You are on page 1of 7

International Journal of Computer Techniques - Volume 3 Issue 2, Mar-Apr 2016

RESEARCH ARTICLE

OPEN ACCESS

Exponential Stabilization of the Coupled Dynamical Neural


Networks with Nodes of Different Dimensions and Time Delays
Jieyin Mai, Xiaojun Li
(Department of Mathematics, Jinan University, Guangzhou Guangdong 510632, China)

----------------------------------------************************----------------------------------

Abstract:
A class of coupled neural networks with different internal time-delays and coupling delays is
investigated, which consists of nodes of different dimensions. By constructing suitable Lyapunov
functions and using the linear matrix inequality, the criteria of exponential stabilization for the coupled
dynamical system are established, and formulated in terms of linear matrix inequality. Finally, numerical
examples are presented to verify the feasibility and effectiveness of the proposed theoretical results.
Keywords----nodes of different dimensions; coupled dynamical system with time delays;
exponential stabilization; linear matrix inequality; Lyapunov function
----------------------------------------************************---------------------------------I.

INTRODUCTION

With the rapid development of science and


technology, neural networks are gradually and
widely applied to the optimal control, combinatorial
optimization, pattern recognition, and other fields.
Due to the finite speeds of transmission and traffic
congestion, time delays are common in actual
systems, and the delays may be constant, timevarying with parameter or mixed (see [1]).
In the past two decades, stability or stabilization
problems have been the hot topics for neural
networks with time delays, including isolated neural
networks and coupled neural networks (see [2]-[17]
and references therein). For instance, the
stabilization problem of delayed recurrent neural
networks is investigated by a state estimation based
approach in [2]. The exponential stability problem
for a class of impulsive cellular neural networks
with
time
delay
is
investigated
by
Lyapunovfunctions and the method of variation of
parameters in[3]. Some studies on exponential
stability or stabilization problem are about neural
networks with time-varying delays (see [5]-[8]).
Some sufficient conditions to ensure the existence,
uniqueness and global exponential stability of the
equilibrium point of cellular neural networks with
variable delays are derived in[5]. For the problem
of exponential stabilization for a class of nonISSN: 2394-2231

autonomous neural networks with mixed discrete


and distributed time-varying delays, some new
delay-dependent conditions for designing a
memoryless state feedback controller which
stabilizes the system with an exponential
convergence of the resulting closed-loop system are
established in [6]. A new class of stochastic CohenGrossberg neural networks with reaction-diffusion
and mixed delays is studied in [7]. Since the form
of coupling is various, there exist different coupled
neural networks, e.g. coupled term with or without
time delays, or hybrid both. Some researches on
stability and synchronization of different coupled
neural networks are investigated (see [9]-[11]).
Some global stability criteria for arrays of linearly
coupled
delayed
neural
networks
with
nonsymmetric coupling are established on the basis
of linear matrix inequality method, in which the
coupling configuration matrix may be arbitrary
matrix with appropriate dimensions in [9]. Since
there may be some chaotic behaviours as the
dynamics performance of time-delay neural
network in certain situations, chaotic neural
networks are formed (see [12]-[14] and references
therein). Some researches on stability and
synchronization of chaotic neural network are
investigated in [12]-[14]. Furthermore, sufficient
condition for exponential stability of the

