87vol.2. N.1, Fall2008, Control Engineers Magazine, Iran

1
ANFIS , 0$ 456 /0 ()* +,-. %& ' !" # $
PSO :" $ '9,
?(@0 & >!<$ = ! ' & ';
20
':5 FE. B. '$- ' CD$ ( E

G! , 5 '(0 .
26
5%$ M; 5 L! B. +" , I( ' J$ D0K ':E!
NO ' PE :!0 6
40
. , EE (C$ '( $ 9, /.$ . $ ':5 0 6R
= E(, S &
49
>& UV<$ W; E. ; (

0N J 'X@ N ' ,E0 =
56
J* -! -IN! Z+E " FE. CD$

(D& = <5 WCC6 :(5
65
/.$ ) ':" N ',: & + ; , BP $. \6 ' ( 9, ]
N^ F(L!
N: 5 (X@ \! _.

JournalofControl,2008,V.2,N.1
1 2 1387

.; /$ _ V a & ' , '4N #CC=$ b(N J($:$ N W. <,` & '

:N (: +$
J( 4N d6 + : . ?S I ,
.FE. ':.; , ':. (. ':E! (1
; , a,C CD$ 'FE. D0K , D0 'FE. +g 5< 'FE. 6R , \ =$ (2
.h., P 'FE. $ , < EE 'FE. 5%$ 'FE. 'FE.
.$ , N,$ (3
.'E. #&iR , >$ 'FE. , ?S I (4
+: 'FE. D0 UV<$ , ( 'FE. J +EN / E 'FE. +g P@ .$ (5
.'I(,. 'FE. , CC6
:N (: +$ + +" : . (5 jN N N$ , h.,
J( Si& '
',N , W( 'FE. (1
.'^!$ , ( '(5 \ P@ '(5 (2
.NP ',5 , kV. '(5 (3
.0 ',0 , \CN , \6 'FE. (4
.l ',N h(:$ , !$ (5
.. , WE(: /= .; (6
.!$ '^!$ .; (7
.! , '%S 'FE. (8
.P@ ' , $V $&iR 'FE. (9
.I .; (10
. 9:" 'FE. (11
J(0" b(N , #)C $ (" \P #& ?S I , .; /$ ': P5 +. , +L<,` :
B" N,! #@ 0 #)C W. <0 .(N .
J( 0 <,` , & ',.

LN0 =X@ N$ #)C . , ;$ =N W5( , < #&iR >E ' .(5 . control@isice.ir
.(N P www.isice.ir :B"
1 2 1387

JournalofControl,2008,V.2,N.1
1-19 =X@ 1 2 1387

01 /)*% + ,-. &' (

!"# $%

PSO 6#
(5-2 ANFIS - /1 2! 34
?(@0 & >!<$ = ! ' & ';
.RJ(!%N 0 P@ L<N l <N
M_aliyari@eetd.kntu.ac.ir, Teshnehlab@eetd.kntu.ac.ir, Sedigh@kntu.ac.ir
_____________________________________________________________
' &. FN ': , N: ';(. #@ !" ' !C J( : %+
+" 9:" J(N , $ , ANFIS ':5 %& , 0$ 456 () * +,-. %&
I!N" ' .F(N +( : :" , #m , ':. ; F(L! B.
';, J( W(;N , F(N !" aS n!l5 ' , 0 '(n
ANFIS 9:" ' J! 5 , PSO B. $ ';< 9, F( +<N , F( E(C F
.W. & ; . (N$
- %& - 0$ /0 456 !" N: ';(. () * +,-. %& ' : :' '^,
.5( P.$ J! 5 , #m , ':. ; ANFIS
Abstract: In this study we predict air pollution data by using Multi Layer Percepteron, Time
Delay Line, Gamma and ANFIS by gradient free learning methods. This paper, using real data
for Arak city during Oct 2003, the following pollution parameters are analysed: Co (Carbon
Monoxide), PM10 (Particulate Matter).
This analysis is carried out in two stages: Predictability analysis using Lyapanov, Exponent,
Correlation Dimension and Rescaled Range Analysis (R/S), Prediction using Multi layer
perceptron, Time delay line, Gamma and ANFIS. Also, a comparative study is performed using
the different methods employed and prediction results are provided to show the effectiveness of
the predictions.
Keywords: Multi Layer Percepteron Neural Networks, Time Series, Air Pollution, Time Delay
Memory, Gamma Neural Network, ANFIS, Particle Swarm Optimization, and Extended Kalman
Filter.
______________________________________________________________________
ao. o. +:o. :,o , ';,
/o.$ o o;$ o( o '<CN ', . , d= ' W.
#@ (;LE( N : ';LE( ' ' I
, N S a& jiR W; " , W5 ?R !
( (N
5 'E( /.$ J( G-. . K , ' ';LE(
o o!" o o o aS , 5 S X.
$ , 0
o o& o$( qoPDS +" 'o= +oEN \! 9, J(
()o oE o IoN o 'D0 +" ,i& , W 0
.
oP o. o o F(oN P. !C J( #N J( $
oE No: '. #@ J( #@%0 . ' (
.F jN ;* N" G-. , F(:-
r ; s 5 S X. (
'o oNN W&o. FoN W&. FN #@ 'i 2001
Journal of Control, 2008, Iranian Society of Instrument & Control Engineers
%9 .1
\& : 'E E, g E ("5 ( !"
oD0 oK o( ' o o oNL( o J(
()" <N , LNL* E( , W. \<E E
o0 o5I E '(I IN > D J( N UV<
'o("5 oN #)oP \o6 , J( G-. W
\S . V ' ':E! aS ' ' #@ '
o oN WoR o<5 o +g ' \& #Op$ W. m
oo oo oo;$ 'oooo<CN , #)ooP Joo( JooV( Foo doo& ooK,
' \& J( : . !" LNL*
, ':.; 'P ,0 N;N #!$ (I5 5 4N IN
Jo : o 5 4N +" 0 d& L( \& 'E
.[2 1] I( F , ,
http://www.isice.ir
+( ?S I , +.; J

N 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
?(@0 & >!<$ = ! ' & ';
<6 ( ( .2

q6iDo@ +o: oR #< : ' + S F . W<
, PM-10 ?o P o( o!" #m o 'o oNN 'o .oN

#m 'o NN 6, W.m \S . CO J E
. 1ppm J E , P g ? P
+o 'o& ( , ' ' #@ N$ N N: '.

.[12]
m3
J( ) C P X ' ? P ( !" #m
\CoE o Noo so ( N: ';(. \ =$ , (I

$
o$ d& q)P #m N J( .( N$ IN h( #@ #m
o o (oo( o. 'o , oNoE, F ! R , N
;o 'o < #m J( C * , ; ' + uN
FoE. o ,0 FNo o q5@ , N & \S ',, N,
Opo$ LNoL* +o \%X$ [4 1]h .. 4N $!"
Wo!6 .Fo N (N$ ~ +" ',, ', , F . , F
? P #m .W. W=@ +" #w L( , N vn ()" J(
4" 3' D0 K +g @0 N$ N: '. (

Wo. Jo 6 , 6' %5 N ,$ , 5'E( K
Jo( #ox$ J(o .oNo ? P * $ NO * J q)P

J(o$ \o< : o( , oo ' g ( ' +:
.[12] F 'I(N
. J( s
, oN 'o :o6 : 'oE ; CN N: ';(.
Jo oE +I .W. FPR , uN +, ': JE
oP ) o ,0 -',, o o +o$ o W!6 . .;
+" %20 o oo : JoV$ o. J$ + 350 +; &%
, on o ',, o N , \ =$ (;. J* (.
N!oo. >oo$$ J(oo , ooNoo oo!$ oo< Woo. 'oo0oo. /oo.$
o .[13] oE No: 'o. o(" IoN +" o ,0 , FoE. 6
Jo(
( >. J E . ,I5 CO W4 K 0.03ppm
'ooP!loo5 Woo : 'oo. oo 'oo Joo LooE, Noo: ';(oo.
Do. +o 'o Wo5 :oN +È ;$ i0

( d& N
zo%0 0 & '" d6 : 'E W. 0
Jo E \<$ , +È \6 J #S
o Jo LoE, L$L$ s$ J( ':E! .W.
o()" J( : N #@ n O 'I v%& E . ',
L !" $ '%S '("5 : P., E ':
oEXS J(`N" z%V S,& , S ';( W. NE <
(oC o No: 'o. o( !o 'o NN W!6 .,
o do& : J( ') W4 K {P J5 S .N '@
.W. E, 0 n
. 'W!P5 a

N , '( W S ' +$ ( WS
, J LE, . @ ' : (
oE( o Wo. F; $,| !" l5 $
W ( ' hS, . + (n
#D0 : , (n #@ $?S J( * . S $
r" PDS . 3 W J( .L S .
oo do= o 2 V !C J( .W. +$ !"
.N FEC$ 5%$ ,
oX ' (;, , F(: +" #,I , N: ';(.
>o D J(o; F(,o oo , ':oE! o. +" :
' 'P , %& ' 3 V .F(N . E
qoPDS o Xo. I(oN o 'o o +o* .oo jN .
LNoL* , ANFIS o o 4 oV . F0 . 4N
o N:" ( J( . #@ ;N" ', +$N NLg
+o& o o o & o; 'o ( $ ';, , +" 9:"
ox$ \o =$ +o:" +& +" : WPS, J( UV<
PSO F(oL! o 'o%V }o|$ 5 V 0 J
Wo(I WoE$ Jo( o! . aN 7(R/S \ =$) #x$ :6 '
'oo 'o o ':o.o 0" V W(;N , F0
o o IoN oo Wo; a:) 'o oP$ \S6 'L(
.F( a
N 5P
o 'o joN F E( F; 5 J( + UV< : G .

+:" : ;N" g +I jN : 4 W. NL*
o* o$ o F E( W(;N , W. X. 8LEP JV$
Time Series
Nonlinearity
Chaotic
7
Nonstationary
8
Seasonality
9
Rescaled Range Analysis
10
Correlation Dimension
5
6
Journal of Control, 2008, V.2, N.1
Particle per million
1 2 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
3
?(@0 & >!<$ = ! ' & ';
&' ( .3
o oo +o$ (" #@ +$ ', C5
.F X. 1 N! 'N +:" : 0 ( #@ #
o( o !" N: ';(. F(( V DN
Jo( o; d& D0K ';! : X. J( N () D0 K
(o o ) oV( Fo; , o@ 'o. , 'o (2) , (1) ';
o0 o o0 ( (;! J(; : ( , 0
'. ' F DN W. F.$ (5%$ #@
Wo. o 'o FE. W. +<N , (L! ':E!
, oE oo \S G @ '. : V( F;
. %& ' \
+o 5o%$ +Io o oI H oC o* .oN X. I(N
o" o , o +oEN + : JL! ., %& '
v,o$ , UV<$ W; N H , H (C J(< , W.
\o<$ oNo0 N L(( (;! . : +EN + N W.
oP N H , Wo. 'oN LNo H o .W. " N IN L!
.,E , ,< , ("5 6 , %& ' .W.
9o , JLNo +" oC J(o< o o W. (R/S \ =$)
,o< #o@ o ,0 ',, () : ',, %& '
.(" W. v,
o ,0 _ V '() : & : G , v| '+:,

, 'VE ,0 ( , _ V '() J . \<$
.(" , (,E) < (nx$
';o, +o" o do& o%& 'o '( \S 0.
.oo N J( 9:" W; +N ' F%$ , +N
aoL, o o 9:o" ',, : '. (
Wo; oPS, o ,0 o o(I J(o(IN, J(o; ,< W6
. ( D0 +I
( z%V! & N: ';(. %& ' :
'o o$ 'o ' 'L( /( ( (E( \S : D(
o( o oL Jog , ,o Jo : F5 ( .i
.[15]W. X. E (: N: ';(.
#o@ 'o o o o +$ !" +" :
joN 5 ' [24] 4GRNN ,3 MLP 2RBF ' /.$ 5
: .[32-25] o MLP 9, Bo. ;, < . aN !" jN
?X $ '" #&iR CD$ RBF : +$ L( ' jN
o o5 o oN:, o!" Do. o ' aN
.[52]W.
oD0 ';o, : o; qo& o%& ';!o : Xo. o
o o o.$ oE(C oN . [29 37-33] 9I
oL( o( .[30] %& ' '" ';$ ( [25] %& '
oV( Fo 'o o +" oE(C , LoE P +:" &
oP( oo oL o( : oI '(C W. <
o 'o. Jo( oE(C F; N NL" z0 @ N: '.
o< LNLo" ' oV( Fo 'o. W. V( F '.
Joo LooE aoo& \oo! oo LNLoo" (Ioo5 Joo( oo Woo.
, (3) '; oo : Joog .oooo (ooV( Foo 'oo.)(oo
oE( o= o WoEN LE P + X a& $ 4 N)(4)
o (o 5 4N +" ? D S ; #& ( =! +C
o o" J( , N _P| EN r" z0 F(
. x +: $
'oN Wo oC (R W. N: '. W LN N! 'N
oC , 0 , , ("5 + r" LN N!
o 'o o6 o \S , UV< W LN N! 'N X
FoE. (n- X@ (IN W (C N! 'N . #
o o Wo. (n + # $ , W. _P| r"
';(o. 'o o so N! 'N .W. J K #
Wo. o F. 'N $ W. F.(6) $ (5) ';
#o o$ oo LNo Jo( , Wo. o(IN Xo@ N! 'N
Jo( o o o +<NN J( .W. # (,,
oP$ oP() *o No: :o oE( # $ #@
.(n #@ (
'o o 5SOM ' 5 ' :
.[38] ( . b(N E N!.
39] oN$ 'o o$ ' )P MLP '
o X. J! ' : X!. E ' ' . [51-41
Radial Basis Function 12

Multi Layer Perceptron 13
Generalized Regression Neural Networks 14
15
Self Organizing Map
11
Lyapanov Exponent
1 2 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
?(@0 & >!<$ = ! ' & ';
x t ',, P( .( W. (8)\ $ F %&
o456 o( x t p $ x t 1 P( +" S ',,

oo oo%& Joo oo oo oo oo0m
o456 Wo LN !C J( b(N , < ( W.
.[40] J %& '
p oo : 0oo$ /oo0
aoo o( 'oo oo oo oo ooC oo oo \oo@6 x t 1 $oo
o oD0 #P ':. F B. 9:" 0.
( . $
o o( o!C J( #)C +" 5(P.$ L( ' :

'oo #)oC o&
Jo( .oN [87-85] +$
.N5 !" '. ( %&
,o< (8)\ . 6D0 <N G , !,IN +(

o5 o4N o 'o IN B( C +<N 7N: 0$
. [8 7] W.
o+ I!N 'E( aS ; '
'o o o 'o;(,, 'o F(o" o Fo(oN ,0 , ',,
',, +& IN ! C , \|X$ 456 0$ /0 : X.
o& o o IoN xt xt 1 ',, P( F( &
:W. (: #@ ',, F(N
X
> x (t )
, x (t ) x (t 1) , x (t 1) , x (t 1) x (t 2) " @
(1)
o +" ! + ( : X. &N ( J(
o;$
() S x t x t C xt xt 1 '
.( +,N b N (), F( 54N ',, ;*
)* % + ,-. &' ( .1 .3

o() o* o o o() o* +,-o. o( MLPo%& 'o
Wo. oD0 q)oP o , ',, () N IN V<
Jo( o
N" : , oNo5 oS o ,0 , ',, o() J N (')()
(o)o() J( s$ ,0 , ',, FCE K R ()()
joN o oE IoN o ,0 o() o(o IoN +; ( XV (')()
. D0K ( D0 (=$ h$ F0 ,0
W!6 .( IN 1/$ qi ( J*
o >o$$ Jo , N:. a, () '',, , () ; ,0
#oo@ J(oo, oo0" oo!, aoo. oo() 'oo',, a, oo() oo ,0
$o& o .oo \o<$ o PS, . 0" () (') ,0
#o@ Vo< oE o( o !Lo. +o( N, ,
'oR [9] F( #@ ' 2
?o< oP( . xt 1 xt C +" ,0 F( 9:"
'; b(N F(N
N: '. ' (MLP1),
.(L( () () : W. * :)

W. ; , N X. L. jN , q& MLP%&
Wo6 W5 E E (;!L. jN ( N I$ F :
x t 1 2 x t x t 1 F(o" G-o. .W. " (10) , (9)
ooL( oo. , (,oo< ooE oo Woo. oo oo* Woo. :) oooo
o F5o o4N vo D o ,0 +o& o (MLP2)a, ?< P(
o W. W. :) W6 W< E E (L.
l. . 0 ';N:, BP 9:" 9, 'R
a, o. o , 2Po$ ';!Lo. , o. o .(VoEE *
b(N . x t 1 2x t x t 1 C ,0
'o() : 'o W5 E P$ L. .( 3 D0 ';!L.
o 'o () a$ W. m \S) W. " (12) , (11)';
o< o W< E D0 ';!L. , N h(:$ L( ()
.(W. X. DS $ (L. h$ : ',, I

. [4 3] N
2! 34 (<6- .3 .3
2! 34 ( .2. 3
0o$ o$ o( + S X \S WES DN
+o& o o S 'o;N: 'o : oE( o%& o jN J(
oE 9, +" o! F(oN + 456 aS &N ',,
#@ $ J( F (7) \ DN F X. ',,
'o;N:, o ',, : o456 Jo( o F( %S WES J( .
oS Xo. (: $ N J( . \& 0$ /0 (
o( 'o o Jo( () E W S : FN $ FP$ FE.
o$ o456 $ ';5 J($ X. , J($. : ( ,
.F X. N
. [3]( 5 >P<0$ /0 456 4#
o;N" ',, o o() o* 'oo 'o o W!6 ( (
o$ +o oN F & ',, '

(
'o o o o (13) \o #o@ F( '0$ /0 $
o ',, J( W. +<N (7)\ 0$ /0
X.
1
h(z ) a (!

( h$ () 5 ( : z 0$ %&
(BP) Back Propagation 21

Time Lagge Feedforward Network (TLFN) 22
Fully Connected 16
Function Signals 17
Error Signals 18
Shot-Term Memory 19
Tapped Delay Line Memory 20
1 2 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
5
?(@0 & >!<$ = ! ' & ';
oN F(oL 4N 0$ /0 456 ( . 5 k

:F( N 5 S (15) \ >$
G(z ) '
(16) \
x ij 1(t ) x ij 0 (t 1)
x ij 3(t ) x ij 2 (t 1)
o '0o..(!oP MLP o o( \) $. D0 K
y ij (t ) w ij 0x ij 0 (t ) w ij 1x ij 1(t ) "w ijk x ijk (t )
o. o . W. 2 ( WLN ( " \ J(
(2)
' !6 N FE( FV #.=
$.F \& NLN F( .= () * +,-.,
:F( (14) \
, x (t 1) , " , x (t m 1) @
GT > X @
Net 1 G1 T X , Net1
mu1
num
nu1
F Net1 F GT > X @ mu1
nu1
num
G 111k G 112 k " G 11nk

1
G 21k G 122 k " G 12 nk
G 1 G 131k G 132 k " G 13nk

#
#
#
#
G 1m 1k G 1m 2 k " G 1mnk
G 2 11 k
G 2 21 k
G 2 31 k , o (t ) G
#
2
G n 1 k
(4)
( 5)
:F( F N ()
o C,
1
2
2T
F(Net 1 )
1un
nu1 (6)
[3] ,-< )2! <6 (? ("% .4 .3

G. ( ';N:, F(( TDLN 0$ 6, + >$
Gk (z ) 5 >$ aS
Fo; N . ', " \S WES '$ + : 0. J(
IIR ( 4 FIR( N$
G(z ) 5 0$ h$ J( W. +"

(
@& '
( D0 5 ( vVN P ( (
. h j
G(z ) ' FIR ' 5 q)P J(
oo ooN$oo oo456 oo( 'oo jooN J(oo$oo. +" ,oo< #oo.= ,
o( o( (o o! . Wo. TDL o( o o
G( z)
z 1
o 'o o ( 5 +EN : 456 FE.
e (t )
i
on o o! .o 5 W. F; ;N" n ';N:
Wo , ( n ( P( N F; (
oN o(" 'o o , +," W. W; ' N*
,o< o.= o|6 o6 . 0 ' W : (
oS
z 1 o 5 G(z ) o 5 'o o (15)\o o J( '
(17) oo$ (14) '; oo oo$ Woo.m \ooS F(:oooo oooo
(7)
oi ' :N PS, C d i +"

.
. . \S WES TLFN : a5
TDL .N vVN
.F X. D0

( n ) o i (n )) 2
o( oN N$ 5P wS J( 0. J( .
qo$m 0o. o( (15)\o 0o. o W. +" F; N 5 IIR
#oP jo
':o.Fo : 9:" W; \S N IN
(
o (nx$ 0 'P $. r , D0 ' 5
Foo o DNo F(N ;N:, J(
[86] TDL 56# .5 .3
(d
oN , ro o Wo. r 2 ' '0. ( /.$
(14)\ W!6 ! '( >$ ( N$
. IN ,0 J(* FP$
1
2
qo(C$ oR o oN$o 3 ( WLN ( N: x$
( 3)
\oS o6 o o Wo. ," ,0 ( ' #.= J( !
: [3] + NL( F; '$ ( 1997
ijl x ijl (t )
l 0
L
j j 1 . . +<N
, ,o< o ( W. (N N (13)\
^h `
. [3] N vVN 1CC6 C '; N :6 : E D0
#
#
x ijk (t ) x ij , k 1(t 1)
> x (t )
( W. (N
' 5 N ( +& N: :6 N* J!N
x ij 2 (t ) x ij 1(t 1)
\o DN P$ r , E, ( 0. J( .W.
Real Valued Kernels 23

Univers Dynamic Mapper 24
Myopic 25
Finite- Duration Impulse Response 26
Infinite - Duration Impulse Response 27
1 2 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
?(@0 & >!<$ = ! ' & ';
'oP o ?oS 'oo456 %& ' 456 0. FN
F(:oo o ,0 () : (;N:,) wij #x$ JP$ , #.= 6
qio .oo &$ +EN 456 < EN $- } D%
:F( F WS (15)\ J( : \S
o q)o6 o ,N ( qCS \S X ' UV ( W. J
.W. :, : ( $. ," R0
m /o. 9, J( .W. r= JLN 456 jN : (

+Ioo qioo m Joo( oo(D oooo oo.= oo4=! oo
FoN o 'o) W. \S W&. m R ()" : ( !"
2
Jo( : ko0 o RoCN 9, Jo( .(Wo. o 'o N W&.
#o& o ooN .. , W=$ E ($
?o& LNo &oN o
m ,=
m o J(o .o 95 #&iR ;

: jN , 456 jN J( .WE IN 456
o.= +o$o W. , NN
m 46i : G ;$ ( , . j,
wy ij
wy ij (t )
wy ij (t )
wx ij 0
wu i
wE
wV
(10)
'V
K
wE
wV
wE
wO
wNet 2
u
u
2
wO wNet
wV
wE
wO
'V
D O
'( O P(S, , 6 .W. x$ \S 456 J( ?& (
Fo h(oS, , oNo So 456 '$N)R # nL +EN ',
o o;$ , o. o o`(, Jo( o
( .N 95 $,: $W
o +o<N (17)\ .W. (n+ #&iR : DCN ( + 0m
Fo4$ o 456 J( ?& 0 aN W. 456 jN J(
.W. x$ \S
P C
J( ' +" d& ; \& : L( (

N,5 >(P ' IIR , FIR N: 0$ 9, ,
ooR (;(oo(N +oo" oo oo( , ':ooE! ,oo= Woo5 ooN
.L( \& , 9:"
o FIR joN 'oo 'o( >o$ W S IN N ' \(, $
, +: I ':6 + F IIR ' W(I
/o.$ oN o \o(, \o< Jo( 'o \o6 o( o( o$ WN
.o 5P ' aN [21] de Vries , Principe
o o 'o(I : oL( o( .Wo. Io ( IN
',, 'o:N W!6 h ':E! J( ';(N$
( ':E! aS [17]h W. on-line #@ ,0
o( o W. x X. I(N ( #x$ +" P
IoN ; o$ >! b(N , W. 9:" \S +"

.W. ," W.
o ,o< o() . jN : 9I J( 5 %& '
+$ '456 .W. XV () (L. D0 K %&
(11)
wO
wNet
Oc
wNet 2
wV
K uG u H
'V
(12)
K u (D O ) uO c u H
m + 0m ' 456 ( :N FE. ,i&
: . $,oX 'o;`(, +EN # $ 456 . 6 +:
(9)
w ij 0
:F( W!6 N () ';N:, F4$ '
.W. UV< , WO 'C 456 ?& ( a, o ; DCN
(8)
x ijl (t )
ww ijl (t )
(13)
2
(D O ) uO
(14)
:F( W!6 ',, () ';N:, F4$ '
wE
wW
'W
K
wE
wW
wE
wW
'W
wE w O wNet 2 wH
wNet 1
u
u
u
u
wO wNet 2 wH
wNet 1 wW
(16)
(D O ) u O c uV u H c u X
(17)
K u G 2 uV u H c u X
(18)
(15)
K uG 1 u X
(19)
G uV u H c
2
V new V old 'V
W new W old 'W
(20)
oo oo ',, , Noo oo() 'oo;N:, oo&

W , V +" oo
.N S G. 456 xij +
X ,H ,
[87 /23-16] ,# -C $D - .6 .3

o o o( 'o n (C a$ X DN
FNo oE( J(o EN 0 (IE W : , N $
'o456 ?o& FNo ; #& ( F W=$ 456
: o( Go F(,o W=$ (| /.$ W.
'o 'o FNo o Fo Wo6 o. TDL . 456 (
_o(P$ 'o oP( Fo ao
N zo0 'no 9: ( n
Wo=$ (o) +" +o qoNO o JoP$ o456 ?& q), F
IoN +" o%& o o 9:o" : o 6 R , S 9:"
.W. +EN 456 FE. : C$ IN J( ! 9:"
1 2 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
7
?(@0 & >!<$ = ! ' & ';
(15) \o o o $ o ) o F(N +:, ( ;$ N

( hS,
W. (14) \ N +" \ .W5 ,0 , XV ()
o( : ,ioP o \oS +:, k 1 (o 'oo oP$ E
oS
P 5
o@ oV o o6 .F(o0 IN 456 ?& W;
.F(: NL( 9:" P(
P
P z 1
o(
#o& G (z ) o%& '
z 1 P
1 (1 P ) z 1
(n #@ (17) \ B. (: #.=
P z 1
x 0 (t )
1 (1 P )z 1
x 1 (t ) (1 P )x 1 (t 1) P x 0 (t 1)
x 1 (t ) (1 P )x 1 (t 1) P x 0 (t 1)
x 1 (t ) G (z )x 0 (t )
56# .2 .6 .3

#oP jo
':o.F : 9:" W; \S N IN
(
F X. D0
1
2
(d
1
2
(n ) o i (n )) 2
e (t )
(24)
J( : \S F(: ,0 () : wij #x$ JP$ , #.= 6
:F( F WS (17) , (15)\
wy ij
(25)
x ijl (t )
ww ijl (t )
: [16] F( (22)D $ Jg,

wy ij (t )
wP ij
wx
ijl
l 0
ijl (t )
wPij
o
o5 o o< x0 (t 1) Wo o* o o <
oooo oo0 oo 'oo< ooC Woo. oo456 ?oo& oo oo ,ooP
1 P >(| \! 0 ' W : x1 (t 1) ,
.W. _!V
oo Woo.
ijl
D ijl (t )
(26)
l 0
+" 'o;DS o oo o( o 5 ( 5 0 P 1

o 1 P 2 , W. Z=X@ W. W. , 6, ( \0
Wo. Z =Xo@ o* Wo. +" 'o;DS o ( 5 +
.W. 0 P 2 G
, FIR 'o 5 o $ >! ( z0 wP : $ 5
+"
wx ijl (t )
wP ij
. [16] IIR 5 $S z0 : wP

ko0 d= @6 : , ' IN '. #.= !
D ijl (t )
(27)
(1 Pij )D ijl (t 1) Pij D ij ,l 1 (t 1)
o. o \o! J EE #@ ' +*, W.
.F(: EE FE.
x ij ,l 1 (t ) x ijl (t 1)
Joo( ooK oo oooo l ! 0 , t ! 0 )oo #oo.= oo

