TELKOM
NIKA
, Vol.12, No
.3, Septembe
r 2014, pp. 7
41~750
ISSN: 1693-6
930,
accredited
A
by DIKTI, De
cree No: 58/DIK
T
I/Kep/2013
DOI
:
10.12928/TELKOMNIKA.v12i3.xxxx
741
Re
cei
v
ed Fe
brua
ry 23, 20
14; Re
vised
May 29, 20
14
; Accepte
d
Ju
ne 12, 201
4
The use of ON-OFF and ANN Controll
ers for Automated
Irrigation System Model Based on Penman-Monteith
Evapotranspiration
Susilo Adi Wid
y
anto
1
, Achmad Widod
o
2
, Achm
ad Hiday
a
tno
3
, Su
w
o
ko
4
1,2,
4
Department
of Mechanic
a
l
Engi
neer
in
g, Universit
y
of Di
p
one
goro (U
NDI
P), Semarang,
Indon
esi
a
3
Departme
n
t of Electrical En
gi
neer
ing, Un
iver
sit
y
of Dip
on
eg
oro (UNDIP) , Semara
ng, Ind
ones
ia
Jl. Prof Sudarto, SH,
T
e
mbalang, Semar
a
n
g
,
Ph/F
ax: +
62 024
74
600
59
ext 1/6
2
02
474
600
59 e
x
t 2
e-mail: susi
lo
7
0
@
y
a
hoo.com
1
,
a
w
id
@u
ndi
p.ac.id
2
, hidyatno@undi
p.ac.id
3
, Su
w
o
k
o
@gma
il.com
4
A
b
st
ra
ct
T
he cli
m
ate ch
ang
e tends to be extre
m
e co
nditi
on t
hat dir
e
ctly affects on decre
asin
g a
g
ricult
ure
production.
Therefore,
applic
ation of automated syst
em in
agric
ulture activities
is
the pot
ential iss
ue which
must b
e
co
ns
ider
ed. T
h
is p
aper
prese
n
ts ON
-OF
F
and ANN contr
o
ll
ers w
h
ich ar
e
app
lie
d to th
e
autom
ated
irrigation system. Contro
lling irrigation sy
stem
used
a calc
ulated
Penm
an-Mont
eith
evap
otrans
pira
tion a
n
d
a r
e
ferenc
e of s
o
il
moisture
as t
he co
mpar
ed
inp
u
t. Input p
a
ra
meters
of t
h
e
evap
otrans
pira
tion i
n
clu
d
e
d
te
mp
eratur
e, he
a
t
radiati
on, at
mosph
e
re pr
es
s
u
re a
nd w
i
n
d
s
pee
d. T
he us
e
of
feed forw
ard
A
NN i
n
clu
d
e
d
1
inp
u
t lay
e
r w
i
th
15
ne
uro
n
s a
n
d
2
hi
dde
n
lay
e
rs w
i
th 1
0
a
n
d
5
ne
uro
n
s a
n
d
1
output
lay
e
r a
n
d
1
in
put
layer,
2
hid
d
e
n
l
a
yer
s
w
i
th 96
a
nd 1
n
euro
n
s an
d 1
o
u
tput layer, errors are 14.3
%
and 3.9%,
res
pective
ly.
Error of the ON-OFF controller wit
h
sam
p
li
n
g
time of 0.05 s
e
co
nd is e
q
u
a
l to t
h
e
error of ANN control
l
er. T
he p
e
rformanc
e of such
contro
ller
s
w
e
re evaluat
ed an
d co
mp
ared bas
ed o
n
er
ror
of both contro
l
l
ers. T
he simulati
on
res
u
lt of ON-OFF controller was us
e
d
as the refer
ence of co
ntro
ller
deve
l
op
ment b
a
sed
on AT
me
ga 8
microco
n
t
r
oller. T
h
e si
mulati
on res
u
lts
show
that the
error of the
ON-
OF
F
controller can be e
a
si
ly a
d
juste
d
by setti
ng t
he sa
mp
li
n
g
time
of the dead
z
o
n
e
d
i
scr
eti
z
a
t
io
n.
Keyw
ords: ON-OFF controller, ANN control
l
er
, irrigatio
n system, ev
apotra
n
s
pirati
on
1. Introduc
tion
Un
controllabl
e exploitation
of natu
r
al
reso
ur
ce
s di
rectly affect
s
on nat
ural
d
e
crea
sing
quality and
al
so
cha
nge
s t
he cli
m
ate te
nds to
be
extreme
co
nditio
n
. In other
ha
nd, the ne
ed
of
food in
crea
ses
pro
portio
nally with
p
opulatio
n
g
r
owin
g, whil
e
agri
c
ultu
re
area
s
extre
m
ely
decrea
s
e
s
ca
use
d
by expa
nsio
n
of hou
sing
a
r
ea.
The
r
efore, ap
plication
of auto
m
ated syste
m
in
agri
c
ultu
re a
c
tivities is the potentia
l issu
e whi
c
h shoul
d be co
nsi
dered.
Variou
s a
u
to
mated a
g
ri
cu
lture
system
s
have be
en
develop
ed in
clud
ed
wee
d
cont
rol
system to
re
duce neg
ative impa
ct to the environm
ent due to ex
ce
ssive
use
of herbi
cid
e
s and
Pesticid
e [1][2], automate
d
irri
gation
system [3]-
[5], agricultural
robot [6][7] a
s
well a
s
sm
art
agri
c
ultu
re [8]
[
9].
