Skip to main content

Theory and Modern Applications

The p th moment exponential stability of stochastic cellular neural networks with impulses

Abstract

This paper studies the p th moment exponential stability of stochastic cellular neural networks with time-varying delays under impulsive perturbations. Based on the Lyapunov function, Razumikhin theory, stochastic analysis and differential inequality technique, criteria on the p th moment exponential stability of this model are derived. These results generalize and improve some of the existing ones. A numerical example illustrates the effectiveness and improvements of our results.

1 Introduction

Since Chua and Yang [1, 2] introduced a cellular neural network in 1988, it has received great attention because of its various applications such as classification of patterns, associative memories and optimization, etc. It should be pointed out that time delays are commonly encountered in real systems, which are the source of oscillation and instability both in biological and artificial neural networks, hence it is necessary and important to discuss the delayed cellular neural networks models. Up to now, many results on the stability of delayed neural networks have been developed [35]. In fact, in real nervous systems, synaptic transmission is a noisy process brought on by random fluctuations from the release of neurotransmitters and other probabilistic causes. Therefore, noise is unavoidable and should be taken into consideration in modeling. Some recent results of stochastic cellular neural networks with delays can be found in [615].

On the other hand, it is noteworthy that the state of electronic networks is often subjected to some phenomenon or other sudden noises. On that account, the electronic networks will experience some abrupt changes at certain instants that in turn affect dynamical behaviors of the systems. Therefore, it is necessary to take both stochastic effects and impulsive perturbations into account on dynamical behaviors of delayed neural networks. In recent years, the dynamic analysis of neural networks with impulsive and stochastic effects has been an attractive topic for many researchers, and a large number of stability criteria of these systems have been reported; see [3, 4, 1012, 16, 17].

In [8], Sun et al. investigated the following stochastic cellular neural networks model with time-varying delays:

d x i ( t ) = [ c i x i ( t ) + j = 1 n a i j f j ( x j ( t ) ) + j = 1 n b i j g j ( x j ( t τ j ( t ) ) ) + I i ] d t + l = 1 m σ i l ( t , x i ( t ) , x i ( t τ i ( t ) ) ) d w l ( t ) ,
(1.1)

where x i (t) denotes the potential (or voltage) of a cell i at time t; Λ={1,2,,n}, n corresponds to the number of units in a neural network; f j (), g j () are activation functions; c i >0 denotes the rate at which a cell i resets its potential to the resting state when isolated from other cells and inputs; a i j and b i j denote the strengths of connectivity between cells i and j, respectively; I i denotes the external bias on the i th unit, τ i (t) satisfies 0 τ i (t)τ and it is a transmission delay. σ(t,x,y)= ( σ i l ( t , x i , y i ) ) n × m R n × m is the diffusion coefficient matrix and σ i (t, x i , y i )=( σ i 1 (t, x i , y i ),, σ i m (t, x i , y i )) is the i th row vector of σ(t,x,y). w(t)= ( w 1 ( t ) , w 2 ( t ) , , w m ( t ) ) T is an m-dimensional Brownian motion defined on a complete probability space (Ω,F,P) with a natural filtration { F t } t 0 .

They investigated the p th moment exponential stability with the help of the method of variation parameter and inequality technique, where p2 denotes a positive constant. More precisely, they established the following fundamental assumptions:

  1. (H)

    For each j=1,2,,n, τ j (t) is a differentiable function, namely, there exists ζ such that

    τ ˙ j (t)ζ<1.

(H1) Functions f j () and g j () are Lipschitz-continuous on with Lipschitz constants L i >0, N i >0. That is, for all x,yR, iΛ,

| f i ( x ) f i ( y ) | L i |xy|, | g i ( x ) g i ( y ) | N i |xy|.

(H2) There exist nonnegative constants l i , e i , such that for all x,y, x , y R, iΛ,

[ σ i ( t , x , y ) σ i ( t , x , y ) ] [ σ i ( t , x , y ) σ i ( t , x , y ) ] T e i | x x | 2 + l i | y y | 2 .

Huang et al. [6] studied (1.1) and obtained the p th moment exponential stability by using Dini-derivative and Halanay-type inequality without assumption (H). When k 1 > k 2 , the equilibrium point of the system (1.1) is p th moment exponentially stable, where

(1.2)

where μ i (iΛ) are positive constants.

