TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol.12, No.7, July 201
4, pp
. 5552 ~ 55
5
8
DOI: 10.115
9
1
/telkomni
ka.
v
12i7.407
9
5552
Re
cei
v
ed Au
gust 5, 201
3; Re
vised Feb
r
uar
y 17, 201
4
;
Accepte
d
March 5, 2014
Measur
e
ment Method of the Grain Quantity Based on
The Ground Pressure
Fan Ch
ao*, Zhang Dexian
, Yang Tiejun, Fu Hongliang
Schoo
l of Information Sci
enc
e and En
gi
neer
ing,
He
na
n Uni
v
ersit
y
of T
e
chnol
og
y,
Z
hengz
ho
u, 45
000
1
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: anfan2
00
3@
gmail.c
o
m
A
b
st
r
a
ct
T
o
meas
ure th
e stored gra
i
n
qua
ntity accura
tely
and rel
i
a
b
l
y
, the meas
ure
m
e
n
t meth
od b
a
sed o
n
the gro
und
pr
essure is
put forw
ard. Accordin
g to
the ra
ndo
mness of t
he gr
anary
pr
essure
distrib
u
t
ion
cause
d
by
the
l
i
mited fl
ui
dity o
f
the gra
i
n, th
e
layo
ut
of the
pr
essure
on
the
grou
nd
is set
u
p
, an
d the
mea
n
pressur
e
o
n
th
e gr
oun
d
is us
ed to
re
prese
n
t
the w
h
o
l
e
gr
oun
d
pressur
e
. At the s
a
me
ti
me, t
he
press
u
re
distrib
u
tion
on
the w
a
ll is
an
al
y
z
e
d
a
nd th
e c
o
mpe
n
sa
tio
n
meth
od
is giv
e
n, w
h
ich can r
educ
e the s
e
n
s
ors
used i
n
this me
thod an
d the cost is
low
.
Lastly, the estimati
on mod
e
l
of the stored grai
n qua
ntity base
d
o
n
the gr
ou
nd
pre
ssure is
p
u
t for
w
ard, to i
m
pr
o
v
e the
pr
edicti
on
accuracy
of
the
grai
n w
e
i
g
ht, the p
a
ra
meters
are esti
mated
base
d
o
n
the
r
a
tio of th
e err
o
r
.
T
he ex
p
e
ri
me
nt results s
how
n that, the
mea
s
ure
m
e
n
t error
i
s
less than
3%
by usin
g this
meth
od,
w
h
ich
can meet the
actual n
eed
o
f
real-time on
li
ne
mon
i
torin
g
the
natio
nal
grai
n storage q
u
a
n
tity
and distri
buti
o
n effectively.
Ke
y
w
ords
: gra
i
n qu
antity, gro
und pr
essur
e
, pred
iction
mo
d
e
l, para
m
eter e
s
timati
on
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
The g
r
ain
is the imp
o
rtant
strategi
c
re
so
urce
of
our country, which
is th
e foun
d
a
tion of
the natio
nal
eco
nomy.
Whose q
uality and
qu
ant
ity is
related
with the
dev
elopme
n
t of
the
eco
nomy a
n
d
the sta
b
ilization of the
so
ciety dire
ctly. But duri
ng t
he g
r
ain
sto
r
age, be
ca
use
of
the temperature, moi
s
ture
and som
e
other ma
n-m
a
d
e
factors, the
grain qu
antity will be lossed,
whi
c
h will d
a
m
age the g
r
a
i
n safety [1, 2]. Thus
, ho
w to learn the quantity ch
ange of the
grain
accurately re
al time is the key pro
b
lem t
o
be solve
d
.
To mea
s
ure the grai
n stora
ge quantity re
al
time, many methods hav
e been u
s
ed, but all
these
method
can’t m
e
a
s
ure the gai
n we
ight dire
ce
tly, the re
sult
s is not accu
rate
and reliabl
e [3-
5]. Thu
s
, the
method
whi
c
h m
ountin
g
som
e
p
r
e
ssure
se
nsors
on the
groun
d an
d the
wall
according to
some
reg
u
lari
ty is propo
se
d in this pap
e
r
.
