TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol.12, No.7, July 201
4, pp
. 5293 ~ 52
9
7
DOI: 10.115
9
1
/telkomni
ka.
v
12i7.401
8
5293
Re
cei
v
ed
Jul
y
27, 201
3; Revi
sed
Jan
u
a
r
y 27, 2
014;
Acce
pted Fe
brua
ry 12, 20
14
SISO P-ILC Algorithm for Output Data Dropouts and Its
Application in Wastewater Biological Treatment Plant
Gu Qun*
1
, Hao Xiaohong
2
, Li Zhuo
y
u
e
1
, Wang Hu
a
1
1
Schoo
l of Ele
c
trical an
d Information En
gin
e
e
rin
g
, Lanz
hou
Universit
y
of T
e
chn
o
lo
g
y
,
Lanz
ho
u, 730
0
50, Chi
n
a
2
Colle
ge of co
mputer an
d co
mmunictio
n,
La
nzho
u Univ
ersi
t
y
of T
e
chnolo
g
y
,
Lanz
ho
u, 730
0
50, Chi
n
a
*Corres
p
o
n
id
n
g
author, e-ma
i
l
: lzgq6
6@1
63.
com
A
b
st
r
a
ct
In order to i
m
p
r
ove the effici
e
n
cy of sew
age
treatme
nt,
w
e
construct P-ILC alg
o
rith
m for
output
data dro
p
o
u
ts.
T
he P-ILC alg
o
rith
m is used
in the aer
ati
on
tank of oxygen
input lin
k, an
d
accordin
g to th
e
actual s
i
tuati
o
n
,
consid
eri
ng t
he d
a
ta
ge
nera
t
ing o
m
issi
ons,
adj
ustin
g
the
alg
o
rith
m c
an
compl
e
tely c
o
n
t
rol
the aerati
on ta
nk of oxygen. After 15
iterations, w
e
can comp
lete
ly cont
r
o
l the oxyg
en i
n
the aerati
on tan
k
volu
me. W
e
kn
ow
mor
e
impor
tant is th
is a
l
g
o
rith
m
ma
y
at
any ti
me acc
o
r
d
in
g to
ne
ed
o
f
aerati
on t
ank
o
f
oxyge
n
su
pp
le
me
nt, w
hen th
e lack
of
oxyg
en, ca
n o
p
e
n
t
he
oxyge
n
fil
l
i
n
g p
u
m
p, w
h
e
n
sufficient
oxyg
en,
close the oxy
g
en fill
ing p
u
m
p
,
to achieve e
n
e
rgy savi
ng g
o
a
ls, ulti
mate
ly mak
e
s the ec
o
n
o
m
ic b
enefits
to
achi
eve the h
i
g
hest.
Ke
y
w
ords
: iter
ative le
arni
ng c
ontrol, outp
u
t d
a
ta, data dro
p
o
u
ts, w
a
stew
ate
r
treatme
nt pro
c
ess, simu
lati
o
n
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
Wate
r refe
rs
to the overall
amount of
water on
eart
h
. Includi
ng the
human
co
ntrol and
surfa
c
e
wate
r and gro
und
water di
re
ctly for irrig
a
tion, powe
r
gen
eration, water
supply, ship
pi
ng,
aqua
cultu
r
e
and othe
r pu
rpo
s
e
s
, as
well as the
rivers, la
ke
s, wells, sp
rin
g
s,
tidal, harb
o
r
and
water a
r
ea.
Wate
r resource
s a
r
e im
po
rtant natural
reso
urce
s in
di
spe
n
sable fo
r the devel
op
ment
of the nation
a
l eco
nomy. With the de
velopment
of
the world e
c
on
omy, pop
ulation growt
h
,
increa
sing a
n
d
expandi
ng
arou
nd the ci
ty, with wate
r rising. Th
e United
Natio
n
s e
s
timate
s, in
1900, the glo
bal wate
r con
s
umptio
n is o
n
ly
11
10
4
cubi
c met
e
rs / year, 1
9
80 is
12
10
3
cubic
meters / yea
r
, 1985
is
12
10
9
.
3
cu
bic m
e
ters /
year. It is ex
pecte
d that
b
y
2000,
dem
and
will
incr
ea
se t
o
12
10
6
c
ubic
meters / year. In As
ia
with the
mos
t
water,
up to
12
10
2
.