Wo. (23)oD 'o ?< : (27) D .
D ijl
D ijl (t )
wPij

wPij
wPij
x ijl (t 1)
x ij ,l 1 (t 1)
wx ijl (t 1)
wPij
wx ijl (t 1)
wPij
Pij
(1 Pij ) x ij ,l 1 (t 1)
x ijl (t 1)
wx ij
Pij
:F( .= J(0" +& W(;N

wx ij 0
wu i
w ij 0
(29)
.(" W. (22) D : W!;. D J(

6 +$ F F $ (29) , (27) (25) /,
.= #@ + F MLP BP :N /, $
\oS \o%5 o oN /o, , o" W. /, $ .F
(t )
(22)
, l 1 (t 1)
wPij
ijl x ijl
l 0
. ($ (27)D +

wy ij (t )
y ij (t ) w ij 0 x ij 0 (t ) w ij 1x ij 1 (t ) " w ijk x ijk (t )
(28)
wy ij (t )
F( (17) , (15) '; ?R
(1 Pij )
wx ij ,l 1 (t 1)
wPij
-C $D .1 .6 .3
0 #@
:F( ("
w (1 Pij )
P +" 'C W6 W;
P ooC Woo. oo. \ooS oo 'ooL( Foo; ooN oo!
(21)
x ijl (t )
,
(1 Pij )x ijl (t 1) Pij x ij ,l 1 (t 1)

l !0
. xijl
u i W. l
(23)
0 L,
Wo. + #o@ \oS MLP 'oo N IN #.= C
: o
( o F( #.= &N IN TDL V !
F E N J( ;$
( , F '0 +" $
'o
P ( : ;$ G. ( k
: h$ ( J5 S
'oho$ Jo( : ao Jg W. X. +" 456 ?& JP$
MLP 'o +:, + CN +:, ( '

. ': 456 +,
1 2 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
?(@0 & >!<$ = ! ' & ';
o%& ' ' !@ 0. ( +& N$ 0. J( J(
Wo. om \oS o

( ; N . ;( E L( #.=
o( +o$ o o Wo. +" oo NL( `(, J(; . D
P 1 , Fo WoS vo0 o . TDL s$ LNL*
+" so o;$ .[54]N _(P$ +<( +( '( aIN
o oP( Wo. TDL o( \($ \S 6 5 F S
. (n ?< ';N,N h$ , (I h$ WE(
'o BP #o.= . TDL o WoEN 'o$ , $h W!6 (
Bo. o oN(no #o@ !,IN +( #@ (;( ?<
o;$ Wo. %6 \S W!;. ' N /, : IN TDL
0 ;N" \%X$ [55] . '

N: ,P JNS
.N
9:" TDL W. J( N
.W.
: F& $i< !,IN +( 9:" 9, J($'( (
N" :
o$ JNoS , ;o, Jang oN 6o = 'o \($ '
Jang .o ;o< 9:o" W&o. +o )o '. 79:"
o $ 9aoL;N o( , 8aoL; o!C , 0 ';< :" ';,
.W. N
'E $ !" ':5 ';E. '
'(nooP$ joo| Joo( \oo! .[84-76] Woo. oo5 ooS +ooCC= :
.oo N: 'I + \($ ' ;N" W S , ':5 ';E.
o!" o: o o0 5 +$ : P J(
o (oIo oN Po. , oN W=o@ $
$ '
o o;N", o oLN"- #i #@ ;N" : N: #@
o VN ?S ' J* qPDS .N $ |( + ':5 FE.
?S $i ( , +: #i W. J 0 5 (:
do= + J( WS ; W; $x$ :N, N
'oP ';o<V N W. ' &| 9:"
.F(N W=@
o o o!" oo o: 5 #@ ' :
: L aN [ 84 80 79 76] : +$ ':5 %& '
o( .NN X. , jN . $ ( , N ':5 ';E.
oN IoN o0 oE(C o o: J( 5 #@ ' J(; :
o o oo o aoS +oE(N !C J( . [77] h
.o ':5 %& ( 0. J( .NN 10ITFFNN
, 11# N! N +N ';:" 0 9, IN W(;N
o$ , +& < #P J( !,IN +(
o0 9, Wo(;N , oN +,: N: '. !" '.
W=o@ $" 'V Jang :" ';, ANFIS
'oo o 9, Jo( E(C L( N .W. N E(C N F0
6 Local Minima
7 Hybrid Learning
8 On Line
9 Off Line
10 Interval Type-2 Fuzzy Neural Networks
11 Levenberg- Marquardt (LM)
ANFIS .4
'o +(
o<N : o( /o.$ 1993 o. ANFIS '
, &S 0. ( +& ANFIS . 5P [53] Jang a :
':o5 ao5 o o;N" oD0 K ':E! , ;E. (. ' >.
.N > 0 '(: #; $
. J( $.= ( , '(nDPN #; $ J( W &
$o.= #o@%0 o oN ()o DPN E (;! )
$o.= #@%0 E ( , (Mamdani) NN .
.(1MLP NN . DPN !, v0
, \o o Jo( aoN ! UV< J( aN : #@%0 J(
:W. V
V J( : N O #.= :Adaptive Network
.1
.W. %& &N

ko. V : IN DPN :Fuzzy Inference Systems
.2
.W. ?< ':5

'o o .F(oN do= \o%X$ o \oS ';<V a, V
_o(P$ #o@ +o o o 2,< ' : 'P.$ CD$
o o +" 3'o;N,N q), o Fo ,< ( N
#o@ o o oN o,I! o oP +o .o o0 5
o0 o o5 o oN$o o ... o( . ( .. 4(L.
o o( o o 9:o" ' ( N$ ;N,N qNO .NL
. 5WO ( (n ?D$ L( #&
o() P( .W. +" ' , #)%$ J( #@%0 ,I
Jo( +o \o o o,I! ~o , \% 0 : $ ' ()
(. \% L( () P() WEN , #)%$
+o$ o J( . ,< ' 5( \$ VEN J(
o .o oC $ No0 Jo( ,I ... , MLP RBF \ ( :
1 Multi Layer Perceptron

2 Feed Forward Network
3 Neurons
4 Sigmoid
5 Fix
1 2 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
9
?(@0 & >!<$ = ! ' & ';
. a - i j ? P$ h$ a - j ',, ? P$ C P

, W. a - i +NS v| k. $ ,0
Ajj
(33)
ifi
i 1
R
+"
fi
i
l
(6G G&' (G G 2! -C $D - 21 .1 .4

ANFIS
:[53] (22) \ ?D () b ' J(
2
i 1
.W. N IN 1Bi- N: '.
wi (31) D
:F( W(;N
o aoS o o!" oo ,i& Jg . MLP%&
(34)
xl b , i 1, 2," , R
i
?o P$ ho$ ( () J( :? P$ h$ () , ',, () : , ()
+o& o . 5 456 6, ( +& , (N
, o o. Io \oS : (o ' o Jo. ?o P$ ho$
aoC oV (n9:" ' +& . GN(,
.N 0
l 1
(6 &' ( 56# (-2 .2 .4

9:o" 'o o Xo. 9, J(o$' ( DN
F4$ W; 3 <NG ' $ . +( ';,
Jo( + .W. 'E ':5 %& ' '
oPS, , v D '; ,0 J #,X$ ,IN B. F(L!
.o _o(P$ : 'D0 +N F ; #& (
o N F4$ 'R F(L! J( ., ;N:, , a
aoN o '(oC : F(oL! Jo( .L S 0 !,IN '. D0
o JoP$ o; sCN W. W6 '; 4:" ' W&.
. X.
9:o" W; 9, J(* 5P [53] 0 ,P !C Jang
o%& 'o (. FP$ W S N ANFIS '
o 0 JNS ' +& () J( :JNS () :a, ()
?o P$ +Io vo| /o.$ +NS $" #S +I WES J( .N
.((31) D ) (" W. L( R ? P$ h$ ',,
o o $ o +NoS o o$" #S () J( :+.I!N () :a. ()
$" #S #@ J( . I!N L( JNS $" ' #S
. FEC$ JNS a$ $" ';$S \ j
+NS
W. JNS : a ,0 () J( ::.':5 K () :a;* ()
I!N $" #S (
f i ) ;(,, : D0 h$ \ J( ("

. v|
o ,0 jo
: () J( ' ,0 : ,0 () :F
()
.((33) D) \@6 F
() '
n ' W. 5 4N MISO W!6 (5) \
. 9:" !,IN +( :!,IN +( -1
k n , ... k 2 k1 \o >o$ $ ',, ',,

o J( ' +NS R 6 +$ j
J( . ? P$
ao LSE : 5#oP J(o JoV$ 6 ( , !,IN +( -2
:+" WN
: IN ':5
oo S X. !$ V !, C ' ,
F(L! a$ ':. :, CN !,IN +( G-. ,
.W 0 ;&
!,IN +( : $ #@ $ J( : LSE , !,IN +( -3
/o.$ oC oV #@ J( . S X. LSE ,
.( 0 9:" LSE /.$ !$ V , !,IN +(
oD0 ':o5 -%& ' a$ $ J( :$$ LSE -4
o o ':o. :,o 'o 65( P.$ J! 5 F(L! : ,
.N %& ' 9, J( [57 56] . X.
- Back Propagation
4 Learning rates
5 Least Square Estimation (LSE)
6 Extended Kalman Filter (EKF)
ho$
(30)
k 1 u k 2 u" u k n
i
- Ajj
j ',, ? P$ +I P PA i j x j
? P$ h$ a -
, #.= LN" F _(P$ a -
j ',, ? P$ h$ &

: a
:W. (: #@ (5) \

wi
P A i1 x 1 u P A i 2 x 2 u " u P A i n x n
1
(31)
+"
i 1 ^1, 2 ," , k 1` ,
(32)
i 2 ^1, 2," , k 2 ` ,
" ,
i n ^1, 2," , k n `
1 Mackey-Glass
2 Member Ship Function
1 2 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
?(@0 & >!<$ = ! ' & ';
10
o = o( #P J( ':.; 'S & \wP J( \6
#o@ [58] No ':o5 'oo EKF 'o J!,
o ';o, +o$o oL( (;, : .W. N ;< (
b(N , E ':. D0 +, 9, J( ! .W. 5
oN^ 'oF(oL! ';, +& D WP B.
,o ';o, : 'E $)C : ( .W. !
66] oE, ';, L( , 6+L* ';N! #m , ':.;
j 9, J(* [59] E ANFIS 9:"
.[67
o o0 o;N" oE(C o , oN 5P aC ' 9:"
oo oo oo< ':oo5 oo%& oo oo( 'oo [67]
o<N o!,IN +( :N +( B. ;, J( $
o W. N ':. :, 9, , J(: $ , GA PSO ';,
F(ooL! oo oo;(L! Joo( .#oo Noo! , 2oo; oo<N 1h(oo.
, oo WoP E, \ 9, J( . h( ,
oo ooE !oo$ ooV 9:oo" 'oo 3oo<: #ooP J(oo
'o(N o : o$)i0 o* W. J (L W&.
, [60] 'o oL( ';o, : . ANFIS ?D$ (N$ W(C$ '.
'oP 'o; oN >. !, ';P (. ,
aoN +o$ NN X. a$ 9:" ' PSO : [61]
io o o Jog , o. o o4N o \o= oE lX$ J(
+" (N$ : N , N ; ! : 0 !C , ! .
.(.
'o( d= [68] [68] [67] ! . aL;N
[67] !, N R 'P * / V 'w5 PSO (L ,
.W. <N ; $ '( j|
-o o o0 ';(:o. ' a. 9, !C Jang
N:o$ )o 'o 9, voVN " : F; $N
'oo aC V ' ':.:, m #N $
o oN oS +$ 5 &

N , 4#.= g J W.
oN o NL( ,0 D0K CN \! ':5 -%&
o !o$ , aC V ' ':. :, $ ';, N" \!
o5P ';o, , W. !$ V ' ':. :, : $\< E
+o$ & J( h(E$ ' J( (n #@ N #@
-o !o$ oV 'o ':. :, < DPN ' q$&
k,Io +o( :oN aC V ':.; W; L( ';, :
.
oD0 oK (;(:o.; [63] h(. <NG [62] a, $ ';N(
".N X. L( ';, , [64]
aC V X. :" ';, F; N J(

',, 'ow5 'o V (IE CN V J( '
9:o" jo ';o, ' F; N D ';, :
. \! !,IN +( $ : N5 (;, & .
J(o 9:o" : !o$ V ;, $ q(C$ aC V
ooN $ioo< oo! .ooNoo Xoo. oo oog , +oo Woo.
Fo; o
N .NN X. J! 5 +" N0 F ( <: #P
9:o" oC(R o Wo.E6 , o = 'oFN ' $ , (L
9:o" 'o(: Wo aoC oV ' WEN" 9i$ J( :
oN J(o F(oo 'L( ';, .[59] ,
o No( #o.= ) g \! ;< ';, , N
PSO oo$ 'oo9, : ':oo5 oo%& 'oooo 9:oo" oo
o4N o >. ,0 aC V ' D0K w6 $
'o N J( . X. RLSE , PSO ( !,IN +(,
..N
: o( !$ V ' , PSO 9, aC V
& \! , +( E, q$& D (;,
.N ':. :, RLSE ( !,IN +( N L( , ';,
o E, (;, .N F : + W. . , >.
oD #=|$ $ 9:" ' !C J( :" ';,
' ':.; ' O (;, +& N F #P J(
:W. (: S
' W. J( F; N N S X. !$ V
'o !,IN +( , aC V 9:" PSO '
.1
[75 73] !$ V

#ooP J(oo , aooC ooV 9:oo" PSO 'oo
\o< o oE( Noo ':o.o; o = #o@ o !$ V
'o& #io< Wo. o FE. $ N 0 Overparametrize
.2
'o , Wo. o o0 \%X$ j| J( [65] .<N W(
[74] !$ V ' <:

1 Quick Propagation (QP)
2 Resilient Propagation (RPROP)
3 Recursive Least Square (RLS)
6 Ant Colony
7 Off-Line
4 Computation Complexity
5 Performance Resulting
1 2 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
11
?(@0 & >!<$ = ! ' & ';
J5o( o o! .oo 0m 'P '$ E(C ' (n
G
. Pbesti , Pi N. :, , ; ';PS
G G
aS , F(L! , x i v i + 5| ( sCN
G
G
IoN o:" ao o( +& v i +$ (N v i F4$
. LN
Po.$ J!o o 5 , aC V 9:" PSO '

!$ V ' 5(
1
o o o 'oD0 : o ao. , a, ';, !C J( !
.F( ; PSO ?D$ , (I h$ +&
-o PSOF(oL! o 'o%V #=|$ ( 'P V

.F < (" F(L! J( $ F(:
(o;$ o m o( !oo #m o&

( : < #m ,
-o PSo ; , W5< WEN !E ( \6 (N$ q(C$

\o6 F(oL! Jo( .o o \oP$ , \oC$ L( #m
o \oC$ Jo6 #m 'o5 : oP o( o( +& !E
.(" , L(
i P + . 5R , q)P WP ( #R$

oL( #m o m o .o o0 i (E IN j j (E
o o 0 NL(E WPS J(; : O +"5 , W. s$
G
G
Pg o o (oE J(o; 'o Pi : Wo. #o& WPS J(; J(
$2H - (6 .5

(&o B<N,) Kennedy /.$ #m , !, ' (
o!$ ',o| 'Io ( +& (l B;) Eberhart ,
o WoEN "&o #ioP$" #i($ : '; $.= 9
.[69]
( " 25 Z0 " ';(N$
: 'oo; o o o [70] 'o : Oo [69] ';(:. J!,
. a
N #m ' +N , '
E 9,
':.; 9, ( $ W5( P.$ [72 71 69] /.$ qP(. 9, J(
' FEC$ =N , L( #m #R$ =N . 0 (N
-o +o #oR$ : NoN ';(^!$

( W. ;N"
.3
. \($ #m , ':.; a $S

ho$ o( o( !E ( '
E 'w5 . m 'P$ PSO
.
oS +" PS , h$ 9: +I : a , N w6
J( F4$ #@ m W&. PSO ("5

G G
F(oL! .o o Pi , Pg ';PS 6 35%$ #N.N d&
: o$ Bo. o o0 'oP Wo6 m o Go .Wo. "
. PSO ( F(L! (:
-o PSo 'P W6 . 'n( 5%$ (|
5o%$ ';&o. , o;PS o WoP !, C
.1

E Z'P D 'w5
o , : m * ( ( 0 n

$ J(; , 0 P5 WPS
a
No. .o N. a
N. 0 S W6 #m a$
Wo. o q)o6 E nK '
E ; +N N ,
C 6 j,
.2
. 0 W6 ; (I h$
. (: (I h$ m '
.3
LNo o D oP o o . : $ B. #m , m
o oE(C o0 s Pbesti 9: m 9:

G
o|6 o6 WoPS o o $ Pi , Pbesti ;
.4
:: $& J( . (N

E 'w5 P
P5 WPS :
. J(IL(
WoPS J(o; o +L(oE WP 'm (.
v i k
v i k 1
(35)
U1 x pbest i x i k
U2
x i k
gbest
x i k
x i k 1 v i k
W&. : v i
oDCN o( LN #%V ( +& N$ P5 WPS
. gbest G(N
:(: ',E$ B. m WPS , W&. x$
.6
o o P5 WoPS F(oL! o$ . =! W.

E 'w5
'o;PS ao$ : o; WoPS J( . (: . ( +&
Jog .oo o0m Pi o +" #o%V S
J(o; +o& o) Pbesti ao 'ox ( Pi (I h$ C
Stochastic Oscillates
S WPS J(; : Pi
.5
o m J(o; +o& o +" +o o=! , Wo.
$
G
xi
Prediction Error
2 Individual Cognitive
1 2 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
?(@0 & >!<$ = ! ' & ';
12
o WoP .Wo. o 5 4N ? P$ h$ 3 ',, ' , W.
(o$ #oP5 o6 ( v0 1?D$ ( q)P) S
. m 10 PSO ' 5
. k0 C 6 : ,"

C 6 +(
-2 J)9 .1. 6

o oo , ':oE! #,oX ';, + E(C , (: '
#oo@ oo oo Woo. 'oo& : 5oo%$ oo oo( LNoo
3 (MNSE)I!N 'D0 #P JLN -
1
N
RMSE
(x (i ) x (i ))
(36)
i 1
o o C
x , ':N C x +"
oo :o ao$ oo 'oD0 #oP JLN n >$$ J(

.N, (: ' 'P +& , .=
o GNo(, #,X ' ' & + E(C '
I!N 'D0 #P JLN +.I!N ' 'P +&
_o(P$ o(: #o@ o I!oN 'D0 #P JLN P .
.
(x (i ) x (i ))
. =!
- / TDL (

()I .6
ANFIS
: Wo. #o& o) No: ';(o. ' :
, o(I
$ , Wo.o >o$ No: >oE6 o o '("5 : #<
oR o , N \CE N s ( N: ';(. \ =$
',, o N, (( . ' , .NE, F !
, Fo . , F FE. ,0 FN q5@ , N & \S
D( ( z%V! & (.F N (N$ ~ +" ',, ',
o.i ' $ ' ' 'L( /( ( (E( \S :
o(: No: ';(o. o( o L, , J : F5 (
(x (i ) x (i ))
.[6]W. X. E
(35)
i 1
N
rV max ,o= oE( m W&. PSO @ VEN
NMSE
Ui
4 (MAE)'D0 ? D S JLN .N _(P$ (: #@ P
.8
Ui ~ U (0,c i ) 5%$ , W0(

%& %& v| LN L &
.N h(:$
. [1011]W. 5 4N (: P . N .

2(RMSE) 'D0 #P JLN n -
.7
qo(C$ (oXV o() ( \S6) +,-. () * %& '

oP$ o +" o s,o< NI JV$ N$ >. >(C$ h$
i 1
' o@5 +Io JLNo oo 'o$.P 0" P W(;N ,
FNo$oN o Wo!6 o* .F(N 4 XV () +,N 5
.W. (: #@ .= C(R , W. (C : @ (C
o() o;N,N :oN o oP$ , a:) oP$ o '4N ; ~
MAE
1
N
x (i ) x (i )
(36)
FNo$ D0 , P. ,

$R F( E( ' W;
.[5]F(," W. v D b(N
i 1
.W. " ;, $ E(C (2) , (1) ,
\o , TDL o o., o o (19) , (18) '\

J( W. 5 #@ ., (21) , (20) '
( < .7
Fo, No o() Fo o o 5 , 0$ /0 @& , () 3 ,
':5 , %& J(* ';(N$ E(C aS !C J(
Jo( .oo oD0 o ,0 o() Jg Nw6 ,0 ()
Bo. o $ :" ';, ' Jg .F(N ANFIS
. 1 ,0 () , 70 N () 4 ',, () ;N,N P$ 9,
.Wo5 #o@ o J( b(N ; P. ANFIS ' PSO
o voVN o , TDL ' 0/005 w ' ;N:, 9:" >(|
o456 o%& 'o o ') (N$ +I N J( Jg
3 oC , o 0/002 9:" >(| 456 5 ' .W.
o No: ';(o. o( ':oE! 456 +, ' WEN
ANFIS . (24) , (23) '; .W. 5 4N
qio (2) , (1) ';!,o (oC o +Io : >o D J( .F(
- : 9, Jo( .Wo. 5 #@ PSO+EKF :" 9,
';(:E! +" (N$ , %& + 456 #S , W. ;<
o Xo. oP 4=! ' ',, +& \S 4=! 4 $ '
Root Mean Squared Error

Normalized Mean Squared Error
4
Mean Absolute Error
3
Fitness
1 2 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
13
?(@0 & >!<$ = ! ' & ';
[12] C. Chatfield, The analysis of time series: An

Introduction, Chapman & Hall, 1989.
[13] O. Nelles, Nonlinear system identification, SpringerVerlag, 2001.
[14] http://www. Aqcc. org, Air Quality Control Company,
Copyright 2001.
[15] G. Dorffner, Neural Networks for Time Series
Processing, Neural Network World, Vol. 4, pp. 447-468,
1996.
[16] T. J. Cholewo, J. M. Zurada and A. Cchoci, Exact
Gradient Calculation in Gamma Neural Networks,
International Symposium on Nonlinear Theory and its
Applications, 1997.
[17] M. Van Veelen, J. A. G. Nijhuis and L. Spaanenburg,
Estimation of Linear Filter Banks for Multi Variate Time
Series Prediction with Temporal Principal Component
Analysis, Report, 1997.
[18] J. C. Principe, J. M kuo and S. Celebi, An Analysis of
the Gamma Memory in Dynamic Neural Networks, IEEE
Transactions on Neural Network, Vol. 5, No 2, pp. 331-337,
1994.
[19] B. de Vries and J. C. Pricipe, The Gamma Model-A new
Neural Model for Temporal Processing, Report 1992.
[20] J. C. Pricipe, B. de Vries and P. G. deOliveria,
Generalized Feedforward Structures: A New Class of
Adaptive
Filters,
Computational
Neuroengineering
Laboratory Report, 1995.
[21] Yu Heeyeol and S. Y. Bang, An improve time series
prediction by applying layer by layer learning method to FIR
neural network, ICONIP, 1996.
[22] S. Celebi, J. C. Pricipe, Analysis of spectrel Feature
Extraction Using the Gamma Filter, Univ Florida, 1994.
[23] A.B. Chelani, C.V.C. Rao, K.M. Phadke and M.Z.
Hasan, Prediction of sulphur dioxide concentrations using
artificial neural networks, Environmental Modelling and
Software, Vol. 17, pp. 161168, 2002.
[24] G. I. Berastegi, A. Elias, A. Barona, J. Saenz, A.
Ezcurra and J. D. Argandoa, From diagnosis to prognosis for
forecasting air pollution using neural networks: Air pollution
monitoring in Bilbao, Environmental Modelling & Software,
Vol. 23, pp. 622-637, 2008.
[25] G. Nunnari, A.F.M. Nucifora and C. Randieri, The
application of neural techniques to the modelling of timeseries of atmospheric pollution data, Ecological Modelling,
Vol.111, pp. 187205, 1998.
[26] G. Soja and A.M. Soja, Ozone indices based on simple
meteorological parameters: potentials and limitations of
regression and neural network model, Atmospheric
Environment Vol. 33, pp. 42994307, 1999.
[27] R. Guardani, C.A.O. Nascimento, M.L.G. Guardani,
M.H.R.B. Martins and J. Romano, Study of atmospheric
ozone formation by means of a neural network-based model,
Journal of the Air and Waste Management Association Vol.
49, pp. 316323, 1999.
[28] J.A. Cannon and R.E. Lord, Forecasting summertime
surface-level ozone concentrations in the Lower Fraser valley
of British Columbia: an ensemble neural network approach,
Journal of the Air and Waste Management Association, Vol.
50, pp. 322339, 2000.
[29] M.W. Gardner and S.R. Dorling, Statistical surface
ozone models: an improved methodology to account for
Jog , o +<N .$ D0 WEN NL" E

?< 9, (N0 (C : L) * (C
oL( Fo; N m ! . v ; , ?< 9, WEN a,
o \oS o4=! 4o;* o;$ ( ' W. ',| IN
o 0 'o;( $ , W. ',, +&
> D .(N \S #4=! >. ';(,, 'n +:, , JP$ aS
on- #@ , !,IN +( ' 9:" L( F;
o!C J( L( F;
N .W. ( !C J( line
ANFIS ':o5-o%& 'oo ' >. :" 9, '

oP! lo5 b(oN PSO Bo. $ ';, '
o oo \@6 , ( %& ' E(C 0
, 'o oL( Fo; oN .Wo. 'oL \oS R , b(N
aoC 'V 9:" ' PSO , 5( P.$ J! 5 ':.+I
(oN$ qio& o:" >o$ Jo( +<N ANFIS !$ ,
.W. ) ( ( B. ' ?D$ .
KL
:" L<N & #<N I , WE(: /= .; -.& .a .1
. 1371 , * i.
2. P. Mlaker and M.Boznar, Percptron Neural Network
Based Model Predicts Air Pollution, IEEE, World Congress,
1997.
3. Haykins, S, Neural Network A Comprehensive
Foundation, Prentice- Hall International, 1999.
o<N , o $o.= 9o o%& 'o N k; . v . a .4
. 1381 I( P@ L<N #<N P@ L<N
5. G.Dorffner, Neural Networks for Time Series Processing,
Report, 1995.
(n DPN %& ' ':. , '( \ =$, (I
$ (" . .6
. 1380-81 . ' !. #CC=$ a & 6, i. :" L<N
7. J.J Kao and S.S.Hung, Forecasting Using Neural Network
versus Box-Jenkins Methodology for Ambient Air Quality
Monitoring Data, Air & Waste Management Association,
Vol. 50, pp. 219-226 , 2000 .
8. A.B.Chelani, D.G.Gaighate, and M.Z.Hasan, Prediction of
Ambient PM10 and Toxic Metals Using Artificial Neural
Networks, Air & Waste Management Association, Vol. 52,
pp. 805-810, 2002.
MATLAB ' (V '@ . , $ ' . a , 'V . a .9
.1380 , * +.0 #<N .; SIMULINK,
10. U.Schlinnk, S.John and O.Herbarth, Transfer function
Models Prediction Ozon in Urban Air, UFZ-Center for
Enviromental Research Leipzig Halle Ltd, Leipzig,
Germany, 2002.
Jo LoE o o. ' N: ';(. +N: . a [11]
.1380 . N +( .R J(!%N 0 P@ L<N
1 2 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
?(@0 & >!<$ = ! ' & ';
[44] M. Kolehmainen, H. Martikainen and J. Ruuskanen,