http://www.ijctjournal.org

Page 141

International Journal of Computer Techniques - Volume 3 Issue 2, Mar-Apr 2016

equilibrium solution of uncontrolled stochastic


interval system is also presented in [15]. The
decentralized continuous adaptive controller can be
designed to make the solutions of the closed-loop
system exponentially convergent to a ball, which
can be rendered arbitrary small by adjusting design
parameters in [16]. Mostof researches on
synchronization problem of neural network are also
based on linear matrix inequality method and
Lyapunovfunctions.
Although there are lots of researches for the
exponential stability or stabilization problems of
neural networks with time delay, most of them
concern with the same dimension of the states. If a
network is constructed by nodes with different state
dimension, the network will exhibit different
dynamical behaviours (see [17]-[19] and references
therein). Dimensions of nodes are actually different
in many practical situations, so such coupled
complex networks need more in-depth study.
Motivated by the above discussion, in this paper
we consider the exponential stabilization problem
of the coupled dynamical system. The dimensions
of states in each isolated network are different, and
the internal time-delays are different from the
coupling delays. In Sec.3, the criteria of exponential
stabilization for the coupled dynamical system are
given and proved on basis of the linear matrix
inequality and Lyapunovfunction. In Sec.4, two
numerical simulation examples are given to
demonstrate the effectiveness of the proposed
theoretical results. In Sec.5, some concluding
remarks are given.

where i = 1, 2,L , N ; xi (t ) = ( xi1 (t ), xi 2 (t ),L , xin (t ))T n

is the state vector of the i th node; A i , Bi n n are


constant matrices that representing the feedback
matrix without and with time delays respectively;
Di = diag (d i1 , di 2 ,L , din ) > 0 is a constant and diagonal
i

matrix; f i () is the activation function; G = ( g ij ) N N is


an outer coupling matrix representing the coupling
strength and the topological structure of the neural
networks; C ij , ij n n are inner coupling matrices
respectively representing the inner-linking strengths
between the cells without and with time
delays,when ni n j , C ij , ij n n are arbitrary
matrices; i , i are the strengths of the constant
coupling and delayed coupling, respectively; i1 , i 2
are the constant and positive delays, i1 > 0, i 2 > 0
and i1 < , i 2 < ; ui (t ) = (ui1 (t ), ui 2 (t ),L , uin (t ))T n is
an external input.
The initial condition associated with (1) is given
as follows:
xij ( s ) = ij ( s ) C ([ , 0] , ) ,(3)
i

where i = max{ i1 , i 2 } , = max{1 , 2 ,L , N } ,


i = 1, 2,L , N , j = 1, 2,L , ni .
For convenience, the notations are givens as
follows:
I denotes M M identity matrix; I n denotes ni ni
i

identity matrix; y = y y denotes the norm of y ;


max () and min () respectively denote the maximum
and minimum eigenvalue of ; M = n1 + n2 + L + nN ;
X (t ) = ( x1T (t ), x2T (t ),L , xNT (t ))T
II. SYSTEM DESCRIPTION AND PRELIMINARIES
U (t ) = (u1T (t ), u 2T (t ),L , u NT (t ))T
Consider a coupled dynamical system with nodes
F ( X (t )) = ( f1T ( x1 (t )), f 2T ( x2 (t )),L , f NT ( xN (t )))T
of different dimensions and time delays:
F ( X (t 1 )) = ( f1T ( x1 (t 11 )), f 2T ( x2 (t 21 )),
dxi (t )

T
T
= D x (t ) + A f ( x (t )) + B f ( x (t ))
dt

i i

j =1

j =1

+ i gij Cij x j (t ) + i g ij ij x j (t i 2 ) + ui (t ),

(1)

dxi (t )
= Di xi (t ) + Ai fi ( xi (t )) + Bi fi ( xi (t i1 )),
dt

X (t 2 ) = ( x1T (t 12 ), x2T (t 22 ),L , xNT (t N 2 ))T

diag (L) denotes

in which each isolated node network with different


dimensions and time delays is considered as:

ISSN: 2394-2231

L , f N ( xN (t N 1 )))

i1

a block-diagonal matrix;
D = diag ( D1 , D2 ,L , DN ) ; A = diag ( A1 , A2 ,L , AN ) ;
B = diag ( B1 , B2 ,L , BN ) ; Wi = diag ( wi1 , wi 2 ,L , win ) ;
i

(2)