3
cubic
meters / yea
r
, followe
d by
North
Ame
r
ica, Euro
pe,
S
outh Ame
r
ica
etc. By 2
000
, Chin
a n
a
tio
nal
water d
e
ma
n
d
is expe
cted
to be
12
10
814
.
6
cubic
meters. Most
of them for the Yangtze
Ri
ver
Bas
i
n,
12
10
166
.
2
cubi
c
meters, follo
wed
by the Y
e
llow
Rive
r a
nd the
Pearl
River
ba
sin.
With
the developm
ent of prod
uction,
the cont
radi
ction bet
wee
n
su
pply and dem
and
of region
al a
n
d
national wat
e
r re
sou
r
ce is
in
crea
singl
y
prom
i
nent.
Along
with
se
wag
e
reu
s
e problem
h
a
s
become a t
opic of
con
c
ern. Acco
rdi
ng to the cl
assificatio
n
o
f
the source
of waste
w
a
t
er,
wa
stewater treatment i
s
ge
nerally
divide
d into th
e
pro
ductio
n
of
se
wag
e
treatme
nt and
sewag
e
treatment. A
c
c
o
rding to the c
l
ass
i
fication of the
sou
r
ce of wa
stewater, wa
stewater treatme
n
t
is
gene
rally divi
ded i
n
to the
prod
uctio
n
of
se
wa
ge t
r
ea
tment an
d
se
wag
e
treatm
ent. Prod
ucti
on
se
wag
e
in
clu
d
ing: in
du
stri
al waste
w
ate
r
, ag
ri
cultural
wa
ste
w
ater
and m
edi
cal
se
wag
e
. Sewage
is sewage
ge
nerate
d
d
a
ily life, is a
com
p
licate
d
mixtu
r
e, ref
e
rs to v
a
riou
s fo
rm
s
of inorgani
c a
nd
orga
nic in
clu
de: the
size
of solid
pa
rti
c
le
s of
floatin
g and
suspe
nded;
gel a
n
d
gel
diffusio
n
in
pure
sol
u
tion.
Acco
rdi
ng to
the quality of
water
po
lluti
on, wate
r poll
u
tion ha
s two
kind
s: on
e ki
nd
is the n
a
tura
l pollution; th
e other i
s
m
an-m
ade
poll
u
tion. The h
a
rm of
water is man
-
ma
d
e
pollution.
Wat
e
r p
o
llution a
c
cordi
ng to t
he different p
o
llution of im
puritie
s is:
ch
emical
polluti
on,
physi
cal poll
u
tion
a
nd bi
ologi
cal poll
u
tion
th
ree
categ
o
rie
s
. The
m
a
in p
o
llutants
are:
the
indu
strial wa
stewater
di
sch
a
rg
e
of
untre
ated; the
discharg
e
of
untreated
se
wa
g
e
; the
extensi
v
e
use of fertilizers, pe
stici
d
e
s
, herbi
cid
e
s
farmlan
d
se
wage; the sta
c
ked in t
he ind
u
strial
wa
ste and
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: 5293 – 52
97
5294
dome
s
tic wa
ste; the
soil
and
wate
r lo
ss; th
e mi
ne wa
stewater. Many
me
th
o
d
s of
waste
w
ater
treatment, g
enerally can
be
divided
into phy
sical
method,
ch
emical
meth
od a
nd
biolo
g
ical
method.. This paper fo
cu
ses on
city life sewage bi
ol
ogical treatm
ent in two sta
ge aeration t
ank
dissolved ox
ygen (DO)
concentra
tio
n
. For the DO pro
c
e
s
s control, incl
ud
ing PID cont
rol,
adaptive
cont
rol, nonli
nea
r
control, ha
s a
large
num
be
r of pa
pers p
ublished.
In rece
nt years, the
resea
r
ch of fuzzy control and neu
ral n
e
twork cont
rol in intelligent control to stabili
ze the DO
value, provid
es some d
e
a
l
with the no
nlinea
r and u
n
ce
rtain me
a
n
s an
d meth
ods of p
r
o
c
e
ss.
Ferrer,
Ro
dri
go et
al. [1]
descri
b
e
s
th
e
co
ntrol
process of
the
m
ould fo
r
biolo
g
ical
waste
w
ater
treatment
pro
c
e
s
s mod
e
lin
g, co
ntrol
an
d optimi
z
atio
n of fuzzy
co
ntrolle
r in
19
98, control
ru
le is
obtaine
d by summing
up t
he op
erato
r
'
s
experie
nce.