2. The Meas
urement Pri
n
ciple of th
e
Grain Quan
tit
y
Becau
s
e of the flat grana
ry is mostly often
used, thu
s
the pre
s
sure distrib
u
tion of which
is re
sea
r
ched
. In order to measure the quantity of
the grain in the
barn, the pre
s
sure tran
sdu
c
e
r
is in
stalled
o
n
the g
r
ou
nd
and
wall
of the ba
rn
. Th
e
sensors a
r
e
e
v
en mou
n
ted
on the
groun
d,
and the n
u
m
ber i
s
N
1
. On
the wall, the
sen
s
o
r
s
are
installe
d with
equal inte
rv
al from botto
m to
top. Suppo
se
d that the hei
ght of
the gra
i
n is h, the in
terval betwee
n
the adja
c
e
n
t sen
s
o
r
s i
s
α
,
thus, the leng
th and width
of the barn i
s
l and w,
the friction coeffici
ent betwe
en the wall a
nd t
he
grain
i
s
μ
, a
c
cording
to th
ose
pa
ram
e
ters,
the
gross q
uantity of
the g
r
ain
in t
he b
a
rn
can
be
cal
c
ulate
d
:
G=G
1
+G
2
(1)
Whe
r
e G
1
is t
he grai
n wei
g
ht presse
d on
the groun
d, G
2
is the grai
n weig
ht pre
s
sed o
n
the wall.
A
q
w
l
q
N
G
N
i
i
_
1
1
1
1
1
(2)
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Measurem
ent Method of the Grain
Quan
tity
Based on
The Groun
d Pressu
re (Fa
n
Cha
o
)
5553
Whe
r
e,
q
i
i
s
t
he me
asure
m
ent value
o
f
the pre
s
su
re se
nsor i,
q
is the mea
n
va
lue of all th
ese
pre
s
sure se
n
s
ors on the g
r
ound, A is the
basal a
r
ea of
barn.
2
1
2
N
j
j
q
al
G
(3)
Whe
r
e,
q
j
is
the mea
s
ure
m
ent value o
f
the pre
s
sure sen
s
o
r
j m
ounted o
n
th
e wall,
α
i
s
the
interval between the adj
acent sen
s
o
r
s, t
hus:
2
1
1
1
1
1
N
j
j
N
i
i
q
al
w
l
q
N
G
(4)
Based
on ab
ove equatio
n
,
if the pressure valu
e
q
i
,
q
j
is
me
as
ur
e
d
r
e
sp
ec
tive
ly, th
e
gro
s
s qua
ntity of the grai
n G in the
b
a
rn
can
be
computed. T
h
us, the
key p
r
oble
m
is
ho
w to
measure the pre
s
sure valu
e of all these
sen
s
o
r
s.
3. The Distri
bution Char
acte
r
istic of
the Pres
sure
on the Grou
nd and Wall
To re
sea
r
ch the distri
butio
n cha
r
a
c
teri
st
ic of
the pre
s
sure of the ba
rn, the experi
m
ent is
done at the Li
aochen
g grai
n depot in Sh
ando
ng provin
ce, the p
r
essure dist
ributi
on is an
alyze
d
based on the
experim
ent re
sults in the fol
l
owin
g.
3.1. The La
y
out of the Pr
essur
e
Sens
or
The si
ze of th
e barn a
nd th
e pre
s
sure po
sition mou
n
te
d on the gro
u
nd and
wall a
r
e
sho
w
n in Fig
u
re 1, the zon
i
ng of the sen
s
or i
s
listed in
Table 1.
Table 1. The
Zoning of the
Pressu
re
zone
Number of
the pr
essure
B1
B2
B3
B4
BC
WL
WR
1-7
9-18
20-30
31-37
11
、、、、
、、、
13
15
17
22
24
26
28
45-52
8
、
53-55
、
57
-60
(a) Size and t
he se
nsors p
o
iso
n
on the
grou
nd
(b) Size and t
he se
nsors p
o
iso
n
on the
wall
Figure 1. The
Size and the
Pressu
re Sen
s
ors Poi
s
on o
f
the Barn
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 7, July 201
4: 5552 – 55
58
5554
3.2. The Rela
tionship be
tw
e
e
n
the Gr
ain Weigh
t
a
nd Ground P
r
essur
e
The a
c
curacy of the pre
s
sure sen
s
o
r
i
s
ab
out 0.5
%
, and the
measurement
rang
e is
100KPa, the relation
ship
s b
e
twee
n the o
u
tput of pre
s
sure
sen
s
o
r
s i
n
every zo
ne
and the loa
d
e
d
grain
weig
ht are sh
own in Figure 2,
we can
se
e that, the output of the pressure se
nsors
increa
sing
wi
th the grain
quantity raisi
ng, becau
se
of the limited liquidity of the grain, the
pre
s
sure on the gro
und is
uneven, and
the influenc
e
of the wall friction is no
n-uniform, whi
c
h
indu
ce the ou
tput of the senso
r
is differe
nce a
nd ra
nd
om.