Neural networks and periodic components in air quality
forecasting, Atmospheric Environment, Vol. 35, pp. 815825,
2001.
[45] Harri Niska, T. Hiltunen, A. Karppinen, J. Ruuskanen
and M. Kolehmainen, Evolving the neural network model for
forecasting air pollution time series, Engineering
Applications of Artificial Intelligence, Vol.17, pp. 159167,
2004.
[46] S.M. Shiva Nagendra and Mukesh Khare, Artificial
neural network based line source models for vehicular
exhaust emission predictions of an urban roadway,
Transportation Research Part D, Vol. 9, pp. 199208, 2004.
[47] J. Hooyberghs, C. Mensink, G. Dumont, F. Fierens and
O. Brasseur, A neural network forecast for daily average
PM10 concentrations in Belgium, Atmospheric Environment,
Vol. 39, pp. 32793289, 2005.
[48] J.B. Ordieres, E.P. Vergara, R.S. Capuz and R.E.
Salazar, Neural network prediction model for fine particulate
matter (PM2.5) on the USMexico border in El Paso (Texas)
and Ciudad Juarez (Chihuahua), Environmental Modelling
and Software, Vol. 20, pp. 547559, 2005.
[49] G. Grivas and A. Chaloulakou, Artificial neural network
models for prediction of PM10 hourly concentrations, in the
Greater Area of Athens, Greece, Atmospheric Environment,
Vol. 40, pp. 12161229, 2006.
[50] A. Zolghadri and F. Cazaurang, Adaptive nonlinear
state-space modelling for the prediction of daily mean PM10
concentrations, Environmental Modelling and Software, Vol.
21, pp. 885894, 2006.
[51] T. Slini, A. Kaprara, K. Karatzas and N.
Moussiopoulos, PM10 forecasting for Thessaloniki, Greece,
Environmental Modelling and Software, Vol. 21, pp. 559
565, 2006.
[52] W. Wang, W. Lu, Xiekang Wang and Andrew Y.T.
Leung, Prediction of maximum daily ozone level using
combined neural networks and statistical characteristics,
Environment International, Vol. 29, pp. 555562, 2003.
[53] J-S.R. Jang ANFIS: Adaptive-Network-Based Fuzzy
Inference System. IEEE Trans. on Systems, Man and
Cybernetics, Vol. 23, pp. 665-685, 1993.
[54] P. Werbos, beyond regression: New tools for prediction
and analysis in the behavioral sciences, PhD thesis, Harvard
University, 1974.
[55] J. S. R. Jang, C.-T. Sun. and E. Mlizutani, Neurofuzzy
and soft computing: a computational approach to learning and
machine intelligence, Prentice Hall, 1996.
[56] S. Shah, F. Palmeieri, and M. Datum, Optimal filtering
algorithms for fast learning in feedforward neural networks
Neural Networks, Vol. 5, 779787, 1992.
[57] S. Singhal and L. Wu, Training multilayer perceptrons
with extended kalman filter algorithm, In Davis S. Touretzky,
editor, Advances in neural network information processing
systems I, pp. 133-140. Morgan Kaufman Publishers, 1989.
[58] D. Simon, Training fuzzy systems with the extended
Kalman filter, Fuzzy Sets and Systems, Vol 132, 189 199,
2002.
[59] M. S. Chen, A Comparative Study of Learning methods
in Tuning Parameters of Fuzzy Membership Functions, IEEE
Conf. on Sys. Man and Cybernetics, 1999.
[60] D. B. Parker, Optimal algorithm for adaptive networks:
Second order back propagation, second order direct
14
nonlinear behaviour, Atmospheric Environment 34, pp. 21

34, 2000.
[30] Jehng-Jung Kao and Shang-Shuang Huang, Forecasts
using neural network versus BoxJenkins methodology for
ambient air quality monitoring data, Journal of the Air and
Waste Management Association, Vol. 50, pp. 219226, 2000.
[31] W.G. Cobourn, L. Dolcine, M. French and M.C.
Hubbard, A comparison of nonlinear neural network models
for ground-level ozone forecasting, Journal of the Air and
Waste Management Association, Vol. 50, pp. 19992009,
2000.
[32] A.L. Dutot, J. Rynkiewicz, Fredy E. Steiner and J.
Rude, A 24-h forecast of ozone peaks and exceedance levels
using neural classifiers and weather predictions,
1269, 2007.
[33] M.W. Gardner and S.R. Dorling, Artificial neural
networks (the multilayer perceptron). A review of
applications in the atmospheric sciences, Atmospheric
Environment, Vol. 32, pp. 26272636, 1998.
[34] A. Comrie, Comparing neural networks and regression
models for ozone forecasting, Journal of the Air and Waste
Management Association, Vol. 47, pp. 653663, 1997.
[35] G. Spellman, An application of artificial neural
networks to the prediction of surface ozone concentrations in
the United Kingdom, Applied Geography, Vol. 19, pp. 123
136, 1999.
[36] U. Schlink, S. Dorling, E. Pelikan, G. Nunnari, G.
Cawley, H. Junninen, A. Greig, R. Foxall, K. Eben, T.
Chatterton, J. Vondracek, M. Richter, M. Dostal, L. Bertucco,
M. Kolehmainen and M. Doyle, A rigorous inter-comparison
of
ground-level
ozone
predictions,
Atmospheric
[37] U. Schlink, O. Herbarth, M. Richter, S. Dorling, G.
Nunnari, G. Cawleyd and E. Pelikan, Statistical models to
assess the health effects and to forecast ground-level ozone,
558, 2006.
[38] Sameer Sharma, S.V. Barai and A.K. Dikshit, Studies of
air quality predictors based on neural networks, International
Journal of Environment and Pollution, Vol. 19, pp. 442453,
2003.
[39] B.M. Fernandez de Castro, J.M. Prada Sanchez, W.
Gonzalez Manteiga and M. Febrero Bande, Prediction of SO2
levels using neural networks, Journal of the Air and Waste
Management Association, Vol. 53, pp. 532539, 2003.
[40] A.B. Chelani, C.V.C. Rao, K.M. Phadke and M.Z.
Hasan, Prediction of sulphur dioxide concentrations using
artificial neural networks, Environmental Modelling and
Software, Vol. 17, pp. 161168, 2002.
[41] M.W. Gardner and S.R. Dorling, Neural network
modelling and prediction of hourly NOx and NO2
concentrations in urban air in London, Atmospheric
[42] Faizal A. Hasham, Warren B. Kindzierski and Stephen
J. Stanley, Modeling of hourly NOx concentrations using
artificial neural networks, Journal of Environmental
Engineering and Science, Vol. 3, pp. 111119, 2004.
[43] C. Cappa, D. Anfossi, M.M. Grosa and P. Natale, Shortterm prediction of urban NO2 by means of artificial neural
network, International Journal of Environment and Pollution,
Vol. 15, pp. 483496, 2005.
1 2 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
15
?(@0 & >!<$ = ! ' & ';
Networks in time series prediction for the case of air

pollution, NAFIPS 2008.
[78] Tanaka, K.; Sano, M.; Watanabe, H., Modeling and
control of carbon monoxide concentration using a neurofuzzy technique, Fuzzy Systems, IEEE Transactions on
Volume 3, Issue 3, Aug. 1995 Page(s): 271 279.
[79] Sungshin Kim; Jaeyong Kim; Chong-Bum Lee; MinYoung Kim, Fuzzy decision support system to the prediction
of ozone concentrations, ISIE 2001.
[80] Morabito, F.C.; Versaci, M., Prediction and estimation
of atmospheric pollutant levels by soft computing approach,
IJCNN 2001.
[81] Tanaka, K.; Sano, M.; Watanabe, H., Identification and
analysis of fuzzy model for air pollution-an approach to selflearning control of CO concentration, Industrial Electronics,
Control, Instrumentation, and Automation, 1992.
[82] Negnevitsky, M.; Kelareva, G., Air quality prediction
using a neuro-fuzzy system, Fuzzy Systems, 2001.
[83] Yilmaz Yildirim, Mahmut Bayramoglu, Adaptive
neuro-fuzzy based modelling for prediction of air pollution
daily levels in city of Zonguldak Chemosphere, Vol. 63, pp.
1575-1582, June 2006.
[84] Francesco Carlo Morabito, Mario Versaci, Fuzzy neural
identification and forecasting techniques to process
experimental
urban
air
pollution
data
Neural
Networks, Volume 16, , pp. 93-506, April-May 2003.
[85] M. Aliyari. Sh, M. Teshnehlab, A. Khaki. Sedigh,
Climate short term prediction of air pollution using multi
layer perceptron neural networks, Third Regional & First
National Conf. on Climate Change, 2003.
[86] M. Aliyari. Sh, M. Teshnehlab, A. Khaki. Sedigh, Short
term prediction of air pollution using TDNN, Conf on
Intelligent Systems CIS2004.
[87] M. Aliyari. Sh, M. Teshnehlab, A. Khaki. Sedigh, Short
term prediction of air pollution using MLP & Gamma neural
networks , UKACC2004.
-%L - .8
Co #x$ :6 ' x$ \ =$ N :1 \
propagation, and second order Hebbian learning, In proc. Of

IEEE International Conference o n Neural Networks, Pages
593-600, 1987.
[61] A. Chatterjee , K. Watanabe, An optimized TakagiSugeno type neuro-fuzzy system for modeling robot
manipulators, Neural Comput & Applic , Vol. 15, No. 1, 55
61, 2006.
[62] C. Karakuzu, Fuzzy controller training using particle
swarm optimization for nonlinear system control, ISA
Transactions Vol. 47, 229239, 2008.
[63] S. E. Fahlman, Faster-learning variations on back
propagation: an empirical study, Proc. Of the 1988
Connectionist Models Summer School, Pages 38-51,
Carnegic University, 1988.
[64] R. L. Watrous, Learning algorithm for connectionist
network: applied gradient methods of nonlinear optimization,
In Proc. Of IEEE International Conference on Neural
Networks, pages 619-627, 1991.
[65] R. Babuka, H. Verbruggen, Neuro-fuzzy methods for
nonlinear system identification, Annual Reviews in Control,
Elsevier, pages: 73-85, 2003.
[66] C. F. Juang, A TSK-type recurrent fuzzy network for
dynamic systems processing by neural network and genetic
algorithm, IEEE Trans. Fuzzy Systems, Vol. 10,No. 2, pp.
155-170, April, 2002.
[67] C. Juang, Y. Liou, TSK-Type Recurrent Fuzzy Network
Design by the Hybrid of Genetic Algorithm and Particle
Swarm Optimization, IEEE International Conference on
Systems, Man and Cybernetics, 2004.
[68] Clerc M., Kennedy J., The Particle Swarm Explosion,
Stability, And Convergence in a Multidimensional Complex
Space, IEEE Trans. on Evolutionary Computation, Vol. 6,
No. 1, Feb 2002.
[69] Kennedy, J., & Eberhart, R. C., Particle swarm
optimization. In Proceedings of the IEEE international
conference on neural networks, 1995.
[70] Heppner, H., & Grenander, U., A stochastic non-linear
model for coordinated bird flocks, AAAS, 1990.
[71] Eberhart, R. C., & Kennedy, J., A new optimizer using
particle swarm theory. In Proceedings of the sixth
international symposium on micro machine and human
science, 1995.
[72] Eberhart, R. C., Simpson, P. K., & Dobbins, R. W.,
Computational intelligence PC tools. Boston: Academic
Press, 1996.
[73] M. Aliyari. Sh, M. Teshnehlab, A. K. Sedigh, A Novel
Training Algorithm in ANFIS Structure, American Control
Conf., 2006.
[74] M. Aliyari. Sh, M. Teshnehlab, A. K. Sedigh, A Novel
Hybrid Learning Algorithm for Tuning ANFIS Parameters
using Adaptive Weighted PSO, IEEE Int. Fuzzy Sys. Conf.,
2007.
[75] M. Aliyari. Sh, M. Teshnehlab, A. K. Sedigh,
Identification using ANFIS with intelligent hybrid stable
learning algorithm approaches, Neural Comput. & Applic,
2008.
[76] Fujita, S.; Tamura, H., Development of a Short-term
Prediction Model for Predicting Photochemical Oxidants in a
Local Area, SICE-ICASE, 2006.
[77] Castro, J.R.; Castillo, O.; Melin, P.; Rodriguez-Diaz, A.,
A hybrid learning algorithm for Interval Type-2 Fuzzy Neural
1 2 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
?(@0 & >!<$ = ! ' & ';
16
PM-10 #x$ :6 ' x$ \ =$ N :2 \

PM-10 N! 'N N :6 \
. 0$ /0 ( :7 \

Co LE P N :3 \
MLP , 0$ /0 >$ :8 \
PM-10 LE P N :4 \

((n #@ ':. (S = WEN ()
MLP1 Co :9 \
Co N! 'N N :5 \
1 2 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
17
?(@0 & >!<$ = ! ' & ';
MLP1 PM-10 :10 \
. '$ ' 0. :13 \
456 0. :14 \
MLP2 Co :11 \

( Pg , ppm ' NN 6,, W&. FN ' NN)
m3
456 G. ( :15 \
MLP2 PM-10 :12 \

( Pg , ppm ' NN 6,, W&. FN ' NN)
m3
. FIR 5 ( :16 \
1 2 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
?(@0 & >!<$ = ! ' & ';
18
456 :17\
PM-10 :21 \

( Pg , ppm ' NN 6,, W&. FN ' NN)
m3
TDL Co :18 \
ANFIS ( 'N : 22 \
( Pg , ppm ' NN 6,, W&. FN ' NN)
m3
ANFIS(PSO+EKF) Co :23 \

( Pg , ppm ' NN 6,, W&. FN ' NN)
TDL PM-10 :19 \
m3
( Pg , ppm ' NN 6,, W&. FN ' NN)
m3
ANFIS(PSO+EKF) PM-10 :24 \

( Pg , ppm ' NN 6,, W&. FN ' NN)
Co :20 \

( Pg , ppm ' NN 6,, W&. FN ' NN)
m3
1 2 1387

... 0$ 456 /0 ()* +,-. %& ' !" # $
19
?(@0 & >!<$ = ! ' & ';
? P #m ';, E(C :2 ,
PM10
JE ' ;, E(C :1 ,

Co
AR
MLP1
MLP2
TDL
2/3287
0/4960
0/4446
0/4159
Gamma
0/4003
RMSE
AR
26/6211
MLP1
14/5388
MLP2
16/9112
TDL
13/5487
Gamma
13/5472
NMSE
0/5764
0/2128
0/2880
0/1915
0/1915
NMSE
0/8168
0/1194
0/0959
0/0786
0/0729
MA
20/3643
10/0029
12/0472
9/5925
9/5920
MA
1/0454
0/3860
0/3278
0/3006
0/2898
PM10
ANFIS
(GD+RLS)
RMSE
NMSE
RMSE
JE ' ;, E(C :()1 ,
? P #m ';, E(C :() 2 ,
MA
ANFIS
(PSO+GD)
ANFIS
(PSO+RLS)
ANFIS
(PSO+EKF)
12/9738
14/5879
4/9365
3/8831
0/1215
0/2643
0/0537
0/0273
8/7631
11/6875
5/5673
4/4365
Co
ANFIS
(GD+RLS)
ANFIS
(PSO+GD)
ANFIS
(PSO+RLS)
ANFIS
(PSO+EKF)
RMSE
0/4048
0/4236
0/08159
0/0734
NMSE
0/0743
0/0825
0/0051
0/0034
MA
0/2961
0/3143
0/0983
0/0882
1 2 1387

I
S
I
C
E
20-25 =X@ 1 2 1387

(6 NJ O (.

( 2! 9M )J
G!, 5 '(0 .
+;$ L<N .; , 5 <N
s.khodayari@gmail.com, Lucas@ipm.ir
________________________________________________________________
aL 5$ . #@ (E \& ':5 FE. ( !C J( :%+
W5( : $&iR ,$ R 4N ':5 FE. . h(:$ +: R 'N ', (E
;(:. b(N +( ..5 $5$F ' , 5 F%$ 5$ WP|, +" B. ,
: FE. J( ( .W. E(C L( 9, * 4N ':5 FE. , 5 S . , \ =$
.W. 0 () $ & +$ , ("
'$- ' CD$ (E (E ':5 FE. :' '^,
Abstract: In this paper, routing in computer networks is performed using a fuzzy logic strategy.
The present network traffic is evenly distributed during the routing. The fuzzy based system will
periodically receive the information from the network and based on this information will divert
the load to other parts of the network. Simulation results are used to show the performance of the
proposed method and comparison results are also provided.
Keywords: Fuzzy system, Routing, Adaptive Routing, Computer Networks.
________________________________________________________________
%9 .1
)J "PJ (2- .2

)J ()2Q" .2.1
'oo (oE ao. , vo0 'F(L! : ( :,
J(o$ F(L! J( .W. E J($ + R$
J(o$ 'oF(oL! .[5]o .= 4! (I B. E
o5 o4N WoO o! o(I o \ J( E $. E
F(oL! W. +" L( .N .$ WPS, !
$ ( ;! (I \ J( .W5 4N $. N
: G , {5 WO +" N: $ @5 ( , N .=
#ox$ o$ o :,o oo(I
UV< N: @5 Wn
o=! o o5$ Jog , o ';0 ( (^!$ !6
'o FEC$ . , & R E J($ 'F(L! .
, (Wo. Bellman-Ford F(oL! +o o) 5 o@5 :N
.W. X. @5 F(L! : ^, J( .[5]6! W!6
(Io5 , '$-o ' :,:, 9E , W5< $

, J( : .W. W I]6 1(E d= : .. ;N" :N
.oo oX( o: J( .. CN s ';$
o 'o;$ o.. , Fo; \(oE : o( 2 'E (E
oCN !oE . +,o5 Opo$ o ("o ', o o Wo. o
ooE o5$ Jg , '$- ' 'E
o : , Wo. o Fo & JoCC= o $ o , Fo; \(oE : (
o: +og o ao
N N,5 #CC=$ , W. (E
. , +" '(: E ';. , ?C=$
o \ '^!$ : jiR +, ( (E v0 F(L! (
WPo|, Bo. o oE J(o; J( . E J($ , J(;
oo oo
oo oo +, 3#ooR$ +oo voo0 IooN , oo oo5$
. x$ +, (^!$ #x$
Link Cost
Distance-Vector
6
Link-State
Routing
Packet
3
Link
http://www.isice.ir
+( ?S I , +.; J

N 1387

G!, 5 '(0 . ':5 FE. B. '$- ' CD$ (E
Wo S (oFoE. Jo* oN kVo. o0 5 :
#o&iR o! Wo!6 F(oL! o oE(C o@5 F(L!
, > o@ 'o,o= : ':o5 ?oD .[2] +<N 0 : 0
, @5 F(L! . 0m 0 'E
oLN o , o ', o o oLN , o vo UV<
+" o$ V$ @5 %C N #&iR ' N (E
'o&S ',6 +" k. FE. ':5 FE. ( .W. .
ao o 'oN Jog , Wo. oN +" +. (I + N
o ':oX( 'o; ,0 , ':5 ';(,, W(w& h$ k.
o o@5 J(o$ o.= .[5]W. 0m , +" ( 'P
. 0
o< #o]I .Wo. o%C o$ ! (I F + B.
Wo; (1)\o o Fo.$ ':o5 FoE. . !pE
.N 46i [5,2] +$ E J($ F(L! :
o \o%X$ +" #%V< W. 5 4N 5$
( NO 1 ! (I !C J( a

N ':.
. S d=
.N :,
'^!$
_@ I(.
)J (2 .2.2

o4N (oE WoX o(: ' +$ ' J($ .
+ ':5
;(,,
0$
+ k. FE.
JNS &

+" Bo. (E LNL* ' F; ';%$ Jg , W5

F(L!
'o .o oE o o Wo. (;oS P$ a
N
+
, o! 1_@ I(. ! N {& ! 0p$ \S : 'L(
J($
E
#&iR
aS
'P
+ ':X(
(z) ;! (I
':5 FE. ;! (I +," W. -1 \
J(o$ +o o ,o 9, : (oE \o& 0

(
Jo( o #,oX$ Jo( o .o #@ 4(E , . , E
o : " W. ( #&iR . ( NO ( ,
Wo; 'oFo%$ o5$ #o&iR Jo( Bo. , N :,
'o F%$ J( . #@ E ' 'P aS J
+o& ! _@ I(. 0p$ 6 W. #@ J(
':o5 Wo. \ DN .N ':5 FE. ',,
Bo. o , o5 oS \o =$ o k. FE. ;(,, J(
':o5 FoE. o ,0 +o& o ;! (I 6R k. &S
',, +oP , o ':oX( o( o .= (I .N !$
(oE o=N (" N$ , E h(:$ 0p$ (

', o No: 'o>oE* 'o. o( + S .n (IE O$
>o(C$ +o$o o '-. %C +. ' N: , E
.," Wo. Wo. o o #oR$ '0p$ : 0
o o5 o4N _o@ o( o! ' a
N ';(:.
voVN oE +o.5 'o o! o( 'N L .W.
o4 oCN 'o , oo o! +" o s _@ , E
o , Fo; !poE o( IoN o! W5 .. , WN $
.W. $
o( : oE +.5 ,i& (E 'F(L! )P
Jo aI E J( IN ;! ', F
6 F4$ $
o o o o (o
N" : .Wo. o 5$ aIN
o aINo #,X IN %C , , W. x qi
. ; (E (" $ ( qi ( 5$
o oo a
N ':5 FE. ( 5$
(
9, Jo( /o.$ .oo 0 _ V /(
(Io5 ho >. ; W & IN (" , 2$ & +$
(
J(o$ F(oL! .oo E J($ + F(L!

o 'ô!$ Jo , o( o5$ #&iR J( W5( E
. vVN +: , (I J( 'P aS J(;

L (6 NJ .3
\o& NoEN o
$, oN Bo. o o ':o5 \oS : (oFE.
.N . , $ \S & : 'E
#o&iR 'o. o( +o$o IN , , WPDS a& (
3
4
Inference System
Routing Table
1
2
Queue Size
Throughput
1 2 1387

G!, 5 '(0 . ':5 FE. B. '$- ' CD$ (E
5$ ' ':5 &S -1 ,
0
I
/.
(_!)
0
0p$
*
_@ I(.
I
/.
*
F
F
F 0
F 0
0
*
/.
/.
F
F
F 0
*
(:
/.
/.
F
F
/.
/.
F
I
(: 0
(: 0
(:
(:
/.
0
0
0
(:
(:
(:
0
I
(6 NJ .3.1

1
o o5 o; No k. 9, : D ':5 FE.

, 'o Fo Bo. o and # & N k. 9, .W.
FoE. .(noo #o@ 2\oCO Io 9, . + ':X( \&
(v)
: + (: #@ N k.
max(min(P , P W ( Z )))
h$
for all z
P W (Z ) , ',, W(w& h$ &

P
>$ J(
Ioo5 aooN oo oo oo4N oo ':oo5 FooE. .Woo. oo ,0 W(oow&
. ':. , 6R MATLAB
)?' K .3.2
( k)
(I W(w& h$ (k)0p$ W(w& h$ (v) _@ I(. W(w& h$ (_!) -3\
!
,0 , ! 0p$ , _@ I(. ';(,, s W(w& h$

ho$ Jo oD .No voVN o W(w& h$ L ! (I
.W. ," (2)\ ,0 , ;(,, W(w&
(6 %'R ' .3.3
o kVo. 0 5 ':5 FE. X. &S

o;X( ,o Jo( .oN ," (1), &S J( .[2] W.
FoE. o ,0 . (N ! 0p$ +I ;N. , _@ I(.
.N , \0 W.;! (I ':5
N)2Q" $SIL .4
!
2
TU .4.1
,0 , ;(,, W(w& h$ J D -2 \
Xo. [6] , [2] o NN ' +& , ;(:.
Fo.$ (5) , (4) \o o o5 oS (o:" o oN
(oN o5 ,0 , ;(,, W(w& h$ $ (3) \
.N
.N
1
3
Rule Base
Mamdani Inference Method

Centroid Defuzzifier
1 2 1387

G!, 5 '(0 . ':5 FE. B. '$- ' CD$ (E
, o o.= o! 0p$ \ J( . S ! 0$

, ,, F ! _@ I(. . ':5 FE. ',, +P
(oE ,o +o :, aL , :, E k,0
Jo( Bo. o ':o5 FoE. . ':5 FE. ( NO
', o +" & , ," W. 5$ WP|, #&iR
.n Op$ N ', h(:$ jN ;! (I ?(R :
)62 (
V .4.2
N 8 (:" -4 \
';o, o 'no9: Wo; N$ NN (: ';i

'o jN E ;i J( : a ., (E _ V
_o@$ o _o V '`(, +" @ W ,
Jow$ !poE o4N ( ' 0 +& .
Jo* J(o .Wo. $ F; 'L( : %C E +.
o(I o o J( ( !, W.E +N. @ I$ '
Wo. Jo 'oL( 'o .(no #o@ N)R +: , )
a:) o o6 J(o .o W , +. W&.
.o(N J(o #o&iR CN W&. ( , W. : 'C
N 14 (:" -5 \
+o. +o: /o. $o & +o$ ;Xo@ I(. N (;i \
(o;i oN$o 0po$ , %C $ ';S P$ E
h(:$ 0p$ , 1.5MBPS ';! N {& ,
+o: /o. B. (:

( . & (: '
'o Jo I(o. F(I .W. 5 4N NO 1.5 1!
. . $ & +$ IN , %C E +.
oNO 30 Fo ':o. o +o: \o . vE= W( 50 F _@

o .o Xo. <(:" 'E : ! 0p$ .= ' .
WD - C)6# X)< .4.3

'o;! o(I ," _ V ';, 5 #@ ';<(:"
[2] oogN" ?ooR .Woo5 ooS (oo:" oo oo5$ oo , oo
, _o@ oR : o$ B. ;! (I .= 5 #@
o oE(C 'o +" joN o. o;$o$ J : . a
N ;! 0p$
'ooE .Wo. o vVN 5$ ' 4N ':5 FE.
> N vVN '@C * J : 5%$ R
zo0 #)o6 Jo( . .N< @0 'E 5$ (I5
_o V ';o<(:" . FP$ IN $ /( ' \6 +$
(: F : $&iR 'E :N D , ',
5 4N WO ':. '. E :N .5 S WE$
o 2 o , (4) \ 1 X$ ;<(:" J( b(N .
.Wo. ," (5) , (4) (3) (2), F.$ (5) \
o oR ':o5 FoE. : o ( $ & +$ (
;o, (o. : o , , $ q(C$ N X.
Foo E +. +: /. E(C Jg .W. E '$
\oC o$h(o. ooE o Xo. ':o5 FE. : '
0po$ +," Wo. ' N ( NO 0.05 \ J(
Wo( 100 o;N" :N <(:" 'E 0 \% '!
No: >E* <(:" E ..5 0 (E 'N W.
oE J( N: +I . ,I5 +" +: Wn
o 5R : .W. 5$ WP|, LN<N N 4 ! _@
o 1000 o$ 800 J ;N" I(. 'E F ( NO 0.005
o o!$ 5o%$ o@C , x ':. W(
o o o 'oaoS J 'E .N ,
o o;! : o& J6 'E .N. %C 0 $
+o: Jo( , o (: %C +" +. +: 0 5$
Wo. o(: ' : ( 2%C +. +: /. ',
NN C$ ?6 N ~ <(:" 'E 'E . Op$
\o J( .N0 NE( W(!, : _@ J5 ' ' ,
o5$ WPo|, : o. JoV$ N$ 0 <(:" 'E
+. +: ', NO ( N: : N .N," W.
+o& o No: /o. Jo( , N ' /. <(:" 'E
1
2
Propagation Delay
Average Delivery Time
1 2 1387