In automa
t
ion of i
rrig
a
tion
system, th
e init
ial step can
be
inst
a
lled
a w
a
ter
va
lve
controlle
r
in irrig
a
tion system
ba
sed o
n
time
fun
c
tion. Ho
weve
r, this control
l
er
can
ca
use
difficulty in achievin
g the optimal growi
ng co
nditi
on.
The next techn
o
logy is a
pplication of soi
l
humidity co
ntrol sy
stem. Actually, this
controle
r
a
ppli
c
ation
wa
s o
n
ly su
cces
o
n
wate
r full l
and
con
d
ition. Th
e high cost in
investment a
nd the
difficul
t
y of maintenance of su
ch
sen
s
o
r
s
be
co
me
con
s
id
eratio
n
s
of the rea
s
o
n
s why these
sen
s
o
r
s
we
re
not widely ap
plied.
In the l
a
st
d
e
ca
de, a
ppli
c
ation
of
soil
humi
d
ity se
nso
r
s
spread
s
widely
and
it wa
s
correl
ated to the rea
s
oni
n
g
of low co
st in in
vestmen
t
and mainte
nan
ce. Cu
rre
ntly, the goal of
resea
r
ch acti
vities in this area a
r
e aim
ed parti
cula
rl
y to reduce water
co
nsu
m
ption ba
se
d o
n
sma
r
t irri
gatio
n co
ntroll
er
system
[3]-[5]. With the
s
a
me reasons
,
smart
controlle
r sy
stem b
a
sed
on wate
r co
nse
r
vation in
irrigatio
n area wa
s deve
l
oped. Thi
s
method u
s
ed
rain wate
r
as
a
sup
p
leme
nt of the irri
gati
on sy
stem, so that
the sa
ving wate
r consumption
can be i
n
crea
se
d
arou
nd of 6
7
% [4]. Application of a
u
tomated ir
ri
gation sy
ste
m
clo
s
ely correlate
s
to the
topographi
c condition a
nal
ysis an
d the
water
re
sou
r
ces. Re
se
arch
of smart irrig
a
tion appli
c
a
b
le
for agri
c
ultu
re
sand
are
a
was al
so devel
oped in [10].
The devel
ope
d cont
rolle
r system ba
sed
on
evapotra
nspiration
con
d
itio
n an
d level
o
f
wate
r
con
s
umption
was limited
by th
e availa
bility of
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 12, No. 3, September 20
14: 74
1 – 750
742
w
a
ter resourc
e
s
.
The ke
y
of
the
m
e
thod
i
s
sche
duling
irrigati
on
p
r
o
c
e
s
s based on
th
e
cal
c
ulatio
n of water b
a
lan
c
e of the plant [10].
1.1 Penman Montei
th ev
a
potra
nspira
tion
Theo
retically, water
co
nsu
m
ption of ag
ricult
u
r
e
are
a
is propo
rtio
nal to the
water lo
st
cau
s
e
d
by grou
nd eva
p
o
ration
and
plants tran
spiration o
n
this area.
This
calle
d
as
evapotra
nspiration. Evapotranspira
tion
con
d
ition ref
e
rs to the referen
c
e ev
apotra
nspiration
(ETo). In the
appli
c
ation,
ETo is rarely meas
ured,
but it is co
m
m
only used i
n
mathem
atical
equatio
n su
ch as Penma
n
-Mo
n
teith [11], in whic
h the input pa
rameters in
clu
de tempe
r
atu
r
e,
radiatio
n, at
mosp
he
re
pressure a
n
d
wind
spee
d.
Practi
cally, d
e
termin
ation
of ETo i
s
sh
own
in
Figure 1.
Figure 1. Sch
e
matic dia
g
ra
m of measu
r
e
m
ent
of weat
her data to
ca
lculate refere
nce
evapotra
nspiration (ETo
).
Penman Mo
n
t
eith equation
is described i
n
Equation 1
[12][13].
0.404
/
1
0.34
(1)
whe
r
e :
ET
0
= refe
ren
c
e e
v
apotran
s
pi
ra
tion [mm/day],
∆
.
.
∗
.
.
exp
= 2.718
3 (ba
s
e of natu
r
al logarith
m
),
T
mean
= mean d
a
ily air tempe
r
atu
r
e at 2 m heig
h
t [°C],
= net ra
diatio
n at the cro
p
surfa
c
e [M
Jm
-2
/day],
G
= soil h
eat flux density [MJm
-2
/day],
=0.000
665P
= win
d
sp
eed
at 2 m height [ms
-1
],
= satu
ration v
apor p
r
e
s
sure[kPa],
=
ac
tual vapor press
u
re [kPa],
=
(T
)=
saturation vapor p
r
e
s
sure d
e
ficit [kPa],
P
= atmo
sphe
ri
c pre
s
su
re [kPa],
λ
= latent heat
of vaporizatio
n, 2.45 [MJ/kg],
Cp
= sp
ecifi
c
he
at at consta
nt pressu
re, 1.0
13x10
-3
[MJ
/
(k
g°C)],
ε
= ratio mol
e
cular weight of
water vapo
ur/dry air =0.6
2
2
.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
The use of ON-OFF and ANN
Controllers
for Au
tomated Irrigation ... (Sus
ilo Adi Widy
anto)
743
Based
on th
e Penman
-M
onteith equ
ation revi
sed
by the FAO (199
9) [12],
the pro
c
e
s
s of
evapotra
nspiration was
si
mulated a
n
d
the re
sult
wa
s u
s
ed
as referen
c
e in
put of ON-O
FF
controlle
r in a
u
tomated irrig
a
tion system.