Very recently, Li [11] generalized (1.1); he considered a stochastic cellular neural network under impulsive perturbations. The condition k 1 > k 2 is also needed to ensure exponential stability in mean square.

We have a question whether the condition k 1 > k 2 in the theorems [6, 11, 12] is an essential condition or not for the equilibrium point of (1.1) to be p th moment exponentially stable.

In this paper, we solve this question and obtain the improved version of the p th moment exponential stability by applying Lyapunov functions, Razumikhin theory and inequality technique. An example is also provided to illustrate the effectiveness of the new results.

2 Preliminaries

In this paper, we study stochastic cellular neural networks with impulses described by the delayed differential equations

{ d x i ( t ) = [ c i x i ( t ) + j = 1 n a i j f j ( x j ( t ) ) + j = 1 n b i j g j ( x j ( t τ j ( t ) ) ) + I i ] d t d x i ( t ) = + l = 1 m σ i l ( t , x i ( t ) , x i ( t τ i ( t ) ) ) d w l ( t ) , t t k , Δ x i ( t k ) = p i k ( x ( t k ) ) = x i ( t k ) x i ( t k ) , k Z + , i Λ ,
(2.1)

where { t k } is the time sequence and satisfies 0= t 0 < t 1 < t 2 << t k < t k + 1  , lim k t k =; x t (s)=x(t+s), s[τ,0]. For k=1,2, , p i k (x( t k )) represents the abrupt change of the state x i (t) at the impulsive moments t k .

System (2.1) is supplemented with the initial condition given by

x t 0 (s)=ψ(s),s[τ,0],
(2.2)

where ψ(s) is F 0 -measurable and continuous everywhere except at a finite number of points t k , at which ψ( t k + ) and ψ( t k ) exist and ψ( t k + )=ψ( t k ).

Let P C 1 , 2 ([ t k , t k + 1 )× R n ; R + ) denote the family of all nonnegative functions V(t,x) on [ t k , t k + 1 )× R n which are continuous once differentiable in t and twice differentiable in x. If V(t,x)P C 1 , 2 ([ t k , t k + 1 )× R n ; R + ), define an operator LV associated with (2.1) as

L V ( t , x ) = V t ( t , x ) + i = 1 n V x i ( t , x ) [ c i x i ( t ) + j = 1 n a i j f j ( x j ( t ) ) + j = 1 n b i j g j ( x j ( t τ j ( t ) ) ) + I i ] + 1 2 trace [ σ T V x x ( t , x ) σ ] ,
(2.3)

where

V t (t,x)= V ( t , x ) t , V x i (t,x)= V ( t , x ) x i , V x x (t,x)= ( V ( t , x ) x i x j ) n × n .

Throughout this paper, the following standard hypothesis is needed:

(H3) p i ( x i ( t k ))= β i k ( x i ( t k ) x i ), where x i is the equilibrium point of (2.1) with the initial condition (2.2), β i k satisfies |1 β i k | d k , d k is a positive constant.

Let y i (t)= x i (t) x , then (2.1) can be written by

{ d y i ( t ) = [ c i y i ( t ) + j = 1 n a i j f ˜ j ( y j ( t ) ) + j = 1 n b i j g ˜ j ( y j ( t τ j ( t ) ) ) ] d t d y i ( t ) = + l = 1 m σ ˜ i l ( t , y i ( t ) , y i ( t τ i ( t ) ) ) d w l ( t ) , t t k , Δ y i ( t k ) = p ˜ i k ( y ( t k ) ) , k Z + , i Λ ,
(2.4)

where

f ˜ j ( y j ( t ) ) = f j ( y j ( t ) + x j ) f j ( x j ) , g ˜ j ( y j ( t τ j ( t ) ) ) = g j ( y j ( t τ j ( t ) ) + x j ) g j ( x j ) , σ ˜ i j ( t , y i ( t ) , y i ( t τ i ( t ) ) ) = σ i j ( t , y i ( t ) + x i , y i ( t τ i ( t ) ) + x i ) σ i j ( t , x i , x i ) , p ˜ i k ( y ( t k ) ) = p i k ( y ( t k ) + x ) p i k ( x ) .

In the following, for further study, we first give the following definitions and lemmas.

x denotes a vector norm defined by

x p = i = 1 n | x i ( t ) | p , ψ τ = sup s [ τ , 0 ] ψ ( s ) .