Yu and Li
aw
prop
osed [2]
real
-time cont
rol
method fo
r in
199
8. Bong
a
r
ds M
et al. [3] devel
op
ed
a combin
atio
n of fu
zzy
co
ntrol a
nd
neu
ral
netwo
rk
cont
rol
schem
e i
n
20
01. Pu
ñ al
A, Ro
cca E, 20
02 [
4
] and
20
04 [
63] p
r
op
osed
the
expert contro
l system ap
p
lied in se
wa
g
e
treat
ment i
n
the ca
se.
Bonga
rd
s M and Ebel A [5],
descri
bed
in
2004 fo
r the
DO
cont
rol b
a
se
d on
anot
her
kin
d
of n
eural
net
work pre
d
ictive
co
ntrol
scheme. Piot
rowski et al. [6] is the method u
s
ing
hiera
r
chi
c
al
model to im
plement effe
ctive
control of
DO
, and d
e
si
gn
a hie
r
archi
c
al
co
ntrolle
r to
guarantee
th
e DO con
c
ent
ration
accu
rat
e
ly
track a
de
sire
d traje
c
tory, e
n
su
re that th
e ni
trog
en a
n
d
pho
sp
horus in se
wa
ge
can be
effectively
remove
d. Hol
enda et al. [7] using mod
e
l
predi
ctiv
e co
ntrol (MP
C
)
method is u
s
ed to control the
s
e
w
a
ge
tr
ea
tme
n
t
pr
oc
es
s o
f
a
e
r
o
b
i
c
poo
l o
f
DO
con
c
entration, and
achieved go
od
results.
Th
is
pape
r will
co
nsid
er d
a
ta l
o
ss situ
ation,
the P ty
pe iterative lea
r
ni
ng control
al
gorithm i
s
u
s
ed to
control the DO co
nce
n
trati
on. Whe
n
the
lack
of
oxygen, can
ope
n the oxygen filling pum
p, wh
en
sufficie
n
t oxygen, clo
s
e th
e oxygen filling pum
p, to
achi
eve ene
rgy saving g
o
als. Figu
re 1
is a
simplified bi
ol
ogical se
wag
e
treatment flow chart.
Figure 1. A Simplified Biolo
g
ical Se
wag
e
Treatme
nt
2. Rese
arch
Metho
d
2.1. Calculati
on Metho
d
s
of Diss
olv
e
d
Ox
y
g
en Aeration Tan
k
The time rate
of chang
e of con
c
e
n
tration
：
)
(
,
,
o
sat
o
L
o
out
in
o
in
o
S
S
a
VK
S
Q
S
Q
dt
dS
V
(
1
)
Whe
r
e
V
is the volume of aerati
on tank
, It is
meas
ured in
3
m
,
in
Q
is the amo
unt of water,
out
Q
is
a
water flow, It is
meas
ured in
h
m
/
3
,
o
S
is the
con
c
e
n
t
ration of di
ssolved
oxyge
n
, It is
measured i
n
L
mg
/
,
in
o
S
,
is the
concentration of
dissolved
oxygen in water,It is measured in
L
mg
/
,
sat
o
S
,
is the
satu
ration of di
ssolved oxygen
con
c
e
n
tratio
n, It is mea
s
ured
in
L
mg
/
,
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
SISO P-ILC Algorithm
for Output Data
Dro
pout
s
and
Its Applicatio
n in Wa
stewa
t
er… (G
u Qu
n)
5295
L
K
is the absorption co
effici
ent,
a
is the ra
tio of area a
nd volume
( On an op
en
no aeration
pool, its a
r
ea
refers to the
area
of co
ntact of air
and
water, volu
m
e
refe
rs to th
e volume of t
h
e
pool. In an a
e
rated p
ond,
the conta
c
t a
r
ea with ai
r a
nd wate
r of b
ubble
size, bubble
size is a
function
of a
e
ration
eq
uip
m
ent an
d ai
r flow),
a
K
L
for the di
ssolved
oxygen m
a
ss tran
sfer
c
oeffic
i
ent, It i
s
meas
ured in
h
/
1
.
