TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol.12, No.6, Jun
e
201
4, pp. 4664 ~ 4
6
7
0
DOI: 10.115
9
1
/telkomni
ka.
v
12i6.544
9
4664
Re
cei
v
ed
De
cem
ber 2
9
, 2013; Re
vi
sed
March 5, 201
4; Acce
pted
March 19, 20
14
Ultrasonic Automatic Trackin
g
System
Dai Juan
*, Zhao Zhihong
, Yan Minglia
n
Schoo
l of Ener
g
y
and El
ectro
n
ical E
ngi
ne
eri
ng,
Nanj
in
g Institute of Industr
y
T
e
chn
o
lo
g
y
, P.R. China, 21
00
2
3
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: Daij@
niit.e
du
.cn
A
b
st
r
a
ct
Ultrasonic ranging system
has
m
any
char
acteristi
cs, suc
h
as well-dir
ec
tionality, small volume,
and
low
pr
ice,
but fai
l
s to r
e
ali
z
e
a
lo
ng
di
stance
meas
ur
ed. In
order
to
real
i
z
e
exactl
y locati
ng
moti
on
obj
ects, multip
le
ultraso
n
ic
s
ensors
are
a
d
opted
to
meas
ure th
eir
moti
on
orie
ntatio
n
so that t
heir
2
D
coord
i
nates sh
oul
d be loc
a
te
d and p
o
sitio
n
ed, w
h
ich re
su
lts in calcul
atio
ns complic
ated
and meanw
h
i
l
e
me
asur
e
m
ent
subj
ect to errors. T
h
is study
addr
esse
s a
me
asur
e
m
ent
meth
od to rea
l
i
z
e
auto
m
atica
lly
tracking
moti
on
objects. It has a reaso
n
a
b
le d
i
stributi
on
struc
t
ure of the ultra
s
onic
se
nsors ado
pted,
critica
l
tracking c
a
lcu
l
ation
metho
d
, w
h
ich makes
it possi
ble
t
o
so
lve co
mplex
pr
obl
e
m
s of e
q
u
a
tions. As it
h
a
s
simple
co
ntrol
instructio
n s
e
que
nce,
it
is
abl
e to
precis
ely co
ntrol ti
ming
per
iod
an
d effective
l
y s
o
lv
e
me
asur
e error
cause
d
by irr
e
gul
ar de
lay fro
m
co
mmands.
T
hese pr
ogra
m
s and
met
hods
are w
e
ll d
e
p
l
o
y
e
d
in the par
am
et
ers collecti
on and status contr
o
l syst
em
for a
m
o
v
i
ng object.
Ke
y
w
ords
: ultr
ason
ic, rang
er findi
ng, trackin
g
, asse
mbl
e
r in
struction
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 moving objects mo
nitori
ng system, a
great
numb
e
r
of data are
usually req
u
i
r
ed for
dynamically trackin
g
,
su
ch
as lo
cation,
dista
n
ce
, an
d traj
ecto
ry.
Usual te
ch
no
logie
s
u
s
e
d
i
n
rackin
g syste
m
includ
e Zig
B
ee techn
o
lo
gy, ultr
asoni
c directio
nal range findi
ng
techn
o
logy, a
nd
microwave d
i
stan
ce m
e
a
s
uri
ng te
ch
n
o
logy.
Amon
g them, Zig
bee te
chn
o
lo
gy ha
s a
great
advantag
e in netwo
rk
com
m
unication
s.
Dista
n
ce
ca
n be
calcula
t
ed by a
nal
yzing
RSSI para
m
eters. Ho
wever,
th
e
a
c
tual
measure erro
r is large eno
ugh to
rea
c
h
approximatel
y one meter. This te
chnol
o
g
y also re
qui
res
three o
r
mo
re ZigBee m
o
dule
s
in com
bo when lo
ca
ting the obje
c
t’s po
sition. F
u
rthe
rmo
r
e, the
tested obj
ect
may also be
required to ha
ve a ZigBee module.