Based o
n
ab
ove re
sults, the mea
s
u
r
e
m
ent
method
of the grain
quantity by using the
mean
val
ue of
the sen
s
o
r
s mou
n
ted
i
n
some sp
e
c
ial zo
ne
i
s
put
forwa
r
d, the
rel
a
tion
ship
betwe
en the
mean val
ue o
f
sen
s
o
r
s i
n
t
h
ree
middl
e a
r
ea
s of the
ground
and
the
grai
n weight
is
s
h
ow
n in
F
i
gu
r
e
3
.
O
b
vious
ly, c
o
mp
ar
ed
w
i
th th
e
z
o
n
e
B2
a
n
d
B3
, th
e
z
o
ne
BC
is
fa
r
t
he
r fro
m
the wall, who
s
e line
a
rity b
e
twee
n the p
r
essu
re a
nd
grain
wei
ght is better. Th
u
s
, to mea
s
ure
the
grain q
uantity accurate, the
sen
s
ors a
r
e
better to be
mounted at th
e zon
e
BC.
0
1
00
200
3
0
0
400
500
600
700
0
5
10
15
20
25
30
35
40
Se
nso
r
v
a
l
ue
(
KP
a
)
We
i
g
h
t
(
T
)
B1
B2
B3
B4
B6
B7
(a) T
he sensors o
u
tput of zon
e
B1
0
100
200
300
400
500
600
700
0
10
20
30
40
50
S
e
n
s
or
v
a
lu
e
(
K
P
a
)
W
e
i
ght(T)
B9
B1
0
B1
1
B1
2
B1
3
B1
4
B1
5
B1
6
B1
7
B1
8
(b) T
he sensors o
u
tput of zon
e
B2
0
1
0
0
20
0
3
00
4
0
0
5
0
0
60
0
7
00
0
5
10
15
20
25
30
35
40
45
50
55
60
Sen
s
or va
lu
e(KPA)
W
e
i
ght
(
T
)
B20
B22
B23
B24
B25
B26
B27
B28
B29
B30
(c) The
sen
s
o
r
s outp
u
t of zone B3
0
100
200
300
400
500
600
700
0
5
10
15
20
25
30
35
40
45
Se
ns
or v
a
lu
e (KP
a
)
We
i
g
h
t
(
T
)
B19
B31
B33
B34
B35
B36
B37
(d) T
he sensors o
u
tput of zon
e
B4
Figure 2. The
Relation
ship
betwe
en the
Pressu
re
Sen
s
ors of every
Zone an
d the
Grain
Weight
0
1
0
0
2
0
0
3
00
40
0
5
00
60
0
7
0
0
0
10
20
30
40
50
Sensor
v
a
l
ue(
KP
A)
W
e
i
ght
(
T
)
M
eanB
2
M
e
anB
3
M
eanB
C
Figure 3. The Relatio
n
ship betwe
en th
e Mean Va
lu
e of Senso
r
s i
n
Thre
e Middl
e Area
s of the
Grou
nd an
d the Grai
n Wei
ght
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Measurem
ent Method of the Grain
Quan
tity
Based on
The Groun
d Pressu
re (Fa
n
Cha
o
)
5555
3.3. The Rela
tionship be
tw
e
e
n
the Gr
ain Weigh
t
a
nd Wall Pres
sure
To mea
s
ure the pre
s
sure distrib
u
tion of
t
he grain on
the wall, two colum
n
s
sen
s
ors a
r
e
mounted
on
the
wall
at eq
ual inte
rval, which
is
sho
w
n
in Fig
u
re1(b). The
experim
ent rel
a
tion
sh
ip
betwe
en
th
e grain
weight
and
th
e wall
pre
s
sure
of
zone WR and
WL
i
s
sho
w
n in
Fig
u
re
4,
f
r
om
this figu
re,
we can
se
e th
at: the outp
u
t
of the p
r
e
s
sure
sen
s
o
r
on the
wall i
s
diffe
rent
a
n
d
rand
om b
e
ca
use
of the li
mited fluidity of the gr
ain.