2.2. SISO Pr
oblem
w
i
th P-ILC Algori
t
hm for Outp
ut Da
ta
Drop
outs
For a multipl
e
input multip
le output nonl
inear
system:
)
(
))
(
(
))
(
(
)
(
)
(
))
(
(
))
(
(
)
1
(
t
u
t
x
d
t
x
c
t
y
t
u
t
x
b
t
x
f
t
x
k
k
k
k
k
k
k
k
(
4
)
Whe
r
e
k
is the numbe
r of iteration
s
,
)
(
t
x
k
,
)
(
t
u
k
,
)
(
t
y
k
As the system state variabl
es
, sy
stem input variable
s
and the
output variabl
e of the syste
m
[8].
H
y
pothesis 2.2.1:
T
he
state varia
b
le
)
(
t
x
k
satisfie
s
Lipschitz conditio
n
, the p
r
e
s
en
ce
of
)
(
x
a
,
)
(
x
b
,
)
(
x
c
,
)
(
x
d
su
ch t
hat for
any ti
me
]
,
0
[
N
t
, we have matrix
F
K
,
B
K
,
C
K
,
D
K
, meet
)
(
)
(
))
(
(
))
(
(
2
1
2
1
t
x
t
x
k
t
x
f
t
x
f
f
,
)
(
)
(
))
(
(
))
(
(
2
1
2
1
t
x
t
x
k
t
x
b
t
x
b
b
,
)
(
)
(
))
(
(
))
(
(
2
1
2
1
t
x
t
x
k
t
x
c
t
x
c
c
,
)
(
)
(
))
(
(
))
(
(
2
1
2
1
t
x
t
x
k
t
x
d
t
x
d
d
.
H
y
pothesis 2.2.2:
T
he i
n
i
t
ial co
ndition
s fo
r n
online
a
r
system
s
o
f
)
0
(
k
x
,
)
0
(
d
x
meet
)
0
(
)
0
(
d
k
x
x
. Where
)
0
(
d
x
is the state vari
able
s
in the
k
initial value,
)
0
(
d
x
is the initial
expectatio
n
value.
H
y
pothesis 2.2.3:
Fo
r the
desired outp
u
t for a given
)
0
(
d
y
, the control a
l
gorithm.
)
(
))
(
(
))
(
(
)
(
)
(
))
(
(
))
(
(
)
1
(
t
u
t
x
d
t
x
c
t
y
t
u
t
x
b
t
x
f
t
x
d
d
d
d
d
d
d
d
Whe
r
e
)
(
t
x
d
and
)
(
t
u
d
are respe
c
tively the desire
d
output and t
he de
sire
d st
ate.
For the SIS
O
linea
r sy
st
em (1
), the
P type
iterative learni
ng
control al
go
rithm is
as
follows
:
)
(
)
(
)
(
)
(
1
t
e
t
t
u
t
u
k
k
k
(
5
)
｝
，
｛
1
0
)
(
t
, if
0
)
(
t
indicate
s that the data lost.
1
)
(
t
we kno
w
data integrity.
)}
(
{
}
1
)
(
{
t
E
t
P
,
1
)}
(
{
0
t
E
Whe
r
e
is the learnin
g
gain
factor,
)
(
)
(
)
(
t
y
t
y
t
e
k
d
k
as th
e system tra
c
king e
r
ror.
Theorem 1
:
To meet th
e assum
p
tio
n
s of n
onlin
ear system 2.2.1-2.
2.3
(4), (5
) the
learni
ng cont
rol
al
gorith
m
is
iter
ative, when the
sy
stem outp
u
t da
ta loss, for th
e lea
r
ning
ga
in
fac
t
or
on the numbe
r an
d all the time the
t
and iterativ
e
k
, we have:
1
))
(
(
)
(
1
t
x
d
t
d
.
For any
]
,
0
[
N
t
,
We know that
0
))
(
)
(
(
lim
t
y
t
y
d
k
k
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: 5293 – 52
97
5296
3. Results a
nd Analy
s
is
3.1.