Ultra
s
oni
c
m
odule
s
are
g
enerally u
s
ed
to mea
s
u
r
e
distan
ce. It h
a
s
well
-direct
i
onality,
high
pre
c
i
s
io
n. Unfo
rtunat
ely, it fails to
measur
e lo
ng
dista
n
ce. It can rea
c
h
a hi
gh p
r
e
c
isi
on
of
millimeter when the target i
s
with
in a range of several meters.
In a micro
w
a
v
e rangin
g
sy
stem, the dist
anc
e and o
r
i
entation can
be reali
z
e
d
b
y
means
of frequ
en
cy sp
ectrum
a
nalysi
s
a
nd
cal
c
ulat
in
g th
e ph
ase diff
eren
ce
bet
ween
a tra
n
smitted
electroma
gne
tic wave a
n
d
the
wave
re
flected
by
an
d re
ceive
d
from the
obje
c
t. It has a
hi
gh
pre
c
isi
on a
n
d
is abl
e to realize a lon
g
distan
ce
m
e
asu
r
em
ent. In parti
cula
r, i
t
is the opti
m
al
choi
ce to
det
ect movin
g
thing
s
, espe
ci
ally those
m
e
tal obje
c
ts.
For a
ge
nera
l
motion o
b
je
ct,
there
are t
w
o
comm
on
req
u
irem
ents fo
r detectin
g
a
moving subje
c
t in a p
r
oj
ect
,
i.e. measu
r
i
ng
its traje
c
tory
and
colle
cting
its statu
s
d
a
ta in r
eal-time.
Com
pared with
other dista
n
ce mea
s
u
r
in
g
techni
que
s, the ultra
s
oni
c
rang
e-fin
d
ing
method ha
s so many adv
antage
s, e.
g. 1) No
n-se
nsit
ive
for external li
ght an
d ele
c
tromagn
etic fie
l
ds
with th
e result th
at it can b
e
u
s
ed
in
the d
a
rk, du
sty,
smo
k
e,
elect
r
omagn
etic i
n
terferen
ce, to
xic an
d ot
h
e
r
ha
rsh
cir
c
um
st
an
ce
s;
2
)
S
i
mple
st
ru
ct
ur
e,
small volum
e
, low co
st, less te
chni
cal
difficulty,
simple and
relia
ble in processing info
rmati
on,
easy to
be
m
i
niaturi
z
ed
an
d integ
r
ated
[
1
]. This
a
r
ticl
e mai
n
ly di
scusse
s h
o
w to
use ult
r
a
s
oni
c
sen
s
o
r
s to o
v
erco
me lon
g
distan
ce
measur
i
ng problem
s and
to solve measu
r
em
ent erro
rs
cau
s
e
d
by the prog
ram.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Research of a New No
n-Layer Protocol
Architecture
fo
r Satellite Network (Lixi
a
ng Liu)
4665
2. Sy
stem Structur
e
Tra
cki
ng syst
em
is mainly
con
s
i
s
ted of
tw
o comp
on
ents,
i.e.
ha
rdwa
re and software,
disc
us
sed as
follows
, res
p
ec
tively.Its
hardware ma
inly inc
l
udes
:
s
e
ns
or, turntable
and its
drive
module, moto
r and its drive
module, MCU, interfac
e d
r
ive module,
displ
a
y modul
e, powe
r
su
p
p
ly
and so on. Figure 1 d
e
scri
bes the
stru
ct
ure of its sen
s
or a
nd turntable.
Figure 1. System
Structure
Diagram
The moving
obje
c
t discussed in this
study is
locate
d at the coordinate of (x, y). The
distan
ce
s me
asu
r
ed
by de
tector
1 an
d
2 are
nam
ed
sep
a
rately a
s
L1
and
L2.
The di
stan
ces
betwe
en th
e
turntable
an
d
dete
c
tor 1 a
nd 2
are b
o
th defie
d a
s
a
.
The
schem
atic
circuit fo
r its
control co
re a
nd interfa
c
e d
r
ive modul
e is sh
own in Figure 2.