W = diag (W1 , W2 ,L , WN ) ;
X
*

Y
is
Z

defined as a matrix in form of T


.
Z
Y

http://www.ijctjournal.org

Page 142

International Journal of Computer Techniques - Volume 3 Issue 2, Mar-Apr 2016

In order to study the exponential stabilization of


the coupled dynamical system (1), the linear
controller isdesigned as
U (t ) = KX (t ) ,i.e., ui (t ) = K i xi (t ) , (4)
where K i = diag ( ki1 , ki 2 ,L , kin ) , K = diag ( K1 , K 2 ,L , K N ) .
Hence, equation (1) can be rewritten as
i

dX (t )
= ( D + K ) X (t ) + AF ( X (t )) + BF ( X (t 1 ))
(5)
dt
+ H1 X (t ) + H 2 X (t 2 ),

( E (t )) = (1T (e1 (t )), 2T (e2 (t )),L , NT (eN (t )))T


= F ( X (t )) F ( X * ),
( E (t 1 )) = F ( X (t 1 )) F ( X * ),
E (t 2 ) = (e1T (t 12 ), e2T (t 22 ),L , eNT (t N 2 ))T .

Assumption 1.The activation function


fi ( xi (t )) = ( fi1 ( xi1 (t )), fi 2 ( xi 2 (t )),L , fini ( xini (t )))T

isLipschitz continuous, i.e., there exist constants


wil > 0 , such that
fil (1 ) fil ( 2 ) wil 1 2

where
1 g12 C12 L 1 g1N C1N
1 g11C11
g C

L 2 g 2 N C2 N
2 g 22 C22
,
H1 = 2 21 21

M
M
O
M

N g N 1CN 1 N g N 2 CN 2 L N g NN CNN
1 g12 12 L 1 g1N 1N
1 g1111
g
L 2 g 2 N 2 N

2 g 22 22
.
H 2 = 2 21 21

M
M
O
M

N g N 1 N 1 N g N 2 N 2 L N g NN NN
is
Assume that
X * = ( x1*T , x2*T ,L , x*N T )T

holdsfor any different 1 , 2 , and 1 2 , where


i = 1, 2,L , N , l = 1, 2,L , ni .
Definition 1[13]For given k > 0 , the dynamical
system (1) is said to be exponential stabilization, if
there exist constant Z > 0 such that the following
inequality E (t ) 2 Z 2 e kt holds for all initial
conditions eij ( s )(i = 1,L , N ; j = 1,L , ni ) of system (7)
and any t T (sufficiently large T > 0 ), where

the
equilibrium point of the coupled dynamical system
(1), then itsatisfies

=sup
s 0

ni

e ( s) .
2

ij

i =1 j =1

Lemma 1 [4] For any x, y n and positive


*
dX
= ( D + K ) X * + Af ( X * ) + Bf ( X * ) + H1 X * + H 2 X * , (6) definite matrix Q n n , the following matrix
dt
inequality holds:
i.e., ( D + K ) X * + Af ( X * ) + Bf ( X * ) + H1 X * + H 2 X * = 0 ,
2 xT y xT Qx + yT Q 1 y .
where xi* = ( xi*1 , xi*2 ,L, xin* )T .
Lemma 2[6]The linear matrix inequality
Define the linear coordinate transformation
Q( x) S ( x)
S T ( x) R( x) > 0 ,
E (t ) = X (t ) X * , andthe new dynamical systems can

be described as follows:
where QT ( x ) = Q ( x) , RT ( x) = R ( x ) , is equivalent to
i