Definition 2.1 (Mao [18])

The equilibrium point x = ( x 1 , x 2 , , x n ) T of the system (2.1) is said to be p th moment exponentially stable if there exist λ>0 and M>0 such that

E x ( t ) x p M x 0 x p e λ ( t t 0 ) ,t t 0 , x 0 R n ,x( t 0 )= x 0 .

In such a case,

lim sup t 1 t lnE ( x ( t ) x p ) λ.
(2.5)

The right-hand side of (2.5) is commonly known as the p th moment Lyapunov exponent of this solution.

When p=2, it is usually said to be exponentially stable in mean square.

Lemma 2.2 If a i (i=1,2,,p) denote p nonnegative real numbers, then

a 1 a 2 a p a 1 p + a 2 p + + a p p p ,
(2.6)

where p1 denotes an integer.

A particular form of (2.6), namely

a 1 p 1 a 2 ( p 1 ) a 1 p p + a 2 p p ,for p=1,2,3,.

3 Main result

Theorem 3.1 Assume that (H1)-(H3) hold; furthermore, let

  1. (i)

    there exist σ>0, λ>0 such that k 1 + k 2 e λ τ d k 1 p σλ;

  2. (ii)

    pln d k 1 <(σ+λ)( t k t k 1 ), kN, then the equilibrium point of (2.1) is pth moment exponentially stable.

Proof We define a Lyapunov function V(t,y(t))= i = 1 n | y i ( t ) | p = y ( t ) p . Let t t 0 and t[ t k 1 , t k ), then we can get the operator LV(t,y) associated with the system (2.4) of the form as follows:

L V ( t , y ) = p i = 1 n | y i ( t ) | p 1 sgn ( y i ( t ) ) [ c i y i ( t ) + j = 1 n a i j f ˜ j ( y j ( t ) ) + j = 1 n b i j g ˜ j ( y j ( t τ j ( t ) ) ) ] + p ( p 1 ) 2 i = 1 n | y i ( t ) | p 2 l = 1 m σ ˜ i l 2 ( t , y i ( t ) , y i ( t τ i ( t ) ) ) p i = 1 n c i | y i ( t ) | p + p i = 1 n j = 1 n | a i j | L j | y i ( t ) | p 1 | y j ( t ) | + p i = 1 n j = 1 n | b i j | N j | y i ( t ) | p 1 | y j ( t τ j ( t ) ) | + p ( p 1 ) 2 i = 1 n | y i ( t ) | p 2 ( e i | y i ( t ) | 2 + l i | y i ( t τ i ( t ) ) | 2 ) p i = 1 n c i | y i ( t ) | p + i = 1 n j = 1 n | a i j | L j ( ( p 1 ) | y i ( t ) | p + | y j ( t ) | p ) + i = 1 n j = 1 n | b i j | N j ( ( p 1 ) | y i ( t ) | p + | y j ( t τ j ( t ) ) | p ) + p ( p 1 ) 2 i = 1 n e i | y i ( t ) | p + ( p 1 ) 2 i = 1 n l i ( ( p 2 ) | y i ( t ) | p + 2 | y i ( t τ i ( t ) ) | p ) = i = 1 n [ p c i ( p 1 ) j = 1 n ( L j | a i j | + N j | b i j | ) j = 1 n | a j i | L i p ( p 1 ) 2 e i ] | y i ( t ) | p i = 1 n ( p 1 ) ( p 2 ) 2 l i | y i ( t ) | p + i = 1 n [ N i j = 1 n | b j i | + ( p 1 ) l i ] | y i ( t τ i ( t ) ) | p k 1 V ( t , y ( t ) ) + k 2 sup t τ s t V ( s , y ( s ) ) ,
(3.1)

where

Let γ= inf k Λ 1 d k 1 p , there exist σ>0, λ>0 such that

k 1 + k 2 γ e λ τ k 1 + k 2 e λ τ d k 1 p σλ
(3.2)

and

lnγ+λτ(σ+λ)( t k t k 1 )>0.
(3.3)

Hence, we can choose M1 such that

e ( σ + λ ) ( t 1 t 0 ) Mγ e λ τ .
(3.4)

For convenience, we denote that ϕ(s)=ψ(s) x for s[τ,0).