Several successful attemp
ts for improvi
ng the Penm
an-M
onteith
equatio
n parameters
are
re
po
rted
in th
e follo
wing
stu
d
y. Ha
rri
so
n
(1
963) p
r
e
s
ent
ed that
the
latent
heat
of
vapori
z
ation i
s
given as a l
i
near fun
c
tion
with
air temperatu
r
e [14], while a co
rrelation equ
ation
for
the slop
e of
satu
ration vapor pre
s
su
re curve
wa
s
repo
rted
by Murray [15]. Several e
m
pi
rica
l
equatio
ns fo
r cal
c
ulatin
g t
he saturation
vapor
pre
ssure i
n
term
s
of air tem
perature
we
re
al
so
prop
osed [1
2
][16][17]. Alth
ough t
hese
equatio
ns
ha
ve different
algorith
m
s,
cl
ose
re
sult
s a
r
e
obtaine
d fro
m
them. It is note
d
that
most of
th
ese
co
rrelati
ons
are rest
ricted
withi
n
the
temperature
range fro
m
0 to 50 °C.
1.2 Arti
ficial neural ne
t
w
ork (ANN)
Control sy
ste
m
based o
n
ANN h
a
s
bee
n used in
ma
ny fields such
as mo
nitorin
g
syste
m
on buil
d
ing damage index [18], increasing power
sy
stem stability
[19][
20], adaptive control
of
spa
c
e
ro
bots
[21][22], etc.
Princi
pally, artificial
ne
ural
netwo
rk (A
NN) is a m
e
tho
d
for produ
ci
ng
the output si
gnal from va
riou
s input p
a
ram
e
ters
in
which its correlation is
determi
ned by
activation fun
c
tion. Th
e ANN m
e
thod i
s
sho
w
n in
F
i
gure
2. Each input
sign
a
l
(a) i
s
give
n
the
weig
ht fun
c
tion
(w). Th
e
multiplicatio
n
between
in
put pa
ram
e
te
r a
nd th
e
weight fun
c
tio
n
is
summ
ed
and
sim
u
lated
in
action
fun
c
tio
n
to
deter
mine
th
e o
u
t
pu
t le
ve
l F
(
a,w). If there cons
ists
of n input pa
rameters (it al
so
contai
ning
n weig
ht
fun
c
tion
s), the o
u
tput functio
n
is de
scrib
ed
in
Equation 2 [2
3]:
∑
∗
(
2
)
Figure 2. ANN method [11
]
.
In automated
irrigatio
n system, ANN co
nt
rolle
r uses
comp
uted ev
apotra
nspiration data
as inp
u
t para
m
eter an
d the referen
c
e soil moi
s
ture
as refe
ren
c
e sign
al. The
resulte
d
wei
ght
function
was use
d
to build actual
control
l
er syst
em
in
whi
c
h th
e out
put is u
s
ed to
co
ntrol
a
serv
o
valve.
2. Rese
arch
Metho
d
This
re
sea
r
ch wa
s p
e
rfo
r
med
by modelin
g evap
otran
s
pi
ration
based o
n
revise
d
Penman
-Mot
eith equ
ation.
The i
nput p
a
ram
e
ters
in
clud
ed tem
p
e
r
ature, atmo
sphere p
r
e
s
su
re,
radiatio
n an
d
wind
speed.
Cal
c
ulate
d
evapotra
nspi
ration represe
n
ts the
actu
al soil
moi
s
ture
affected by weather
condition whi
c
h
will be
adjusted t
o
the reference soil moisture.
The output d
a
ta of evapotranspiratio
n
was us
ed a
s
input of ON-OF
F
controller a
nd ANN
controlle
r. The both perfo
rmances
we
re then co
m
p
ared. Th
e block diag
ram
s
de
scribe b
o
th
controlle
rs a
r
e sho
w
n in Fi
gure 3.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
9
30
TELKOM
NIKA
Vol. 12, No. 3, September 20
14: 74
1 – 750
744
2.1 Input par
a
meter
s
Cal
c
ulation
o
f
evapotra
nspiration
condi
tions
wa
s p
e
r
forme
d
in
cy
cle b
a
se of
24 ho
urs
usin
g the
we
ather
data in
put. The
s
e d
a
ta incl
ude:
a
i
r
tempe
r
atu
r
e,
radi
ation, wind sp
eed a
nd
ai
r
pre
s
sure whi
c
h were assu
med a
s
sin
u
siodal si
gnal.
Referrin
g to the data, sim
u
lating data in
put
is de
scribe
d in the followin
g
se
ction.
Heat radiatio
n is a maj
o
r factor that
deter
mi
ne
s the rate of e
v
apotran
spi
r
a
t
ion. Heat
radiatio
n i
s
a
com
pon
ent
on pl
ants en
ergy b
a
lan
c
e
with
re
spe
c
t
to the
net radiation. In
fact,
infrared
radi
a
t
ion is al
so
a compo
nent
in the
net
radiation.
Ho
wever, t
h
e b
a
lan
c
e i
s
always
negative
or
zero
so
that it
ca
n b
e
elimi
nated.
In thi
s
simul
a
tion,
radiation
wa
s
assume
d a
s
a
sinu
sio
dal sig
nal with ampl
itude of 2 MJ/m
2
in range
of 112 MJ/m
2
. The freque
n
c
y is 2
π
/24 or
0.2168 rad/h
our de
rived
from 24 hou
rs cycle.
(a)
(b)
Figure 3. a. ON-OFF contro
ller, b. ANN controller.
Tempe
r
atu
r
e and humidity are
pa
ramete
rs
that
affe
ct
on the
d
r
oug
ht and
the
atmosp
he
re
drying cap
abi
lity.