dei (t )
= Di ei (t ) + Aii (ei (t )) + Bii (ei (t i1 ))
dt

and Q ( x) S ( x ) R 1 ( x ) S T ( x) > 0 .
(7)
Lemma 3 [5]If Q , R are real symmetric matrices,
N
N
+ i gij Cij e j (t ) + i g ij ij e j (t i 2 ) + U i (t )
and Q > 0 , R 0 , then a positive constant exist,
j =1
j =1
such that the following inequality holds:
That is,
Q + R < 0 .
dE (t )
= DE (t ) + A ( E (t )) + B( E (t 1 ))
,(8)
dt
III.
EXPONENTIAL STABILIZATION ANALYSIS
+ H1 E (t ) + H 2 E (t 2 ) + U (t )
Theorem 1.Under the assumption 1, the coupled
where
T
dynamical
system (1) is exponential stabilization,if
ei (t ) = (ei1 (t ), ei 2 (t ),L , ein (t ))
M M positive definite diagonal matrices
there
exist
= ( xi1 (t ) xi*1 , xi 2 (t ) xi*2 ,L , xin (t ) xin* )T ,
P , Q , R and K ,such that the following linear
i (ei (t )) = f i ( xi (t )) f i ( xi ), U i (t ) = K i ei (t ),
matrix inequalities hold: > 0 ,where
R( x) > 0 ,

i (ei (t i1 )) = f i ( xi (t i1 )) f i ( xi ), U (t ) = KE (t ),
E (t ) = (e1T (t ), e2T (t ),L , eNT (t ))T , E (t ) = X (t ) X * ,

ISSN: 2394-2231

http://www.ijctjournal.org

Page 143

International Journal of Computer Techniques - Volume 3 Issue 2, Mar-Apr 2016

= *

*
*

PA PB PH1
Q
0
0
*
R
0
*
*
Q
*
*
*

PH 2
0
0 > 0 ,(9)

0
Q

V&3 (t ) = e (t + )iT (ei (t )) Rii (ei (t ))


i =1
N

e (t i1 + )iT (ei (t i1 )) Rii (ei (t i1 )) (13)


i =1

e (t + ) T ( E (t )) R ( E (t ))
et T ( E (t 1 )) R ( E (t 1 ))

= PD + D T P + K + K T WQW WRW 2Q
Pi = diag ( pi1 , pi 2 ,L , pini ), P = diag ( P1 , P2 ,L , PN ),

From Lemma 1 and Assumption 1, we obtain


2 E T (t ) PA( E (t ))

Qi = diag ( qi1 , qi 2 ,L , qini ), Q = diag (Q1 , Q2 ,L , QN ),

E T (t ) PAQ 1 AT PE (t ) + T ( E (t ))Q ( E (t )) (14)

Ri = diag (ri1 , ri 2 ,L , rini ), R = diag ( R1 , R2 ,L , RN ),

E T (t ) PAQ 1 AT PE (t ) + E T (t )WQWE (t ),

Wi = diag ( wi1 , wi 2 ,L , wini ),W = diag (W1 ,W2 ,L ,WN ).

Moreover, the gain matrix of a desired controller of


the form (4) is given by K = P 1 K .
Proof. From Lemmas 2 and 3, we get the
following conclusion: there exist a positive constant
, such that

*
1 = *

*
*

PH 2

0
0 <0,

0
Q
*

where = + P + (e 1)(Q + WRW ) .


*

PA PB PH1
Q 0
0
* R
0
*
*
Q
*
*
*

E T (t ) PH 2Q 1 ( H 2 )T PE (t ) + E T (t 2 )QE (t 2 )
2 E T (t ) PH 1 E (t )
E T (t ) PH1Q 1 ( H1 )T PE (t ) + E T (t )QE (t )
(11)~(17) into V& (t ) , it yields

+ e WRW + Q + PAQ 1 AT P + PBR 1 BT P


1

(16)
(17)

(18)

+ PH1Q ( H1 ) P + PH 2 Q ( H 2 ) P 2 PK ]E (t )}

From Lemma 2, 1 < 0 is equivalent to


= P P( D + K ) ( D + K )T P + WQW + e Q
+e WRW + Q + PAQ 1 AT P + PBR 1 BT P

V1 (t ) = et eiT (t ) Pe
i i (t )

+ PH1Q 1 ( H1 )T P + PH 2 Q 1 ( H 2 )T P

i =1

2 E T (t ) PH 2 E (t 2 )

(15)