It is obvious that

ϕ τ p ϕ τ p e σ ( t 1 t 0 ) M ϕ τ p e λ ( t 1 t 0 ) .
(3.5)

Now, we should prove

E y ( t ) p M ϕ τ p e λ ( t t 0 ) ,t t 0 .
(3.6)

Firstly, we prove when t[ t 0 , t 1 ),

EV ( t , y ( t ) ) =E y ( t ) p M ϕ τ p e λ ( t 1 t 0 ) M ϕ τ p e λ ( t t 0 ) .
(3.7)

If (3.7) is not true, there exists t ¯ [ t 0 , t 1 ) such that

E V ( t ¯ , y ( t ¯ ) ) > M ϕ τ p e λ ( t 1 t 0 ) > ϕ τ p e σ ( t 1 t 0 ) > ϕ τ p E V ( t 0 + s , y ( t 0 + s ) ) , s [ τ , 0 ] .
(3.8)

Since V(t,y(t)) is continuous on [ t 0 , t 1 ), which implies that there exists t ˆ [ t 0 , t ¯ ) such that

EV ( t ˆ , y ( t ˆ ) ) =M ϕ τ p e λ ( t 1 t 0 )

and

EV ( t , y ( t ) ) EV ( t ˆ , y ( t ˆ ) ) ,t[ t 0 τ, t ˆ ),

then there exists some t ˜ [ t 0 , t ˆ ) satisfying

EV ( t ˜ , y ( t ˜ ) ) = ϕ τ p

and

EV ( t , y ( t ) ) EV ( t ˜ , y ( t ˜ ) ) ,t( t ˜ , t ˆ ).

Hence, for any s[τ,0], t( t ˜ , t ˆ ),

EV ( t + s , y ( t + s ) ) EV ( t ˆ , y ( t ˆ ) ) <γ e λ τ ϕ τ p e λ ( t 1 t 0 ) <γ e λ τ ϕ τ p γ e λ τ EV ( t , y ( t ) ) .

By (3.1) and (3.2), we get

ELV(t,y) ( k 1 + k 2 γ e λ τ ) EV ( t , y ( t ) ) (σλ)EV ( t , y ( t ) ) ,t( t ˜ , t ˆ ).

Then

EV ( t ˆ , y ( t ˆ ) ) EV ( t ˜ , y ( t ˜ ) ) e ( σ λ ) ( t ˆ t ˜ ) < ϕ τ p e σ ( t 1 t 0 ) M ϕ τ p e λ ( t 1 t 0 ) =EV ( t ˆ , y ( t ˆ ) ) ,

which is a contradiction. Hence, (3.7) holds.

Next, we will show

EV ( t , y ( t ) ) M ϕ τ p e λ ( t t 0 ) ,t[ t k 1 , t k ),kΛ.
(3.9)

Assuming (3.9) holds for k=1,2,,m, we shall show that it holds for k=m+1, i.e.,

EV ( t , y ( t ) ) =E y ( t ) p M ϕ τ p e λ ( t t 0 ) ,t[ t m , t m + 1 ).
(3.10)

Suppose (3.10) is not true. Then we define t ¯ [ t m , t m + 1 ) such that

EV ( t ¯ , y ( t ¯ ) ) >M ϕ τ p e λ ( t t 0 ) .

From (H3) we get

E V ( t m , y ( t m ) ) = i = 1 n E | y i ( t m ) + p ˜ i ( y i ( t m ) + x i ) | p = E i = 1 n | 1 β i m | p | y i ( t m ) | p E i = 1 n d m p | y i ( t m ) | p = d m p E V ( t m , y ( t m ) ) d m p M ϕ τ p e λ ( t m t 0 ) = d m p M ϕ τ p e λ ( t ¯ t m ) e λ ( t ¯ t 0 ) < d m p M ϕ τ p e λ ( t m + 1 t m ) e λ ( t ¯ t 0 ) < M ϕ τ p e λ ( t ¯ t 0 ) M ϕ τ p e λ ( t m t 0 ) ,

which implies t ¯ ( t m , t m + 1 ). Let

t ¯ =inf { t ( t m , t m + 1 ) : E V ( t ¯ , y ( t ¯ ) ) = M ϕ τ p e λ ( t ¯ t 0 ) } ,

then for t( t m τ, t ¯ ), we can get EV(t,y(t))EV( t ¯ ,y( t ¯ )).

Hence, there exists t ( t m , t ¯ ) such that

EV ( t , y ( t ) ) = d m p M ϕ τ p e λ ( t ¯ t 0 ) e λ ( t m + 1 t m )

and

EV ( t , y ( t ) ) EV ( t , y ( t ) ) ,t[ t , t ¯ ).