Figure 2. Sensor Sig
nal Tr
ansceive
r
Schematic
Diag
ram
In orde
r to
convenie
n
tly locate th
e ob
ject’s
2D
co
ordin
a
tes, thi
s
sy
stem ad
opts two
ultrasoni
c se
nso
r
s, i.e. de
tector 1 an
d 2 in
Figure 1. For accurat
e
ly locating a
moving obje
c
t,
multiple ultra
s
oni
c
sen
s
o
r
module
s
a
r
e typical
me
asu
r
em
ent d
e
vice
s in the
system [2]. This
system trie
s
to adopt ultraso
n
ic
sen
s
o
r
modul
es
to
conve
r
t the detectin
g
sig
nals of movi
ng
obje
c
t from multiple dire
ctions to the corres
pon
ding
distan
ce
s, and then empl
oy algorithm
s of
cal
c
ulatio
n of
point
s to
d
e
termin
e the
2-D
coo
r
di
n
a
tes
of a
m
o
ving o
b
je
ct, and
mea
n
while
determi
ne th
e be
havior
of a m
o
ving
obje
c
t in
real time. T
h
e two
ultrasonic sen
s
ors are
assembl
ed i
n
a vertical
angle di
re
ction, sho
w
n F
i
gure 1.
We
think abo
ut the feature
that
C4
3
2
1
8
4
U2
A
C2
1
2
EMI
T
R4
R3
C1
Q1
R1
R2
C3
R5
T1
VC
C
co
n
n
ect
o
r
VC
C
MC
U(
U
1
)
ZD
1
ZD
2
B0
A0
A1
B1
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 6, June 20
14: 4664 – 4
670
4666
ultrasoni
c se
nso
r
s a
r
e ab
le to receive
wave from a dire
ction to detect moving obje
c
ts, the
n
based on th
e positio
n da
ta of them, turntabl
e adju
s
ts its di
re
ction in ord
e
r t
o
kee
p
them
in
detecta
ble co
rne
r
in a ce
rtain dista
n
ce for tra
cki
ng.
The
Tra
n
smit
-Re
c
eive
sig
nals of
ultra
s
onic a
r
e
provided
by M
C
U and
a
n
i
s
ola
t
or T
1
,
sho
w
n i
n
Fig
u
re
2. A0, a
cod
ed di
gital
sign
al, is
sup
posed to
be t
r
an
smitted by
MCU an
d th
en
A1 is
achieve
d
by the
amp
lifier Q1. Su
b
s
eq
uently
, th
e si
gnal A
1
g
oes thro
ugh
the i
s
olato
r
T
1
,
and t
r
an
smits the a
m
plified
co
ded
si
gnal
A0 to t
r
an
sd
uce
r
con
n
e
c
ted with wirel
e
ss
inte
rface. B0
is the
sign
al
comin
g
from
the wirele
ss i
n
terfac
e an
d
is put into
M
CU
after run
n
ing throug
h
the
operational
amplifier
B
1
.
The dista
n
c
e of
obj
e
c
t
ca
n be
cal
c
ulate
d
de
pe
nding
upo
n
th
e
input/output signal delay ti
me.
This system’
s
softwa
r
e pro
g
ram
m
a
inly aims
at g
ene
rating the
cod
ed si
gnal
s in
spe
c
ific
time and
re
ce
iving the
wirel
e
ss
sign
als, t
hen j
udgin
g
wheth
e
r it i
s
t
he me
asured
sig
nal o
r
not. If
the answe
r is yes, the delay time between t
he outp
u
t and input sign
als shoul
d be figured
out,
and may be
multiplied by the micr
owave tran
smissio
n
spe
ed. Th
e
n
, the expect
ed dista
n
ce a
nd
its p
r
oje
c
tion
on th
e
coo
r
dinate
axis
of syst
e
m
can b
e
a
c
hi
e
v
ed. Based
on the
p
r
evi
ous
proje
c
tion
dat
a and th
e p
r
in
ciple
of inerti
a, the obje
c
t’
s ine
r
tia flo
w
will re
present
s its track [3]. By
usin
g a lo
cat
i
ng algo
rithm
,
its position
at the
next moment
can
be estim
a
ted
.