V& (t ) et {E T (t )[ P PD DT P + WQW + e Q

V (t ) = V1 (t ) + V2 (t ) + V3 (t ) ,(10)

E T (t ) PBR 1 B T PE (t ) + T ( E (t 1 )) R ( E (t 1 )),

Substituting

Consider the Lyapunovfunction as

V2 (t ) =

2 E T (t ) PB ( E (t 1 ))

(19)

<0

e ( s + ) eiT ( s )Qi ei ( s )ds

It is derived that V& (t ) 0 . Therefore, we know


V (t ) is decreasing from t =0 , and V (t ) V (0) .
t
V3 (t ) = e ( s + )iT (ei ( s )) Rii (ei ( s ))ds .
t
From the initial conditions (3) of the coupled
i =1
Calculating the time derivatives of V1 (t ) , V2 (t ) and dynamical system (1),we obtain the initial
conditions of thenew dynamical system (8) are
V3 (t ) along the trajectories of system (7), we have
N
N
eij ( s ) = ij ( s ) xij* ( s ) C ([ , 0 ] , ) .
t
T
t
T
&
&
V1 (t ) = e ei (t ) Pe
i i (t ) + 2e ei (t ) Pe
i i (t )
(11)
It follows from (10) that
i =1
i =1
i =1

t i 2

i1

= et E T (t ) PE (t ) + 2et E T (t ) PE& (t )

V&2 (t ) = e ( t + ) eiT (t )Qi ei (t )

i =1

i =1
N

V (0) = e0 eiT (0) Pe


i i (0) +

( t i 2 + ) T

(12)

ei (t i 2 )Qi ei (t i 2 )

i =1

e (t + ) E T (t )QE (t ) et E T (t 2 )QE (t 2 )

i =1

i 1

i =1

i (ei ( s)) Rii (ei ( s))ds


N

i =1

+
i =1

i 1

e ( s + ) eiT ( s)Qi ei ( s )ds

( s + ) T

= E T (0) PE (0) +
N

ISSN: 2394-2231

i 2

i 2

(20)
e ( s + ) eiT ( s )Qi ei ( s)ds

e ( s + )iT (ei ( s)) Rii (ei ( s))ds

http://www.ijctjournal.org

Page 144

International Journal of Computer Techniques - Volume 3 Issue 2, Mar-Apr 2016


N

i =1

e ( s + ) eiT ( s )Qi ei ( s )ds


i2

e ( s + ) eiT ( s )Qi ei ( s )ds = e e s E T ( s )QE ( s )ds (21)


i =1

e max (Q ) e s ds


i =1

e 1

max (Q)

e ( s + )iT (ei ( s )) Rii (ei ( s ))ds


i2

e e s E T ( s )WRWE ( s )ds

e max (WRW ) e s ds

(22)

e 1

max (WRW )

Substituting (21)~(22) into V (0) , we get


2

V (0) max ( P ) +

e 1

[max (Q ) + max (WRW )]

(23)

= {max ( P) +

e 1

[max (Q ) + max (WRW )]}

2
e t min ( P ) E (t ) V1 (t ) V (t )

From

V (t ) V (0) ,we have

min ( P)

V (0) .

{max ( P)

e 1

[max (Q ) + max (WRW )]} e t

2
= Z e t
+

where

Z=

min ( P )

0.6
-0.8 ,
-0.2

0.9
,
0.3
-0.2 -0.4
0.9 -0.9 0.7
11 =
, 12 = -0.6 0.1 0.3 ,
0.9
0.4

-0.8 -0.1 0.8


0.3 0.1
C22 = -0.2 -0.3 -0.9 , 21 = -0.2 -0.1 ,
-0.6 0.9 0.5
-0.6 -0.7
0.9 0.1 0.7
0.2 0.6
22 = 0.4 -0.8 0.6 , G =
,
0.7 0.1
-0.3 0.8 -0.4
1 = 0.1 , 2 = 0.2 , 1 = 0.9 , 2 = 0.1 ,