On the other hand, for any t[ t , t ¯ ), s[τ,0], either t+s[ t m τ, t m ) or t+s[ t m , t ¯ ].

If t+s[ t m τ, t m ), we can obtain

EV ( t + s , y ( t + s ) ) M ϕ τ p e λ ( t + s t 0 ) <M ϕ τ p e λ τ e λ ( t ¯ t 0 ) e λ ( t m + 1 t m ) = e λ τ d m p EV ( t , y ( t ) ) .

If t+s[ t m , t ¯ ], we can get

EV ( t + s , y ( t + s ) ) EV ( t ¯ , y ( t ¯ ) ) < e λ τ d m p EV ( t , y ( t ) ) .

Then for any s[τ,0], we get

EV ( t + s , y ( t + s ) ) e λ τ d m p EV ( t , y ( t ) ) e λ τ d m p EV ( t , y ( t ) ) ,t[ t , t ¯ ).

Hence,

LV(t,y) ( k 1 + k 2 e λ τ d m p ) EV ( t , y ( t ) ) (σλ)EV ( t , y ( t ) ) .

Then

EV ( t ¯ , y ( t ¯ ) ) EV ( t , y ( t ) ) e ( σ λ ) ( t ¯ t ) = d m p M ϕ τ p e λ ( t ¯ t 0 ) e λ ( t m + 1 t m ) e ( σ λ ) ( t ¯ t ) .

From the condition (ii), it is obvious that

d m p e λ ( t m + 1 t m ) e ( σ λ ) ( t ¯ t ) <1.

Hence,

EV ( t ¯ , y ( t ¯ ) ) <M ϕ τ p e λ ( t ¯ t 0 ) =EV ( t ¯ , y ( t ¯ ) ) ,

which is a contradiction. Hence, (3.10) holds.

By induction, we can obtain that (3.9) holds for any kΛ, i.e.,

EV ( t , y ( t ) ) M ϕ τ p e λ ( t t 0 ) ,t t 0 ,

which implies that the equilibrium point of the impulsive system (2.1) is p th moment exponentially stable. This completes the proof of the theorem. □

Theorem 3.2 Assume that (H1)-(H3) hold, μ i (iΛ)>0,

  1. (i)

    if there exist σ>0, λ>0 such that k 1 + k 2 e λ τ d k 1 p σλ;

  2. (ii)

    pln d k 1 <(σ+λ)( t k t k 1 ), kN,

where

then the equilibrium point of the system (2.1) is pth moment exponentially stable.

Proof Let V(t,y(t))= i = 1 n μ i | y i ( t ) | p , the proof of the theorem is similar to that of Theorem 3.1 hence it is omitted. □

Corollary 3.3 Assume that (H1)-(H3) hold, μ i (iΛ)>0,

k 1 = min 1 i n { 2 c i j = 1 n ( L j | a i j | + N j | b i j | ) j = 1 n μ j μ i | a j i | L i e i } > 0 , k 2 = max 1 i n { N i j = 1 n μ j μ i | b j i | + l i } ,
  1. (i)

    if there exist σ>0, λ>0 such that k 1 + k 2 e λ τ d k 1 2 σλ;

  2. (ii)

    2ln d k 1 <(σ+λ)( t k t k 1 ), kN,

then the equilibrium point of the system (2.1) is exponentially stable in mean square.

Remark 3.4 In many stability results for stochastic cellular neural networks, ELV0 is an important condition for their conclusions [1315], which means that the origin systems without impulses need to be stable. However, by constructing the impulses, we do not need this condition to ensure the equilibrium point of the impulsive system (2.1) is p th moment exponentially stable. Our results show that impulses play an important role in the p th moment exponential stability for the stochastic cellular neural network with time delay, even if the corresponding systems may be unstable themselves. It should be mentioned that our results develop an effective impulse control strategy to stabilize underlying retarded cellular neural networks. And it is particularly meaningful for some practical applications.

Remark 3.5 It is important to emphasize that, in contrast to some existing exponential stability results, see [6, 11, 12, 19], the condition k 1 > k 2 is needed to ensure the equilibrium point of the system (2.1) is p th moment exponentially stable, while in our paper we omit it and obtain the results.

4 Illustrative example

In the following, we will give an example to illustrate the advantages of our results.