G!, 5 '(0 . ':5 FE. B. '$- ' CD$ (E

Wo. o ;, C : (:" , IN +: J( , N
.N. h(. CN , h(:$ ':5 FE. W S J(
1 _ V ';, ' " W. $ & +$ 2,
$ & +$
$ & +$
_@ $ & +$
(%)0p$0.5+_@0.5
(%)0p$0.2+_@0.8
(%)(;$
94.22
96.34
83.91
59.16
800
90.21
82.64
76.63
51.78
900
82.21
76.54
71.66
47.91
1000
(%) ':5 $ & +$
(W() E I(.
2 _ V ';, ' " W. $ & +$ 3,
$ & +$
$ & +$
_@ $ & +$
(%)0p$0.5+_@0.5
(%)0p$0.2+_@0.8
(%)(;$
95.16
95.55
81.82
55.56
900
91.79
82.93
56.46
950
88.21
87.97
53.33
1000
(%) ':5 $ & +$
77.68
(W() E I(.
1 _ V ';, ' %C +. +: /. 4,
+. +: /.
+. +: /.
+. +: /.
+. +: /.
(NO) ':5
(NO)0p$0.5+_@0.5
(NO)0p$0.2+_@0.8
(NO)(;$ _@
0.76
0.91
1.32
1.67
800
0.86
1.14
1.26
1.58
900
1.03
1.33
1.47
1.75
1000
(W() E I(.
2 _ V ';, ' %C +. +: /. 5,
+. +: /.
+. +: /.
+. +: /.
+. +: /.
(NO)':5
(NO)0p$0.5+_@0.5
(NO)0p$0.2+_@0.8
(NO)(;$ _@
0.78
0.83
1.34
2.01
900
0.79
1.02
1.92
950
0.81
1.09
1.97
1000
1.37
(W() E I(.
ho o$ o #@ $4$ R ':5 FE. ', ( 'P
(L W&. ', ( Jg .N % ; : '<
. +" +: $ ':.
( < .5
( : X. '$- ' (E !C J(
/(o o4N o ':o5 FoE. .W5 S . ':5 FE.
KL
+o$ ;o, (o. o oE(C o _o V D. , ' _ V

o }Do. x$ . +<N h : '$ ; X., $) $ &
[1] L. Peshkin & V. Savova, "Reinforcement

Learning for Adaptive Routing", In Proc. of the
International. Joint Conf. on Neural Networks,
IJCNN, 2002.
ox$ o( o o5$ Bo. o;! s '(I

':o5 FoE. Jo( , o o
( (" '$) ; $
o o$ ';o, (o. o E(C a
N R
' . \@6
N J(; '^!$ IN , }D. x$
1 2 1387

G!, 5 '(0 . ':5 FE. B. '$- ' CD$ (E
[5] D. W. Glazer and C. Tropper, "A New Metric

for Dynamic Routing Algorithm", IEEE Transactions
on Communications, Vol. 38, No. 3, March 1990.
[6] M. Kara, H. Karabelli & N. Duru, "Fuzzy Based
Routing In Packet Switching Networks", International
XII, Turkish Symposium on Artificial Intelligence
and Neural Networks TAINN 2003.
[7] Chiang Kang Tan, "The Use of Fuzzy Metric in
QoS Based OSPF Network",
http://www.ee.ucl.ac.uk/~lsacks/tcomsmsc/projects/
pastproj/ck_tan.pdf, 2000/2001.
[2] A. Pasupuleti, A. V. Mathew, N. Shenoy and S.

A. Dianat, "A Fuzzy System for Adaptive Network
Routing", Proceedings of SPIE Vol. #4740, SPIE's
16th
Annual
International
Symposium
on
Aerospace/Defense Sensing, Simulation, and
Controls, Orlando, Florida USA. 1-5 April 2002
[3] J. Boyan and M. L. Littman, "Packet Routing in
Dynamically Changing Networks: A Reinforcement
Approach", Advances in Neural Information
Processing Systems, volume 7, pages 671-378, 1994.
[4] Bruce S. Davie, Larry L. Peterson, Computer
Networks: A Systems Approach Morgan Kaufmann,
second edition, June 2000.
1 2 1387

26-39=X@ 1 2 1387

Q"
O ,#
- !I)! (! Y2 M1Z (6J"%
!& [ %
2
NO ' PE :!0 6
;< .,5 L<N . 1.R J(!%N 0 P@ L<N ,<N
1
h_khaloozadeh@kntu.ac.ir 2makbarisani@yahoo.com
________________________________________________________________
CN /0 ':E! , +$ $. +:0 !CN W5 (C! $N^ ' J$ ':E! !C J( :%+
W!6 x 13 D0K W!6 'w5 #)P j
W. a
N D0K #% J ,0 '
\& ' ' ,N @ ! ( $ IN , (5 I(N , , ;!6 +N B. $ . \<$
4 >. 5%$ '$ : X. W. +" 5%$ X! \<$ ' X! : ( , 5%$
' (I h$ $ 5%$ M; L! ( , 5%$ M; LV$ ( J ,0 ' 4 J .
5 W!6 5 L! IN , ,0 5 L! +" & , 6R ' $N^$ FE.
J! LV$ E(C 5%$ ; E C 6 FE. . . E(C W!6 LV$ +P 5( FP$ J!
.W. 5%$ M; ! ') ] J #N.N D ( >DS (
5( FP$ J! 5 +$ + ' $N^ $ ,0 , W!6 5 L! :' ' ^,
5%$ M;
Abstract: This paper presents nonlinear modeling, simulation, and controller design for wind
energy conversion system in Dizbad power plant. The wind power plant scheme consists of a
three-phase induction generator that connected to the wind turbine via a fixed ratio gearbox. A
static VAR compensator was connected at the generator terminals to regulate its voltage. The
induction generator was connected to the utility through a double transmission line. The
mechanical power input and terminal voltage output was controlled using blade pitch angle and
firing angle of thyristor. An optimal stochastic control is designed for the system based on the
optimal stochastic estimator. The performance of the proposed scheme is compared with both
output feedback and state feedback controllers with extended Kalman filter. These controllers
are applied to the nonlinear system and their performances are recognized under gust and
different type of disturbances. The responses of the proposed optimal stochastic controller
exhibited a good damping and fast recovery under these disturbances.
Keywords: State and Output Feedback Controllers, Wind Turbo Generator, VAR Compensator,
Extended Kalman Filter, Stochastic Optimal Control.
________________________________________________________________
http://www.isice.ir
+( ?S I , +.; J

N 1387

27
NO ' PE :!0 6 5%$ M; 5 L! B. +" , I( ' J$ D0K ':E!
%9 .1
oD0K FoE. +: W&. '(nx$ W W &
Fo] #ox$ o ,i& (: .(N \< ' !pE ' J$
qio ',, o& >o IoN /.$ 9 M!pE W&.
Jo( \o6 'o NoN 'o 9, .o o FE. W0( K
Io( FoE. ( : X. [1] .W. ;< +$ \<
' '^N \($ FE. 0. (1) \
Mo(,: , SVC oE($ o$" Mo(,: \o l5 FE. ' ',,
o$N^ W&o. 'o :N \S 'x . J$ =N
',, .oo oo ( Pg ,Vs , I s ) +" +oo( , +oo$ , ôo!, , Z m
Boo Moo ôo!, Joog .oo oo v w oo W&oo. FooE. 9oo<K
FoE. o ao( .o o {o5 WoO , a P Vb W(;N
:W. +<N (: \
.W. ;< #,X! LN +& _P| '

Mo(,: o o o , W. X. +,. $N^ ( : [2]
LN $N^ WO W&. (IN J$ W&. :5
oo$N^ oo(=$ ôo!, oo 'oo ':oo5 ooL! oo( [3] .Woo.
+oR Wo S W & .W. 6R E($ $" M(,: , +,.
\(o$ 'o FoE. ]oC! 'o$N^ : X. M(I , $)
'o o|( o o( .Wo. o o< o0 ';!. ' '^N
.W. + [4] '$ (C! $N^ ' '^N \($ FE.
oL! Jo( .Wo. o o& o ', Ix! W!6 [5]
#N.N , N #N.N +N , +N N
FE.
'o$N^ o o SVC : Xo. [6] .o o ,<
FE. a( (2) \
(! -< ]
NJ (6 % .3
'o 'ôN \(o$ FoE. +,o. $N^ : X. ' (C!
(! Y2 % .1.3
.W. +
o o ,0 o( , ',, o. FE. ( #@ ' J$
o,N o( o$N^ J$ > FE. ':E! !C J(
, o W&o. o 'o J$ ( '!$ +$ ,0 +I .
+:oo0 !ooCN W5oo (ooC! oo$N^ 'oo Joo$ : \oo< 'oo
(: \ +$ MD . LE ' J$ :5 M(,: , $ W&.
Jo o ,0 ' CN /0 ':E! , +$ $.
.
ô!, , +$ ( , ' G-. a

N D0K #%
Pm
AR
C pV 3
5o%$ Mo; JV$ M( 5%$ M; 5 L! ( : J
(1)
2
v, }D. AR J$ N #S Pm l5 FE.
C p , o ',, ' o4=! W&. V !L* U $ '
[7] . J$ #S WO
: o( , FCoE 'o :oN ?o(R : +o$ o C p #S WO
Mo(,: : Po$ #oS WO .," W. (,(" #.= ?(R
. (C p (O , E )) o o o W&. rN W&. WEN ,
[1] .W. _(P$ (: #@ #S WO J(
C P (O , E )
S (O 3)
(0.44 0.0167E ) sin
0.00184(O 3) E
15 0.3E
oD0 W&o. WoEN) O
'oI(N FoE. oPS, o( $ JV$ J( . X.

o( o $ . 5 4N FE. ' ' :N , (5
FoP$ J!o o 5 W. D0 K FE. ( ' '^N \($ FE.
6R I(N 'W!6 D0K JV$ W; FE. J( ' 5(
o o6R +" o o $ o >oDS ( W!6 5 L!,
. E(C 5%$ M; L! +" . & : G .
(! 2<\ NJ 21 .2

FE. J( . +<N ' '^N \($ FE. 0. (1) \
: , 'o Jo$ $ R ( : W. (C! $N^ ( \
(2)
.Wo. o \o% o . , CN /0 ( ?(R : L( R
R ZR
, o o :o5 M(,: E +"
V
+o$ o +o o( : (C! J ' $ ^!, F4$ '
0* ' (,: W&. Z R , W. (',, W&. rN
o(: MD : J$ N ,< .W. $ jP R , J$ '
. \@6
o( \o + J( .W. X. (SVC) $.
o$" Mo(,: ox$ o o o Wo. ox _ o. o( , WoO +:0
. x$ +" GN,N E($
1 2 1387

NO ' PE :!0 6 5%$ M; 5 L! B. +" , I( ' J$ D0K ':E!
Rs X rr
D
Mqs
Vqs
1
M

ds Z Vds Z
b
b Rr X M
Mqr
Vqr
D

Mdr
Vdr
Rs X ee
D
0

Rr X M
D
Rs X M
D
Rr X ss
D
Z Zr
Zb
Rs X M M qs

D M ds
Z Zr M
qr
Zb M
dr
Rr X ss
D
0
Tm
Pm
ZR
28
1
UARC pV 2 / O
2
(3)
Mo(,: o W&o. \& . ' J$ N #S (

+$ >(| #x$ (: \ . s$ $ ' (,: W&. ,
. W. +<N
(12)
.oN o \oC $o. Wo. $ 'x $ #)P J(

o oN (: #@ pu 1 #@ $ ERx ,! ,<
.
Te
M dr idr M dr iqr
(13)
"9< % .3.3

Ioo ';ooES oo( : ':oo. oo. 'oo Goo oo
No 'o WoES o oE(C ' 46i \S R N
o.( o ';o qo@%0 J( . 45@ E E"
_ V :5 '(,: ' (O ) WEN +$ >(| #x$ (3) \
Go . l@ WEN a:) ' (;N: , )
)9" 2<\ % .2.3
ao& J5o o4N , oK o( : + 45@ W!6 J(
oK ' o(,: W&o. , Z m o ' o(,: W&. W5 DPN
'o= 'o. +,o. ho L. ]C! $N^
oR o ZG G W&. F R ' (,: W&. , Z H $
N 5 4N \CE x +& <N ' S, d, q
o$N^ o$ No W&. FE. J( . 5 4N
, ' (N \ J( ,"( E( .[8] W. (: #@
Go >(o| RG +" o o o _(P$ (: #@ Zm
o4N Wo!6 ox +o& o +( !6 WEN ':.
.W.
.W. $ >. F(L
M ds
Zb (Vds Rs ids M qs )
#o@ o No , (! M(,: W&. J MD J5 4N
M qs
Zb (Vqs Rs iqs M ds )
p
Zm
2
M dr
Zb (Vdr Rr idr ) (Zb Zm )M qr
M qr
Zb (Vqr Rr iqr ) (Zb Zm )M dr
Zm
RG ZG
(14)
: 0 (: #@ W6 ( M!P Z r
Z m
Zb
2H A
(Tm Te )
(15)
(4)
(5)
(6)
(7)
.W. (: #@ +( MD #)P J(
(oC! $N^ , ' J$ .( j

N: WO H A +"
.W.
M qs
X ss iqs X M iqr

2 , ,61 % .4.3
M ds
X ss ids X M idr
M qr
X M iqs X rr iqr
Mo(,: ox$ o o +$ . F4$ J( .
M dr
X M ids X rr idr
+o( , ô!, J MD WO +:0 ' . \@6 E($ $"
(: W!6 'w5 FE. ' +( P (C + S
(: \ d, q '= '. W!6 'w5 #)P #@
. !$
o o o +$ + WO +:0 ( : !C J(

Xo. o$N^ 'o oN( ^!, J LN WO ' E($ /.$
(8)
(9)
(10)
(11)
:W.
Per unit
1 2 1387

29
NO ' PE :!0 6 5%$ M; 5 L! B. +" , I( ' J$ D0K ':E!
Vqs

Vds
oD0 o .W. D0 DCN 6 (19),(18) (16),(15),

:W. (: \
X
0
Z
b
(16)
Mo(,: (I5 . 90 M(,: \ W( E($ '
AX Bu Gd
(20)
: Xo. o .o o o +o( , o +$ d& $"
.W. (: X , u , d +"
. .= (: #@ (C! +( @& (5 \ =$

2(S D ) sin[ 2(S D )]
Vs
SX L
X ['Mds,'Mqs,'Mdr ,'Mqr ,'Zr ,'Vds,'Vqs,'idl ,'iql ,'idTL,'iqTL]

t
u ['E,'D]
d 'Vw
Wo!6 'ox o soCN ':o. D0 N, W. m a:)
oDCN o oN 'o * : G .N JP E( , ',,
9o<K o( , ',, , , Wo!6 x 11 \ X 014u1 (
.W. 5 4N (: #@
X0
Zb Vqs
i
Zb X c qc
0 Vds
idc
[M ds 0 , M qs 0 , M dr 0 , M qr 0 , Z 0 r ,Vds 0 ,Vqs 0 , idl 0 , iql 0 ,

idTL 0 , iqTL 0 , E 0 , D 0 ,Vw0 ]t
.W. (: #@ X 0 '&
iL
.Woo. oo$" Moo(,: D , ooE($ ooP GNoo
X eq
+" oo
: N (: #@ +,. h L. i L '

iqL

idL
0
Z
b
Zb iqL Zb

0 idL X Leq
Vqs
V
ds
(18)
9< 01 % .5.3

ho Lo. d, q 'o= 'o. oCN /o0 #)P
: N (: #@ W!6 'w5 \ +,.
"4 % Q"

4_ .5
(17)
Wo!6 'ow5 \o o d, q 'o= 'o. \ENX( #)P
X 0 [0.80,0.10,0.74,0.233,323.58,0.09,0.80,0.17,0.02,
0.35,0.61,6,1.66,12.5]
Vs
X eq
iqTL

i dTL
.o o Jo +$ , N( ^!, \ q$& ' J$ ,0
Z b RTL
X
TL
Z
b
i Z V Z V
qs
qb
qTL b
b
Z b RTL i dTL X TL Vds X TL Vdb

X TL
Zb
(19)
o +o$ J!, . , ' J$ ' ,
9, J( I .W. (C! J ' N( ^!, F4$ '
',, ,< J, .W. SVC _ . E($ $" M(,:
:oo5 Moo(,: oox$ 9, Joo( oo Ioo .Woo. 'oo oo$N^ oo
FE. ', N '<5 ' .W. ' J$ '
ox$ oL & 'o r10 deg/ s ,o= o6 ( ' J$ g
#4$ , (! +$ #4$ .F( 4N J$ M(,:
Mo( o oL! Jo( o6R . \@6 I L! ?(R : ^!,
'oooD0 'ooo Looo. ', ooo ooo oooLN , .Wooo. oooD0 ooo
o o Lo. , J( . &
(Vref Vs )
( Pref Pg )
\o@6 o(: ,I5 $ N >$ S W!6 'w5 D0

.
z
)z *u Ed
(21)
. _(P$ (: #@ +"
+
Q ^% M1 % .4
'o 'ôN \(o$ FoE. ( ' FE. (: $
WoEN Go o. , oCN /0 ]C! $N^ ' J$ $)
'w5 \ +$ W(;N ((SVC) +$ + , WO
o oDCN o6 W!6 #)P .," W. (;N FE. W!6
:(" (: #@ * L. N D0
X
AX Bu
[7]:W. (: X , u +"
X ['Mds, 'Mqs , 'Mdr , 'Mqr , 'Zr , 'Vds , 'Vqs, 'idl , 'iql , 'idTL, 'iqTL]t
u ['E , 'D ]t
d
'Vw
MoDCN 6 L #x$ J( .N 5P | W!6 'x

' G($ .E d, q '= +,. h L. ,
.W. " | IN A, B
Wo!6 'ow5 oD0 o M( !C J( L! 6R 9,
(12) #)ooP \oo 'oo oo,N FooE. ooD0 ooK oo .Woo.
1 2 1387

NO ' PE :!0 6 5%$ M; 5 L! B. +" , I( ' J$ D0K ':E!
W!6 5 E C 6 FE. , : C 6 FE. ' >DS (1) ,
FE. `(, (C
C 6 FE. `(, (C
: C 6
W!6 5 E
23.4 j 2888.4
23.4 j 2888.4
23.4 j 2888.4
23.4 j 2888.4
27.28 j 2134.4
27.28 j 2134.4
27.28 j 2134.4
27.28 j 2134.4
22.04 j 375.08
22.04 j 375.08
22.04 j 375.08
22.04 j 375.08
0.34 j 377
30 j 377
0.34 j 377
30 j 377
37.14 j 9.59
37.14 j 9.59
37.14 j 9.59
37.14 j 9.59
0.42
40
5.5
0.8
z13u1
X 11u1
J
,
2u1
'P 'P
ref
g

A
) 13u13 11u11
D 2u11
dt , 'Vref
0
, *13u1
0
30
t
'V s dt
B11u2
0 , E13u1
G 11u1
0
'o 9, o o Wo!6 o5 L! ) ,I5 $

. \@6 (: \ >DS ( 6R
K s z
(22)
o .o o \o@6 K s G($o oE oC 6 'o >oDS vVN

. N (: #@ y ,0
y 6u1
C6u13 z13u1
(23)
. _(P$ (24) MD #@ y +"
y ['Pg , 'Vs , 'I s , 'Z r ('Pref 'Pg )dt ,
('Vref
'Vs )dt ]
(24)
[2] . +<N W!6 5 L! a( (: \
) N` Y"

4_ .6
'o Wo!6 o0 o;$ o Wo!6 2PDS ; '$
\S E ':N \S W!6 $ I(N : '& '&
o( ; ' W!6 JV$ J( : 5%$ 9, .
W!6 5 L! (4) \
. X. 5%$ FE.

o( o ,io& ,0 ( $ ' J$ FE.
(1) ,o o UV< ' + E C 6 ' >DS vVN
J!o o 5 : Xo. .o o IoN I(oN !" N B.
: \@6 (: #@ K s G($
.W. ',|
( W!6 D0K ' FE. ' JV$ !pE
D0K 'FE. ' ' .= F(L! ( . ' .W.
o5( FoP$ J! 5 W(;N W. & >(C$ 0
:W. (: #@ ' :N , FE. .

z
y
f ( z , u , t ) G( t )w
h( z , t ) v
(25)
z( 0 ) | N ( z 0 , p 0 ), w | N ( 0 , Q ), v | N ( 0 , R )
Ks
106.29 195.06 47.54 32.02 71.67 0.821 2.67

0.438 0.073 1.479 4.84 0.804 0.013 0.016
22.008 639.9 3.21 99.82 2746.2 7678.5

0.688 0.616 0.04 0.648 37.11 144.1
"
!
o o. o oN o voVN 'oN o E C 6 v D ' >DS
. a:) (
o Xo@ Mo4=! GNo(, , /o. o 5o%$ ox z +"

FoE. ',,
u , FoE. oD0 K h$ f ( x ,u ,t ) , ( z0 , p0 )

o ,0 y , I(oN +o: o x N:, G($ G , (5 I(N w ,
. ':N I(N v , ':N \S
2
Deterministic optimal control
1 2 1387

31
NO ' PE :!0 6 5%$ M; 5 L! B. +" , I( ' J$ D0K ':E!
o Wo!6 J(o$ (IN , J(; W. $@ z JV$

Wo!6 o . 5%$ , 0N 'I(N w6 z W!6
.W. (: #@ !, '
P (0) p 0 , z (0) z 0
(26)
:W. (: #@ JV$ 'D0 GN( P G($ + :,
P A(z, t)P PAT (z, t) GQGT PHT (z, t)R1H(z, t)P
(27)
o _o(P$ o(: #o@ o (J^ ';E($) H , A D J(
:W.
;!6 JV$ , I(N !" E MC 6 D0 K FE. a( (5) \
5( FP$ J! 5
GG ,GG6 GGM1 a-! [GGL2! GG ["PGGJ .7

(LQG)
: {5 (: #@ FE.
x Ax Bu Gw
(31)
Xo@ Mo4=! GNo(, , /o. o 5o%$ W!6 x x +"
',, u , FoooE. 'ooo G($ooo A, B , x(t 0 ) | N ( x0 , p0 )

G($o G, w | N (0, Q) o. #@ (5 I(N w , FE.
. I(N N:,
: W. (: a, M #@ & U0
J [ x(t 0 ), t 0 ]
1T
1 T
x (T ) S (T ) x(T ) ( x T Qx u T Ru )dt
2
2t
(32)
Q t 0, R ! 0 , +oooC oooN:, G($ooo S (T ) t 0 +" ooo
wf ( z , u , t )
A( z , t )
wz
(28)
wh( z , t )
H ( z ,t )
wz
'oD0 GNo( G($o +o UVo< , $o( Mo!P \6 : G
: .= (: #@ J! ; P(t )
K (t ) P(t ) H ( z, t ) R 1
(29)
:(" (: #@ ;!6 JV$ \ENX( M!P W(;N
T
z f ( z , u , t ) K ( y h( z , t ))
(30)
y h( z , t )
\oS oK Lo. 'o JV$ J(; l5 JV$ WO
FE. ' . D0 GN( J( 4N : z I(N , B.
, W!6 x 13 + \S WES R + ' ,N D0
oo oo( #oo@ oo oo(5 I(ooN . oo , ',, 2 , oo ,0 6
o o 30 db #oS , Xo@ JLNo . 5%$ 13u 1
N:, 'G($ , WNi . ',, , W!6 N:, ' G($
o ooN o< I(oN ( ' W!6 (

N" : .W.
. +: : P$ N$
o(: #o@ o 'DS G($ ( #@ G N:, G($ +$
o Wo. $@ [t 0 , T ] : u (t ) L. JP$
:W5 4N
. \S6 (: M(I h$
j (t 0 )
E{J [ x(t 0 ), t 0 ]}
(33)
.W. WO T +: , :" x(T ) +"
(LQG) 3o. a, Mo oD0 o M!poE aoN o !poE J(
. 0
a` "4
. ':N \S qCS x(t ) W!6 F {5
diag>1 1 1 1 5 4 4 3 3 10 10 10 10@
5%$ 6u1 I(N ( #@ ':N I(N Jg

.W. 30 db #S , X@ JLN .
o $ oE( .Wo. o & !pE J( W!6 5 W(;N
o 5 : & ( z (t ) )#)6 JV$ 5 W
oC 6 FoE. o a( . \@6 D0K 5( P.$ J!
.W. " (: \ 5( FP$ J! 5 , W!6 5 E
: 0 (: #@ D0 5 +NS W!6 J(
u (t ) K (t ) x(t )
(34)
:W. (: #@ \S6 & U0 +" D,
[9]
: JP$ (: #@ 5 ;
3
Linear Quadratic Gaussian
1 2 1387

NO ' PE :!0 6 5%$ M; 5 L! B. +" , I( ' J$ D0K ':E!
':o (I h$ (I h$ +" 'nL( , l5 #)P : X.
32
K (t ) R 1 B T S (t )
(35)
:+"
.W. %6 \S (: \ ; JV$ , ;

j * (t 0 )
1
trace
2
1
1
E [ x T ( t 0 ) S ( t 0 ) x ( t 0 )] trace
2
2
( SG Q cG
) dt
(K
R v KP ) dt
\o@6 o(: \o o $( M!P \6 : S (t ) N: +C G($

:
t0
(45)
t0
\@6 PDS LQ F4$ : , M W. m a:)
S
1 T
A S SA SBR B S Q
T
(36)
j (t0 ) Mo; Wo!6 #@ j (t0 ) M(I h$ , J( $

:("
9<K WPDS a& W & (I (I5 LN a, M .W.

(I (I5 LN 0" .W. (5 I(N MD., x(t ) L.
j * (t 0 )
.W. ':N WPDS a& W &

hoo ',, Lo. oo $o@ o LQG oo Fo4$ o6R
F4$ @ #)P .W. 5 4N +" ' IN r (t )
:W. (: #@
x
( A LH ) x Bu Lz
K opt x r
(46)
(47)
J! 5 , 5%$ ; 5 E C 6 FE. a(

.W. " (6)\
1
1
E[ x T (t 0 ) S (t 0 ) x(t 0 )] trace SGQ cG T dt
2
2
t
(37)
:(" W. (: #@ u * (t ) M; L.