11 = 0.5 , 12 = 0.7 , 21 = 0.6 , 22 = 0.4 ,

By calculating, we know that equilibrium points


of the two isolated node networks (2) are
T

x%1 = 10 7 * ( 0.0111, 0.0034 )

Combining with (23) and (24), we obtain


2

0.3 0.7
-0.6

-0.7 0.3 , C21 = -0.2


0.7
0.4 -0.1
-0.8
0.4 0.3
, C12 =

0.3
-0.7 0.1

x%2 = 10 7 * ( 0.0756, 0.0117, 0.1720 )

e t min ( P ) E (t )

E (t )

and

-0.8
B2 = -0.5
-0.3
0.1
C11 =
-0.1

{max ( P) +

,
(24) respectively. Therefore, it is easy to derive that
wil = 1 , and W = I 5 , for i = 1, 2,L , N , l = 1, 2,L , ni .
By using the MATLAB LMI Control Toolbox,
solving the linear matrix inequalities in Theorem 1,
(25) it yields the following feasible solutions:

[max (Q ) + max (WRW )]} .

P = diag ( 3.7960,3.4276,3.9682, 4.5893,5.0527 )

Q = diag ( 20.2148, 20.3078, 20.4814, 20.5548, 20.2048 )


R = diag ( 20.2106, 20.2726, 20.3884, 20.4373, 20.2040 )

K = diag ( 20.1689,19.9210,19.4580,19.2622, 20.1957 )

That is, the coupled dynamical system (1) with


Then, we have
nodes of different dimensions and time delays is
K = diag ( 5.3132,5.8119, 4.9035, 4.1972,3.9970 ) .
exponential stabilization.This completes the proof.
According to Theorem 1, given the initial
IV.
NUMERICAL EXAMPLES
condition:
X (0) = (0.5, - 0.2, - 0.3, 0.2, 0.5)T ,
a
coupled
dynamical
Example 1.Consider
system (1) is composed of two isolated node the coupled dynamical system (1) can achieve the
networks, in which the parameters are given as exponential stabilization, and the equilibrium point
is
follows:
X = 10 10 * (0.0913, 0.0236, 0.1704, 0.0514, 0.3277)T ,
N = 2 , n1 = 2 , n2 = 3 , f ( xil ) = sin( xil ) , i = 1, 2 , l = 1,..., ni ,
the simulation results are plotted in Fig. 1. All state
8 0
0.2 -0.1
-0.9 0.8
D1 =
, A1 = -0.5 0.9 , B1 = 0.7 0.6 ,
trajectories converge to the equilibrium point, and
0
9

the equilibrium point of the coupled dynamical


3 0 0
0.1 -0.1 -0.2
system (1) is different from that of each isolated
D2 = 0 7 0 , A2 = 0.1 0.2 -0.3 ,
node network (2).

0 0 6

ISSN: 2394-2231

-0.5

0.4

0.9

http://www.ijctjournal.org

Page 145

International Journal of Computer Techniques - Volume 3 Issue 2, Mar-Apr 2016

respectively. Therefore, it is easy to derive that


wil = 2 , and W = 2 I 5 ,for i = 1, 2,L , N , l = 1, 2,L , ni .
By using the MATLAB LMI Control Toolbox,
solving the linear matrix inequalities > 0 in
Theorem 1, it yields the following feasible solutions:

0.6
x11(t)

x 1 (t)

0.4

x12(t)

0.2
0
-0.2

0.5

1.5

2.5
t

3.5

4.5

Q = diag ( 24.8830, 23.2186, 22.5922, 24.6529, 26.5236 )

x 2 (t)

0.5

x 21(t)

R = diag ( 24.6942, 22.3220, 21.7158, 26.2661, 29.3649 )

x 22(t)

K = diag ( 44.4094,53.7464, 62.1294, 47.1680,39.8634 )

x 23(t)

Then, we have

-0.5

P = diag ( 3.5079, 2.2782, 2.0888, 4.6528,3.3211)

K = diag (12.6598, 23.5912, 29.7447,10.1376,12.0031) .