While,
vapor pre
s
sure deficit
(VP
D
)
is a m
e
teorol
ogical varia
b
l
e
that is
used
to
measure the
atmosp
here
drying cap
ability.
VPD sho
w
s th
e
vapor p
r
e
s
sure differe
n
c
e
(co
n
centratio
n
of water va
por) bet
wee
n
plants an
d
air-dried mois
ture. In the modeling, the
air
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
9
30
The use of ON-OFF and ANN
Controllers
for Au
tomated Irrigation ... (Sus
ilo Adi Widy
anto)
745
pre
s
sure is a
s
sumed
a
s
si
nusi
odal
sig
n
a
l with
amplit
ude of
5 kP
a
and th
e con
s
tant of 9
5
kPa.
Freq
uen
cy is
2
π
/24 or 0.21
68 rad/h
o
u
r
.
Tempe
r
atu
r
e
affects
on alt
e
ration
of the
ET
co
rrelati
on
VPD and also adve
ctio
n.
Whe
n
all other fa
cto
r
s a
r
e eq
ual, ET on wa
rm con
d
ition
s
tends to be la
rg
er than the pl
ant temperature
.
ET
in
cr
ea
ses
w
a
r
m
vege
ta
tio
n
be
cau
s
e les
s
ene
r
g
y is
r
e
q
u
ir
e
d
to
e
v
apo
r
a
te th
e
wa
te
r
.
Tempe
r
atu
r
e
also
impa
cts
the rel
a
tive e
ffectiv
eness
of the radia
n
t ene
rgy a
n
d
wind
affect
s
on
evaporating
water. Radia
n
t energy is
more effe
ctively utilized for ET when t
e
mpe
r
atures
are
high. In
contrast, win
d
ha
s more
impa
ct
on ET
w
h
en te
mp
e
r
a
t
u
r
es
ar
e
low
.
In
th
is
s
i
mu
la
tion
,
temperature
i
s
a
sin
u
soid
al sig
nal
with
amplitud
e 5
o
C in
24 h
o
u
r
s
cy
cle,
so t
he f
r
e
q
ue
ncy
is
2
π
/24 or 0.21
68 rad/h
o
u
r
. The tempe
r
at
ure rang
e (offset) i
s
arroun
d of 30
o
C.
The
wind
ha
s two majo
r role
s; firstly, it
transport
s he
at that
build
s up
on
adja
c
ent
surfa
c
e
s
su
ch a
s
d
r
y de
sert o
r
a
s
p
halt to v
egetatio
n which a
c
ce
lerate
s eva
p
o
r
ation
(a
process
referred
to a
s
adve
c
tion
).
Seco
ndly, wind
se
rve
s
to accel
e
rat
e
evapo
ratio
n
by en
han
cing
turbule
n
t tran
sfer of
wate
r vapor from m
o
ist veget
atio
n to the dry a
t
mosph
e
re. In this
case, the
wind
con
s
tant
ly repla
c
e
s
the moist ai
r lo
cated
with
in a
nd just a
bove
the plant can
opy with d
r
y air
from ab
ove. Wind
sp
eed
wa
s a
s
sume
d as a
sin
u
so
i
dal cu
rve
with amplitud
e o
f
1 km/h in ra
nge
of 3.5 km/h.
2.2 Referen
ce
p
a
rameter
Output para
m
eter of the evapot
ra
nspiration cal
c
ulat
ion rep
r
e
s
ent
s actu
al soil
moistu
re
influen
ced
by weat
her pa
rameters. Soil
moistu
re
co
ndition
s
shou
ld be
arran
g
e
d
to ap
propri
ate
the spe
c
ific soil moisture whi
c
h is dete
r
mine
d by
the cultivation plant type by adjusting
wa
ter
irrig
a
tion.
In real
condit
i
ons,
be
side
s the type of
plant, soil mo
isture
is al
so
influen
ced
by age
of
plant and
soil
type. In this
modelin
g, the refe
re
nce so
il moisture was a
s
sume
d as a G
aussi
a
n
sinu
soi
dal si
g
nal. Referen
c
e soil moi
s
ture is assu
med
in a rang
e of 35% with am
plitude of 15
%,
while the fre
q
uen
cy is 2
π
/2
4 followin
g
the 24 hou
r cycle (Figu
r
e 6
)
.
2.3 Design o
f
electr
onic
dev
i
ce of ON-OFF co
ntr
o
ller
Assu
med
inp
u
t sig
nals we
re g
ene
rated
by ele
c
tro
n
i
c
circuit
whi
c
h u
s
ed Atme
ga 8
as
sinu
sio
dal sig
nal gene
rato
r. Due to the sinu
sio
dal
sig
nals have the
same freq
ue
ncy, the circu
i
t
only use
d
a sign
al gene
rator and e
a
ch amplitude
and con
s
tant
s wa
s adj
u
st
ed by operational
amplifier
LM
358. The
co
nfiguratio
n of
circuit is
sh
own i
n
Figu
re 4. The
use
of 1 byte DAC
(083
2)
sho
w
s that the accura
cy of outp
u
t signal
i
s
1
9
.6 mV. Before op
erate
d
, each input
si
gnal
is adju
s
ted th
e op-a
m
p gai
n to appro
p
ri
ate the amplitude of the inp
u
t data cha
r
a
c
teri
stic.
Figure 4. Electroni
c a
r
chitecture of voltage si
nu
siod
al sign
al assu
med as in
put para
m
eters.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
9
30
TELKOM
NIKA
Vol. 12, No. 3, September 20
14: 74
1 – 750
746
3. Results a
nd Analy
s
is
Block
diag
ra
m of the inpu
t paramete
r
s
model
a
nd P
enman
-Mo
n
teith evapotra
nspi
ration
usin
g Simulin
k is illust
rate
d in Figure 5
.