And with th
e
raise of the
g
r
ain h
e
ight, t
h
e
friction
of
wal
l
increa
se
s o
b
viously, the
grou
nd
pr
essure
is affecte
d
serio
u
sly.
Thus,
when
the
grain h
e
ight i
s
high
er, the influen
ce of the friction force must be
co
nsid
ere
d
.
0
100
200
300
400
500
600
700
-1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
Se
ns
or val
u
e
(K
Pa
)
W
e
i
ght
(
T
)
B8
B5
3
B5
4
B5
5
B5
7
B5
8
B5
9
B6
0
(a) T
he sensors o
u
tput of zon
e
WR
0
1
00
20
0
3
0
0
400
500
600
700
-1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
S
e
nsor val
ue (K
Pa)
We
i
g
h
t
(
T
)
B45
B46
B47
B48
B49
B50
B51
B52
(b) T
he sensors o
u
tput of zon
e
WL
Figure 4. The
Relation
ship
betwe
en the
Wall Pre
s
sure and the G
r
a
i
n Weig
ht
The friction fo
rce of the wal
l
can be cal
c
ul
ated by usi
ng the layer-by-layer meth
od whe
n
the sen
s
o
r
s mounted
on the wall
as Fi
gure
1
(b). Su
ppo
sed that t
he se
nsor h
e
i
ght interval i
s
h,
the mea
n
val
ue
)
(
j
q
ca
n b
e
computed
at e
v
ery layer. F
o
r exa
m
ple, i
n
the Fi
gu
re
1(b
)
, the
1
6
sen
s
o
r
s are
divided into t
w
o
colu
mn
s, each colu
m
n
has 8
sen
s
o
r
s,
whi
c
h
co
rrespon
ds to the
z
o
ne
W
R
a
nd W
L
, th
us
th
er
e
a
r
e two
s
e
n
s
or
s
s
j
1
an
d
s
j
2
at th
e laye
r j, the
me
an
value of
layer j
can
be calcul
ated a
s
2
/
))
(
)
(
(
)
(
2
1
j
q
j
q
j
q
and the mea
n
frict
i
on force of e
v
ery layer
ca
n be
comp
uted:
)
(
ˆ
2
j
q
hC
G
B
j
(5)
Whe
r
e, C
B
is the pe
rimet
e
r of th
e ba
rn, h is
the
h
e
ight inte
rval
of the
sen
s
or,
is
fric
tion
c
oeffic
i
ent. Thus
, the fric
tion force is
:
)
)
(
)
(
(
2
ˆ
2
1
2
j
j
j
j
B
S
q
S
q
hC
G
(6)
The relatio
n
ship betwe
en the output of the wall
pressure sen
s
ors a
nd the grain
weig
ht is
sho
w
n
in
Fig
u
re
5,
whe
r
e
the SumL
an
d Sum
R
i
s
th
e sum
of the
pre
s
sure valu
es
of all
the l
e
ft
and rig
h
t col
u
mn sen
s
ors respe
c
tively, MeanSum
d
enote
s
the m
ean value of
the both, we
can
see
that, the
re
gula
r
ity of the Sum
L
a
nd Sum
R
i
s
simila
r, the
relation
ship
b
e
twee
n the
wall
pre
s
sure and
the grai
n wei
ght is no
nline
a
r. But the su
m of the all p
r
essu
re o
n
th
e wall i
s
linea
r
with the sq
ua
re of the mea
n
pre
s
sure of the grou
nd, that is:
2
2
q
G
(7)
Whe
r
e th
e
q
is the
mea
n
p
r
essu
re
of
zo
ne B3, the
re
lationship b
e
twee
n the
bot
h is sho
w
n in
Figure 6, whi
c
h me
an
s that, the influence of
the
wall pre
s
sure can b
e
comp
ensated by the
grou
nd
pre
ssure, thu
s
th
e
sen
s
o
r
s on
the
wall
can
b
e
re
moved, a
nd the
grain
quantity ca
n
be
measured onl
y by mounting sen
s
o
r
s o
n
the grou
nd.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 7, July 201
4: 5552 – 55
58
5556
0
100
200
300
400
500
600
70
0
0
10
20
30
40
50
60
Wa
l
l
Pre
s
s
u
re
(
KPA
)
We
i
g
h
t
(
T
)
Sum
L
Sum
R
M
eanSum
0
500
10
00
15
00
20
00
0
10
20
30
40
50
