Abstract
In this paper, based on the exponential dichotomy theory, contraction mapping fixed point theorem and inequality analysis technique, we obtain some sufficient conditions ensuring the existence and global exponential stability of pseudo almost periodic solutions for a new generalized cellular neural network model with continuously distributed leakage delays. Our results complement with some recent ones. Moreover, an illustrative example and its numerical simulation are given to demonstrate the effectiveness of the obtained results.
Similar content being viewed by others
Explore related subjects
Discover the latest articles, news and stories from top researchers in related subjects.Avoid common mistakes on your manuscript.
1 Introduction
In the last three decades, the dynamical behaviors of delayed cellular neural networks (DCNNs) have received much attention due to their potential applications in associated memory, parallel computing,pattern recognition, signal processing and optimization problems (see [1–5]). In particular, a neural network usually has a spatial nature due to the presence of an amount of parallel pathways of a variety of axon sizes and lengths, it is desired to model them by introducing continuously distributed delays over a certain duration of time (see [6–8]). Recently, a typical time delay called Leakage (or “forgetting”) delay may exist in the negative feedback terms of the neural network system, and these terms are variously known as forgetting or leakage terms (see[9–13]). Consequently, the dynamic behaviors of cellular neural networks with continuously distributed leakage delays have been extensively and intensively studied. We refer the reader to [14–18] and the references cited therein.
On the other hand, the variation of the environment plays an important role in the dynamics of DCNNs. As pointed out in [19, 20], periodically varying environment and almost periodically varying environment are foundations for the theory of nature selection. Compared with periodic effects, almost periodic effects are more frequent, and many phenomena exhibit great regularity with being pseudo almost periodic which allow complex repetitive phenomena to be represented as an almost-periodic process plus an ergodic component. Therefore, many researchers have focused their attention on the study of existence and stability of almost periodic solutions and pseudo almost periodic solutions for DCNNs (see [21–25] and the references cited therein). Most recently, Zhang [26] considered the following CNNs with continuously distributed delays in leakage terms:
in which \(n\) corresponds to the number of units in a neural network, \(x_{i}(t)\) corresponds to the state vector of the \(i\)th unit at the time \(t,\, c_{i}(t)>0\) represents the rate with which the \(i\)th unit will reset its potential to the resting state in isolation when disconnected from the network and external inputs at the time \(t\). \(a_{ij}(t)\) and \( b_{ij}(t) \) are the connection weights at the time \(t,\, h_{i}(s)\ge 0,\, K_{ij}(u) \) and \(\tau _{ij}(t)\ge 0\) denote the leakage delay kernel, transmission delay kernel and transmission delay, respectively, and \(I_{i}(t)\) denotes the external inputs at time \(t\). \(f_{j}\) and \( g_{j}\) are activation functions of signal transmission, \(i,j=1, 2, \ldots , n\).
By using Lyapunov functional method and differential inequality techniques, in [26], it has been established some sufficient conditions to guarantee that all solutions of (1.1) converge exponentially to the almost periodic solution. Moreover, it is well known that the global exponential exponential convergence behavior of solutions plays a key role in characterizing the behavior of dynamical system since the exponential convergent rate can be unveiled (see [27–30]). However, to the best of our knowledge, few authors have considered the exponential convergence on the pseudo almost periodic solution for (1.1). Motivated by the above discussions, in this paper, we shall establish the existence and uniqueness of pseudo almost periodic solution of (1.1) by using the exponential dichotomy theory and contraction mapping fixed point theorem. Meanwhile, we also shall give the conditions to guarantee that all solutions and their derivatives of solutions for (1.1) converge exponentially to the pseudo almost periodic solution and its derivative, respectively.
The initial conditions associated with system (1.1) are of the form
where \(\varphi _{i}(\cdot )\) and \(\varphi _{i}'(\cdot )\) are real-valued bounded continuous functions defined on \((-\infty , 0]\).
In the following part of this paper, given a bounded continuous function \(g\) defined on \(\mathbb {R}\), let \(g^+\) and \(g^-\) be defined as
For convenience, we denote by \(\mathbb {R}^{n}(\mathbb {R}=\mathbb {R}^{1})\) the set of all \(n-\)dimensional real vectors (real numbers). We will use \(x=(x_{1}, \ x_{2}, \, \ldots , x_{n})^{T}\in \mathbb {R}^{n}\) to denote a column vector, in which the symbol \((^{T})\) denotes the transpose of a vector. We let \(|x|\) denote the absolute-value vector given by \(|x|=(|x_{1}|, \, |x_{2}|,\ldots , |x_{n}|)^{T} \), and define \(\ \Vert x\Vert =\max \limits _{1\le i\le n}|x_{i}|\). A matrix or vector \(A\ge 0\) means that all entries of \(A\) are greater than or equal to zero. \(A>0\) can be defined similarly. For matrices or vectors \(A\) and \(B,\, A\ge B\) (resp. \(A>B\)) means that \(A-B\ge 0\) (resp. \(A-B>0\)).
The paper is organized as follows. Section 2 includes some lemmas and definitions, which can be used to check the existence of pseudo almost periodic solutions of (1.1). In Sect. 3, we present some new sufficient conditions for the existence and uniqueness of the continuously differentiable pseudo almost periodic solution of (1.1). In Sect. 4, we establish sufficient conditions on the global exponential stability of pseudo almost periodic solutions of (1.1). At last, an example and its numerical simulation are given to illustrate the effectiveness of the obtained results.
