eri
ng.
2012; 1
0
(8): 1
948-
195
5.
[5]
T
i
e W
,
Long
Q, Hong
C, Ji
ng W
.
Simu
lat
e
Stud
y
of Au
tomatic Parki
n
g S
y
stem.
TELKOMNIKA
Indon
esi
an Jou
r
nal of Electric
al Eng
i
ne
eri
ng.
2013; 1
1
(12):
732
4-73
30.
[6] Gillesp
ic
T
D
.
Fund
atncnta
l
s of Vchicle Dyn
a
tnics.
SAE International. 1992.
[7]
W
u
RH, Z
h
an
g GR. D
e
rivati
on
and
e
x
-
peri
m
ental
veric
a
ti
on
of the
veh
i
cle traj
ector
y
f
o
r b
a
ck
w
a
r
d
motion.
Mech
a
n
ical
and In
dus
trial man
age
ment.
200
6; 274:
94-10
2.
[8] Liu
JK.
Intelli
ge
nt Control
. Pu
b
lishi
ng h
ous
e o
f
electronics i
n
dustr
y
.
2
005.
[9]
Shi
X
M
, Hao ZQ.
F
u
zz
y
co
ntrol an
d MAT
L
AB simu
lati
on
. T
s
ing
hua U
n
iv
er
sit
y
Press. 20
0
8
.
[10]
W
ang P. Study on Par
a
ll
el
Parkin
g
S
y
ste
m
Based on
F
u
zz
y
Lo
gic
Contro
l.
Journ
a
l of Jia
m
usi
Univers
i
ty ( Natural Scie
nce E
d
itio
n)
. 201
2; 30(1): 28-3
7
.
[11]
Z
hang JR, W
ang L. F
u
zz
y
co
ntrol simul
a
tio
n
s
y
stem base
d
on MAT
L
AB.
Z
i
don
ghu
a Yu Yiqi Yib
i
a
o
.
200
3; 150(
1).
[12] W
u
B.
Rese
ar
ch o
n
p
a
th si
mu
lati
on
and
moti
on c
ontro
l
for auto
m
atic
parki
ng
. He
Fe
i U
n
i
v
e
r
si
ty
o
f
T
e
chnolog
y. 2
012.
[13]
Verón
i
ca S
R
,
F
r
ancisco
RM, Ishan
i C, J
o
s
é
MM
HR. Esti
mation
of Ack
e
rman
an
gl
es
for front-a
xl
e
steered ve
hicl
e
s
.
Artificial Intel
ligence Resear
ch
. 2013; 2(
2): 18-2
7
.
Evaluation Warning : The document was created with Spire.PDF for Python.