Me
a
n
Su
m
(
M
eanB
3
)
2
Figure 5. The
Relation
ship
betwe
en the
Grain
Weig
ht and the Wall Pressure
Figure 6. The
Relation
ship
betwe
en the
MeanSum a
n
d
(Mea
n B3)
2
4. Parameter
Estimation of the G
r
ain
Weigh
t
Mod
e
l Based o
n
the Ratio of
Error
Acco
rdi
ng to
the Figu
re
2 and
Figu
re
3, we
kno
w
that, to mea
s
ure the
grai
n wei
gh
accurately, th
e sen
s
o
r
s on
the
gro
und
are
better to
be m
ounte
d
i
n
the
zone
B3 an
d
zon
e
BC,
thus, the sen
s
ors 20, 22, 24,
26, 28 and 30 are only
used to mea
s
ure the gro
u
nd pre
s
sure, the
mean value
q
of these six sensors ca
n b
e
comp
uted:
6
6
1
i
q
q
(8)
Whe
r
e the
i
q
qi
denotes the
pressu
re val
ue of thes
e six senso
r
s. Base
d on the Figure 6 and
Equation (7
), the
total
wall pre
s
sure
ca
n
be exp
r
e
s
se
d by the
sq
u
a
re
of the
q
, thus
, acc
o
rdin
g
to the Equation (1
), the grain
wei
ght G can b
e
estim
a
ted:
)
(
ˆ
2
2
1
0
q
a
q
a
a
A
G
(9)
Whe
r
e th
e A
is the
a
r
ea
of the b
a
rn, A=l×w, l
an
d w i
s
th
e l
ength
and
wi
dth of the
b
a
rn
r
e
spec
tively,
0
a
,
1
a
and
2
a
are th
e paramete
r
s of this mo
del
.
Thus,
the key
probl
em which will
be re
solved i
s
ho
w to get the paramete
r
s.
The re
gre
s
si
on is the co
nventional m
e
thod to esti
mate the pa
ramete
r [6, 7], th
e
optimize
d
target function o
f
which i
s
:
n
i
i
i
G
G
E
2
)
ˆ
(
(10)
But this met
hod
will indu
ce the
estim
a
tion erro
r
is large
wh
en
the grai
n wei
ght is
small.
To
resolve this p
r
oble
m
, the param
eter e
s
timation meth
o
d
based on th
e ratio of erro
r is put forward,
whi
c
h is:
n
i
i
i
i
G
G
G
E
1
2
ˆ
(11)
Whe
r
e
E is the ratio of error, the
i
G
is the actual loa
ded
grain wei
ght for the sampl
e
i,
i
G
ˆ
is the
predi
cted
wei
ght, n is th
e
sampl
e
nu
mb
er for mod
e
l
estimation. T
he Equ
a
tion (11)
ca
n en
su
re
that the opti
m
ization ta
rg
et of the mod
e
l is
con
s
i
s
te
nt with the
re
quire
ment of
the grai
n qu
a
n
tity
measurement
, and the pre
d
iction a
c
cu
racy ca
n be i
m
prove
d
.Sub
stitute the Equation (9) int
o
the
Equation (11), we can g
e
t:
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Measurem
ent Method of the Grain
Quan
tity
Based on
The Groun
d Pressu
re (Fa
n
Cha
o
)
5557
n
i
i
i
i
i
G
q
a
q
a
a
G
E
1
2
2
2
1
0
(12)
For optimi
z
ed
model de
scri
bed in ab
ove equatio
n,
to make the E is lest, the para
m
eters of
0
a
,
1
a
and
2
a
can b
e
cal
c
ulate
d
as:
B
A
1
a
(13)
Whe
r
e,
2
1
0
.
a
a
a
a
,
ii
i
i
i
i
i
i
i
ii
i
i
i
i
i
i
i
ii
i
i
i
i
i
i
G
q
G
q
G
q
G
q
G
q
G
q
G
q
G
q
G
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
1
2
4
2
3
2
2
2
3
2
2
2
2
2
2
2
Α
,
i
i
i
i
i
i
i
i
G
q
G
q
G
2
1
B
i
q
is the mean v
a
lue of the ground p
r
e
s
sure for t
he sam
p
le i, which is compute
d
as Equation
(8), the
2
i
q
,
3
i
q
and
4
i
q
is the squa
re, cubic a
nd qu
ar
tic
of which res
p
ec
tively.