Therefo
r
e, the
turntable
will
be no
w turn
ed to the location m
entio
ned ab
ove in
orde
r to achieve the obj
ect
tracking
[4]. The
movin
g
obj
ect
sta
t
us
can
al
so
be
sa
mple
d in
re
al ti
me by th
e
other
transmissio
n device
s
an
d sen
s
o
r
s in th
e system.
If such m
o
ving o
b
ject ne
ed
s to be co
ntroll
ed,
the corre
s
p
o
n
d
ing si
gnal m
a
y be transmi
tted by
those comm
uni
cati
on devices in
system.
3. Realiz
ation Method
Among the fu
nction d
e
si
gn
ing progress
of tr
acking
sy
stem, there may be man
y
factors
having effe
ct on its tra
c
kin
g
perfo
rma
n
ce. Therefor
e,
the most im
p
o
rtant p
r
oble
m
s in thi
s
stu
d
y
are focused
on sele
cting
one pe
rfect a
u
tomatic-
tra
c
king meth
od and greatly redu
cing the
error
gene
rated by
the prog
ram.
These two p
r
obl
em
s will b
e
discu
s
sed b
e
low in d
e
tail.
3.1. Automa
tic Tracking
Metho
d
In Figu
re 1
mentione
d a
bove, the a
n
g
le is 90
°
,
the di
stan
ce
s from the
co
ordin
a
tes
cente
r
(0,0
) to these two detecto
rs
are
both
a. We
define the d
i
stan
ce
s bet
wee
n
su
ch t
w
o
detecto
rs a
n
d
the detected
obje
c
t as l1 a
nd l2
. Based
on the Pythagore
an Th
eorem, we will g
e
t:
2
1
2
2
2
2
2
2
)
(
)
(
l
x
a
y
l
a
x
y
.
(
1
)
From
(1
), we
may calcula
t
e the value
s
of
x and
y, i.e. the proje
c
tion valu
es
on the
coo
r
din
a
te ax
is of sy
stem. Obviou
sly, the sol
u
tion
of t
h
is formula i
s
quite comple
x. Howeve
r, if
l1 an
d l
2
ca
n be
me
asu
r
ed in
adva
n
ce, the
system
turnta
ble i
s
now rotate
d t
o
some
exten
t
so
that we’ll
get
2
1
l
l
. Based
on
th
e pri
n
ci
ple
of
con
g
ru
ent tri
angle
s
, the
o
b
ject i
s
l
o
cat
ed o
n
the
coo
r
din
a
te ax
is of
system
at this mo
me
nt, i.e.
the projectio
n
s
of the
obj
ect o
n
both the
axis x
and y are eq
u
i
valent. We m
a
y calculate x and y by:
2
2
1
2
2
2
2
2
2
2
)
(
)
(
l
l
x
a
x
or
l
l
a
y
y
.
(
2
)
The sol
u
tion
of (2) is exp
r
e
s
sed a
s
:
2
-
2
2
-
2
2
2
2
2
a
l
a
y
a
l
a
x
(
3
)
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Research of a New No
n-Layer Protocol
Architecture
fo
r Satellite Network (Lixi
a
ng Liu)
4667
Whe
n
l ha
s
two solution
s, their locatio
n
s a
r
e
re
spe
c
tively on th
e two p
o
ints whi
c
h a
r
e t
h
e
intersectio
n
p
o
ints on the circle
s an
d line
x
y
. Such circl
e
is defined by
cente
r
s of (a,
0) or (0,
a), radiu
s
of l. The minimu
m sum of dist
ances
bet
we
en the obje
c
t and the two d
e
tectors shall
be
vertical li
ne
s
from two d
e
tection
poi
nts
to line
x
y
. Acco
rding
to
right
triangle
cha
r
a
c
teri
stics,
those
verti
c
al
line
s
shall
b
e
the
conn
ection
line
bet
ween dete
c
tor 1
an
d dete
c
tor 2,
sho
w
n
in
Figure 3.
Figure 3. The
Shortest Di
st
anc
e amon
g the Tra
c
king
Process
The minimu
m length of l is
a
2
2
, i.e.
a
45
sin
. We’
also get
2
a
l
, i.e. x and y
can
not have i
m
agina
ry sol
u
tions.