A simulated evapotran
s
p
i
ration si
gnal
is a sinu
soi
dal
sign
al where
its frequ
en
cy is same a
s
th
e frequ
en
cy of referen
c
e
sign
al (G
au
ssian
sin
u
soid
al).
The am
plitud
e of evap
otra
npiratio
n
outp
u
t sign
al i
s
lo
wer than th
e
referen
c
e
sig
nal (Fi
g
u
r
e 6
)
. It
implies th
e d
e
sig
ned O
N
-OFF controll
e
r
as
well a
s
ANN
controll
er mu
st have
the amplifica
t
io
n
function.
Figure 5. Block di
agram of
Penm
an-Mo
nteith evapotranspiratio
n
.
Figure 6. Co
mpari
s
o
n
bet
wee
n
refe
ren
c
e si
gnal
a
n
d
calculated ev
apotra
nspiration sig
nal.
3.1 ON-OFF
co
n
t
roller
In pro
c
e
s
sing
the sim
u
late
d evapot
ran
s
piration
sig
n
a
l
, ON-OFF
controlle
r u
s
e
s
de
ad
zon
e
circuit
and
me
mo
ry integ
r
atio
n a
s
fe
edb
ack sign
al on the outp
u
t of calcul
ated
evapotra
nspiration (Fi
gure
3a). Outp
ut si
gnal is
a fo
rm
of ON-OFF
configuration
with variatio
n
of
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
The use of ON-OFF and ANN
Controllers
for Au
tomated Irrigation ... (Sus
ilo Adi Widy
anto)
747
pulse wi
dth [24]. Whe
n
si
gnal i
s
hig
h
, sole
noid valv
e will b
e
op
e
ned
so that t
he irrigatio
n
water
will be
disch
a
r
ged
to en
ha
nce th
e
soil h
u
midity. Re
fe
ren
c
e
sig
nal i
s
ap
proximated by tria
ngul
ar
sign
al. Actual
ly, the increa
se of
soil
moi
s
ture
cau
s
ed
by discha
rgin
g wate
r i
s
det
ermin
ed by t
h
e
soil type, lev
e
l of den
sity and g
r
ain
particle si
ze. In
t
he control
sy
stem, the
soil
cha
r
a
c
teri
sti
c
is
integrate
d
into a con
s
tant
whi
c
h is u
s
e
d
to co
rrect the
slope of the i
n
crea
se of so
il moisture.
The o
c
curre
n
ce
of error (the ratio
b
e
twee
n maxi
mum erro
r a
nd the am
plitude of
referen
c
e
sig
nal) of O
N
-O
FF
controller
can
be
set b
y
adju
s
ting
sampling
time
of dea
dzone.
By
assumin
g
the
co
nsta
nt of
soil
cha
r
a
c
teristic i
s
1
an
d
adju
s
ting
sa
mpling time
of dead
zo
ne
are
0.5 and 0.00
5 se
cond, th
e controlled
sign
al erro
rs are in ra
nge
of 0-10 or 2
8
.
6% and 0-5
or
14.3% (Fig
ure 7), re
sp
ect
i
vely [25]. Howeve
r,
adju
s
ting a ve
ry small the
sampling time
of
dead
zo
ne ca
n
affect on the
pe
rform
a
nce
of c
ontrol
sy
stem caused by
the pa
ssivity of
mech
ani
cal system, su
ch the sol
enoid v
a
lve.
Figure 7. Performa
nce of ON-OFF c
ontro
ller (referen
ce sign
al is bl
u
e
, controlled
sign
al is re
d,
and O
N
-OFF
con
d
ition of solenoi
d valve is bla
c
k) with
dead
zon
e
sa
mpling time o
f
0.005
s
e
c
o
nd [25].
3.2 Hard
w
a
r
e
perfo
r
man
ce of O
N
-OF
F
controller
Based
on th
e
simul
a
tion result, the
ON-OFF
co
ntroll
er
wa
s mad
e
by usin
g AT
mega
8
microcontroll
er. The
pe
rfo
r
man
c
e i
s
sh
own
by di
spl
a
ying the in
p
u
t and th
e ou
tput sign
als u
s
ing
oscilo
ssope.
The refe
re
nce sign
al an
d
the cal
c
ulat
ed evapot
ran
s
piration (ET
o
) of the si
g
nal
pro
c
e
ssi
ng
is sh
own i
n
Fi
g
u
re
8a
an
d 8
b
, re
sp
ectivel
y
. The n
eed
e
d
am
plificatio
n fun
c
tion
of t
he
ETo sign
al to
appropri
a
te the refe
ren
c
e
sign
al is a
r
ou
nd of 3 times.
It equals wit
h
the simul
a
tion
result in Figure 6.