N
u * (t ) R 1 BT S (t ) x(t )
(38)
. , #.= F
6 W; I 5 :N ' T '
Xo. o(: \ K F] W!6 WO : +$ L! &
:
K f R 1 B T S f
(39)

( S 0) .W. $( M!P JP W , ,= \6 S f +"
a`< "4
\oS qoCS x(t ) 'oWo!6 o$ {5 W!6 J(

o(: Mo!P o z (t ) 'o:oN o +" 'o o .N ':N
:
z (t )
Hx v
(40)
Jo( .o o o. 'o:N I(N v | N (0, Rv ) +"
'o:oN ', : x (t ) JV$ ;$ u(t ) L. W!6
ooo(I hooo$ ':ooo. ooo; ooo .ooo o o oooE, ooo \ooo@6 ooo
>oC& , 4v, 9, : X. , j (t 0 )
E{J [ x(t 0 ), t 0 ]}
. (& , JV$ (5 ' $( E($ !P . ,
5 JV$ , ; E C 6 D0 FE. a( (6) \
J!
(! -< !& [ Q"

4_ .8
K opt 5o%$ o; o5 o; o m $( #)P \6
:: W. #& .=
Kopt
0.0021 0.036 0.033 0.016 0.108 0.0003 0.001

0.248 1.146 1.185 0.084 0.0068 0.036 0.396
[11] : (: #@ JV$ #)P
P AP PA GQ cG T PH T ( Rv ) 1 HP
P(t 0 ) P0
(41)
T
PH T ( Rv ) 1
L
x
& ' K (t ) ; ; /( J( .W. W!6 JV$
. , x (t ) JV$ L. x(t )
: 0 (: #@ D0 5 +NS
u (t )
K (t ) x (t )
(44)
. 0 (36),(35) #)P + K (t ) s #)P
K opt 'nL( . +<N >DS \= (2) , .N

4
(42)
Ax Bu L( z Hx )
(43)
x (t ) , J! 5 ; L(t ) , D0 GN( G($ P +"
1.99u10 5 0.00037 0.031 0.00059 9.998 0.177

22.181 0.0182 0.1163 0.1706 9.98
1.6562
.= E C 6 ' >DS 5%$ ; W!6 5 ; J(
t t t0
Sweep method
1 2 1387

33
NO ' PE :!0 6 5%$ M; 5 L! B. +" , I( ' J$ D0K ':E!
+$ (>DS ( ', : " W. ) K s 5 ; '

. E(C , < ;V.
(6 .9
: Xo. o o om /o, ?R " W. K s &
oC 6 FoE. o5 ,Io5 FE. W!6 5 & , >DS (
o oD0K FoE. I(oN +,o oD0K FE. W!6 . E
o 5 : Xo. o oD0K FoE. IN , (':N , (5 I(N)I(N
o FoE. o. o4 J(o .o o . 5( P.$ J!
Jo( o (o DCN) !, W&. .W. 5 4N $ &
(oNO o o 12) N W&. (IN W. NO 12/5 !pE
+o: .Wo. oNO o o 16/2 o$ W&. J($) .W. ' J$
L( MDCN ( W&. $ : G .W. NO 1/7 ,6 $
#No.N 'o. o( o !o6 , o W. !, C : $)
oO +o& NO 1/4 #x$ M , NO 2 v,$ , ..
o$ ko \o . oS +" ', Jo$ 'o o /.$ 9
(13) o$ (7) ' \ E C 6 FE. '',, , FE. ,0
.W. +<N a,: #@ , PS, :N
oD0K FE. 5( FP$ J! 5 & + '
D0K E MC 6 FE. J 'E(C (: '\ I(N !"
o$ + B. {5 , I(N J5 4N +, W!6 5
J!o JV$ FE. , I(N !" FE. , $ & ;!6
o o +o<N o \o R + .W. 5 #@ 5( FP$
#No.N , o vo D 'o( , ( ' E C 6 FE. .
C 6 FE. . . C \S6 J$ +$ , ^!, , W&.
'o( , o( ' o( o ,io& o5( FoP$ J! 5 E
S C \S6 J$ +$ , ^!, , W&. #N.N , v D
+o . I(N +, D0K FE. . (IN NV
o5( FoP$ J! 5 : X. +<N k \ R
.N + I(N : N #N.N +$ D0K FE.
1 2 1387

NO ' PE :!0 6 5%$ M; 5 L! B. +" , I( ' J$ D0K ':E!
34
5%$ ; 5 , W!6 5 E C 6 FE. : C 6 FE. ' >DS (2) ,
6 94 NJ c)- )!9
% J 94 NJ c)- )!9
% J 94 NJ c)- )!9
"4
!&
23.4 j 2888.4
23.4 j 2888.4
18.106 j 2356
23.4 j 2888.4
23.4 j 2888.4
18.106 j 2356
27.28 j 2134.4
27.28 j 2134.4
17.48 j1728
27.28 j 2134.4
27.28 j 2134.4
17.48 j1728
22.04 j 375.08
22.04 j 375.08
26.24 j313.58
22.04 j 375 .08
22.04 j 375 .08
26.24 j313.58
0.34 j 377
30 j 377
3.69 j19.75
0.34 j 377
30 j 377
3.69 j19.75
37.14 j 9.59
37.14 j 9.59
2456
37.14 j 9.59
37.14 j 9.59
61.46
0.42
40
7.22
5.5
0.586
.8
0.442
1 2 1387

35
NO ' PE :!0 6 5%$ M; 5 L! B. +" , I( ' J$ D0K ':E!
+: >E6 W&. #x$ N (7) \
+: >E6 $N^ +$ , $. ^!, #x$ N (10) \
+: >E6 ' $N^ W&. #x$ N (8) \
( a,:) +: >E6 ' $N^ W&. #x$ N (9) \
+: >E6 $N^ +$ , ^!, #x$ N (11) \

( a,:)
1 2 1387

NO ' PE :!0 6 5%$ M; 5 L! B. +" , I( ' J$ D0K ':E!
36
oE C 6 FE. '',, , FE. ' ,0 k \ .W. 5

oD0 oK FoE. o. .Wo. o +<N (19) $ (14) ' \
, 5( FP$ J! 5 E(C 5%$ ; E C 6
'oR o .o v D '( , ( ' >DS ( 5
o $o. ô!, 'o , %25 ,6 F] W!6 #N.N +$ '
Fo0 F] W!6 #N.N %85 , , #N.N %60
%50 ,o6 F] W!6 #N.N $ ' (,: W&. ' Jg .W
%40 ,o6 Fo] Wo!6 #No.N $o. +o( 'o .Wo. 5(
.W. 5(
G($o + S W & .W. (: E #,X$ ',,
o .Wo. o5( E ',, #N.N u (t ) L. ' R N:,
o , , #N.N %88 $" M(,: ' 'R
Jo$ :5 M(,: ' Jg .W F0 F] W!6 #N.N %97
F] W!6 .W. ; E , NO 1 n . ( ,i&
.W. 5( %80 ( MDCN 6 #N.N IN
FE. ' :5 M(,: , $" M(,: #x$ N (12) \
+: >E6 D0 K
+: >E6 ' $N^ W&. #x$ N (14) \
+: >E6 $. ^!, #x$ N (15) \
FE. ' :5 M(,: , $" M(,: #x$ N (13) \
( a,:) +: >E6 D0 K
FoE. 5%$ ; , J! 5 & + '

I(N !" D0K FE. J 'E(C (: '\ I(N !"
+: >E6 $N^ +$ #x$ N (16) \
>oDS o( o Wo!6 o5 o o o 5( FP$ J! JV$
#o@ 5o%$ o; Wo!6 o5 FE. + , ([8] h )
1 2 1387

37
NO ' PE :!0 6 5%$ M; 5 L! B. +" , I( ' J$ D0K ':E!
JoV$ o oE(C 5o%$ o; o o oE oC 6 FE. .
Jo o o #No.N o o D ( >DS ( J!
+o* W. N J( ( IN , W. 5%$ M; L! ') ]
' L! vVN W. 5%$ ' X! ' P!D FE.
.W. 5%$ ; '$
KL
+: >E6 $N^ $. +( #x$ N (17) \
o( GNo5 ô!, o PoE o0 oC$ o= . (=D [1]
.ox 0o. 9, o : \CE I( o $ * ,N
242o237 M=X@ 1377 W<;( 24 o22 l .; GNX J<
[2] J. R. Winkelman and S. H. Javid, Control design
and performance analysis of a 6MW wind turbine
generator, IEEE Trans on PAS, Vol. 102, No. 5, pp.
1340-1347, May 1983.
[3] R. Chedid and F. Morad and M. Basm , Intelligent
Control of a Class of Wind Energy Conversion
Systems, IEEE Trans on energy conversion, Vol. 14,
No. 4, Dec 1999.
[4] B. S. Borowy and Z. M. Salameh, Dynamic
response of a stand-alone Wind Energy Conversion
Systems with battery energy storage to a wind gust,
IEEE Trans on Energy Conversion, Vol. 12, No. 1, pp.
73-78, Mar 1997.
[5] H.D. Battista, R. J. Mantz, Sliding mode control of
torque ripple in wind energy conversion systems with
slip power recovery, IEEE 1998.
[6] Y. H. Song, A. T. Johns, Nonlinear thyristorcontrolled static VAR compensation, The European
Power Electronics Association, Brighton, Sept. 13-16,
1993.
o ,0 5 L! 6R PE NO ' 6 :!0 [7]
GNoX Jo
.oI( 'o J$ D0K ( FE. '
.1382 ; +;$ +( l ! J
+: >E6 D0 K FE. $" M(,: #x$ N (18) \
>E6 D0 K FE. ' :5 M(,: #x$ N (19) \
+:
( < .10

, og FoE. o( : oCS oEN oD0K ':oE! !C J(
Wo!6 o5 oL! o6R 6 :!0PE NO ' [8]
o$N^ 'o Jo$ \o oI( 'o ,N $N^ -J$ >
GNooX Joo:, ooI( 'oo Joo$ ooD0K oo( FooE.
a
N CN /0 , +$ $. +:0 !CN W5 (C!
.1383 W<;( ;< (ICEE 2004) +( l .;
o \o<$ Wo!6 x 13 D0K W!6 'w5 #)P j

[9] M. S. Grewal, A. P. Andrews, Kalman filtering:

theory and practice, John Wiley, Press 2001.
Mo; o5 oL! o6R PoE NoO 'o 6 :!0 [10]
.
Mo; o5 oL! o6R o ao
N ':oE! Bo. G-.
J:oN oI( 'o Jo$ oD0K o( FoE. 'o 5%$
o( : Xo. o (oC! Jo 'o o$ ^!, F4$ ' 5%$
.1383 ; +;$ +( l ! J GNX
: Xo. o o(! +o$ F4$ IN , $. +$ +
[11] A.E. Bryson, Y.C. Ho, Applied Optimal Control,

Hemisphere, 1975.
B. $ . a
N sCN : P., bN :5 M(,:
JoV$ o( Jo ,0 ' (5 I(N , , ;!6 +N
FoE. 'o o(I h$ $ 5%$ M; ! ( , 5%$ M;
o5( FoP$ J!o 5 +" & , ( 6R ' $N^$
. E(C >DS ( W!6 5 !
1 2 1387

NO ' PE :!0 6 5%$ M; 5 L! B. +" , I( ' J$ D0K ':E!
38
I( ' J$ 5 #%V<
:e
C5 = ' .
J$ jN
43
$ DS
(Stall) hDS ( F4$
+$ F4$
NO 3-4
J( hDS W&.
NO 12
N W&.
NO 25
') hDS W&.
#, 660
$N^ N +$
CS , 1500
$N^ +,. W&.
19/75
N: WO
1:56
G >(|
:5 'x WE!

pu>E6
d, q '= '. $. <N
pu>E6
' = '. $ <N
d, q
pu>E6
d, q
i ds , i qs
pu>E6
d, q
' = '. $ +(
i dr , i qr
pu>E6
d, q
' = '. +:0 +(
i dc , i qc
' = '. E($ +(
i dL , i qL
pu>E6
d, q
' $N^ 5 #%V<

() K jX$
690
(W!,)N ^!,
0/91
N +$ >(|
50
(I$) N GN5
1500
+,oo. W&oo.
2/88
GRooox GNooo
(pu)
(RPM)
0/0059
(pu) $ W,C
0/0063
(pu) $. W,C
0/0058
(pu) $ GN
0/0489
(pu) $. GN
M dr , M qr
' = '. $. +(
pu>E6 d, q ' = '. CN /0 +(

60
M ds , M qs
i dTL , i qTL
pu>E6 $. +( , ^!,
I s , Vs
pu>E6 (C! $N^ N +$
Pref
pu>E6 (C! $N^ +$
Pg
pu>E6 $., $ W,C
Rs , Rr
pu>E6 CN /0 GN , W,C RTL , X TL

pu>E6 SVC GN , GN,N
X L .X C
pu>E6 d, q ' = '. $. ^!,
V ds , V qs
pu>E6 d, q '= '. W(;N B ^!,
V db , V qb
pu>E6 $. N ^!,
V ref
(314rad/s) N W&.
Zb
(rad/s)' $N^ $ W&.
Zr
(pu) \($ ' ( (C

FE. a$ ' 660 KVA : ( +$
FE. a$ ' 690V : ( ^!,
:(SVC) +$ +
XC
3.8 pu
, XL
4.0 pu
:CN sD0
X TL
0.15 pu
:GN . RTL 0.015 pu :W,C
1 2 1387

39
NO ' PE :!0 6 5%$ M; 5 L! B. +" , I( ' J$ D0K ':E!
Z b R ss X ss

D
Zb
c XM
Z b R rr
Z b X M I qr 0

2H AD
c
Z b X c X rr

Zb
2H
0
0
0
0
0
0
0
0
0
0
wTm
wE
Zb

Z b R ss X M
D
Z b R ss X rrc
D
0
Z b Rrrc X M
D
Z b R ss X ss
D
(Z b Z m 0 )
Z b X M I dr 0
Z b X M I qs 0
2H AD
2H AD
Zb X c X M
D
0

Zb X c X M
D
0
0
0
Zb
Z b R ss X M
Zb
I qr 0
I dr 0
D
(Z b Z m 0 )

Z b R rrc X ss
wT
m
wZ m
Zb
Zb X c
Zb X c
Zb X c X M
Zb
Zb X c
Zb
Zb
Zb
Zb
Z b X M I ds 0
Zb
2H AD
2H A
D
0
X TCR 0
0
Zb
X TL
0
Zb
X TCR 0
X TL
Z b RTL
X TL
Zb
Zb X c
Zb
Z R
b TL
X TL
0
0
0
0
0
0
0
wf
wD
wg
wD
0
0
1 2 1387

40-48 =X@ 1 2 1387

)9 (!)

5-2 0
(6
2
!1 4_
- JJ
2
= 1E(, S &

+ ; L<N l ,
a_gharaveisi@uk.ac.ir , 2kashki@gmail.com
____________________________________________________________________________
9, . ] BP uN," D0 K FE. ' . $ ':5 6R W; ( , !C J( :%+
, \ n 6R 'L! .N W=$ FE. ( #)P ':N , 0 #@ ] 6R
W. . $ >(| >. : , EE (C$ '( '$$ , 6 6
$ ]
N" : . JP$ >(| ; C , . (C$ '( '$$ a, 6 , ,"
. IN D0K ';E. L( ' +" +$ 6 + & #@ 6R \6
. $ ':5 . (C$ '( '$$ EE (C$ '( '$$ :' ' ^,
Abstract: In this paper, an optimal Takagi-Sugeno fuzzy control is designed with a novel
method called combined discrete and continuous action reinforcement learning algorithm
(CDCARLA). The proposed method is implemented for a nonlinear system that is Cart-Pole
system. Simulation results show that the proposed method has significant performance. The
advantage of CDCARLA is that it does not need system dynamics as well as any other
information of power system. It can be said that, this method will consider nonlinear features of
power system. It is shown that CDCARLA method can be considered as one of the automatic
design technique for designing of controller parameters.
Keywords: Continuous Ation Rinforcement Larning Atomata,
Reinforcement Learning Automata, Takagi-Sugeno Fuzzy Control.
Continuous
Action
_________________________________________________________________
.= ;, J( @ ( .W. " =$ ';, #CC=$
%9 .1
FE. W!6 #)P D0 \ ', : . $ >]|
';, .W. ( , a,C +," W. ' W=$
N,5 ' ':5 ';E. 0 ';!.
5 S X. . $ ':5 6R IN
N ':5 : BP uN," FE. ':.( W; [7] .N
.W. X. N^ F(L! /.$ 6R
.N z%0 0 g , D0 K ';E.

$ jN ':5 ':5 ';E. F; jN : (
0 ( . $ ':5 ';E. * .W. .
(C$ '( $ 9, : X. !C J( ';< (
' N ':5 i0 N D0 K ';E.
9, .W. . $ ':5 6R ' . , EE
W=$ FE. #)P ( . $ ':5 6R
\6 ' , a
N 0 #% 6R ]
, Sugeno [2] , Wang , Tanaka [1] . a P \ D
a PN ;N" #x$ : IN , (: F%$ 'x P$ ]E
: X. . $ ':5 6R 9, Tanaka
[9] , [8] . (C$ '( 9, . . E W.
(. [6] $ [3] .N ] W=$ FE. D0 #)P
http://www.isice.ir
+( ?S I , +.; J

N 1387

41
= E(, S & . , EE (C$ '( $ 9, /.$ . $ ':5 0 6R
.W. X.

( 5 6R , PID 6R >$$
* ' : F%$ x ; C J5( ' h , J( 9,
, .i 6R 9, ( j, ' , >.

W=$ FE. ' ( , #)P 6R
N
2l
.WEN >. $$ 6R ' n ' 9, n! . a:)
, 6 W. 6 , \ !C J( ] 6R 'L!

( JP$ >]| >. : EE (C$ '( : X.
JP$ : +" C J(; . (C$ '( : X. G-. ,
( ( W=$ FE. 6R 'L! J( .(" W.
',, , BP uN," FE. (1 \
.N +" ';( , #)P ':N

N , LN . P

(6
.3
. $ ':5 ( N,, uN," FE. '
@5 P( W!6 x ;* W=$ FE. . X. ,
T =N ' (,: W&. , T
=N (,: x C5 W&. x C5
:E (: #@ x J( ,6
x [4,4]
x [3,3]
T [80q,80q]
T [2,2]
D0 K ';E. J(5,P : ( BP uN," FE.

. ' Ni +& $ d& +" z0 D0 K WPR
X V ';, . 5 4N 6R ';, ]
J( ';< 9, .W. ] BP uN," FE. '
: 6 ':. b(N , & BP uN," FE. ', !C
FN$ 9, J( + & $ , 9, + "
.F( D0 K ';E. : P., E ' +"
: (: #@ . $ ':5 a-i &S
+<N W=$ FE. +& BP uN," FE. (1) \
Rule i :
1
2
3
4
if ( x is MF ) AND ( x is MF ) AND (T is MF ) AND (T is MF
p
q
r
s 2)
Then
O` f<-# (6J"% .2
a x b x c T d T e
i
i
i
i
i
F +"
x , '& = WEN uN," =N (,: T uN,"
R _%N l ',, , , ',N

WPS
FE. J( : . WO h ( WEN C5
: 'ooo 'ooo .Wooo. i 1,2, " ,16 , p, q, r , s 1,2 ooo

,= W(w& h$ , W!6 'x : ( ' !E g
':5 FE. &S L( J( . 5 4N x _(P$
'; o Wo!6 ox o;* W(w& h$ .W. ':5 &S 16 '
.W. +<N (5) $ (2)
i D uN," !6 W.
x* C5 WPS +" +N (
\S (1) D #@ N,, uN," ( #)P . S &
cos T F mlT 2 sin T

g sin T
mM
T
4 m cos 2 T
l
3 m M
F ml T 2 sin T T cos T
x
mM
.W. +
(1)
m 9.8 m / s , <N v g l5 /,

' . $ ':5 6R ' .W. a M ,
uN," a
5 4N M
( MF21 , MF11 ) WPS x W(w& h$ (2 \
1kg , m 0.1kg l 1m #% J(

.N
1 2 1387

= E(, S & . , EE (C$ '( $ 9, /.$ . $ ':5 0 6R
42
L<* d& l5 6 , 6R 'L! X$

.W. (: ;N" F%$ 'x P$ ]E g
F%$ 'x #x$ ,6 ?S JEN L! J( : X. Jg
. N :N
JJ )9 (!) ( .5
( MF22 , MF12 ) C5 W&. x W(w& h$ (3 \
$$ \ 6R ';< 9, , 6

: FEC$ V * W. EE (C$ '(
N& (g (% - :- a .1.5
ai ) < F%$ 'x #x$ ,6 ( $
a$ FN$ n! +E( ':5 ( ei ... bi
L( #& .F FEC$ 4F%$ x , b F%$ 'x
( MF23 , MF13 ) (,: x W(w& h$ (4 \
+" D. F(L 4N G($ ( #@ x a$

, ( G($ J( +. LN" ':5 &S ( >.
F%$ x , b BP uN," FE. ' J( . x
:: $& W. _(P$ \S
(i
1,2,",16 ) ai 'x :, , -1
1,2,",16 ) bi 'x :a, , -2

( i 1,2,",16 ) ci 'x :a. , -3
(i
1,2,",16 ) d i 'x :a;* , -4

( i 1,2,",16 ) ei 'x :F
, -5
(i
\ 1) ( MF24 , MF14 ) ' (,: W&. x W(w& h$ (5 \
J( . 1F%$ x b \ ':5 &S (2) D $

.N C JP$ F%$ x 80 ( \ 6R '
g - h ( 6 Y` :a-! a .2.5

. JP$ ( , ' : x ';, _(P$ : G
%

2
!1 4_ (! C (Q" .4
+* L P(. 6R a, 6 * : R
Jg . 0 N)R 6R , 6 0 (: ;N" P$
L P(. 6R , 6 vVN F : P$
, : P$ J( .
N 0 R < 6R a, 6
. 6R 6 , (L W&. : ' =!% ;N" R
N
',E : 11 55 $ -55 : x , b #x$ ,6
. +<N F%$ x ';, ' : (1) , . FEC$

(6
'( F(L! : X. /C5 !C J( ';< 9,
, (N,, uN,") W=$ L. N , (C$
$ ':5 ' 6R E C 6 FE. (
: (: 6 , ' 6R 9, . a
N .
'x : ( ' >. : JP$ :- 4
2
EE (C$ '( '$$ : X. F%$
JP$ : F%$ x C J(; J5( :a-! 4

3
. (C$ '( '$$ : X.
Discrete Action Reinforcement Learning Automata

Continuous Action Reinforcement Learning
Automata
1 2 1387

43
= E(, S & . , EE (C$ '( $ 9, /.$ . $ ':5 0 6R
{l1 , l 2 ," , l n } % / 01
J '.$

'2 #$ % 3*4 !"
2
'.$ #,+ % E '( & )*+,- &

!" #$ % / 01 # J

!" #$ %
f 1( k ) o f 1( k 1)
f 2( k ) o f 2( k 1)
E '(
#
f n( k ) o f n( k 1)
EE (C$ '( $ (6 \
F%$ x ' , ' : -1 ,
, 6 !L* h$ W. +<N (6) \ DN
W(C$ L. : X. x
#x$ ,6
:
:(: N a

N \CE #@ _ V ' , ' ; , x$
[-55, -45]
( x$ 6, W(C$ L. ( W=$ 6 !L* h$ a$ -1
[-45, -35]
[-35, -25]
[-25, -15]
[-15, -5]
[-5, 5]
[5, 15 ]
[15, 25 ]
[25, 35 ]
[35, 45 ]
10
[45, 55 ]
11
J( W $ ( .( ;
&
+$ O$ E,
W(C$ L.
N , J (I -2
. vVN :

(6 % "
-
D Q! .1.3.5
W=$ L. ! #% . $ ':5 FE.
E C 6 FE. \<$ BP ,N FE. + P(
.W. ':. MATLAB /= FE. J( ,
6 . F%$ x 80 E, FE. 0.

x 16 , vVN : : FE. 0. \$ '
. vVN 5%$ #@ (x G($ : +. () F%$
)I K .2.3.5
!i g - 6 ,!2-# % :a a .3.5
JJ )9 (!) 6
. _(P$ (,: 'D0 , WPS 'D0 : X. (I h$

LN j
. 4N >. (I h$ +& , , gN"
:P( .W. (,: , WPS 'D0 h
~
J
e
0
6 !L* h$ x ' , : ( '

!L* h$ !, C . WEN
2
x
(t )dt eT2 (t )dt

0
vVN I 5 :N ( , W. ':. \ +: W

(,: , WPS 'D0 >$$ eT (t ) , e x (t ) . W
f i (n) i 1,2,",5
:P( vVN W0( R 6
10sec i
(0)
f i ( n)
1
i
nl
0
n 1,2,", nl
other
'( $ .W. a i , ' _(P$ ' : P$ nl

.W. +<N (6) \ x ,
(3)
n ' EE (C$
Reinforcement Signal
1 2 1387

= E(, S & . , EE (C$ '( $ 9, /.$ . $ ':5 0 6R
0
J ! J mean
J
J
(7)
J min d J d J mean
E ( J ) mean
J mean J min
1
J J min
.W. S ' (I F , JLN >$$ J min , J mean
. (8) D (7) D E #@
D J( h (,: . _(P$ (5) D #@ , W.
.W. X@
e x (t )
x(t ) x
(5)
eT (t ) T (t )
(,: \ . N 5 (I h$ +& (4) D vVN
, 30 !, =N (,: BP uN," FE. WPS , =N

. $ ':5 ( /.$ 1 h WPS
J mean J
J mean J min
E ( J ) min 1, max 0,
44
(8)
. W. +<N (8) , (7) '; >$$ 6R
4 "Q+ K g .4.3.5

WEN 6 !L* h$ 6 vVN : ':
vVN J( P J( ( x$ vVN J( + v0
x$ J( . < ( 'P \6 +" vVN 6 v0
. a
N ( W=$ FE. N) W(C$ L. : X.
. & (9) D : X. 6 !L* h$ x$
fi
( k 1)
( n ) D i f i ( n ) E ( J )Q ( n , i )
(k )
30 !, =N (,: BP uN," =N (,: (7 \
(9)
(10) #@ , W. a i : I$ ]N L. Q(n, i) h$

. _(P$
(10)
Q(n, i ) 2 ( ni ) n, i 1,2,...,11
: D i ( 6 !L* h$ #)6 j
N" '
2
. JP$ (11) D
Di
1 h WPS BP uN," WPS (8\
1
11
(k )
i
(n) E ( J )Q(n, i )
(11)
n 1
4 "Q+ K O %)%L 6 !& jk< .5.3.5

vVN 5%$ #@ '( : 6 !L* h$ x$ : G
a n : vVN P
$ 6 4 J( .
D : ( i
1,2,",5 ) a i , : ( K 1,2,",11)

. .= (" W. (12)
n
Ci ( n)
f (k )
i
(12)
k 1
D ', : , W. 0 d r 1 r 5%$ & : X. G-.