0

0.5

1.5

2.5
t

3.5

4.5

the
initial
condition
T
X (0) = (1, 1, 2, 2,1) ,according to Theorem 1,the
Fig. 1.The state trajectories of coupled
coupled
dynamical system (1) can achieve the
dynamical system in Example 1.
exponential
stabilization and the equilibrium point
Example 2. Consider a coupled dynamical
system (1) is composed of two isolated node is
X = ( 0.0376, 0.1092, 0.1181, 0.1190, 0.0145)T ,
networks, in which the parameters are given as
the simulation results are plotted in Fig. 2. All state
follows:
trajectories converge to the equilibrium point, and
N = 2 , n1 = 2 , n2 = 3 , f ( xil ) = sin( xil )+1 ,
the equilibrium point of the coupled dynamical
i = 1, 2 , l = 1,..., ni ,
system (1) is different from that of each isolated
29 0
1 -1
-3 1
D1 =
A
=
B
=
node network (2).
, 1
, 1
,

3 -3
21 = 2 1 ,
-3 5

-4

1 4
-5 -5 4
1 -1 , B2 = 1 2 4 ,
2 -4 3
1 1
4 -2 -4
=
,
-2 -4 2

-2
3 -5 -5
, 12 =

,
2
-1 -3 1
4
-4 2 1

-2 , C22 = 3 -5 4 ,
-4 4 4
4
-4 3 -3
0.9 0.7
22 = -5 2 3 , G =
,
0.1 0.4
-2 4 3

1 = 0.1 , 2 = 0.4 , 1 = 0.9 , 2 = 0.6 ,


11 = 0.5 , 12 = 1 , 21 = 0.8 , 22 = 0.4 ,

1.5

x 1(t)

5
-3
A2 = 1
-2
-1
, C12
-2

x 11(t)

0.5

x 12(t)

0
-0.5
-1
-1.5

x 2(t)

0 36
33 0 0
D2 = 0 21 0 ,
0 0 34
-5
C11 =
1
-1
11 =
-1
1
C21 = -5
2

Given

0.5

1.5

2.5
t

3.5

4.5

x 21(t)

x 22(t)

x 23(t)

0
-1
-2
0

0.5

1.5

2.5
t

3.5

4.5

Fig. 2.The state trajectories of coupled dynamical


system in Example 2.

V. CONCLUSION
By calculating, we know that equilibrium points
In this paper, the exponential stabilization
of the two isolated node networks (2) are
problem has been investigated for the coupled
T
neural networks. Thedimensions of states in each
x%1 = ( 0.0645, 0.2690 )
isolated network can be different, and dimensions
T
x%2 = ( 0.3305, 0.3848, 0.0039 ) ,
of nodes are actually differentin many practical

ISSN: 2394-2231

http://www.ijctjournal.org

Page 146

International Journal of Computer Techniques - Volume 3 Issue 2, Mar-Apr 2016

situations, so the derived results will have wider


applicability.
The
criteria
of
exponentialstabilization for the coupled dynamical
system are established on basis of the linear matrix
inequality andLyapunovfunction. The effectiveness
of the proposed theoretical results has been
demonstrated by two numerical simulation
examples. In addition, finite-time control could be
an
alternative
method
for
the
exponentialstabilization of such kind of coupled
dynamical system, which deserves our further
investigation.