5. Experiment and Analy
s
is
The ba
rn 45#
of the Liaocheng g
r
ain d
epot is
used to verify this measurement
method,
the length
an
d width
of whi
c
h i
s
2
3
.2m a
nd 7.3m
, a
n
d
six sen
s
ors
a
r
e mo
unted
o
n
the g
r
ou
nd
at
the position 2
0
, 22, 24, 26, 28 and 30 as shown in Fi
g
u
re 1, the gra
i
n is loade
d by divided into 9
times, the da
ta of 4 times are sel
e
cte
d
to es
timate the para
m
ete
r
s, the grain weig
h and th
e
mean p
r
e
s
sure value in every time is sh
own in Ta
ble
2.
Table 2. The
Experiment Data
Experiment
data
in batch
Modeling data
The loaded g
r
ain
w
e
ight
i
G
(
T
)
Mean pressure v
a
lue on
the groun
d
i
q
(KPa)
The loaded g
r
ain
w
e
ight
i
G
(
T
)
Mean pressure v
a
lue on the
ground
i
q
(KPa)
65.29
128.82
204.402
280.729
355.469
427.963
521.033
582.683
649.69
3.82833
8.36417
13.62133
18.99868
24.563
29.471
34.901
39.079
43.176
65.29
128.82
204.402
649.69
3.82833
8.36417
13.62133
43.176
Based o
n
the
s
e four time
s
data, the mod
e
l is
built by using the
conv
entional reg
r
e
ssi
on
method which descri
bed i
n
Equation (1
0) is:
)
00013
.
0
0816
.
0
0705
.
0
(
ˆ
2
q
q
A
G
(14)
Whe
n
u
s
ing t
he metho
d
b
a
se
d on the
ratio of er
ror,
whi
c
h is
expressed by Eq
uation (11), t
h
e
model can be
built as:
)
000135
.
0
0813
.
0
0719
.
0
(
ˆ
2
q
q
A
G
(15)
Acco
rdi
ng to above two m
odel
s, the pre
d
icted g
r
ain
weig
ht and th
e estimation
error is
sho
w
n in
Table 3.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 7, July 201
4: 5552 – 55
58
5558
Table 3. The
Predi
cted Re
sults by u
s
ing
Two Kind
s M
odel
s
Predicted results b
y
the
regr
ession
Predicted results b
y
the
ratio of er
ror
The actual loade
d grain
w
e
ight
i
G
(T
)
Predicted w
e
ight
i
G
ˆ
(T)
Percent of the
error
(%
)
Predicted w
e
ight
i
G
ˆ
(T)
Percent of the e
r
ror
65.29
128.82
204.402
280.729
355.469
427.963
521.033
582.683
649.69
65.16959
129.06601
204.25936
282.43302
364.66767
438.33633
521.07835
585.62789
649.67378
0.18442
0.19097
0.06978
0.607
2.58776
2.42388
0.0087
0.5054
0.0025
65.26136
129.0226
204.10257
282.20856
364.42452
438.12
520.93901
585.58161
649.74735
0.04387
0.15728
0.14649
0.52704
2.51935
2.37334
0.01804
0.49746
0.00883
Comp
ared th
e data
listed
i
n
Tabl
e 2,
we can
se
e th
at the me
an
error i
s
0.73
% and th
e
maximum error is 2.59% f
o
r the conve
n
tional
re
gre
ssi
on metho
d
,
but when u
s
ing the met
hod
based o
n
the
ratio of e
r
ror, the mean
error i
s
0.6
99%
and the
maximum erro
r is 2.51%, whi
c
h
mean
s that the erro
r is l
e
ss fo
r the n
e
w
mod
e
li
ng
method. At the same tim
e
, whe
n
the
grain
weig
ht is light
, the predicte
d
method ba
sed
on the rati
o of erro
r is
more reliabl
e.