(a)
(b)
Figure 8. Oscilloscop
e disp
lay of a. Referen
c
e
si
gnal
as Ga
ussia
n
sinu
sio
dal, b. ETo sign
al.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 12, No. 3, September 20
14: 74
1 – 750
748
Princi
pally, O
N
-OFF
co
ntroller i
s
de
sig
ned to
tran
sf
orm th
e ETo
sign
al into
pu
lse
width
modulatio
n
si
gnal
(PWM)
whi
c
h i
s
used
to a
c
tivate a
sol
enoi
d val
v
e. Accordi
n
g
to the
mo
del
led
system
in Fi
gure
3a,
O
N
-OFF
cont
roll
er
wa
s
built
by inputting
prog
ram
m
ing
functio
n
into
the
microcontroll
er mem
o
ry (embed
ded
al
gorithm
)
a
s
the si
gnal p
r
oce
s
sing. P
W
M outp
u
t signal
whi
c
h is p
r
od
uce
d
by adju
s
ting time sa
mpling of dea
dzo
ne of 10 ms is sho
w
n in Figure 9
a
. By
this PWM
sig
nal, the
reference
signal
i
s
a
pproxim
ated by tri
ang
u
l
ar
sign
al whi
c
h i
s
sho
w
n
in
Figure 9b.
(a)
(b)
Figure 9. a. Pulse
width mo
dulation
sign
al to
activate sole
noid valve (dea
d zo
ne
sampli
ng tim
e
of 10 ms), b. Approximatio
n of t
he reference sig
nal with triangula
r
sign
al as
cont
rolled
sign
al of
ON-OFF
cont
rolle
r (de
ad zone samplin
g
time of 10 ms).
3.3.
ANN con
t
roll
er
Acco
rdi
ng to
the bl
ock
d
i
agra
m
in
Fi
gur
e
3b,
the
input
sig
nal
s of
neu
ral
netwo
rk
controlle
r are
the calcul
ate
d
evapotra
nspiration a
nd the refe
ren
c
e
sign
al. Based
on these inp
u
ts
sign
al co
nfig
uration, n
e
u
r
al network u
s
ed fe
ed forward meth
od
. Theoreticall
y
, feed-forwa
r
d
neural netwo
rks can app
roximate any
nonline
a
r functio
n
, and
thus the b
a
ckpropa
gati
on
algorith
m
s a
r
e popul
ar for
training fee
d
-f
orward neu
ral
networks [26
]
.
The
weig
ht factor obta
i
ned from t
r
ai
nin
g
p
r
o
c
edure d
epe
nded
on th
e network
architectu
re.
As the mode
l 1, the desi
gn of neur
al
network a
r
chitecture is d
e
scrib
ed a
s
the
followin
g
: 1 i
n
put laye
r
with
15
neu
ro
ns
and
2 hi
dde
n
layers
with 1
0
an
d 5
n
eurons an
d 1
o
u
tput
layer. In Matlab pro
c
e
d
u
r
e
,
the network
co
n
s
tru
c
tion i
s
expre
s
sed
as the follo
wi
ng:
jarn
et=
new
ff(min
m
ax(P),[15 1
0
5 1],{'
t
ansig'
'
t
ansig'
'
t
ansig'
'
pure
lin'
}
);
From the trai
ning procedu
re, the ANN
controll
ed d
a
ta is clo
s
e to the referen
c
e
signal
(Figu
r
e 10
a). The erro
r occurs
whe
n
sig
nal dire
ction
cha
nge
s extremely (up
-
do
wn or d
o
wn-u
p),
while
on
the
co
ntinuo
us
alternatio
n e
r
ror tend
s to
be
low.
Error
co
ntrolle
d
sig
nal
of A
NN
controlle
r a
c
h
i
eves i
n
ran
g
e
of
0-3
o
r
1
4
.
3%. It equal
s with t
he
co
ntrolled
si
gnal
error of O
N
-O
FF
controlle
r wit
h
setting de
a
d
zo
ne sampli
ng time of 0.005 se
co
nd (Fi
gure 7
)
.
By modifying the network
architectu
re,
the
error
ca
n
be re
du
ced t
o
3.9% (Fig
u
r
e 10
b).
Erro
rs o
c
curs in rang
e of
0-0.82
which
wa
s di
stri
but
ed in l
o
catio
n
s
whe
r
e
the
sig
nal di
re
ction
cha
nge
s. Th
e modified
n
e
twork
archit
ecture i
s
de
scribe
d a
s
t
he follo
wing:
1 input lay
e
r, 2
hidde
n layers with 96 and 1 neuron
s an
d 1 output la
yer. Based o
n
literature re
view, the error
level ca
n al
so be
re
du
ced
by u
s
ing
ca
scad
e
correlati
on a
r
tificial
n
eural
net
work (CA
N
N) m
o
del
with embe
dd
ed Kalman le
arnin
g
rul
e
[26].
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
The use of ON-OFF and ANN
Controllers
for Au
tomated Irrigation ... (Sus
ilo Adi Widy
anto)
749
.
(a) (b)
Figure 10. Error of ANN
co
ntrolled
sign
a
l
resulte
d
by modifying net
work a
r
chitecture, a. 1 inpu
t
layer with 15
neuron
s and
2 hidde
n laye
rs
with 10
an
d 5 neuron
s a
nd 1 output la
yer, b. 1 input
layer, 2 hidde
n layers
with 96 and 1 n
e
u
r
on
s and 1 o
u
tput layer.
3.3 Weigh
t
function o
f
ANN
Applica
b
le to the indepe
nd
ent control sy
st
em, the wei
ght function
should be d
e
cl
ared in
an e
m
piri
cal
formul
ation
or
a con
s
tants if
the
weig
ht
fu
nction
is not
continue.
The
weig
ht fun
c
tio
n
is then u
s
e
d
in the ope
rating sy
stem
algorithm
i
n
stalled in
micro
c
o
n
troll
e
r
memory a
s
t
h
e
embed
ded co
ntrol
sy
stem.