Whi
c
h
solutio
n
is
m
o
re
suitabl
e
betwe
en
su
ch two
solutio
n
s?
Gene
rally, it is req
u
ired to analyze its st
atus of t
he previous
se
con
d
in orde
r to sele
ct a suita
b
le
solutio
n
ba
se
d on the p
r
e
v
ious tra
c
k [4]. For the p
u
rpo
s
e
of this sy
stem ru
nning fa
st a
nd
meeting th
e functio
n
of re
al-time tracki
ng movi
n
g
o
b
ject, the t
r
a
cki
ng
step
s
are
de
signe
d
as
follows: the t
u
rntabl
e
rotat
i
on i
s
controll
ed by
mean
s of me
asurin
g
l
1
an
d
l
2
so
that the t
r
a
c
ked
obje
c
t is loca
ted on the p
e
rpe
ndi
cula
r
bise
ctor
of conne
ction lin
e betwe
en th
e two dete
c
t
o
rs’
points, i.e.
l
1
=
l
2
. Controll
e
d
by the mecha
n
ical driv
e system, the system
mo
ves forward and
backward in orde
r to co
ntinuou
sly ke
ep
the sho
r
test
distan
ce fro
m
the object, which me
ets:
2
2
1
a
l
l
l
.
(
4
)
Duri
ng the p
e
riod
of syste
m
moving, the angle
of turntable rotatio
n
and movin
g
statu
s
are
re
corded
and ta
ken
on
as e
s
timatio
n
po
sition val
ue for th
e ne
xt test. Then, let the syste
m
kee
p
s trackin
g
and the val
ues of
l
1
and
l
2
may be arri
ved at time t. The sy
stem’
s
motion
can
be
adju
s
ted by the step
s liste
d above.
Con
s
id
erin
g
of the limitati
on of ult
r
a
s
o
n
ic d
e
tectio
n distan
ce, we do
sele
ct
the
trackin
g
method of ke
eping
sho
r
test distance. The detecta
ble range of ultra
s
onic
sen
s
o
r
mainly depe
n
d
s
on
its wave
length
a
nd freque
ncy. As
the wa
v
e
length
be
comes lon
ger, its freq
ue
ncy
corre
s
p
ondin
g
ly become
s
smalle
r, an
d its dete
c
table
rang
e be
co
mes la
rg
er. F
o
r exam
ple, the
detecta
ble ra
nge of a co
m
pact sen
s
o
r
’s with
millimeter-scale i
s
ab
out 300
~50
0
mm, and the
one
of the
sen
s
o
r
with
wavele
ngths greate
r
t
han
5mm
may be
up t
o
8m [5].
When th
e
sele
cted
ultrasoni
c’s frequ
en
cy is
up to
40KHz, its wavele
n
g
th is ab
out
8.5mm
(i.e. the velo
city of
ultrasoni
c (34
000m/s) i
s
di
vided by fre
q
uen
cy (4
000
0
H
z)). T
heo
ret
i
cally, its me
a
s
uri
ng
dista
n
c
e
sho
u
ld b
e
le
ss tha
n
5m. A
c
cordi
ng to
(4
), wh
en
we
d
e
fine the val
u
e of a
as 1m,
the sy
stem m
a
y
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 6, June 20
14: 4664 – 4
670
4668
kee
p
t
he sh
ort
e
st
di
st
an
ce
2
1
I
, less than 1m. According to the maximum de
tectable
distan
ce of 4
m
, it can be
measured wh
en the obj
ec
ts
mo
ve
fo
rw
ar
d
or
b
a
c
k
w
a
r
d
w
i
th
a
sp
ee
d
less than 3m/
s
.
3.2. Errors G
e
nera
ted b
y
the Progr
am
Gene
rally, ultrasoni
c sen
s
ors are kno
w
n a
s
tran
sceivers
wh
en
they both se
nd an
d
receive. The
s
e sen
s
o
r
s l
aun
ch a
series of
ultra
s
onic b
eam
at a certai
n
time, and the
corre
s
p
ondin
g
re
ceive
r
wi
ll start
a
s
so
on a
s
th
e
se
nding
op
erati
on e
n
d
s
. Su
ch
se
ndin
g
a
nd
receiving op
e
r
ation
s
are
co
mpleted un
de
r cont
ro
l of the microcomp
u
ter
program.