. vVN x , : (13)
Indexi
r d Ci (1)
Ci ( j 1) r d Ci ( j )
(13)
WPS (,: 0 & J(
'x$ (4) D (I h$ ( J( .W. ] W!6 'D0 '
h$ . (I (I5 d& IN WPS ] W!6 'D0 F
(
. (: D #@ >. (I
~
J Mxess
(6)
:N W & (
M , WPS ] W!6 'D0 xess

.W. M
1000 i I 5
)I (-2 6 )9 Q ,!2-# % .3.3.5

+," W. >. 6 !L* h$ #x$ N" '
:F $ (: ( (I ', : W(C$ L.
N X@ W(C$
) EN ' (I -1
N ( W(C$
J] EN ' (I -2
. ( , X@ J l5 6 , J ' (I W(C$ -3
>. h$ : ( . S (I (C LE EN WX@
. (7) D W(C$ L.
1 2 1387

45
= E(, S & . , EE (C$ '( $ 9, /.$ . $ ':5 0 6R
g - 6 6 Y) lk 6.3.5

, o5( (Io5 ox ,o o :o J(; 6 ) N, $
ho$ #x$ N, (13) $ (9) '; .( : (. 6
. +<N x , b 6 !L*
a;* , EE 6 !L* h$ #x$ N, (12 \
, , EE 6 !L* h$ #x$ N, (9 \
F
, EE 6 !L* h$ #x$ N, (13\
z%0 0 6 J(< W. ' : , ; :
. +<N x , ; : (2) , .W.
6R 'L! , 6 a
N : G x , ; : -2 ,
; :
; : G(N
x ,
[35, 45 ]
10
[5, 15 ]
[-5, 5]
[-5, 5]
[-5, 5]
a, , EE 6 !L* h$ #x$ N, (10 \
)9 (!) ( .6

" W. ' : !C ';< 6R 9, a, 6
: J( FE. C J(; , X. , 6
+" $ + ( gN" BP uN," FE. ' .(" W.
a. , EE 6 !L* h$ #x$ N, (11\
. S x , FE. 0. : x 16 W.
1 2 1387

= E(, S & . , EE (C$ '( $ 9, /.$ . $ ':5 0 6R
$ 100 ': b(N J( . +<N 6R a, 6 : G

.W. \@6 g w
0.05 , g h
46
(C$ '( '$$ N . (C$ '( '$$ $
.W. +<N (6) \ W. EE
0.8 ': , $
h WPS : ; 5 + ,= ' 6R : 6 J(

.F _(P$ (: #@ (I h$
W
~
J 10 e x2 (t )dt eT2 (t )dt Mxss
(14)
'$$ $ N (I ', : W(C$ L. +," W.

!L* h$ .W. .= \S (8) D ', : , EE (C$ '(
E . #@ FE. 6R : 6 J( 6
.N _(P$ W0( a5
( a 3 ) F:( ! >(| . 6 !L* #x$ N, (14 \
U i Li
0
f i (0) ( x)
x [ Li ,U i ]
i 1,2," , m
other
vVN : ') 6 , J( 6 >$$
(15)
U i , Li +"
h$ x$ . ? P$ +" a i FE. x W. 'x ,
. & #,X !, EE W!6 < . 6 !L*
. a
N (16) D #@ , W(C$ L. : X. x$ J(
f i ( k 1) ( x ) D i f i ( k ) ( x) E ( J ) H ( x, xi )
(17) D #@ W. . h$ (
(16)
H h$ l5 D

. _(P$
FIN >(| . 6 !L* #x$ N, (15\
gh
( x xi ) 2
(17)
exp
2
U i Li
2 g w (U i Li )
>$$ g h , g w a i vVN x xi D J(
H ( x, xi )
. . h$ jX$ , {&

6 !L* h$ : ( N (: W6E ( ' Jg
. 4 , .= (: #@ D i 1
Di
1
Ui
(k )
i
( x) E ( J ) H ( x, xi )dx
(18)
Li
, . #@ 6R a, 6 IN x 5%$ vVN

. a
N .= (19) D : P
$ 6
x
a, , WE >(| . 6 !L* #x$ N, (16\
Ci ( x )
f ( x)dx
i
(19)
Li
W. 'C . $ ':5 >(| C J(;
(I5 >(| C J(; 6 !L* l5 (5 $
\ .W. z%0 0 6 !L* C J(<
>(| . 6 !L* #x$ N, (17) $ (14) ' \ .(
.W. +<N (I h$ #x$ (18)
1 2 1387

47
= E(, S & . , EE (C$ '( $ 9, /.$ . $ ':5 0 6R
',N #x$ (22 \
6R ':5 (" l5 '; : DN

. ,= IN '^N ,i& , W. 0 E &
h WPS !6 ' (,: , WPS 'N IN (23) \
.W. +<N v, P L.
(I h$ #x$ D0 (N (18\
(.
(6 X)< 2 .7
9, 6R . $ ':5 &
N
30 !, =N (,: BP uN," FE. !C J( ';<
WPS , X@ ' (,: W&. , C5 W&. X@ !, WPS
.W. +<N (19) \ 2 h
BP uN," 6R ; & (19\
" (22) $ (20) ' \ ',N , (,: WPS 'N
uN," (,: , C5 WPS #x$ (23 \
.W.
( < .8
' . $ ':5 6R W; , !C J(
#@ ] 6R 9, . ] BP uN," D0 K FE.
'L! .N W=$ FE. ( #)P ':N , 0
'( '$$ , 6 6 , \ n 6R
W. . $ >(| >. : , EE (C$
C , . (C$ '( '$$ a, 6 , ,"
6R \6 $ ]
N" : . JP$ >(| ;
h WPS +" 'S , C5 WPS #x$ (20 \
';E. L( ' +" +$ 6 + & #@

. IN D0K
KL
& WPS 'S , uN," ' (,: =N #x$ (21 \
[1] K. Tanaka, H. Wang, Fuzzy Control Systems

Design and Analysis, John Wiley & Sons Inc., 2001.
1 2 1387

= E(, S & . , EE (C$ '( $ 9, /.$ . $ ':5 0 6R
48
[2] K. Tanaka, M. Sugeno, Stability analysis and

Design of Fuzzy Control Systems, Fuzzy sets and
systems, vol. 45, pp. 135-156, 1992.
[3] G. Ren, Z. Hung-Xiu, Analytical Design of TakagiSugeno Fuzzy Control Systems, American Control
Conference, pp. 1733-1738, 2005.
[4] M. Akar, . zgner, Decentralized Parallel
Distributed Compensator Design for Takagi-Sugeno
Fuzzy Systems, Proceeding of 38th Conference on
Decision & Control, pp. 4834-4839, 1999.
[5] A. Jadbabaie, M. Jamshidi, A. Titlie, GuaranteedCost Design of Continuous-Time Takagi-Sugeno Fuzzy
Controllers via Linear Matrix Inequalities, IEEE, pp.
268-273, 1998.
[6] K. Tanaka, T. Hori, H. Wang, New Robust and
Optimal Designs for Takagi-Sugeno Fuzzy Control
Systems, International Conference on Control
Application, pp. 415-420, 1999.
[7] M.Y. Shieh, C.W. Huang, T.H. Li, A GA-based
Sugeno-Type Logic Controller for Cart-Pole System,
IEEE, pp. 1028-1032.
[8] M.N. Howell, M.C. Best, On-line PID tuning for
engine idle-speed control using continuous action
reinforcement learning automata, Control Engineering
Practice 8 147-154, Elsevier Science Ltd., 2000.
[9] M.N. Howell, T.J. Gordon, Continuous action
reinforcement learning automata and their application to
adaptive digital lter design, Engineering Applications
of Artificial Intelligence 14 54956, Elsevier Science
Ltd., 2001.
1 2 1387

49-55 =X@ 1 2 1387

]' mkC L 2J )L

3
0N J 2 'X@ N 1' ,E0 =
(INRIA) EN5 , $XN #CC=$ I 3 (" \( - JENE(, L<N 2 I($ ;. P@ L<N 1
1
khosrowjerdi@sut.ac.ir 2safarisn@uwplatt.edu 3ramine.nikoukhah@inria.fr
_____________________________________________________________________________
(LTI) +: (nN x$ D0 ' FE. >& UV<$ W; E. ; ( !E !C J( :%+
!E : 5X |( _@$
N< _(P$ (nN" ';& ( '(,
( .L S .
X. . J( . 0 L2 'w5 ,= '^N ';<K w6 >& UV<$ W; E. ; (
9, '& ( $0 .," W. l5 ':. ; !E ;(: v + 5( 0. (L! :
.< +<N !C J( (
$( ' !P E. ( >& UV<$ :' '^,
Abstract: The problem of optimal sensor location for fault detection in Linear Time-Invariant
(LTI) systems is considered. A novel characterization of undetectable faults is shown to give
rise to a new optimal sensor location problem for fault detection in the presence of
L2 disturbances. An algorithm is presented whereby suboptimal solutions to this problem can be
determined. A numerical example is presented to illustrate the theory andexecution of the
algorithm.
Keywords: Fault Detection, Sensor Location, Algebraic Riccati Equation (ARE)
____________________________________________________________________________
, $ . >& hS UV<$ W; ;N" #&iR >%N
%9 .1
: ,` 0 ';!. =! J( . 0 $O

E. ; ( !E >& UV<$ ';, (
P.$ , /E ' FE. + $ g 0 ';!.
+.; , +6R $ IN FE. & ?S #4N
+L<,` $ +N |( ';! >& UV<$ ';,
JV$ ( : !E J( [5] $ [3] +P .W. VLN
', 1\ ?D +L<,` & .[2] , [1]W. 5 S
' FE. [6] .W. 5 S . FE. ';!6 >.
I N ! a. ' 5 6R W; " ';, (
( : [7] .W. vVN l5 !E R E +P D0K
N W. `(, 0. L(, ( B. 5 J( .W.
0. #4N & W; E. ( !E : '" (
,0 , u ',, #&iR : X. $ 6R '
( !E [8] .W. X. x * FE. ( `(, (C
$ (N !$ Q N L. a. !L. w 9<K w6 , y
FE. CN 'X@ +N ( !E . ( ( E.
'I f ';& jS, hS N L. 9: +
'(, B. [9] a

N. , .W. s$ x * '
?(R J( : $ N ' F%$ E. , ;= 4N ("5 _ V
FE. ( >. '; = JP$ ' , W.E6 (4N
', ;N" vV , v DN #O$ W; . #;$
( >& UV<$ W; E. +S W; E C 6
UV<$ ';, r< , >$$ J( . 5 ("5 &
.W.
! 5 0 _ V 'E. W. $&iR 9: >&

. '; = E. J( gN* W. ;( . S N
http://www.isice.ir
+( ?S I , +.; J

N 1387

0N J 'X@ N ' ,E0 = >& UV<$ W; E. ; (
e)2 (6< - !< .2

B , A . N. i !C J( 5 'N
E +" LN >$ A ! B , A t B LN" +C ';E($
50
|( _@$ + $ 5 a

N ' i0 !C J(
>& UV<$ W; E. ; ( !E : 'X%= , X
(;<K O$ W=$ (LTI) +: (nNx$ D0 ' FE. '
;X ( _@$ J( L.0 . ( E ,= '^N
';E($ . W ';E($ , W JP N ';E($ A-B
N _(P$ ( 'E W. (nN" ';& a.
R T A
; .W. +:, '(n" N." a ; P ( +"
N
$( ' !P :.(

E. J ';PS a$ ', P J( ') + ( ':.
A P PA PRP Q 0
IN A RP ' N CC6 +C G($ P W.
':. ; !E J( ;(: v J5( W; 5(0. (L!

5 (, FP$ + ( F0 N* J( ,i& .
( N 6 : + g : v W; . I(,
'I , ;= ; ( !E 4N 'L( \(E ,I J(
#% +" W!6 'w5 ?C=$ ';E($ >E6 G \($ h$ G($
.N \6 , _(P$ IN >& UV<$ W; ("5 L(
QT , R
v P G($ F( N Q

T
\($ h$ ';E($ 'w5 . (N G : ( A, B, C , D)
.N @i0 (: >$ +$ ( F0 !C J( gN"
aN .
N . +<N RH f ( , . ( CC6
!C ( >!D |( ':N , N ',
: r N r ! 0 !. ': .W. N.
+" N , (nN" ';& &

+" N .W F0
r jP $
: +<N (: #% H 2 'w5
:r
w 2 d r}
{w :
UV< L2 N >0, f : ', P (n LN h$ 'w5
G(N . , (C$ '
a.e '%0 N .N
_(P$ (: #% G \($ h$ G($ ( ' H
G
inf V (G( jZ ))
Z >0,f
X. .N _(P$ 9<K w6 ;& '(n"

( !E X _@$ W; a:) |( E ( _(P$ :
: X. 0 +<N , " E. ;
+" ;(: v +$ $( ' !P (L!
':. b(N ; '& ( $0 .W5(
J( : ' @i0 W. m +( . +<N ( (4N
.W. ( [10] !C
" [11] (: F! .W. \S6 (. C LN V +"

E $( ' !P : X. H G(N .= + W.
.W5 0 !C , :.
G : ( A, B, C , D) RH f , D ! 0 !. .1 N"
G($ , V (D ) ! D /C5 , G
! D LN" .N
EC , Q +C
Q ( A BR
QBR
1
1
D C ) ( A BR 1 D T C ) T Q
T
B T Q C T ( I DR
= ', ' `(, C A BR
1
1
D T )C
D C BR
1 T
B Q G($ ,
.. R : D 2 I D T D 0 +" N jZ
)n<2# (' .3
D0 ' FE. ' (nN" ';& &
V J(
E
L2 'w5 ? P ';<K O$ W=$ +: (nNx$

(L 4N (: D0 FE. .F _(P$
>& UV<$ 6, I;

E C 6 FE. (1 \
1 2 1387

51
0N J 'X@ N ' ,E0 = >& UV<$ W; E. ; (
(: #% +" ,0 _@$ >& +, FE. WP|,

W.
G1u G 2 w1
y1
(2) D L2 'w5 f z 0 ': 6 . w1 :1 +"

_@$ >& FE. WP|,
G1u G 2 w 2 G 3 f
y2
'w5 (nN" ';& +$ ( . w2
:1 +"
.N _(P$ (: #% L2
(2) FE. ' L2 'w5 (nN" >& f >& .1 T)`

D . y1
y 2 F
w1 , w2 :1 ': (
(: &
? P W. L2 'w5 (nN" >& f P
Ax B1u B 2 w B3 f
y C T x D1u D 2 w D3 f
(1)
0 ',, u R m FE. W!6 x R n +"

l
!6 ';& f R q 0N 9<K ',, w R

' :N ,0 y R p , ("5 'I ( , E. r=
G($ +<N CT N . >& UV<$ W;
LE W. +<N T E. WPS ' :N
( ? P T {5
( W. m \S .
C i , Bi A WO ';E($ . T 4 R s <5 &

.N >. P

D { f L2 : G2 w G3 f , a.e. w : 2 }
';|5 (1) FE. W. J( aI E ,` J( ( b(N
J(I :
( ; . PR D D &
_(P$
:N(: < :. ," ',=
W; W. a N (nN"N$ D &

f >&
+:, '(n" N." a. 'P F; . J( (LV.
.W. . T 4 R s a$ ': CT G($ (I)
.F _(P$ D &

': ( , W. (n jZ = ', ( A, B2 ) k,: (II)
. \ 'D. $ jZI A B 2 G($ Z R

.W. (n" (CT , A) k,: T 4 ': (III)
'w5 q u q \($ h$ G($ W (N {5 .2 T)`

LN" RH f
J T *
sup Wf
f D
. \ 'D. $ D 2 G($ (IV)
.W. D &
+:, '(n" N."
+ J( 2 _(P$ W N:, G($ , W. m \S
( L2 'w5 (nN" >& J(I .= $ :. E
') + .= W; , (: F! .(n a
N V! EN5 :
. ( J T *
, N jZ = ', 'X@ G 2 {5 .2 N"
-F h$ ( N +" G 2 M 1 N #% ' (I
$
RH f W N:, G($ Jg .E 0 h$ ( M , 0
J T d J T LN" . MG3W 1
*
JT
z 0 vVN ' N
+"
2
MG3W 1
(3)
n! W. 0

f
MM 1 N
dim( y ) t q
dim( f ) (V)
' ; v ( , Jw$ W; (I) {5 W. $ \S

s d6 (IV) $ (II) ';|5 .W. ',| E. WPS
:N Jw$ (V) {5 .E N. [12] J! 5
.N , >& UV<$ W; (5 : 5 ';(
5 4N E. ; ( !E !C J( * .1 <
, ;= ; ( 4N 'L( ':. ; \(E +$ W.
4 J( ' N _(P$ >& UV<$ W; IN FE. 'I
T E, (1) FE. D1 , B1 , A ';E($ W. 5

.N & $x$ !C J( ( 9, , W5 4N
+ (: #% GN5 :6 (1) FE. ,0 +$ 6
N
y
G1u G 2 w G3 f
(2)
F( i 1,2,3 ': +"
-F h$ ( N (

N" : [12] 13-35 wS ?D .$o
MG 2
.p
Gi
CT ( sI A) 1 Bi Di
(2) D L2 'w5 (nN" ';& _(P$ W; (

(2) D f 0 ': .F( 4N _ V WP|, , W=$
1 2 1387

0N J 'X@ N ' ,E0 = >& UV<$ W; E. ; (
2J )L "J .4
F( w : 2 ': LN" f D {5
MG2 w
W; L(, N ! ( ( 3 _(P$ ?D
F( , MG 2 w
J T (;PS E. + S , >& UV<$

: D &
'(, J* WCC6 .( ; \S6
. vE= E. ; ( !E : '( _(P$
MG3W 1
' N T 4 .(L 4N (1) FE. .1 "J
L( #P ( , \S6 J T

2
MG3W 1
T 4
MG3 f
MG 3 f
2
MG3W 1
(: ':. ; !E P 1 !E ( 6 (
JT *
. W.
sup Wf 2 d
f D
' N T 4 .(L 4N (1) FE. .2 "J

sup GT
(5)
T 4
. GT
MG3W
1
MG3W 1
#% !6 'w5 ?C=$ MG3 +<N +$ , W. I(,
(:
V 0.5 (CT x y )
D3 )
F( W RH f >. N:, G($ ( ': n!

0 GT d GT f ,
T 4
!E 2 \ ?D . ! P$ 2 !E ,2 \ $
GT CN G($ ; J( ':.6 !I E. ; (

<KK X! Q 0
GT Wf \ J( .W. 4 &

',
. vE= Q N L.
.[12] (: #% !6

( A LCT ) x Ly
CT , V
'w5 ?C=$ J T M 0 h$ +<N +$ .2 <
MG 3 : ( A LCT , B3 LD3 , V
JT
+"
0.5
J(
A LCT (: MG3 RH f 2 !E W. m +(
0.5
W F0 n!
Wf 2 d MG3W 1Wf d 2
2

N
Wf 2 d
(4)

MG 2 w d MG 2 w d MG 2 f w d 2
2
2
2
J( ,i&
W; 'S ILN j| J( . 9E a,C (n" ';&
inf
52
( B 2 D 2 T PCT T )V 1 +"
,iP V
D2 D2T , L
v P
P T t 0 , V 2 ! 0 (IV) $ (II) ';|5 ?R
$( ' !P X%= :.(

( A B2 D2 V
T
PC T V
1
1
CT )P P ( A B2 D2 V
T
CT P B2 ( I l D2 V
T
1
D2 )B2
1
CT )T
( 0. W!6 'w5 ?C=$ J( W. J( >! N .
W; N L. Q +" L(, N !

.[3] W. >& jS, UV<$
>& '(n" : CS _(P$ $ W. " I* (
( W. >& a,C '(n" a. L2 ';<K w6
.[10] F(N
&
? P f W. a,C (n" f >& .3 T)`
N L. <KK X! !$ W; N ! FE. (2 \
v .= W; 5(0. (L! 1 F! : X. 6

.F ( 2 !E ;(:
.N ( A, B 2 , B3 , D 2 , D3 ) ' : I N)2Q"
:(N >$ (: \6
. D { f L 2 :
Wf
! JT }
_(P$ P l5 _(P$ W. +<N [10] (DN

( >& UV<$ #)C ;& '(n" ' W. 'L(
.W.
!E : 5X |( _@$ l5 _(P$ : X. (" V
.F(N ( E. ; (
1 2 1387

53
0N J 'X@ N ' ,E0 = >& UV<$ W; E. ; (
k,: (N vVN '=N T 4 : 1 4
1
0
0
0 0.86939 43.232 x
0 0.99335 1.3411
. (n" (CT , A)

h$ G($ (N vVN 'R W N:, G($ : 2 4
1 0 0 0 0
0 0
0 1 0 0 0 w 1 1 f
0 0 1 0 0
0.5 0
0 c12 (T ) c13 (T ) 0 0 0 1 0
x
w
1
0
1 0 0 0 0 1
+"
c12 (T ) 0.137 0.86939T ,

c13 (T )
$ IN D T G($ , GT : ( AT , BT , C T , DT )

. \ N.
T 0 4 ': D ! 0 !. ', $ : X. : 3 4

D C J( . GT 0
! D ' N D C 6
. D *
(N vVN '= T 1 4 &
: 4 4
. $ 3 , 2 \6
, T1 z T 0
27.61 43.223T
#5 >E6 b. v WPS T 4
MG3W 1 \($
#% !6 'w5 ?C=$ RH f 'w5 GT
vVN D * C J(< T v D C : 5 4
[0.15758 , 0.63878 ] ,
, ) =N ' FE. !6 ';& f

( .W. (ft)
UV<$ W; T WPS JP$ @ . +<N -;
.W. vVN (: #% W G($ .W. f >& jS, hS
0.1
I
s 0.1
\S V I F(L! Scilab /= N ( : X.
W ( s)
.(N
4N N. '& 'I5 aN /.$ 6 l5 F(L!
' NN 'P V .W. ':. \S Matlab ( [13] Scilab
. +<N & ( ', l5 F(L! :
(!%' p ) .5
o( !E ,` J( " W. b(N + +<N W;
.W. F. 4 \ T WPS >E6 D * , ':.
AFT/F-16 o 'o ( >& UV<$ W; ;

. v
T
0.2 WPS (C$ J T +<N \ J( (DN
+S ' . vVN N T 0.2 n! , W. 6
F( 4N 3 \ ?D W. _@$ [8]
UV<$ W; Q 2 N L. 9 $ 6 '; . ;

. v
WP|, ; J( .N +<N T _ V (C ': >&
1o N;N >& ( jS, ': Q 2 N L.
J( (DN . +<N -; FE. 5 \ ?D
f2
WEN Q 2 N L. W.E6 < ( ':. b(N

.( (I5 T o T
0 . 2 S,
f 2 >&
( < .6
FE. >& UV<$ W; E. ; ( !E !C J(
';& .W5 S . (LTI) +: (nNx$ '
_(P$ +:, '(n" N." a. 'P ( (nN"
.( ( a,C (n" ';& ' X(P$ +" : X. , N
';<K w6 E. ; ( W; '( !E
AFT/F-16 ( >& UV<$ W;

. v WPS (3 \
N ':, /( ( AFT/F-16 ' D0

W. (: #% M=0.6 & , h=3000 ft jX$
.( ( +" v .= W; 5(0. (L! ; L2
1 2 1387

0N J 'X@ N ' ,E0 = >& UV<$ W; E. ; (
54
[8] D. Kim and Y. Kim, Optimal Sensor Location for
\6 , _(P$ W; +$ !C J( ( (, J( ,i&
Robust Fault Detection Filter, IMechE, Vol. 213 pp.

13-21, 1999.
FE. L( 'I ( , ;= ; ( !E 4N 'L( F; \(E
[9] Y. Xu and J. Jiang, Optimal Sensor Location in
( )@ W. F; . J( . |6 ,` PR
Closed-Loop Control Systems for Fault Detection and

Isolation," in Proceeding of the American Control
Conference, Chicago, Illinois, pp. 1195-1199, June
2000.
N." P ', 'O$ * E C 6 FE.
[10] M. J. Khosrowjerdi, N. Safari-Shad and R.
W 0 E C 6 FE. & ', 'O$ *
!N N$ 'L( F; j| . >& UV<$ W;
E. ( L( #P ( , J T * +:, '(n"

j| J( \S6
Nikoukhah, Optimal Sensor Location for Robust Fault

Detection," in Proceeding of European Control
Conference (ECC), Kos, Greece, July, 2007.
[11] L. Jian, L. W. Jian, and Y. Guang-Hong, An LMI
Approach to Minimum Sensitivity Analysis with
Application to Fault Detection, Automatica, vol. 41,
pp. 1995-2004, 2005.
J T * >& UV<$ ; W; ' N

(2IQ.
/.$ 30/940 < 'S ?R E(N J!, ,` J(
, ! W(6 I($ ;. P@ L<N 'I( N , <,` WN,P
.W. $I;
$
[12] K. Zhou, J. C. Doyle and K. Glover, Robust and
KL
Optimal control, Prentice-Hall: Englewood Cliffs, NJ,

1996.
[1] Y. Chen and R. J. Patton, Robust Model-Based
[13] S. L. Compbell and R. Nikoukhah, Modelling and
Fault Diagnosis for Dynamic Systems, Kluwer

Academic Publishers, London, Prentice Hall, 1999.
Simulation in Scilab/Scicos, Springer, 2005.
[2] R. J. Patton, P. Frank, and R. Clark, Issues of Fault

Diagnosis for Dynamic Systems, Springer Verlag,
London, 2000.
[3]
K.
Raman,
Optimization
of
measurement
Schedules and Sensor Designs for Linear Dynamic

Systems, IEEE Transaction on Automatic Control,
Vol. AC-21, No. 1, pp. 55-64, Feb. 1976.
[4] M. Morari and ODowd, Optimal Sensor Position
in the Presence of Nonstationary Noise, Automatica,
Vol. 16, No.5, pp. 463-480, 1980.
[5] A. Arbel, Sensor placement in Optimal Filtering
>E6
F. (4 \
and Smoothing Problems, IEEE Transaction on

Automatic Control, Vol. AC-27, No. 1, pp. 94-98, Feb.
1982.
[6] K. Watanabe, M. Sasaki, and D. M. Himmelblau,
Determination of Optimal Measuring Site for Fault
detection of Nonlinear Systems, Int. Journal System
Science, Vol. 16, pp. 1345-1367, Nov. 1985.
[7] M. Basseville, A. Benveniste, G. V. Moustakides
f 2 r= >&
(5 \
and A. Rougee, Optimal Sensor Location for

Detecting Changes in Dynamical Behavior, IEEE
Transaction on Automatic Control, Vol. AC-32, No.
12, pp. 1067-1075, Dec. 1987.
1 2 1387

55
0N J 'X@ N ' ,E0 = >& UV<$ W; E. ; (
0 . 4 ': Q 2 N L. (8 \
0.2
': Q 2 N L. (9 \
0 . 6 ': Q 2 N L. (6 \
0.5
': Q 2 N L. (7 \
1 2 1387

56-64 =X@ 1 2 1387

Y+ -" -I<" q,J # NJ 9M
(D& = 1<5 WCC6 :(5
+;X@ L<N N,! , (. 1+;X@ L<N #S -l . \%=!5
mataei1971@yahoo.com 2fr_haghighat@yahoo.com 1
____________________________________________________________________________
: $ 4 J( . ] J* -! -IN! Z+E " FE. ' CD$ L! ( !C J( :%+
W!6 5 vVN . J( . S X. CD$ D0K 5 , CD$ , WO '; D0 ';5
DCN ( J* -! -IN! Z+E JPN FE. N
'( , #@ _N! FCE 9, : X. D0K
FE. D0 5 ';!L. +,I5 . Jw$ F! ( >!S +" P$ sCN & * : 5%$
#O . \@6 n * : V! DCN (;N FE. N
'( WEVN 6
';< L! >. & + +<N ' +( . #@ _N! FCEK 9, : X. IN > D J(
D( 6 _ V /( FE. L! J( & ':. b(N J* -! -IN! Z+E " FE. ':.(
. ] x$ a$ D v D P$ DCN , FE. jN
_N! '( '$ J* -! -IN! Z+E " FE. CD$ :' '^,
Abstract: In this paper an adaptive controller for the Unified chaotic system is presented. For
this purpose, a combination of linear feedbacks with constant and adaptive gains and a nonlinear adaptive state feedback is used. In this way, the Lyapunovs direct method is used to
select the suitable non-linear state feedback. Moreover, the asymptotic stability guarantee of the
uncertain Unified chaotic system in a stochastic point on the general manifold of its equilibrium
points is proved in a lemma by using non-linear feedback controller. Then, the asymptotic
stability of the final controlled system in any desired point on the mentioned manifold is
guaranteed by adding the linear feedback signals to the prior controlled system. This subject is
proved, by using the Lyapunovs indirect method. Finally, the simulation results related to
applying the controller to the Unified chaotic system, in different cases, even when
simultaneous-changing of the system type and the desired equilibrium point occurs are provided
to show the well-acceptable performance of the proposed controller.
Keywords: Adaptive Control, Unified Chaotic System, Lyapunovs Stability Theory.
_____________________________________________________________________________
, [5] (ô! ';oE. [4] oI o.; [3] #oS N,!
UVo< qi v" a,I! J( ., [6] ( ';<,
, 2o 1# /o.$ OGY 9, ] : G B. J( , (
o , vo" o [7] o" ';oE. o : 3r(
';oE. : ( .W. 5 S JCC= $ E D 0
NLo" o5 'o;`(, (N +" >. W S >. "
1
Ott
2
Grebogi
3
Yorke
%9 .1
* a & _ V ' 0 +" jS, . , v" (

o : o0 :,o .W. 5 S P!D E D 0
, [1] (ooo 'oooI!$ ooo & ooo" ooo ooo(=$ oooN
.o hoS, X N$ ( J( [2] J #V " +.),
5 \ =$ W; 'I +P N$ v" P!D > K !,
N a & : X V '0 ;E. : 'E D0K
http://www.isice.ir
+( ?S I , +.; J