To calculate
the ANN wei
ght functio
n
,
activation fun
c
tion i
s
n
eed
ed, in thi
s
ca
se the
choi
ce
a
c
tivation fun
c
tion
a
r
e
purelin
an
d tan
s
ig. In
Matlab
pro
c
e
dure,
the
cal
c
ulation
of wei
ght
function of
ANN with
n
e
twork ar
chite
c
ture of 1 i
n
p
u
t layer, 2 h
i
dden l
a
yers with 96
and
1
neuron
s and
1 output layer is expre
s
sed
as the followi
ng:
Y(i) =
purelin(t
ansi
g
(in
put(:,i)*
w
e
ight_l
ayer
1 +
w
e
ight_bias
1
)
* w
e
ight_lay
e
r
2 +
w
e
ight_bia
s
2)
4. Conclusio
n
This pa
per p
r
ese
n
ted ON-OFF and ANN cont
rolle
rs
whi
c
h are ap
plied to the automated
irrig
a
tion
syst
em. Penma
n
-Monteith eva
potran
s
pi
ra
tio
n
and
a
reference of
soil
moistu
re a
s
t
h
e
comp
ared in
put are u
s
e
d
in this co
ntrolling
i
rrig
a
tion syste
m
. The perfo
rmance of such
controlle
rs
were eval
uate
d
and
co
mpa
r
ed b
a
sed o
n
error of bo
th cont
rolle
rs. The sim
u
lat
i
on
results
sho
w
that the error of the O
N
-OFF
co
nt
roll
er can b
e
e
a
sily adju
s
te
d by setting
the
sampli
ng tim
e
of the dead
zon
e
discretization.
Ackn
o
w
l
e
dg
ments
This
work was support
ed
by Hibah Penelitian of KKP3N from
the Ministry of Agriculture
Republi
c
of Indonesi
a
under
Contract no 105/PL.220/I.1/3/2014.
K, March 10, 2014. The
out
hors
thanks to Mr. Vighorm
e
s a
nd Rom
adh
o
n
for their hel
p in prep
ari
n
g
the experime
n
ts.
Referen
ces
[1]
Sampur
no RM
, Seminar KB,
Suhar
noto Y.
W
eed Co
ntro
l Decis
i
on S
u
pport S
y
stem
Based
on
Precisio
n Agric
u
lture Ap
pro
a
c
h
.
T
e
lkomnik
a
, 201
4; 12(2): 47
5-48
4.
[2]
Luck JD, Z
and
ona
di RS, L
u
c
k
BD, Shearer
SA. Reduci
n
g
Pesticide Ove
r-Applic
atio
n
w
i
th Map-
Based A
u
tomatic Boom Secti
on C
ontrol
on
Agricult
ural S
p
ra
yers.
Transactions of the
A
SABE,
201
0;
53(3): 68
5-6
9
0
.
[3]
Card
enas
LB, Dukes MK,
Miller GL.
Se
nsor-Bas
ed C
ontrol
of Irrig
a
tion
in Ber
m
uda Grass
.
Procee
din
g
s of
the 2
005
Ame
r
ican S
o
ciet
y
o
f
Agri
cultur
al E
ngi
neers
Ann
u
a
l Intern
atio
nal
Meetin
g.
T
a
mpa, F
l
orida
.
2005.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 12, No. 3, September 20
14: 74
1 – 750
750
[4]
Dukes MD. W
a
ter Conserva
t
i
on
Pote
ntial
of
La
ndsc
ape
Irri
gatio
n Sm
art C
ontrol
l
ers.
Transaction of
the ASABE.
20
12; 55(2): 4
–25
.
[5]
Mecham
BQA.
Practical
Gui
de to
Usi
ng S
o
il M
o
istur
e
S
ensors
to C
o
n
t
rol L
andsc
a
p
e
Irrigatio
n.
W
o
rld W
i
de W
eb. 200
5, http://
w
w
w
.
n
c
w
c
d.
or
g/ims/ims_inf
o
/practi1
d.pdf.
[6]
T
illett ND, Hague, March
ant
JA. A Robotic S
y
stem f
o
r Plant Sca
l
e Husb
an
dr
y
.
Journa
l of
Agricult
ural En
gin
eeri
ng R
e
se
arch
. 199
8; 69:
169-1
78.
[7]
Johns
on DA, N
a
ffin DJ, Puhal
l
a
JS, Sanches
J,
W
e
llingto
n
C
K
. Develo
pme
n
t and Implem
entatio
n of
a T
eam of Rob
o
tic T
r
actor for
Automon
ous P
eat Moss H
a
rv
ersting.
Jo
urna
l of F
i
eld
Ro
bot
ics
. 200
9;
26: 549-
67
2.
[8]
King RE, Sigr
i
m
is N, Compu
t
ationa
l
Intell
ig
ence i
n
Cro
p
Producti
on, Guest Eds
.
Co
mp
uters an
d
Electron
ics in
Agricult
ure
. 20
01; 31(1).
[9]
Murase H. Artif
i
cial Intelli
ge
nc
e in Agr
i
cultur
e
,
Guest Eds.
Computers and Electron
ics
in Agricult
ure
.
200
0; 29(1-
2).
[10]
Micha
e
l D, Mar
y
l, Shed
d, Davis SL.
Smart
Irrigation Co
nt
rollers: O
per
ati
on of Evapotra
nspir
a
tion-
Based C
ontrol
l
e
rs
. Universit
y
of F
l
orida.
Rep
o
rt number: AE
446. 20
09.
[11]
Bro
w
n
PW.
Conv
erting
R
e
ferenc
e Ev
a
potrans
pirati
on
into
T
u
rf W
a
ter Us
e
. T
u
rf Irrigation
Mana
geme
n
t Series N
o
.2. Arizona C
o
o
perati
v
e Exte
nsio
n. 200
0.