Therefo
r
e, the
prog
ram o
p
timized m
a
y affect the mea
s
urem
ent accu
racy.
Becau
s
e l
a
n
guag
e C m
a
y prod
uce quit
e
long
cod
e
s,
it is not re
co
mmend
ed
wh
en test
cod
e
a
r
e d
e
s
ign
ed. Inste
ad of C lang
uage, a
s
sem
b
ly langu
age
is ad
opted.
It may dire
ctly
operate o
n
pi
ns,
su
ch
as
cl
r Px.x and
se
tb Px.x.
The time they
spe
nd i
s
up
to o
n
l
y one m
a
chi
ne
cycle, i.e. 12 times of the crystal oscillato
r cycle,
which can be preci
s
ely cal
c
ul
ated.
In the design
of receivin
g p
r
og
ram, jnb Px
.x command
s are al
so recommen
ded. It takes
pre
c
ise time-two ma
chin
e
cycle
s
. Interrupt meth
od
is not
sele
ct
ed ju
st be
ca
use th
e inte
rrupt
respon
se tim
e
rem
a
in
s un
certai
n. The i
n
terrupti
on re
spo
n
se
time is
un
ce
rtain. The
rea
s
on of
this is d
ue to
sampli
ng op
eration
perfo
rmed every m
a
chi
ne cy
cle
and then
su
ch interruption
sampl
ed
que
ried i
n
the n
e
xt machi
ne
cycle. If t
here exists an i
n
terrupt
flag,
the sy
stem
will
prod
uce a lo
ng call inst
ru
ction (LCA
LL
double
cycl
es),
whi
c
h m
a
y control CPU to ru
n th
e
corre
s
p
ondin
g
interrupt service
pro
g
ra
m. Howe
ver,
this call m
a
y be blocke
d
by one of three
ca
se
s li
sted
belo
w
: an
eq
uivalent o
r
hi
gher p
r
iori
ty
prog
ram
is b
e
ing
pro
g
re
ssed; th
e
current
(que
ry)
cycle
is not the l
a
test on
e in
whi
c
h t
he
cu
rre
nt instructi
on is
executed; the curre
n
t
instru
ction
i
s
the exa
c
t return
(RETI)
co
mmand
o
r
i
s
to re
ad
and
write
in
stru
ctions on
IE an
d IP
regi
sters (Ex
e
cutin
g
anot
her in
structio
n is ne
ce
ssary). If the blockin
g
co
n
d
ition ha
s b
een
can
c
ell
ed an
d the interrup
t flag still exists, then the
interru
pt servi
c
e prog
ram
wil
l
be activated
.
In
interruption p
r
ocess, it takes ab
out 3-8
machi
ne cy
cl
es from
starti
ng the sig
nal
to executing t
h
e
first instruction, whi
c
h will
brin
g
un
co
rre
c
table
erro
rs
[6]. Here, we
take
an
exa
m
ple fo
r
rang
ing
prog
ram,
sho
w
n in Figu
re
4.
unsi
gne
d int distan
ce
()
{ ... .
.
/
/
timer initializ
ation
#pragma as
m
/
/
in
s
e
r
t
a
s
se
mb
le
pr
ogr
a
m
c
l
r P1.0
// mic
r
ocomputer
pin o
peratio
n, sen
d
ultrasoni
c d
e
tecting p
u
lse sign
al
nop
s
e
tb P1.0
Nop
… ..
// sendin
g
the requi
re
d ultrasoni
c coded
sign
al
c
l
r P1.0
/
/
the ultr
as
onic
detec
t
ing s
i
gnal ends
s
e
tb TR0
/
/
ac
tivati
ng the timer to s
t
art work
jnb P1.1,$
// determine wh
ether th
e ultrasoni
c resp
on
se si
gn
al is re
ceived
or not
c
l
r TR0
//
freez
ing the timer
mov R0,TH0
/
/
dealing with the timing value
mov R1,TL0
#pra
gam e
n
d
a
sm // end of the assembl
e
r
....