Example 1 Consider the following model:

{ d x i ( t ) = [ c i x i ( t ) + j = 1 2 a i j f j ( x j ( t ) ) + j = 1 2 b i j g j ( x j ( t τ j ( t ) ) ) ] d t d x i ( t ) = + σ i ( t , x i ( t ) , x i ( t τ i ( t ) ) ) d w ( t ) , t t k , x i ( t k ) = ( 1 β i k ) x i ( t k ) k Z + , i Λ = ( 1 , 2 ) ,
(4.1)

where f i (x)= g i (x)=tanh(x), 0 τ i (t)τ=0.5, t k t k 1 =0.1, x 1 ( t k )= x 1 ( t k ) 3 , x 2 ( t k )= x 2 ( t k ) 4 . C 2 × 1 = ( 2 2 ) , ( a i j ) 2 × 2 = ( 0.5 0.8 0.3 0.6 ) , ( b i j ) 2 × 2 = ( 0.2 0.3 0.1 0.4 ) .

Obviously, L i = N i =1, e i =0, l i =1 (i=1,2), β 1 k =2/3, β 2 k =3/4.

Let d k =2/3, σ=3.4, λ=0.2, μ 1 = μ 2 . Then for p=2, we can get k 1 =min(1.4,1.2)=1.2>0, k 2 =max(1.3,1.7)=1.7, k 1 + k 2 e λ τ d k p =1.2+1.7× e 0.2 × 0.5 ( 2 / 3 ) 2 =3.027<σλ=3.40.2=3.2, 2ln d k 1 =2ln(2/3)=0.811<(σ+λ)( t k t k 1 )=3.6×0.1=0.36.

All conditions of Corollary 3.3 are satisfied, then the equilibrium point is exponentially stable in mean square.

Remark 4.1 If μ 1 = μ 2 , we have computed k 1 =1.2, k 2 =1.7. If μ 1 μ 2 , set μ 2 μ 1 =α, then k 1 =min{1.70.3α,20.8/α}, k 2 =max{1.2+0.1α,1.4+0.3/α}. If μ 1 < μ 2 , then α>1, we can compute k 1 <1.4 and k 2 >1.4; if μ 1 > μ 2 , then 0<α<1, k 1 <1.2 and k 2 >1.4. Hence, in either case, we always have k 1 < k 2 , so the exponential stability in mean square of the system (4.1) cannot be derived by applying the corresponding exponential stability result for cellular neural networks given in the literature [6, 11, 12, 19], since k 1 > k 2 is not satisfied.

Remark 4.2 Since ρ[ C 1 (M M 1 K+M M 2 K+N N 1 +N N 2 )]=33.635, where ρ[ C 1 (M M 1 K+M M 2 K+N N 1 +N N 2 )] was defined in [8], the condition ρ[ C 1 (M M 1 K+M M 2 K+N N 1 +N N 2 )]1 is not satisfied. Hence, the results in [8] are useless to judge the exponential stability of the system (4.1).

References

  1. Chua LO, Yang L: Cellular neural networks: theory. IEEE Trans. Circuits Syst. 1988, 35: 1257-1272. 10.1109/31.7600

    Article  MathSciNet  Google Scholar 

  2. Chua LO, Yang L: Cellular neural networks: applications. IEEE Trans. Circuits Syst. 1988, 35: 1273-1290. 10.1109/31.7601

    Article  MathSciNet  Google Scholar 

  3. Wu B, Liu Y, Lu J: New results on global exponential stability for impulsive cellular neural networks with any bounded time-varying delays. Math. Comput. Model. 2012, 55: 837-843. 10.1016/j.mcm.2011.09.009

    Article  MathSciNet  Google Scholar 

  4. Ahmada S, Stamova IM: Global exponential stability for impulsive cellular neural networks with time-varying delays. Nonlinear Anal. 2008, 69: 786-795. 10.1016/j.na.2008.02.067

    Article  MathSciNet  Google Scholar 

  5. Zhao H, Cao J: New conditions for global exponential stability of cellular neural networks with delays. Neural Netw. 2005, 18: 1332-1340. 10.1016/j.neunet.2004.11.010

    Article  Google Scholar 

  6. Huang C, He Y, Huang L, Zhu W: p th moment stability analysis of stochastic recurrent neural networks with time-varying delays. Inf. Sci. 2008, 178: 2194-2203. 10.1016/j.ins.2008.01.008

    Article  MathSciNet  Google Scholar 

  7. Huang C, Cao J: Almost sure exponential stability of stochastic cellular neural networks with unbounded distributed delays. Neurocomputing 2009, 72: 3352-3356. 10.1016/j.neucom.2008.12.030

    Article  Google Scholar 

  8. Sun Y, Cao J: p th moment exponential stability of stochastic recurrent neural networks with time-varying delays. Nonlinear Anal., Real World Appl. 2007, 8: 1171-1185. 10.1016/j.nonrwa.2006.06.009

    Article  MathSciNet  Google Scholar 

  9. Wu S, Li C, Liao X, Duan S: Exponential stability of impulsive discrete systems with time delay and applications in stochastic neural networks: a Razumikhin approach. Neurocomputing 2012, 82: 29-36.