[8]

[9]

[10]

[11]

[12]

[13]

REFERENCES
[1]

[2]

[3]

[4]

[5]

[6]

[7]

J.Zhang, X.Wu, Y.Li,Adaptive control of nonlinear time-delay systems


based on cellular neural networks (National Defence Industry
Press,Beijing, 2012).pp. 1-21
H.Huang, T.Huang, X.Chen, et al.,Exponential stabilization of
delayed recurrent neural networks: a state estimation based approach,
Neural Networks,vol. 48, pp. 153-157, 2013.
Q.Wang, X.Liu,Exponential stability of impulsive cellular neural
networks with time delay via Lyapunovfunctionals, Applied
Mathematics and Computation,vol. 194,pp. 186-198, 2007.
M.Tan, Y.Zhang, New sufficient conditions for global asymptotic
stability of Cohen-Grossberg neural networks with time-varying
delays, Nonlinear Analysis: Real World Applications,vol. 10,pp.
2139-2145, 2009.
M.Tan, Y.Zhang, W.Su, et al.,Exponential stability of neural networks
with variable delays, International Journal of Bifurcation and
Chaos,vol. 20(5),pp. 1541-1549, 2010.
M.V.Thuan, L.V.Hien, V.N.Phat,Exponential stabilization of nonautonomous delayed neural networks via Riccati equations, Applied
Mathematics and Computation,vol. 246,pp. 533-545, 2014.
Q.Zhu, J.Cao,Exponential stability analysis of stochastic reactiondiffusion CohenGrossberg neural networks with mixed delays,
Neurocomputing,vol. 74,pp. 3084-3091, 2011.

ISSN: 2394-2231

[14]

[15]

[16]

[17]

[18]

[19]

G. Zhang, X. Lin, X.Zhang,Exponential Stabilization of Neutral-Type


Neural Networks with Mixed Interval Time-Varying Delays by
Intermittent Control: A CCL Approach, Circuits systems and signal
processing,vol. 33(2), pp. 371-391, 2014.
Z.Wang,
H.Zhang,Synchronization
stability
in
complex
interconnected neural networks with nonsymmetric coupling,
Neurocomputing,vol. 108,pp. 84-92, 2013.
G.Wang, Q.Yin, Y.Shen,Exponential synchronization of coupled
fuzzy neural networks with disturbances and mixed time-delays,
Neurocomputing,vol. 106,pp. 77-85, 2013.
J.Cao, G.Chen, P.Li,Global synchronization in an array of delayed
neural networks with hybrid coupling, IEEE Transactions on Systems,
Man, and Cybernetics-Part Cybernetics-Part B: Cybernetics,vol.
38(2),pp. 488-498, 2008.
S.C.Jeong, D.H.Ji, Ju H.Park, et al.,Adaptive synchronization for
uncertain chaotic neural networks with mixed time delays using fuzzy
disturbance observer, Applied Mathematics and Computation,vol.
219,pp. 5984-5995, 2013.
G.Zhang, T.Wang, T.Li, et al.,Exponential synchronization for
delayed chaotic neural networks with nonlinear hybrid coupling,
Neurocomputing,vol. 85,pp. 53-61, 2012.
X.Huang, J.Cao,Generalized synchronization for delayed chaotic
neural networks: a novel coupling scheme, Nonlinearity,vol. 19,pp.
2797-2811, 2006.
U.Udom,Exponential stabilization of stochastic interval system with
time dependent parameters, European Journal of Operational
Research,vol. 222,pp. 523-528, 2012.
C.Hua, Q.Wang, X.Guan,Exponential stabilization controller design
for interconnected time delay systems, Automatica,vol. 44,pp. 26002606, 2008.
Y.Wang, Y.Fan, Q.Wang, Y.Zhang,Stabilization and synchronization
of complex dynamical networks with different dynamics of nodes via
decentralized controllers, IEEE Trans. Circuits Syst. I,vol. 59,pp.
1786-1795, 2012.
Y.Fan,Y. Wang, Y.Zhang, Q.Wang,The synchronization of complex
dynamical networks with similar nodes and coupling time-delay, Appl.
Math. Comput.,vol. 219,pp. 6719-6728, 2013.
M.C. Tan, W.X. Tian,Finite-time stabilization and synchronization of
complex dynamical networks with nonidentical nodes of different
dimensions, Nonlinear Dyn.,vol. 79,pp. 731-741, 2015.

http://www.ijctjournal.org

Page 147

You might also like