N 1387

57
(D& = <5 WCC6 :(5 /J* -! -IN! Z+E " FE. CD$
.oN IN! FE. WEN #,X '^!$ < " '(
o" FoE. o5 oS +,o5 ?C=$ , P!D 0 . *
V >(| 'S J5 4N 2002 . N. , ! q(;N
ox$ o oo +o<N 0 #P!D .[14 -8] W. J* -! -IN! Z+E
' '6, a5 N ?5 n! l5 FE. . W!6 #)P D0
: o@0 'o;`(, :,o >o n FE. ' ' \
W=$ >$ " FE. .[20] ;< FE. . J( #)P
o J0o J(o .[10 -8] oo FoE. J( NL" '5
. 5P (: D , J* -! -IN! Z+E FE. +&
+" 'o #ox$ o( PoN o , o FoE. Jo( o d=
x
y
z
.W. n , X j|
(25D 10)( y x)
(28 35D ) x xz (29D 1) y
D 8
xy
z
3
(1)
+" ,i& , E PDS K ( , 0N FE. ' : 0
oS, .o N IN! & FE. D [0, 0.8) S, FE.
0 .8
. N0 J* & FE. D (0.8, 1]

:(: /, ?D d , c b a ( ' _(P$
a
b
o( kVo. J(o .o ox$ N;N D ( , LE" N$

0oN 'o , " ';E. ' CD$ L!
JoCC= : o0 o $ +$ , W I]6 E PDS K (
.[14 -11] W. 5 S
'o oCD$ oL! o( o] 4N !C J( gN"
25D 10
oV! oDCN J* -! -IN! Z+E " FE. N

':.(
28 35D
(2)
29D 1
D 8
3
: o* Jo( .W. FE. W!6 'x 'w5 z0 * ( :

4N FE. .FN FE. P$ sCN & * P J(
o oD0 o5 ';!Lo. , oCD$ oD0K 5 >$ : X.
' ';< L! (D .(n #@ CD$ , WO ';
#o% J* -! -IN! Z+E " FE. W!6 #)P +$

.N E(N: (3) #)P
x
y
z
kVoo. 'oo oo.. {oo5 oo( FooE. oo 'oo ?ooS ooC
o( ao$ oPS, /(o o .[16 , 15 8] oo L! I" WC5
Jo( .Wo. " 5 ' D [0, 1] ': l5 " FE.
oooS, , oooooo ooo! ooo& FoooE. ooo \(ooo$ ooo D
+o ao P vo" o o: o < #)C : 'E
FoE. 'o ox$ , PoN WEN (n ?D$ , a,C &
.W.
a( y x)
bx xz cy
xy dz
(3)
#)oP : Xo. o o;N" o( 5 (;E. : 'E

/, b _@$ J* -! -IN! Z+E FE.
o5 o +oP .Wo. oD o \oS o ( +P (3) , (2)
_@$ IN! #)P ( W5 ( jS, aL; #). "
o! 'o;(,, voVN o=N , oP!D Wo=$ FE.

oo6R : oo aoo. ooV .oooo oo5P a, ooV FooE.
ooL! oo6R \oo6 aoo;* ooV .oooo oo ooL!
o o FE. '( Jw$ s d6 , D ';<
o b(oN F
o oV . ] ;N" #O F! , >!S
] P!D W=$ FE. ', _ V (:" * a
N s ':.
. #@ F< V '
N , ' h ,
o& ) m b J( .[16 , 15]
o o o& #o6 o >. IN! FE. . a5 4 (
()!-2- jk< - `"M D NJ % .2
: o( ';< L! 6R ,( : . 0 . N=$ N
"
o$ (;, > K ?D . ,I5 b ! ';!L.

+NoS Wo. o ;o< Jo* -! -IN! Z+E FE. ' +
.[17] oN _< .i " ( J!, 1IN! 1963 .
Lo. o!C Jo( .o ho FoE. ';!6 ?< N$
FoE. , [18] o(, , J* /.$ 1999 . 2J* FE. +" : G ;!.
ox ?o< o Wo. WO ; D0 5 L. ( a,
' , oo oo oo( oo5P [19] Joo* , oo! /oo.$ 2002 oo. 3oo!
. h FE. x W!6

1
Lorenz
Chen
3
L
2
Rayleigh
1 2 1387

(D& = <5 WCC6 :(5 J* -! -IN! Z+E " FE. CD$
oN , 5o%$ oDCN o( \o@6 FoE. o FE. a, W!6
.o No
o( )o o om o& a5 ' V!
o o FE. >. D0 5 ';!L. +,I5 a, 6
o Jo* -o! -IoN! Z+oE FE. : (;N WEVN 6
oL! 6R #]I . ?C= a. V
. ] +" #O F! , >!S , a ,
58
Q"
4_ 6 r% .3
o6R : o WoES Jo( oP!D Wo=$ FoE. o5P : Go
FooE. ooP$ soCN JooP$ Woo; o. Joo( .ooo +oo oL!
o o oS X@ FE. W!6 'x N: ?< WE5
o(: ';(,oE$ o (3) oD : a. , , #)P $ #%(
.F.
+" W!6 #)P J* -! -IN! Z+E FE. :1 N"
';!L. vVN FE. J( .F( 4N W. (6) D
#% u2 , u1
u1
u2
y , z
x2
d
(4)
-o! -IoN! Z+oE FoE. P$ sCN & * (4) D ',E$ ,
(7)
zg
D ?D g +"
g
x xy
2
(oC +o UVo< #o@ .oo _@$ 3 'w5 J*

o. ?S + (3) D a, !P : X. FE. ' ?S
#o% o* J( hS, J* -! -IN! Z+E FE. (N P$ DCN
. UV< (:
(8)
oP$ soCN o& o* o hoS, 5o%$ oDCN ( :,
0, 0, 0
Lo. , o (7) oD o o o +NoS :$o
E1
E2
E
3
o FoE. a, Wo!6 o!P D0K V n6 X, u 2 !
Wo. +" Jo* -! -IN! Z+E FE. CD$ : !C J(
(b c) N C g ! , . ;&
(4) oD o * : V! DCN FE.
. ( N

D (4) D FE.
g1
o FoE. a, , , !P$ ';!6 ':. Z+E + 0
o
( FoE. 'o + JPN \! J!, . F5
o FoE. 'ox : Xo. o + .W. a PN g1 ?S C
>o$ g~ , w o 'ox x, y, z, g '; P(
d (b c) , d (b c) , (b c)
d (b c) , d (b c) , (b c)
X V! C ': x
yx
(9)
g~
g1 g
(10)
.F( 4N (: W!6 #)P FE. G-.
w
~
g
f ( x, y , z , g )
g
h ( x, y , z , g )
(11)
o(: _No! ho$ 'o(N (11) oD o o _o@$ FE. ' ,
.F(N vVN
V ( w, g~ )
>
1 2 ~2
w g
2
(12)
X, z
X 2
DCN N
d
. ( N

D
9M Q"
4_ .4
.F _(P$ (10) , (9) /,

w
X, y
(5)
'o;(,, voVN zo%0 a, oV $=|$ $

o FE. W!6 #)P J* -! -IN! Z+E FE. !
.N E(N: (: #% +$ n
x
y
z
a( y x) u1
(b u2 ) x xz cy
(6)
xy dz
oDCN o :q), o Fo 6R ' NL ! ';!L.
X 2
oP$ oDCN o( x X , y X , z
o& a5 V!
so :qoNO , o voE= (6) oD o o E C 6 FE. '
o0 o(: #% (12) $ (6) /, J5 4N V h$ N: ?<
,"o no oP$ oDCN FE. N

'( Jw$
.
oD0K o5 L. ( vVN , 6 4 J( .
V ( w, g~) (c a) w2
(13)
o!P oD0K oV no6 J| F5 D( CD$
1 2 1387

59
(D& = <5 WCC6 :(5 /J* -! -IN! Z+E " FE. CD$
oooo D [0, 1] ': oooo (2) ooooD oooo oooo $
, k1 'o o; o oN {5 +$ #O : 6 J( :$o
.W. (c - a) (4D
.oE oO (oC (15) oD oD0 o5 'o;$ s k2

9, : Xo. o no 'o ; ' S \S ' ,= G-.
oooo oooo
V ( w, g~) ho$ w
o6 FE. ( 4 J( .F JP$ _N! FCEK
0 oDCN L(oE o W. }|, J(

w o Wo!6 x n! .W. JP X V (w, g~) h$ , JP W
ooooooooDCN ooooooooP( 4 'oooooooow5 oooooooo0 ooooooooP$ ooooooooDCN
oL Xo@ oC o No

oD (11) oD _@$ FE.
X2
, g (b c)
. D0 E x X , y X , z
oDCN 6 D0 FE. %V< !P , J^ G($ #@ J(
o oo _o@$ (6) oD o o FE. G .
. 0 (18) , (17) /, ?D >$ E P$
a
a k1
c
c k2 X
X
X
X
X
0
0
d
0
y , x ';!6 (8) , (7) /, ';!L. : X.
.N L Y N 5%$ C ( N

D , FE.
o4=! : (8) , (7) /o, o +NS N" {5 +
t0 : I 5 6 t1 gN* , & (6) FE. t = t0 V!
0
X
0
(17)
o( o FoE. y , x 'o;!6 No

(L $ vVN
+o$o (6) D : a. ENX( !P \6 Y N 5%$ C
#o%( oN J(IL(o Y 5o%$ C n x , '
.. F0 (: D FE. z W!6 x '
(O d ){O3 (a k1 c k 2 X )O2
(18)
[(a k1 )(c k 2 X ) ac X 2 ]O k1 X 2 } 0
z (t ) e dt z (0)
oP$ DCN FE. '( s I(, -, P ,
Jow$ (19) oD o oS +o ,"o #o@ 4N
. 0
k ! 0
1
k 2 X ! c a k 1 Q 1
(k1X 2 )
ac X 2
(a k1 c k 2 X)
c
k 2 X !
a k1
Y2
(1 e dt )
d
(14)
Y2
z (t ) t o f N: (14) D $
d
o5 J* -! -IN! Z+E FE. +<N j| J(
.oo oL
o +NS : X. + (6) D +" (
(19)
Q2
o hoS, x
Y, y
Y, z
Y 2
5o%$ DCN (8) , (7) /,
d
. ( N

D FE. P$ sCN & *
+" W!6 #)P J* -! -IN! Z+E FE. :2 N"
';!L. vVN FE. J( .F( 4N W. (6) D
#% u2 , u1
Jow$ 'o (19) D " W. ';(,EN $ J(

o( o o k1 o; WoE5 4No oP$ DCN FE. '(
oC + (X) W #@ , 5 4N V! W C
Q Q
Q Q
min 1 , 2 :) ooooI max 1 , 2 : k 2 oooo; X
X
X
X X
JPoN Q2 , Q1 (oC o Jo( o o $ . vVN (*
o o vVN I CN" k2 ; ? D S W. a:) E
o6 : o o; voVN .F +R 0 S + ," :
J5 j d& \& W. J , W. '%S $ S5 I
':o.( 'o (16) oD ?R k2 ; F4$ =N n! . L!
FoE. oP$ soCN o& * hS, V! P$ DCN FE.
oX) X +o Wo #o@ k2 ; F4$ =N J( .W. >.
u1
u2
k1 ( X x)
(15)
z g k2 ( X y)
?D g CD$ , W. W V! C ( k1 ; +"

D ?D k 2 ; , (8) D
k2
Jsign( X ) X y 2 , J ! 0
(16)
x X , y X , z X ooV! ooDCN oooo :, oo
N
D (4) D FE. P$ sCN & * hS,
. (
1 2 1387

(D& = <5 WCC6 :(5 J* -! -IN! Z+E " FE. CD$
60
25
o FoE. +o. o$ (<) <(I5 ; ( J

( X +
20
. 4N P$ DCN ( P$ W!6
15
oL! o6R N, 2 F! , 1 F! >$ J5 4N >$$ J(
10
Oo o 'o .o UV< D0K 5 : X. CD$
5
x (t)
oo _o@$ (6) oD o +" o( o5 o oE C 6 FE.
+NoS J( .N X. (15) D +NS : WE5
-5
-10
oD o ?oD k2 oCD$ o; , W. V! W C ( k1 ;
-15
. :, (8) D ?D g CD$ , (16)
-20
0
10
15
20
25
30
35
(6 X)< .5
Tim e(s )
FE. x W!6 N: . (1 \
b(oN ';o< CD$ L! (" + +<N ' WES J(
o5 o4N ' oNL ;(:. J( . ] '& (:" *
40
+o a,oC Jog , FE. jN x$ aL L! &
30
o a:) .o o(: vo D P$ DCN N;N x$ WEN L!
y (t)
20
+NS k1 ; C a

N ';<(:" W. m
(16) D J C , k1
10
20 #% (15) D
oCD$ 'o o!, (C , FE. !, /( , J
1 #%
.W. 5 4N (: #% , +E( ';< L!
-10
x(0)
-20
0
10
15
20
25
30
3, y (0)
2, z (0)
3, g (0)
0, k 2 (0)
35
Tim e(s )
NJ s< g aQ Q"

!' 2 .1.5
FE. y W!6 N: . (2 \
oDCN FoE. ':o.( oL! ("o o WES J(

o J* -! -IN! +E FE. 4 J( . . v D P$
60
o5 o4N o o0 IoN! o& FE. P D
50
0.4 ':
oooDCN FoooE. vooo D oooP$ oooDCN oooooo {ooo5 .oooooo

z (t)
40
\o& , t = 5(s) o4=! oL! , o E
10, 10, 35.71
#ox$ (3) o$ (1) '; o FoE. W!6 'x N: . .
30
>o$ u2 , u1 o! ';!Lo. , (4) \ g ! N:
20
; oo Joo( oo DNoo .Woo. oo +oo<N (6) , (5) '; oo
10
o W&o. o FoE. ';!6 +NS & : G 46i
0
0
10
15
20
25
Tim e(s )
30
35
: ogN" ?oD IoN g oCD$ o .No L v D P$ W!6

oooP( 4Nooo oooN oooC ooo ,ooo ooo4N ooo oooD 'ooo$
FE. z W!6 N: . (3 \
. L g1
(b c)
1 2 1387

61
(D& = <5 WCC6 :(5 /J* -! -IN! Z+E " FE. CD$
FoE. z oP$ Wo!6 D o ox$ \!o ;LN WO
X
D ?D
(D 3) / 8
46i DN .N +<N (12) $ (7) '; >$$ IN
W&o. o +og FoE. jN x$ FE. ';!6
-5
g(t)
(o:" J( s b(N .( x$ z
-10
oCD$ o +" ,i& .N L v D P$ W!6 0
-15
. L g1 N C 4N ?R , \S N IN g
-20
-25
0
25
10
15
20
25
30
35
Tim e(s )
20
g CD$ N: #x$ (4 \
15
10
250
x (t)
5
200
0
-5
150
-10
100
U (t)
-15
50
-20
0
10
15
20
25
30
35
Tim e(s )
-50
FE. x W!6 N: . (7 \
-100
-150
0
40
10
15
20
25
30
35
25
30
35
Tim e(s )
30
u1 ! L. (5 \

20
y (t)
30
10
25
0
20
U (t)
-10
15
-20
0
10
15
20
25
Tim e(s )
FE. y W!6 N: . (8 \
30
35
10
0
0
10
15
20
Tim e(s )
u2 ! L. (6 \
: D o ox$ o lo5 /(o +o Wo=$ WoES J(
t1 = ' 4=! >$ D
0.9 G-. , D
0.8 D
0 .4
o& FoE. IN! & FE. : FE. jN t2 = 25(s) , 15(s)
Jo( o Wo o $ o( .oo x$ J* & FE. G-. , !
X = 10 oL( #oP o( ooN x$ v D P$ DCN * W!6
1 2 1387

(D& = <5 WCC6 :(5 J* -! -IN! Z+E " FE. CD$
62
60
30
50
25
40
z (t)
u (t)
20
15
30
20
10
10
5
0
0
0
0
10
15
20
25
30
10
15
35
20
25
30
35
30
35
Tim e(s )
Tim e(s )
FE. z W!6 N: . (9 \
u2 ! L. (12 \

5
M9< - NJ s< at g Q"

!' 2 .2.5
!`
-5
t1 = 'oo4=! D o o oo {5 4 J( . x$

G-ooo. , D
0.8 ooo D
g(t)
vo D P$ DCN F , FE. jN F {5 6 J(

0.4 : >ooo$ t2 = 25(s) , 15(s)
o& FoE. : >$ IN FE. jN #%( .( x$ D
-10
-15
0 .9
-20
Jog .oo x$ J* & FE. G-. , ! & FE. IN!
-25
0
>$ IN FE. v D P$ DCN n ' 4=! {5
10
15
20
25
Tim e(s )
10, 10, 34.1 oooooo E 10, 10, 35.7 :

(o:" , o o< IN J( . x$ E 12.3, 12.3, 51
G-oooooo. , E
g CD$ N: #x$ (10 \
vo D P$ DCN FE. ';!6 (L ( ':. b(N S
250
Jo( b(oN .Wo. o4N o oN oC o g oCD$ (L ,
200
.W. +<N (16) $ (13) '; m (:"
150
100
u (t)
25
20
50
0
15
-50
10
-100
-150
x (t)
10
15
20
25
30
35
Tim e(s )
-5
-10
u1 ! L. (11 \
-15
-20
0
10
15
20
25
30
35
Tim e(s )
FE. x W!6 N: . (13 \
1 2 1387

63
(D& = <5 WCC6 :(5 /J* -! -IN! Z+E " FE. CD$
( _N! FCE 9, : X. WEVN 6 +" : G .(
40
FoE. o oS ( 6R CD$ D0K 5 L!
30
20
, WoO 'o o; o oD0 o ';!L. + 5| P 6
10
& * hS, V! DCN FE. N

':. ( CD$
y (t)
.:. ( N

D +" P$ sCN & X : 5%$ DCN (
, oD F! , >!S n! l5 6 , .( \@6 +" P$ sCN
-10
a,oC , ';< L! (" + +<N W; +( .( #O
-20
o* b(oN vo D P$ DCN , FE. ' x$ WEN +" +
-30
.( ] '& (:"
-40
0
10
15
20
25
30
35
30
35
30
35
Tim e(s )
KL
FE. y W!6 N: . (14 \
[1] Nieuwenhuys, B. E., Gluhoi, A. C., Rienks, E. D. L.

and Weststarte, C. J., Vinod, C. P., Chaos, oscillations
and the golden future of catalysis, Catalysis Today,
vol. 100, no. 1-2, pp. 49-54, February 2005.
60
50
z (t)
40
[2] Wang, X. and Zhang, J., Chaotic secure

communication based on nonlinear autoregressive filter
with changeable parameters, Physics Letters A, vol.
357, no. 4-5, pp. 323-329, September 2006.
20
[3] Zhang, H., Ma, X., Xue, B. and Liu, W., Study of
intermittent bifurcations and chaos in boost PFC
converters by nonlinear discrete models, Chaos,
Solitons and Fractals, Vol. 23, no. 2, pp. 433-444,
January 2005.
10
0
0
10
15
20
25
Tim e(s )
FE. z W!6 N: . (15 \
[4] Orel, V. E., Romanov, A. V., Dzyatkovskaya, N. N.

and Melnik, Yu. I., The device and algorithm
forestimation of the mechanoemission chaos in blood
of patients with gastric cancer, Medical Engineering
and Physics, vol. 24, no. 5, pp. 365-371, June 2002.
-5
g(t)
[5] Sprott, J. C., Vano, J. A., Wildenberg, J. C.,

Anderson, M. B. and Noel, J. K., Coexistence and
chaos in complex ecologies, Physics Letters A, vol.
335, no. 2-3, pp. 207-212, 2005.
30
-10
-15
[6] Shabunin, A., Astakhov, V., Demidov, V., Provata,

A., Baras, F., Nicolis, G., and Anishchenko, V.,
Modeling chemical reactions by forced limit-cycle
oscillator: synchronization phenomena and transition to
chaos, Chaos, Solitons and Fractals, vol. 15, no. 2,
pp. 395-405, 2003.
[7] Ott, E., Grebogi, C. and Yorke, J. A., Controling
Chaos, Physical Review Letters, vol. 64, no. 11, pp.
1196-1199, 1990.
[8] Lu, J. and Wu, X., A unified chaotic system with
continuous periodic switch, Chaos, Solitons and
Fractals, vol. 20, no. 2, pp. 245-251, April 2004.
[9] Wu, X., Wang, J., Lu, J. and Iu, H. H. C.,
Hyperchaos behavior in a non-autonomous unified
-20
-25
0
10
15
20
25
Tim e(s )
g CD$ N: #x$ (16 \
( < .6
J* -! -IN! Z+E JPN FE. CD$ !pE !C J(
* : V! DCN n FE. O 4 . 5 4N
vooVN FooE. 'oo >oo. oo! ',, , +" ooP$ sooCN oo&
1 2 1387

(D& = <5 WCC6 :(5 J* -! -IN! Z+E " FE. CD$
[15] Hwang, C. C., Fung, R. F., Hsie, J. Y. and Li, W.

J., A nonlinear feedback control of the Lorenz
equation, International Journal of Engineering
Science, vol. 37, no. 14, pp. 1893-1900, 1999.
[16] Richter, H., Controlling the Lorenz system:
combining global and local schemes, Chaos, Solitons
and Fractals, vol. 12, no. 13, pp. 2375-2380, 2001.
[17] Lorenz, E. N., Deterministic non-periods flows,
Journal of atmospheric Science, vol. 20, pp. 130-141,
1963.
[18] Chen, G. R. and Ueta, T., Yet another chaotic
attractor, International Jpurnal of Bifurcation and
Chaos, vol. 9, no. 7, pp. 465-466, 1999.
[19] L, J. and Chen, G. R., A new chaotic attractor
coined, International Jpurnal of Bifurcation and
Chaos, vol. 12, no. 3, pp. 659-661, 2002.
[20] L, J., Chen, G. R., Cheng, D. and Celikovsky,
S., Bridge the gap between the Lorenz system and the
Chen system, International Jpurnal of Bifurcation and
Chaos, vol. 12, no.12, pp. 2917-2926, 2002
64
chaotic system with continous periodic switch, Chaos,

Solitons and Fractals, vol. 32, no. 4, pp. 1485-1490,
May 2007.
[10] Ge, Z. M. and Yang, K. W., Chaotic range of a
unified chaotic system and its chaos for five periodic
switch cases, Chaos, Solitons and Fractals, vol. 33,
no. 1, pp. 246-269, July 2007.
[11] Liu, J., Chen, S. and Xie, J. Parameter
identification and control of uncertain unified chaotic
system via adaptive extending equilibrium manifold,
Chaos, Solitons and Fractals, vol. 19, no. 3, pp. 533540, February 2004.
[12] Chen, B., Liu, X. and Tong, S., Adaptive fuzzy
approach to control unified chaotic system, Chaos,
Solitons and Fractals, vol. 34, no. 5, pp. 1180-1187,
2007.
[13] Lu, J., Huang, B. and Wu, X., Control of a
unified chaotic system with delayed continuous
periodic switch, Chaos, Solitons and Fractals, vol. 22,
no. 1, pp. 229-236, October 2004.
[14] Hua, C. and Guan, X., Adaptive chaos control for
chaotic systems, Chaos, Solitons and Fractals, vol.
22, no. 1, pp. 55-60, October 2004.
1 2 1387

65-73 =X@ 1 2 1387

(-6 !' ,!

- O` W4 ( %)%L 5-2 S2
<\ N)2Q" 0 * (!6# L2! <
2
2
1
N: 5 1(X@ \! _.
L<N (. 1 N,$ B
seif.dalilsafaei@gmail.com, 2Fbrazandeh@aut.ac.ir
____________________________________________________________________________
( ] ) ':" N ',: BP $. \6 W; h(. , " E , !C J( :%+
+. J6 = ,< , 5% '^N \@X W6 E + S 9, J( ,iP .W.
: \@6 b(N ;N . S (: P$ 0 S 9, J( N^ F(L! ') W S $ J| .
' \6 ') W&. 9, J( '(I : .W. E(C W. ' 'w5 !L* h$ ' , b(N ':.
uNi W6 \%X jN 9, +N ,= S (: P$ +N | # & +N ; ':" (: P$
. +. W; W S , +. J( 'D0 UV< E +N R W; #
N^ F(L! , ':. ; S N ',: :' '^,
Abstract: In this Paper, an efficient and quick method for inverse kinematic solution of hyperredundant manipulators is presented. In addition, it minimizes the joints displacement, energy
consumption, and actuators torque while the end effecter reaches the target. Due to the flexibility
of the genetic algorithm, it is capable of handling any number of constraints. At the end, the
simulation result is compared with a method based on workspace density function. The
advantages of new method are fast, optimization of robot function, handling many constraints,
independent of joint design, real time, minimum error and ability to control the direction of
reaching the target.
Keywords: Hyper redundant manipulator, Constraint, Optimization, Genetic algorithm
_____________________________________________________________________________
WP@ X. +$ # J( '
%9 .1
[6,5]6 [4]' E ' X. [2,3]+.$

[8]L /= (. # & [7]: , WXN ' !! .:
[10]( J # & [9] # [5]E ' \ .:
[12] (w5 ' LE( J0. , [11].N (w5 #0 #
.
DCN +. 5@ ) ':" N ',: :
E R J6 N JP E ', W6 ( , UV< '
'S . , +$ S J( .N | IN (S WE(
' W(,= I WV. 'S , N JP$ /.$ aN
\ WV. 'S . ' FEC$ # J( $m
.. S P@ ' # ':" + * \!