..
//
c
a
lc
ulat
e the dis
t
ance
return l;
/
/
return to the detec
ted value
}
Figure 4. The
Shortest Di
st
anc
e amon
g the Tra
c
king
Process
3.3. Ultrasonic Wav
e
Sen
s
or Controll
er
Ultra
s
oni
c is
referred to a mech
ani
cal
wave with a freque
ncy greater than th
e human
heari
ng
ran
g
e
, i.e. app
rox
i
mately 20
kHz. Th
e
comm
on ultrasoni
c
freque
nci
e
s a
r
e from d
o
ze
ns
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Research of a New No
n-Layer Protocol
Architecture
fo
r Satellite Network (Lixi
a
ng Liu)
4669
of kHz to M
H
z.
The
co
n
t
rol pa
rt i
s
mainly u
s
ed
to dete
r
min
e
tho
s
e thi
n
gs
se
nt from
the
transmitter, such
as pul
se
chai
n fre
quen
cy, duty ra
tio,
spa
r
se m
odu
lation, co
unti
ng an
d di
stan
ce
detectio
n
. In orde
r to achi
eve a better pre
c
isi
on in a
wide ra
nge,
we may ado
p
t
two ultraso
n
ic
pulses
with
d
i
fferent freq
u
enci
e
s
se
nt a
l
together, i
n
whi
c
h the
on
e with
a g
r
ea
ter freq
uen
cy
is
use
d
to measure the ne
ar t
a
rget, whil
e the one
with a
less freque
n
c
y is used for the far one [4].
In this trackin
g
system, we
adopt two ul
tr
aso
n
ic
wav
e
sen
s
o
r
s wit
h
50kHz and
80KHz
freque
nci
e
s,
betwe
en whi
c
h the o
ne a
c
ts a
s
the tr
a
n
smitter a
nd
the other a
s
the re
ceiver.
In
actual
process, we sele
ct the
ultrasoni
c
module
integ
r
ated
co
ntrol
with dete
c
tin
g
fun
c
tion, well
comp
atible wi
th the micro
c
ompute
r
. The
micro
c
om
put
er provide
s
two serie
s
(3-5) pul
se
s trig
ger
sign
al wh
ose
width is 1
0
u
s
(hi
gh an
d low leve
l a
r
e
both 10u
s) a
nd anoth
e
r
width is 6u
s (b
oth
high
and
lo
w level
a
r
e
6us), i.e. th
e fre
que
ncie
s a
r
e
6
10
2
10
1
(ab
o
ut 50K
Hz) a
n
d
6
10
2
6
1
(About 80K
Hz).
After
e
a
ch
seri
es
pul
se has be
en sent, the correspon
ding timer
sho
u
ld be im
mediately act
i
vated (setb
TR0
)
. Once
the re
spo
n
se sign
al is dete
c
ted (jn
b
P1.1,$),
then the tim
e
r i
s
fro
z
e
n
. By measuri
ng the
difference in time
betwe
en th
e pul
se b
e
in
g
transmitted from the parti
cula
r directio
n and t
he correspon
ding
echo b
e
ing
received, it is
possibl
e to ca
lculate the ex
pecte
d dista
n
c
e by:
T
v
607
.
0
5
.
331
(
5
)
Whe
r
e,
v repre
s
e
n
ts the velocity of ultrasoni
c wa
ve in the air;
T re
presents the
enviro
n
m
ent
temp
erature. Its val
ue
ca
n
be
measured du
ring
the
initialization or spare time.
Therefore, th
e expecte
d di
stan
ce l ca
n be cal
c
ul
ated
by:
2
0
1
t
t
v
l
(
6
)
Whe
r
e,
l represents d
e
tected di
sta
n
ce;
t1 represents
ultrasoni
c re
spon
se re
ceiv
ed time(p
rovi
ded by timer),
and
t0 represents
ultrasoni
c pul
se tran
smi
s
si
on time (set this value a
s
0
)
.