    Article  Google Scholar 

  10. Chen L, Wu R, Pan D: Mean square exponential stability of impulsive stochastic fuzzy cellular neural networks with distributed delays. Expert Syst. Appl. 2011, 38: 6294-6299. 10.1016/j.eswa.2010.11.070

    Article  Google Scholar 

  11. Li X: Existence and global exponential stability of periodic solution for delayed neural networks with impulsive and stochastic effects. Neurocomputing 2010, 73: 749-758. 10.1016/j.neucom.2009.10.016

    Article  Google Scholar 

  12. Li X, Fu X: Synchronization of chaotic delayed neural networks with impulsive and stochastic perturbations. Commun. Nonlinear Sci. Numer. Simul. 2011, 16: 885-894. 10.1016/j.cnsns.2010.05.025

    Article  MathSciNet  Google Scholar 

  13. Mao X: Exponential Stability of Stochastic Differential Equations. Marcel Dekker, New York; 1994.

    Google Scholar 

  14. Li X, Fu X: Stability analysis of stochastic functional differential equations with infinite delay and its application to recurrent neural networks. J. Comput. Appl. Math. 2010, 234: 407-417. 10.1016/j.cam.2009.12.033

    Article  MathSciNet  Google Scholar 

  15. Yang Z, Zhu E, Xu Y, Tan Y: Razumikhin-type theorems on exponential stability of stochastic functional differential equations with infinite delay. Acta Appl. Math. 2010, 111: 219-231. 10.1007/s10440-009-9542-1

    Article  MathSciNet  Google Scholar 

  16. Peng S, Zhang Y: Razumikhin-type theorems on p th moment exponential stability of impulsive stochastic delay differential equations. IEEE Trans. Autom. Control 2010, 55: 1917-1922.

    Article  MathSciNet  Google Scholar 

  17. Li X, Zou J, Zhu E: P th moment exponential stability of impulsive stochastic neural networks with mixed delays. Math. Probl. Eng. 2012., 2012: Article ID 175934. doi:10.1155/2012/175934

    Google Scholar 

  18. Mao X: Stochastic Differential Equations and Applications. Horwood Publication, Chichester; 1997.

    Google Scholar 

  19. Wan L, Zhou Q: Attractor and ultimate boundedness for stochastic cellular neural networks with delays. Nonlinear Anal., Real World Appl. 2011, 12: 2561-2566. 10.1016/j.nonrwa.2011.03.005

    Article  MathSciNet  Google Scholar 

Download references

Acknowledgements

The authors would like to thank the editor and the anonymous referees for their detailed comments and valuable suggestions which considerably improved the presentation of this paper. This work was supported in part by the National Natural Science Foundation of China under Grant No. 11101054, the Hunan Provincial Natural Science Foundation of China under Grant No. 12jj4005, the Open Fund Project of Key Research Institute of Philosophies and Social Sciences in Hunan Universities under Grant No. 11FEFM11 and the Scientific Research Funds of Hunan Provincial Science and Technology Department of China under Grant No. 2012SK3096.

Author information

Authors and Affiliations

Authors

Corresponding authors

Correspondence to Xiaoai Li or Enwen Zhu.

Additional information

Competing interests

The authors declare that they have no competing interests.

Authors’ contributions

XL completed the proof and wrote the initial draft. JZ gave some suggestions on the amendment. XL then finalized the manuscript. Correspondence was mainly done by EZ. All authors read and approved the final manuscript.

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 2.0 International License (https://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Reprints and permissions

About this article

Cite this article

Li, X., Zou, J. & Zhu, E. The p th moment exponential stability of stochastic cellular neural networks with impulses. Adv Differ Equ 2013, 6 (2013). https://doi.org/10.1186/1687-1847-2013-6

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/1687-1847-2013-6

Keywords