\! ) ':" N ',: . N sCN : 'E
& 'E : NS ) ':" , \%X (: P$ +
' + : # & d& `(, J( ;( W. UV< (
. hN 0 +, F0 , ~
\ aR0 # ) ':" N ',: N #
# jN J( _ V ';N ) ':" # , . #
1
+
( * /.$ J!, ' ) ':" # +& .[1]
.W. # jN J( ' aN J($ >. ( D
W(,= , \@X W&. W(,= \@X +, (,: W(,=

i0 + \S6 \ aN 'S . \@X , ,<
Chirikjian
http://www.isice.ir
+( ?S I , +.; J

N 1387

N^ F(L! /.$ ) ':" N ',: & + ; , BP $. \6 ' ( 9, ]
N: 5 (X@ \! _.
r% .2.2
66
'^N + \S6 \@X , ,< + \S6 \@X #6
'(,: : ' J5( , BP $. \6 !C J( @
+<N 1 \ . ... , hN 0 a& \@X ,
$ # . @5 + ( h h$ ' N \@X
$CC=$ L<(:" 0. ) ':" N ',: &N
.
.[9] JPL
(1)
( xe xg ) ( ye y g ) 2 ( ze z g ) 2
2
h( x )
G-. . 0 E , ,: . !C J(

\6 \@X 'w5 ; vVN: $ \ ( 9,
#%V xg , y g , z g , (O ';N) #. #%V xe , ye , ze
=! v J(; J5( , N^ F(L! /.$ BP $.
F (IN RCN vVN , R /.$ DCN
#O 4 ;N . ,< + ;
.N JP$ E ',
: \@6 b(N , ':. : \@6 b(N ( 9, + "
$2 !' ,!

.3.2
'w5 !L* h$ ' n D ' 9, : (
:: W. #& # & + ;
. E(C W. (1WSDF 9,)'
i0 WE( 4 J( ' :# W6 + uNi -1

(: W. (N (: D \%X a$ #6
. \@X ';6 +I FCE D +:
n
min( | (T i , x T i , x ) | & min | (T i , y T i', y ) |)

'
i 1
i 1
i \%X (NO (,:

(,: T
(2)
Ti, x
'
, i \%X !, (,: T i, x 2 D
i \%X !, (,: IN T i , y . x 0* '.

JP$ WPS : # . y0* '. i \%X (NO
'
i, y
S +" # = , L( JP$ WPS '

WPS WPS , # !, WPS (# ' =! )
. N (NO
\@X , ,< WE( 4 J( ' :,< ':. ; -2
W6 W&. N" $ . .= 3 D :
[9] ) ':" N ',: :1\
uJ v - -6 2 .2

uJ T)` .1.2
5 4N ,< . 45@ v O : +$ J( E
\@X '(,: J5( , BP $. \6 5@ !C J( :
> D J( \! . +. WPS FE. ,<
J| W. , ] . N UV< +. W;
$ . E R # , ,< +N .= \S
,< N ( ' 0 S BP $. \6
('^N + ; 4 )J W6 \S6 # \@X N"
+N | , '^N , E R +: =! # & +N ; ,
; +& %C ,< J5 4N . v D WPS
. \@X +, (,: S , ,< S N # , 'S
. W. |5 ':.
v : +$ W. a" 'S W&. #x$ {5
Ti 2
(3)
. 45@
S , # & +N ; \ .
i 1
('& , .) x,y '= '. \@X , ,< Ti
. 0
N ,< WE( \%X , ,< .= ' .

+" ' 5 S ' ! )\%X +" O ' ! a$ :
+<N ,< .= =N 4 D .N .= (# . $ \%X
.
Ti
Z i x ( Ri u Fi )
(4)
9, J( !C 27 : ) Work space density function 1

( .W.
1 2 1387

N^ F(L! /.$ ) ':" N ',: & + ; , BP $. \6 ' ( 9, ]
67
N: 5 (X@ \! _.
+ .N , WE(
Ti , 'T i '^N + ; '
EN a I $ \%X J Ri +, = Z i ) D
+ ;$ WES J( , \S WES ,<

; 10 D WE(
(5D) ' m ',N Fi , Ci ()!
.
'T i + ' . 0 'T i
W6 J( # W. J( D J( a;X .

.[12].
n
i 1
i 1
(10)
min( (T i , x T 'i , x ) 2 (T i , y T i', y ) 2 )
%R ,!< e2 .4.2

,< .= ' +! (N :2\
, ,< ' N WE( # ' :,< -1
.(11 D)N :,
$ ':
6 : \@X
Ti Tmax(i)
(11)
i 1,2,3
( \%X /.$ \=$ \S ,< F(I IN
Tmax(i ) D
Fi
m g
(5)
j i
: IN ! EN a I . ! a m 6 D

.(" W. 6 D
. UV< '. $ /.$ & \S ,< F(I

+, (,: WE( 4 J( ' :\%X +, (,: W(,= -2
.LN < ':
6 : \%X
T min(i , j ) i, j max(i, j)
0,1,...n
(12)
\%X +, (,: 6 , \S6 IN T max(i , j ) , T min(i , j )

. UV< '.
W4 5-2 .3
( .: ';< 9, ' =N , V J(
F(L! /.$ ':. ; , ';, ] 9, X
Ci
0
i
(6)
A ci'
0
'
JL #i($ G($ i A , ! a I ci 6 D

. .= (: \ , $ W(,N /, ?(R :
0
i
0
i 1
i 1
i
(7)
(8)
i i
i A
cos Ti
sin Ti cos D i
sin Ti sin D i
0
sin T i
cos T i cos D i
sin D i
cos T i cos D i
cos D i
T i , @5 di
cos D i d i
1
ai 1
sin D i d i
# , S (: P$ J( \! . #,X E N^
J +, (,:
D0 , \6 +: (: (I5 >. , ';, : X.
(N 3\ . g (,: D i ,
. IN F(L! ' >$$ +& (: !D .
r< & R ai ) D
X i , X i 1 '=
.W. " J(
J Y` .1.3

/.$ J( . _(P$ # ' W6 E WE(
N ';5 +& E ', F (IN sCN JP$ ,
$ # . @5 + \S6 F(L! X, .(n #@ E
.W. sCN J(
s- è- Y` .2.3
( WE( (N :K" 0 W6 N # N" '
a$ {5 6 J( .W5 4N +" ' j, WP|,
5@ WP|, J( J5 4N . J: & , F ;!,
W6 j, ' 'L( WP|, +$ , W. . '
[12] $ W(,N 'nLN : 3\
: G . N j, WP|, E, F(L! & . vVN
(: D + '^N + : :'^N + -3
' j, WP|, +& \S 6 # WP|, W6 j,
.
. 5 4N 'P WPS +.
Ti 'T i
(9)
1 2 1387

N^ F(L! /.$ ) ':" N ',: & + ; , BP $. \6 ' ( 9, ]
N: 5 (X@ \! _.
68
\@X +, (,: N^ F(L! 'x .

." W. ;N" : WES J(
<\ N)2Q" 5-2 O` W4 .4.3
JP$ , # . J @5 + \S6 V J( :
W. (: , 'S J5 4N
h
(x e x g ) (y e y g ) 2 (z e z g ) 2 R
2
Ti Tmax(i)
i 1,2,3
T min(i , j ) i, j max(i, j)
(14a)
(14b)
(14c)
0,1,...n
* , I) +" x$ $ # . @5 + R

/, 14 D 'x C . x$ +. WS (+
J( ':. ; : X ( .W. 12 , 11 1
JP$ N h$ ' ; ? D C ( +. V
R /.$ ':
&: $ ;N" @5 W. ( v a$
* . * ( : 2 +& ) JP$
. x$ IN +. WS & J( + I ,
F s ( v $ a
N ':. ; _(P$ $
( , kV. N^ F(L! /.$ (N ," @5 +
#& .N ';LN 'P # & a
N ' $&iR N
N _S = ; v ( N^ F(L! WX +$ L(
J ' v a$ kV. ,
E \6 'w5 a$
.:
' v kV. , N^ F(L! /.$ S ':. ; '
x R '( 'x :, [13]$ WX vVN 9, : >.
W P$ ';,:, : F ':" ' .W. X.
) ' ':" ' . F5 >. WS (W5)F
.(" W. >. WS $ ( (I5 W P$ WE(
N: h$ .= =N L( ' 9, 9, J( #,X$ ;$
\6 ., \6 N: h$ .= ' . vVN (5 ,
S ( 6 ( \6 , (N ," S a$ (
D : , , N: h$ .N ' FEC$ N ,"
.(" W. 16 D : a, , , 15
F ( x)
(15)
h( x )
j
F ( x)
f max g i ( x)
(16)
j 1
W. UV< DN . .= 1 D : h(x) 15 D

S h$ g i (x) 16 D .W. @5 h$ ;$ , J( N: h$
.= ' . S \S ' v a$ C 6 f max ,
# W6 WP|, j, (N : 4 \
$2 !' ,!

.3.3
% + ; , # W6 + uNi 9, V J(
# \@X , ,< + ; 9, , '^N
$ #6 i0 + ; . . 5-3 WES
'^N +
N , 10 D + ; >. (2 D)\@X
,n
N ; \@X $ #6 i0 $@ (: .
\@X #6 i0 ':. ; . ; IN ;N" j

+& S a$ WE( #@ N^ F(L! /.$
0 N^ F(L! + >. N ':. ; E ,
' 'w5 U%V$ +& W=$ 'L( 9, : \! J .
. X. \%X
\%X 5 4N \%X ' (w5 4 J( '
+, (,: .= ' .W. JP$ 'w5 :N +, S ;$
+. ' # uNi W6 W; N: \%X :

+, (,: G-. .(NO 0/2 +& ) 5 4N UV<
.(" W. (: D : X. \@X
W6 W&. * 5 4N +: =
(13)
\@X +, (,:

vw @6 $ W&. FEC$ : \@X W6 W&. 13 D
S ;$ \@X N" $ .(" W. N0* $$ WEN
:,
$ JP$ +: : ;N" r6 +: E ' ,= +,
vVN ' N # 'I5 WV. #%V< $ +: J() N
," + uNi s .((N W6 uNi FE.
( # uNi W6 ' " W. '(,: (C .
'^N + ( 10 D
N * E | T i , j
Ti, j |
'
'^N + ; : X ( . ; 'C IN 5%
; ' ,= J5 S N ; 'C +.
N^ F(L! : X. BP $. \6 P WES .
1 2 1387

N^ F((L! /.$ ) ':"

' N

',: & + ; , BP $. \6 ' ( 9,, ]
699
N: 5
(X@ \! _.
(6 6
6 WU4 X)<< 2 .4
II5 aN " W. b(N ( 9, + " #O 4
a$ v ,), , ' \6 a$ N: C f maxx
f max +& +" C J(< , .= (:.. ,"

S
. 0 ;N" . (: W. ':.

Matlab
DiU 2! $22
4 è-- 2 .1.4
. 5PP
\ \ J( .W. J( '< }|$ 5 \
# ( _ V ' WP|, F(L

L! & . 4
IN \6 'w5 .

W. +<N uN ' ( \6 6
+<N (6$ UV<

U
E R J6 ':" 13
1 ' =X@
, (N ," S a$ ++" (w5) S

\S . ,
" 1 , N^
F(L! ' , # #%V
V< .W.
' FEC$ ( N ," SS ( 6 (w

w5) J K 'w
w5
1 , N
F(L! , # #%V< IN C .W.
. m =
=(@ N" L
3 N: h$ . w5 J( X$
f max WO CC , g (x) j
[13] N: JP$

LNL* . :5 \
F(L! F; ' :1 ,
a 1
! +:,
N. 1
R
2
\@X +, +I
1000
!, SX$ ' v P$
'(
!,
v jN
' CE vVN
vVN L &
' DCN ( >$
>$ L &
0/9
>$ 6
15
F(L!
$ #P5 +I
N^
NO 0/2
0 .N

* W. \6
6 . N: h$$ , F(x) /0 J( W. W. \6
6
.
UV< E ', ':" 13 #

# W6 =N . :6 \
g ( x)
' 5 4N +:
. .
#@ (: 9,, : IN vVN (5
: (:. ," S a$
a )J v -1
.W. ;
J K v

(: . vVN
v
N: \6 J v
v , J -2
.
+
C vVN
IN J K v
v , J -3
. '* & +" N:

( , a,:, , >$ ' ' DCN ( >$ L & : .W. XX. ( a,:,
2-C Y) W4 jk< .5.33
: : # . @5 ;N" ( v $ \S 6
\@X W6
vVN ( JP$ 6RR /.$ & J()*

':
&
NO 0/6
9
9:
W&.
.= \6
\ J( $ ,<< 3 D $ 6 J( .W..
3 :
D0
. vVN v +& ,< J( \6 ,
J$N 2 :
:
,<
W. +" (LN

LN , h$ FCE R ,,< N" \!
[12] 9, > D J( . , # +.N +
.W. .
Jou
urnal of Control, 20008, V.2, N.1
1 2 1387

N^ F(L! /..$ ) ':" N '

',: & + ; , BP $.

\6 ' ( 9, ]
N: 5 (X@ \! _.
70
r ,%2 2! o (< (2 .2.4

r%
+.. O ';N W; JP$ W S 9, J( '(I :
: (
'
S #@ N$ F; J( . UV< +,,
. , NN^ F(L!
+. _ V ' W; . :7 \
, . ) : W. .0 W6

W
J. ' 7 \
.. UV< J( : $$ .0 'P '
(%` (? 2! N))2Q" !' 2

.3.4
w5 JP$ +. ' ':" 20 #
# W6 . : 8 \
.
W6
6 ':" 2\@X $ w
w5 W6 . '
F : ' y , x '= '.. ;N"
5 4N
(\< ,%
, 2 .4.4
; W!6, # W6 WP|, '^N + ; . W;

Jg , F.. (J* /0)+ ; a& , (uN sD0)'^N +
.(\ N &
)W. ..= ;N" 5% '^N
' E # '^N + ; W!6 W. UV< N +
.
R
UV<
U

+& DCN 'P$ W!6 J( ' .W.
, j, WPS : # W6 , W.
W 5 4N w5
, (3.2V ); # & ' N J( +.
W!6 < |, 4 .(4.2 V)N | +" , 'S
8 \ .W. ( F.$ _ V \
\ 2 # _ V
_
'
' WPS
UV<< .N +<<N 3 $1 ' # _ V '
.W. .

JP$
: F E @5 # . W.
. 2 : D0 +I , NO 1/4 \6 +: W!6
6 J( '
. 0/2 0/14 0/19 >$$ 3 $1 ' W!6 'D0
,< + ; . :9 \
Journal of Coontrol, 2008, V.2, N.11
1 2 13887

N^ F((L! /.$ ) ':"

' N

',: & + ; , BP $. \6 ' ( 9,, ]
711
N: 5
(X@ \! _.
WSDF 5-2 J)9 - W4 ,6 2 .5.4

(* 9, +:
+ =! W. UV< 10 \ DN
(,: JP$ 9,
9 J( '(I :
( . '$ ( 9, +

+ (* 9,
9 . uNNi # W6
6 W. ' N
.W. <N # W6 + uNi =@
@
WSDF 9, (a \6 +: . : 10 \

(
9, (b
WSD
DF 5-2 J)99 - 2-C ,% 2 .6.4
E(C WSDF
F 9, 9, J
J( ( 9, W S
W #O '
' 'w55 !L* h$ : X. ' 9,, J( .W. (
.( (w5 W!6 ,:: $@ 6*6 P P$)
# . hS W(
W
W. UV<
U
DN 11-a \
WE( N # . .(NN & +" : , 0
# +. >." + + > @ #@
@ (: 0
, \@X WPS +<N y )':" 10
1 # 6 WP|,, E(C :11 \
=N (b , WSDF 9, ?R # W6 =
=N (a . WPPS +<N
; +, # W6 =N (c # & + ; ( 9, # W6
# & +
+
,
{5 ( 9, ,< + ; #OO ' .
= \6 $ ,,<
C .=
1kg
g ,l
0.1m
UV< DDN .W. N W!6 : a ', ;NN"

' ,<< ,: ; # &
& !6 W..
.( b-11\).
Jou
urnal of Control, 20008, V.2, N.1
1 2 1387

N^ F(L! /.$ ) ':" N ',: & + ; , BP $. \6 ' ( 9, ]
N: 5 (X@ \! _.
WUi s< %)%L 5-2 ,!< J- 2 .8.4
72
WSDF 5-2 J)9 - M1 2 .7.4
';N #%V JP$ \@X jN ( F(L! LE, ;$
(I5 $ # . @5 +I + D0
P ' 9, : X. IN # . #%V JP$ . #
';, D0 +I . : \@6 b(N .( (I5 ':"
x$ n! . \6 \S $ W(,N N # FCE $.
'(I : ( .W. > D J( '( IN (WSDF9, : K ) n
: # . @5 )N: h$ .= =N x$ d& ;$ \%X jN
! R ( , ':" (I5 D0 (I5 a& 9, J(
jN 2 ' BP $. E 7, 6 ' \ . (
S \S ' v vVN ., W. +" > D J( \! .
.W. ( \6 ':" 2,1 \@X (w5 , '

X@ #
: ;N" @5 N vVN ( v N^ F(L! /.$
jN E, , N \6 \%X : @0 jN ' ;$ L( ' 9,
' N. 1 # . +& ) P C
.E \@X
\%X jN , ':" N # /( x$ > D J( .(
Q)! (%R (2 R .9.4

_ V 'S + 5| W S 9, J( '(I : L( (
5 9, , S ' N^ F(L! ') (N$ \!
\@X ; '(,: JP$
( < .5
BP $. \6 ' " , h(. ( , !C J(
'E i0 W. UV< 12- a \ $ . N x$

.( D0 +I , P$ (I5 L( ';, :
E, D0 +I W5
N +$ 12-b \ .
5 D0 : ':" 35 D0 +& ) N ':"
+" F; N . F E IN +" C , ( ':"
+. ' 5 4N @5 ) P C : +" +I W.
.( N (I5 ( # .
9, . ] ) ':" N ',: & + ; ,
# W6 + \S6 W; +, (,: ; vVN \ \6
$. \6 +" + uNi , '^N + ; >.
, 'S J5 4N J ' v kV. , # BP
,< J( \6 vVN W(;N , N^ F(L! /.$ +"
) W&. \@X jN F(L! + E, a& J( 'D0 .
BP $. \6 W S uNi #@ # W6 W S
; S '(: P$ +N | W S (w5 , ' =X@ ' #
: _ V #; : +. W S , # & +
hN 0 a& +$ N" $ . 9, J( ';(I
W S J( + 5| : J( (" W5 4N S ( IN
. F(L!
KL
[1] Sofge D, G.Chinag, Design, implementation and
cooperative coevolution of autonomous/teleported control
system for a serpentine robotic manipulator.
[2] Nakamura, Y., Advanced robotics redundancy and
rptimisation, Addison-Wesley Pub. Company, 1991, Reading.
[3] Ma S, Hirose S., Nenchev D. N., Improving local torque
optimization techniques for redundant mechanisms, Journal
of Robotic Systems, Vol. 8, No. 1, pp. 75, 1991.
WSDF 9, 'D0 N (a D0 . :12 \

( 9, 'D0 N (b
[4] www.voronoi.com and www.snakerobots.com .
1 2 1387

N^ F(L! /.$ ) ':" N ',: & + ; , BP $. \6 ' ( 9, ]
73
N: 5 (X@ \! _.
[9] Gravange A and Woodfine R, mine-sniffing robotic

snakes and eels, IEEE Int Conf on Robotics and Automation,
2000.
[5] Sastry S and Kohn M, Milli Robotics for remote

minimally invasive surgery, Robotics and Automation
systems, 1997.
[10] Suthakorn J, Chirikjian G S, A new inverse kinematic

algorithm for binary manipulators with many actuators,
Advanced Robotics, Vol. 15, No. 2, pp. 225 244 2001
[6] Paljug R , Ohm T and Hayati S, The JPL serpentine robot

: a 12 DOF system for inspection ,IEEE International
Conference on Robotics and Automation, ,1995.
[11] Fromherz M P J, Hogg Tad, Shang Y, Jackson W B,

Towards constraint-based actuation allocation for hyperredundant manipulators, Published in IJCAI-01Workshop on
Modelling and Solving Problems with Constraints, Aug.
2001.
[7] Paap k .l, Christaller T and Kirchner F, A robot snake to

inspect broken building , IEEE Int,Conf on Intelligent Robots
and systems, 2000.
[12] Deb.k , An introduction to genetic
[8] Wolf A, Brown H.B, Casciola R, Costa A, Shammas E

and Choset H , A mobile hyper redundant mechanism for
search and rescue tasks ,International Conference on
Intelligent Robots and Systems, Las Vegas , Nevada, 2003.
1 2 1387

87vol.2. N.1, Fall2008, Control Engineers Magazine, Iran

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

87vol.2. N.1, Fall2008, Control Engineers Magazine, Iran

Uploaded by

Copyright:

Available Formats

':5 FE. B.  '$-  '   CD$ ( E

>& UV<$ W; E. ; (

J* -! -IN! Z+E " FE. CD$ 

1  2 1387  

1  2 1387  

1-19 =X@ 1  2 1387

01 / )*% + ,-. &' (  

Journal of Control, 2008, Iranian Society of Instrument & Control Engineers

+( ?S I ,  +. ; J

<6 ( (    .2

, PM-10 ?o P o( o!" #m  o 'o oNN 'o  . oN  

+o ' o& ( , ' ' #@  N$    N  N : '.

J( ) C P     X '    ? P ( !" #m

\CoE o Noo so (  N : ';(. \ =$ , (I

o o (o o(  o. 'o , oNoE, F  ! R  , N

 FoE. o ,0 FNo o q5@ ,  N & \S ',,  N,

Opo$ LNoL* +o  \%X$  [4 1]h   . . 4N  $!"

? P #m .W.   W=@ +" #w L( , N vn   ()" J(

4" 3' D0 K +g @0 N$  N : '. (  

Jo( #ox$ J(o . oN o ? P    * $ NO * J q)P

.[12]  F 'I(N

.   J(  s   

Jo oE +I .W. FPR ,  uN +,  ': J E

oP ) o ,0 -',, o o +o$ o   W!6  . . ;

, on o ',, o N   , \ =$  (;. J* (.

N!oo. >oo$$ J( oo , ooNoo oo!$ oo< Woo. 'oo0oo. /oo.$

'ooP!loo5 Woo : 'oo. oo 'oo Joo LooE, Noo : ';(oo.

Do. + o 'o Wo5 :oN  +`E ;$  i0 

Jo  E \<$ , +`E \6  J  # S 

 o()" J( : N # @ n O 'I v%&  E . ', 

 L   !" $ '%S ' ("5 : P., E ' : 

oEXS J(`N" z%V S,& ,  S ';(  W. NE   <

.W. E, 0 n

.  'W!P5 a

,    J LE, .  @ ' : (

oE( o Wo. F; $,|  !"  l5    $ 

. N FEC$ 5%$ ,

oX  '  (;, , F(: +" # ,I , N : ';(.

>o D J(o; F(,o oo , ':oE! o.   +" : 

' 'P , %& ' 3 V  .F(N .  E

LNoL* , ANFIS o o 4 oV  . F0 .  4N

PSO F(oL! o  'o%V }o|$  5 V  0 J

'oo 'o o  ':o.o 0" V  W(;N , F0

o 'o joN F E( F; 5 J( +  UV< : G . 

Journal of Control, 2008, V.2, N.1

Particle per million

1  2 1387  

?( @0  & >!<$ = ! ' & ' ;

&' (  .3

o0 o  o0 (  (;! J(; : ( ,  0 

'. ' F  DN W.   F.$ (5%$ #@  

, oE oo \S  G   @ '. :  V( F;

.  %& ' \  

+o 5o%$ +Io o oI H  oC o* . oN X. I(N

 o " o , o +oEN +  : JL!  .,   %& '

v,o$ , UV<$ W; N H , H (C J(< , W.  

\o<$ oNo0 N L( (   (;! . : +EN +  N W.

.,E , ,<   , ("5  6 , %& '  .W.  

,o< #o@ o  ,0  ',, () : ',, %& ' 

 o  ,0 _ V '() : & : G ,   v| '+:, 

Radial Basis Function 12

Journal of Control, 2008, V.2, N.1

1  2 1387  

':5 FE. B. '$- ' CD$ ( E

>& UV<$ W; E. ; (

J* -! -IN! Z+E " FE. CD$

1 2 1387

1 2 1387

1-19 =X@ 1 2 1387

01 /)*% + ,-. &' (

+( ?S I , +.; J

<6 ( ( .2

, PM-10 ?o P o( o!" #m o 'o oNN 'o .oN

+o 'o& ( , ' ' #@ N$ N N: '.

J( ) C P X ' ? P ( !" #m

\CoE o Noo so ( N: ';(. \ =$ , (I

o o (oo( o. 'o , oNoE, F ! R , N

FoE. o ,0 FNo o q5@ , N & \S ',, N,

Opo$ LNoL* +o \%X$ [4 1]h .. 4N $!"

? P #m .W. W=@ +" #w L( , N vn ()" J(

4" 3' D0 K +g @0 N$ N: '. (

Jo( #ox$ J(o .oNo ? P * $ NO * J q)P

.[12] F 'I(N

. J( s

Jo oE +I .W. FPR , uN +, ': JE

oP ) o ,0 -',, o o +o$ o W!6 . .;

, on o ',, o N , \ =$ (;. J* (.

N!oo. >oo$$ J(oo , ooNoo oo!$ oo< Woo. 'oo0oo. /oo.$

'ooP!loo5 Woo : 'oo. oo 'oo Joo LooE, Noo: ';(oo.

Do. +o 'o Wo5 :oN +`E ;$ i0

Jo E \<$ , +`E \6 J #S

o()" J( : N #@ n O 'I v%& E . ',

L !" $ '%S '("5 : P., E ':

oEXS J(`N" z%V S,& , S ';( W. NE <

.W. E, 0 n

. 'W!P5 a

, J LE, . @ ' : (

oE( o Wo. F; $,| !" l5 $

.N FEC$ 5%$ ,

oX ' (;, , F(: +" #,I , N: ';(.

>o D J(o; F(,o oo , ':oE! o. +" :

' 'P , %& ' 3 V .F(N . E

LNoL* , ANFIS o o 4 oV . F0 . 4N

PSO F(oL! o 'o%V }o|$ 5 V 0 J

'oo 'o o ':o.o 0" V W(;N , F0

o 'o joN F E( F; 5 J( + UV< : G .

1 2 1387

?(@0 & >!<$ = ! ' & ';

&' ( .3

o0 o o0 ( (;! J(; : ( , 0

'. ' F DN W. F.$ (5%$ #@

, oE oo \S G @ '. : V( F;

. %& ' \

+o 5o%$ +Io o oI H oC o* .oN X. I(N

o" o , o +oEN + : JL! ., %& '

v,o$ , UV<$ W; N H , H (C J(< , W.

\o<$ oNo0 N L(( (;! . : +EN + N W.

.,E , ,< , ("5 6 , %& ' .W.

,o< #o@ o ,0 ',, () : ',, %& '

o ,0 _ V '() : & : G , v| '+:,

1 2 1387

x t ',, P( .( W. (8)\ $ F %&

o456 o( x t p $ x t 1 P( +" S ',,

p oo : 0oo$ /oo0

o o( o!C J( #)C +" 5(P.$ L( ' :

,o< (8)\ . 6D0 <N G , !,IN +(

, x (t ) x (t 1) , x (t 1) , x (t 1) x (t 2) " @

)* % + ,-. &' ( .1 .3

N: '. ' (MLP1),

.(L( () () : W. * :)

x t 1 2 x t x t 1 F(o" G-o. .W. " (10) , (9)

l. . 0 ';N:, BP 9:" 9, 'R

b(N . x t 1 2x t x t 1 C ,0

'o() : 'o W5 E P$ L. .( 3 D0 ';!L.

.(W. X. DS $ (L. h$ : ',, I

2! 34 (<6- .3 .3

2! 34 ( .2. 3

0o$ o$ o( + S X \S WES DN

o$ o456 $ ';5 J($ X. , J($. : ( ,

.F X. N

. [3]( 5 >P<0$ /0 456 4#

o;N" ',, o o() o* 'oo 'o o W!6 ( (

o$ +o oN F & ',, '