The Desi
gn of ILC-DO Con
t
rol Sy
stem
Acco
rdi
ng to
the dem
and
of aeration ta
nk of
oxygen quantity,
we usin
g
P-IL
C
a
l
gorithm
to control the
oxygen quan
tity. The P-ILC algo
rith
m i
s
used in the
aeratio
n tank of oxygen input
link, ca
n with
the aeration
tank of oxygen dem
a
nd i
s
very good,
accordi
ng to
the amount of
oxygen aeration tank an
d decid
ed to give the am
ount of oxygen aeration
tank inp
u
t, and provid
e
a theoreti
c
al
basi
s
for the
energy savin
g
our pla
n
,
an
d ensure that
the aeratio
n tank of oxyge
n
is
sufficie
n
t, the sewage tre
a
tment the high
est efficien
cy.
Each in
put ox
ygen process, given a de
si
red oxyge
n
d
e
mand
d
y
, looki
ng for the
co
ntrol
input
)
(
t
u
k
, made in the co
ntrol of the
ac
tual in
put
amount of o
xygen,
1
k
y
and
d
y
.
Con
s
id
erin
g the po
ssibility
of the actual operatio
n, it
can b
e
con
s
i
dere
d
, dema
nd co
ntrol in
put
sele
ction of a
e
ration tan
k
of oxygen, each in
put
to the aeration tank of oxyge
n
pro
c
e
ss, when
the amo
unt o
f
oxygen inp
u
t
oxygen de
m
and
set a
rrive
d imme
diatel
y clo
s
ed
oxygen inp
u
t valve,
the am
ount
of oxygen
a
nd the
de
si
red traje
c
to
ry
of the
inp
u
t
(i.e. setpoi
nt) exi
s
t e
r
ror
:
d
k
k
y
y
e
1
1
,Where
k
is th
e input
oxygen nu
mbe
r
.
Figure 3
de
scrib
e
s the
structure a
nd
pro
c
e
ss of IL
C co
ntrol met
hod.
Figure 2. The
Structure an
d Pr
ocess of ILC Control M
e
thod
3.2.
Simulation
We u
s
e the si
mulation hyp
o
thesi
s
.
Figure 3. First and Secon
d
Control Inp
u
t
Fi
gure 4. 20 Iterative Lea
rn
ing Co
ntrol th
e
Tra
cki
ng Curve
Figure 3 is
a first and a
second
cont
rol i
nput (th
e
amount of o
xygen aeration tank
requi
re
d), Fig
u
re 4 for the twentieth iteration lear
ning
control alg
o
rithm tracking
curve, we can
see
that afte
r
20 iteration
s
,
the le
a
r
nin
g
control, the
target curve
t
r
a
c
king, ca
n co
mpletely
cont
rol
0
5
10
15
20
25
0
0.
2
0.
4
0.
6
0.
8
1
1.
2
1.
4
i
y(
i
)
y
d
(i
)
y
k
(i
)
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
SISO P-ILC Algorithm
for Output Data
Dro
pout
s
and
Its Applicatio
n in Wa
stewa
t
er… (G
u Qu
n)
5297
the numbe
r of oxygen in the aeration
tank, this
algorithm
can
control the o
xygen effective.
Figure 6 is the curve of error co
nvergence curve, it can be seen th
at the algorithm will converge,
the algorith
m
is prove
d
to conver
g
e
, the algorith
m
is reasona
ble.
4. Conclusio
n
(1) In o
r
d
e
r t
o
ma
ke
the
se
wag
e
tre
a
tment effici
en
cy is hig
h
e
r
, the a
e
ration
tank of
oxygen we
re analyzed, an
d the P-ILC al
gorithm is
u
s
ed in the aera
t
ion tank of oxygen input link,
and a
c
cordin
g to the actu
al situation,
consi
der
i
ng th
e data ge
nerating omi
ssi
o
n
s, adju
s
ting
the
algorith
m
ca
n
completely contro
l the ae
ration tank of
oxygen.
(2)
Throug
h
the simul
a
tio
n
experi
m
ent
to validate the mod
e
l an
d its corre
s
p
ondin
g
algorith
m
, the
results a
s
sh
own i
n
Figu
re
4, sh
o
w
n in
Figure 5, not
only ca
n the
aeratio
n oxyg
en
deman
d go
o
d
trackin
g
, and
can a
c
hieve pe
rfect
tracking
oxygen dem
an
d goal afte
r 15
regul
ation. M
o
re imp
o
rtant
is this alg
o
rit
h
m may
at any time according to nee
d
of aeratio
n tank
of oxygen supplem
ent, whe
n
the la
ck of oxy
gen,
can
ope
n the oxygen fil
ling pum
p, when
sufficie
n
t oxygen,
clo
s
e th
e oxygen
filling p
u
mp, to
achieve
ene
rgy
saving
g
oals,
ultimate
ly
make
s the e
c
onomi
c
ben
efits to achieve
the highe
st.