6. Conclusio
n
To me
asure
the
store
d
g
r
a
i
n qua
ntity accura
tely an
d
reliably, the
method
ba
se
d on th
e
grou
nd p
r
e
s
sure i
s
put forward. Accordi
ng to t
he exp
e
rime
nt, the layout of the pre
s
sure on t
h
e
grou
nd is giv
en, and the random of the
pressu
re
sen
s
or
can b
e
eli
m
inated an
d comp
en
sated
by
usin
g the g
r
o
und me
an p
r
essure, whi
c
h ca
n re
du
ce
the sen
s
o
r
s
use
d
in this
method a
nd t
h
e
cost is low. Lastly, to improv
e the predi
ction a
c
curacy of the
grain
weig
ht, the para
m
eters a
r
e
estimated
ba
sed
on the
ra
tio of the erro
r. The
exp
e
ri
ment re
sult
s
sho
w
n th
at, the mea
s
u
r
e
m
ent
error is le
ss than 3% by using this meth
od, wh
i
c
h can
meet the measu
r
em
ent re
quire
ment of the
grain q
uantity.
Ackn
o
w
l
e
dg
ements
This p
r
oje
c
t suppo
rted by the state 863
proj
e
c
ts of China (No.20
1
2
AA1010
08 ),
national
natural
sci
en
ce foun
dation
of China (No. 61071
197
), and the ed
ucatio
n dep
a
r
tment of He
nan
provin
ce scie
nce a
nd tech
nology key project (No.14A
5100
19).
Referen
ces
[1]
F
an Ch
ao, Z
h
ang
De-
x
i
an,
F
u
Hon
g
-li
ang.
Desig
n
of the
Measur
ement
Nod
e
of the
Grain Qua
n
tity
Monitori
ng
Sys
t
em B
a
se
d o
n
the CAN-
bus
. I
n
ternati
o
n
a
l C
onfere
n
ce
on
Chal
le
nges
i
n
Enviro
nmenta
l
Scienc
e an
d C
o
mputer En
gi
n
eeri
ng. W
uha
n. 2010; 1: 21
1-2
14.
[2]
F
an Ch
ao, Z
h
ang Y
u
a
n
, Z
h
a
ng D
e
-
X
ia
n.
R
e
searc
h
on t
h
e Mon
i
torin
g
S
ystem of th
e
Grain Qua
n
tity
Based
on th
e
CAN-b
u
s
. 2
nd
Internatio
na
l Confer
ence on
Comp
uter
En
gin
eeri
ng and T
e
chnolog
y,
Che
ngd
u. 201
0; 1: 458-4
60.
[3]
S T
r
abelsi
，
SO Nelso
n
．
Study on
the
acc
u
racy of
bulk
dens
ity an
d
moisture c
onte
n
t pred
ictio
n
i
n
w
heat from ne
ar field free sp
ace meas
ure
m
ents
．
T
e
chnisc
hes Messe
n. 2007; 74(
5): 280
-289.
[4]
MS Venk
atesh
,
GSV Rag
hav
an. An
Overvi
e
w
of Mi
cro
w
a
v
e
Proc
essin
g
and
Di
el
ectric Properti
es o
f
Agri-food Materials
．
Biosystem
s Engineering.
2004; 8
8
(1): 1-18.
[5]
BS Sabri
na, K
T
Sebastian,
MF
Louis
e
. Va
riatio
n i
n
the
Measur
ement
of
Crani
al V
o
l
u
m
e
a
nd S
u
rfac
e
Area Usi
ng 3D
Laser Sca
n
n
i
n
g
T
e
chnolo
g
y
.
Journ
a
l of F
o
re
nsic Scie
nces.
2010; 5
5
(4):
871
–8
76.
[6]
Z
hang
Xinfen
g
,
Z
haoYan. A
pplic
atio
n of S
upp
ort
Vector
Machi
ne to R
e
lia
bi
lit
y
M
i
cro
-
area S
heet
Resistance.
T
E
LKOMNIKA Indon
esia
n Jour
nal
of Electric
al
Engin
eeri
ng.
2
013: 11(
7): 355
2 – 356
0.
[7]
Z
h
ihu
i
Z
hu, Xin
f
u Liu, Ru
nli Z
han
g,
etal
. Pro
bes Vaci
ll
ating
and Ma
pp
ing T
e
chn
i
qu
e at T
e
sting Micro-
area Sh
eet R
e
sistanc
e.
T
E
LKOMNIKA Indon
esia
n Jour
nal of
Electric
al Engi
ne
erin
g
Indones
ian
Journ
a
l of Elec
trical Eng
i
ne
eri
n
g
. 11(4): 1
828
-183
4.
Evaluation Warning : The document was created with Spire.PDF for Python.