[12]
Alle
n RG, Pereira LS, Ra
es D. Smith M. C
r
op
eva
potrans
pirati
on. Guide
lines for comp
uting cro
p
w
a
t
e
r req
u
irem
ents, Rome: F
A
O. 1998.
[13]
Zotarelli
L, Du
kes MD, Romero CC, Migl
ia
cci
o KW
, Morgan KT
. Step by Step C
a
lcu
l
a
t
ion of th
e
Penma
n
-Mont
eith Ev
apotr
a
n
s
pirati
on (F
AO
-56 M
e
t
hod).
Univers
i
t
y
of
F
l
orida.
Re
por
t numb
e
r:
AE459. 2
013.
[14]
Harriso
n
LP.
F
undam
enta
l’s
conc
epts
an
d d
e
finiti
ons
r
e
lati
ng t
o
h
u
m
idit
y
.
In W
e
xl
er A
Editor
Humid
i
t
y
and
moisture, 3, Re
inh
o
ld Pu
bl
ishi
ng Co., Ne
w
Y
o
rk, 1963.
[15]
Murra
y
,
F
W
.
On the comput
ation of satur
a
tion va
por pr
essure.
J. Appl. Meteor
. 196
7; 6: 203-2
04.
[16]
W
a
rd, Andre
w
D,. Environme
n
tal H
y
dr
ol
o
g
y
.
Second E
d
itio
n,Le
w
i
s Publ
is
hers, 200
4.
[17]
Ya
w
s
CL, Yan
g
HC. T
o
estimate vapor pr
es
sure eas
il
y.
Hy
drocar
bon Pro
c
essin
g
. 198
9; 68(1
0
): 65.
[18]
Mardi
y
o
no, Sur
y
a
n
ita R, Adn
a
n
A. Monitorin
g
S
y
stem on Pr
edicti
on of Buil
din
g
Dama
ge Inde
x us
in
g
Neur
al-N
et
w
o
r
k
.
T
e
lkomnik
a
.
2012; 1
0
(01):
155-
164.
[19]
Harikris
hn
a D, Srikanth NV. D
y
n
a
mic Sta
b
ilit
y Enh
anc
e
m
ent of Po
w
e
r S
y
stems Usi
ng Ne
ural-
Net
w
ork C
o
n
t
rolle
d Static-
C
omp
ensator.
T
E
LKOMNIKA Indon
esi
a
n Jour
nal
of
Electrica
l
Engi
neer
in
g
. 2012; 10(
1): 9-1
6
.
[20]
Gunad
in
IC, A
bdil
l
a
h
M, S
o
e
p
rija
nto A,
Pe
nan
gsa
ng
O. Stead
y-State
Stabi
l
i
t
y
Asses
s
ment Us
in
g
Neur
al Net
w
o
r
k Based on N
e
t
w
ork Eq
uiv
a
le
nt.
T
e
lkomnika
.
2011; 9(3): 41
1-42
2.
[21]
W
enhu
i Z
,
Yamin F
,
Xia
opi
n
g
Ye. Adaptiv
e
Neur
al N
e
t
w
ork Rob
u
st Co
ntrol for Spac
e
Robot
w
i
t
h
Uncertaint
y
.
T
E
LKOMNIKA
T
e
leco
mmunic
a
tion
Co
mputi
ng E
l
ectro
n
ics
an
d
Contro
l
.
20
13; 11(
3):
513-
520.
[22]
Shuh
ua Z
,
Xi
a
opi
ng Y,
Xi
ao
ming
J, W
e
nh
ui Z
.
Ada
p
tive
Contro
l
of Sp
a
c
e Ro
bot Ma
ni
pul
ators
w
i
t
h
T
a
sk Space B
a
se o
n
N
eura
l
Net
w
ork.
T
E
L
K
OMNIKA T
e
leco
mmu
n
icati
o
n Co
mputin
g
Electron
ics
and C
ontrol
. 2
014; 12(
2): 349
-356.
[23]
Krogh A. W
hat are artifici
al ne
ural n
e
t
w
orks?.
Nature Biotec
hno
log
y
. 2
008;
26(2):19
5
-1
97
.
[24]
Umair SM, Us
man R. Autom
a
tion
of Irri
gati
on S
y
stem Usi
ng ANN
bas
ed
Control
l
er.
Internati
ona
l
Journ
a
l of Elec
trical & Co
mp
u
t
er Sciences IJ
ECS-IJENS
. 2010; 10(
02): 41
-47.
[25] W
i
d
y
anto
SA,
W
i
dod
o A,
Hid
a
y
atno
A.
Si
mu
latio
n
of
Auto
mated Irri
gati
on
ON-OF
F
Controller
Bas
e
d
on Evap
otrans
pirati
on An
alys
is
. Proceed
in
g
of 2014 IAES
Internat
io
nal Confer
ence on
Electrical
Engi
neer
in
g, Computer Sci
e
n
c
e and Inform
a
t
ics (EECCS 2014), Yog
y
a
k
ar
ta, 2014.
[26]
Diama
ntop
ou
lo
u MJ, Georgi
o
u
PE, Pap
a
mi
chail
DM. Per
f
ormance Ev
al
uatio
n of Artifi
cial N
eur
al
Net
w
orks
in E
s
timating
Refe
rence
Evap
otransp
i
ratio
n
w
i
t
h
Min
i
mal
Met
eoro
l
og
ical
Da
ta,
Globa
l
NEST
Journa
l
. 201
1; 13(1): 18
-27.
Evaluation Warning : The document was created with Spire.PDF for Python.