Usi
ng M
CU
capture fe
ature, we
can ea
sily rea
d
the timer value
--t
1, base
d
on e
quation
s
above, a soft
ware progra
m
can be d
e
signed
a
nd the
distan
ce l ca
n be cal
c
ul
ated.
4. Conclusio
n
The metho
d
prop
osed in t
h
is sy
stem is suit
able fo
r trackin
g
the o
b
ject
s slo
w
ly moving
on the
surfa
c
e. Whe
n
tra
c
ked
obje
c
ts
moving faste
r
, Doppl
er
sig
nal problem
s sho
u
ld b
e
to
ok
into con
s
ide
r
ation. When i
t
is used to re
alize a
ste
r
eo
tracki
ng, the cal
c
ulatio
n wi
ll becom
e mo
re
compl
e
x, and
the result is not good en
o
ugh. Ther
e is a problem d
u
ring tra
c
king
progress: wh
en
a new obj
ect
inse
rted bet
ween the sy
ste
m
and the or
i
g
inal tra
c
ked
obje
c
ts, the system will turn
to tra
c
k the
o
b
ject i
n
serted
. Lack
of ima
ge
colle
ction
and
analy
s
is
equipm
ent, h
o
w to
lo
ck th
e
obje
c
ts tracked exa
c
tly ha
s be
ca
me a
challen
ge.
We
have attemp
ted to acqui
re features
of the
obje
c
ts, or to
use i
nertia
p
r
inci
ple to a
n
a
lyze t
he track of dete
c
ted
obje
c
ts, an
d then to filter
out
method
s for trackin
g
o
b
je
cts. Such
met
hod
s have
so
me of effec
t
s
,
but fail to
s
o
lve it c
o
mpletely
.
In the further research,
we’ll
seek for the
se
nsors, which
can acquire the characteristi
cs
informatio
n, in orde
r to improve pe
rformance of
tra
cki
ng sy
stem
and achi
eve reliabl
e feature to
track spe
c
ific
moving obje
c
ts.
Ackn
o
w
l
e
dg
ements
This
wo
rk
wa
s finan
cially
suppo
rted by t
he Pr
in
cip
a
l
Found
ation o
f
Nanjin
g Inst
itute of
Industry Te
ch
nology, proj
e
c
t No. CX1
2
-SCXJ-04.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 6, June 20
14: 4664 – 4
670
4670
Referen
ces
[1]
Li Maos
ha
n. T
he pri
n
ci
ple
a
nd tech
nol
og
y
practice of u
l
trasonic d
i
stan
ce measur
em
ent.
Practical
me
asur
e
m
ent techn
o
lo
gh
. 19
94; 1: 12-2
0
.
[2]
W
A
NG Runtia
n. Dou
b
le
freq
uenc
y
ultras
ou
nd d
i
stanc
e m
easur
ement.
A
c
oustic tech
ni
q
u
e
. 19
96; 3:
116-
118.
[3]
He Yin-
nia
n
, Li Kai-ta
i. Hu
ang a
i
-
x
i
ang.
T
he
stabil
i
t
y
and co
nver
g
ence of i
nerti
al al
gorit
hm.
Mathem
atic
a numerica sinic
a
. 1998; 2
0
(3): 2
39-2
50.
[4]
LIU Yu
e, W
A
NG Yon
g
-tian,
HU
Xiao-m
i
n
g
. Stud
y
on th
e 3td
of tracki
ng
alg
o
rithm t
o
me
asure
th
e
motion of the ri
gid b
o
d
y
.
Co
mputer
meas
ure
m
e
n
t & control
.
2002; 1
0
(6): 3
63-3
65.
[5]
T
eng Yanfei
Che
n
Sha
ngs
ong. Stud
y
o
n
precis
ion o
f
ultrasonic r
ang
ing.
F
o
re
i
gn el
ectroni
c
me
asur
e
m
ent techn
o
lo
gy
. 200
6; 25(2): 22-2
5
.
[6]
Z
hou Yuca
i. MCS51. Sout
hea
st Universit
y
Pr
ess. Nanj
ing: 2
002.
Evaluation Warning : The document was created with Spire.PDF for Python.