Ackn
o
w
l
e
dg
ements
This re
se
arch
was su
ppo
rted
by The Nation
al
Nature
S
c
ien
c
e
Found
ation o
f
Chin
a
No.61
263
008
, the
Nation
al Natural
S
c
ien
c
e
Fou
n
dation
of G
ansu Provin
ce
(G
rant
NO.
1112
RJZA02
3, 1212
RJYA
031).
Referen
ces
[1]
J F
e
rrer, MA
Rodri
go, A
Se
co, et a
l
. Ener
g
y
sa
vi
ng
in
the aerati
on pr
ocess b
y
fuzz
y log
i
c
co
ntrol.
W
a
ter Science
and T
e
ch
no
log
y
. 1998; 38(
3): 209-
217.
[2]
Rue
y
-F
a
ng Y
u
, Shu-L
i
a
ng
Li
a
w
,
C
hen
g-N
a
n C
han
g, et
al.
App
l
yin
g
r
eal-
t
ime co
ntrol t
o
en
ha
nce th
e
performa
nce of
nitrog
en rem
o
val in th
e co
nti
nuo
us-flo
w
SB
R s
y
stem.
W
a
t
e
r Scienc
e a
n
d
T
e
chno
logy
.
199
8; 38(3): 27
1-28
0.
[3]
Bong
ards M. Improvin
g the
efficienc
y of a
w
a
st
e
w
ater tr
eatment pl
ant
b
y
fuzz
y co
ntrol an
d ne
ura
l
net
w
o
rk.
W
a
ter
Sci T
e
chnol.
2
001; 43(
11): 18
9-19
6.
[4]
A Puñal, E Ro
ca, M Lema. System for moni
tori
ng a
nd di
ag
nosis of an
aer
obic
w
a
ste
w
at
er treatmen
t
plants.
Water Research
. 20
0
2
; (36): 265
6-2
666.
[5]
EF
Carrasco,
J Rodri
g
u
e
z, A Puña
l, E R
o
ca, JM
Lem
a. Dia
gnos
is
of acid
if
icatio
n
states in a
n
ana
erob
ic
w
a
s
t
e
w
a
t
er tre
a
tment p
l
ant
usin
g a fuzz
y-
bas
ed
e
x
pert s
y
s
t
em.
Control
Engi
neer
in
g
Practice
. 200
4;
12(1): 59-6
4
.
[6]
Bong
ards
M, E
bel
Hi
lmer T
．
Predictiv
e co
nt
rol
of w
a
stew
at
er w
o
rks by
n
e
u
ral
netw
o
rks
. Procee
din
g
s
of the W
o
rld Automatio
n
Co
n
g
ress. 200
4; 1
7
: 397-4
02.
[7]
R Piotro
w
s
k
i
, MA Brd
y
s, K K
onarcz
ak, et al.
Hiera
rc
hica
l d
i
ssolve
d
o
x
yg
e
n
contro
l for ac
tivated sl
udg
e
process
e
s.
Co
ntrol Eng
i
n
eeri
ng Practice
. 2
0
08; (16): 11
4-1
31.
[8]
Hao Xiao
ho
ng,
Gu
Qun, Xia
n
Jun Du, Xu
W
e
it
ao, Du
Xianj
un. C
l
o
nal
Selecti
on Al
go
rithm Bas
e
d
Iterative Le
arni
ngC
ontrol
w
i
th
Ran
dom Dist
u
rba
n
ce.
T
E
L
K
OMNIKA Indones
ian J
our
n
a
l of Electric
a
l
Engi
neer
in
g
. 2013; 11(
1): 443
-448.
[9]
Sutikno, T
o
le. F
P
GA for ro
botic a
ppl
icati
ons:
F
r
om an
droi
d/huma
n
o
i
d rob
o
ts to artificial me
n
.
T
E
LKOMNIKA T
e
leco
mmunic
a
tion C
o
mputi
n
g Electron
ics a
nd Co
ntrol.
20
11; 9(3): 40
1-4
02.
Evaluation Warning : The document was created with Spire.PDF for Python.