2 Preliminary Results
In this section, we shall first recall some basic definitions, lemmas which are used in what follows.
In this paper, \(BC(\mathbb {R},\mathbb {R}^{n})\) denotes the set of bounded continued functions from \(\mathbb {R}\) to \(\mathbb {R}^{n}\). Note that \((BC(\mathbb {R},\mathbb {R}^{n}), \Vert \cdot \Vert _{\infty })\) is a Banach space where \(\Vert \cdot \Vert _{\infty }\) denotes the sup norm \(\Vert f\Vert _{\infty } := \sup \limits _{ t\in \mathbb {R}} \Vert f (t)\Vert \).
Definition 2.1
(see [19, 20]) Let \(u(t)\in BC(\mathbb {R},\mathbb {R}^{n})\). \(u(t)\) is said to be almost periodic on \(\mathbb {R}\) if, for any \(\varepsilon >0\), the set \(T(u,\varepsilon )= \{\delta :\Vert u(t+\delta )-u(t)\Vert <\varepsilon \; \hbox {for all } \; t\in \mathbb {R}\}\) is relatively dense, i.e., for any \( \varepsilon >0\), it is possible to find a real number \(l=l(\varepsilon )>0,\) for any interval with length \(l(\varepsilon )\), there exists a number \(\delta =\delta (\varepsilon )\) in this interval such that \(\Vert u(t+\delta )-u(t)\Vert <\varepsilon , \, \hbox {for all }\, t\in \mathbb {R}.\)
We denote by \(AP(\mathbb {R},\mathbb {R}^{n})\) the set of the almost periodic functions from \(\mathbb {R}\) to \(\mathbb {R}^{n}\). Besides, the concept of pseudo almost periodicity (pap) was introduced by Zhang in the early nineties. It is a natural generalization of the classical almost periodicity. Precisely, define the class of functions \(PAP_{0}(\mathbb {R},\mathbb {R}^{n})\) as follows:
A function \(f\in {BC(\mathbb {R},\mathbb {R}^{n})}\) is called pseudo almost periodic if it can be expressed as
where \(h\in {AP(\mathbb {R},\mathbb {R}^{n})}\) and \(\varphi \in {PAP_{0}(\mathbb {R},\mathbb {R}^{n})}.\) The collection of such functions will be denoted by \(PAP(\mathbb {R},\mathbb {R}^{n}).\) The functions \(h\) and \(\varphi \) in above definition are respectively called the almost periodic component and the ergodic perturbation of the pseudo almost periodic function \(f\). The decomposition given in definition above is unique. Observe that \((PAP(\mathbb {R},\mathbb {R}^{n}),\Vert .\Vert _{\infty })\) is a Banach space and \(AP(\mathbb {R},\mathbb {R}^{n})\) is a proper subspace of \(PAP(\mathbb {R},\mathbb {R}^{n})\) since the function \(\phi (t)=\cos \pi t+\cos t+e^{-t^{4}\sin ^{2 }t}\) is pseudo almost periodic function but not almost periodic. It should be mentioned that pseudo almost periodic functions possess many interesting properties; we shall need only a few of them and for the proofs we shall refer to [19].
Lemma 2.1
(see [19, p. 57]) If \(f\in PAP(\mathbb {R},\mathbb {R})\) and \(g\) is its almost periodic component, then we have
Therefore \(\Vert f\Vert _{\infty }\ge \Vert g\Vert _{\infty }\ge \inf \limits _{x\in \mathbb {R}}|g(x)|\ge \inf \limits _{x\in \mathbb {R}}|f(x)|\).
Lemma 2.2
(see [19, p. 140]) Suppose that both functions \(f\) and its derivative \(f'\) are in \( PAP(\mathbb {R},\mathbb {R})\). That is, \(f=g+\varphi \) and \(f'=\alpha +\beta \), where \(g,\alpha \in AP(\mathbb {R},\mathbb {R})\) and \(\varphi ,\beta \in PAP_{0}(\mathbb {R},\mathbb {R})\). Then the functions \(g\) and \(\varphi \) are continuous differentiable so that
Lemma 2.3
Let \(B^{*}=\{f|f,f'\in PAP(\mathbb {R},\mathbb {R}) \}\) equipped with the induced norm defined by \(\Vert f\Vert _{B^{*}}=\max \{\Vert f\Vert _{\infty }, \Vert f'\Vert _{\infty }\}=\max \{\sup \limits _{t\in \mathbb {R}}|f(t)|, \sup \limits _{t\in \mathbb {R}}|f'(t)|\}\), then \(B^{*}\) is a Banach space.
Proof
Suppose that \(\{f_p\}_{p=1}^{+\infty } \) is a Cauchy sequence in \(B^{*}\), then for any \( \varepsilon >0\), there exists \( ~N(\varepsilon )>0\) such that
By the definition of pseudo almost periodic function, let
From Lemma 2.2, we obtain
On combining (2.1) with Lemma 2.1, we deduce that, \(\{g_p\}_{p=1}^{+\infty },\{g_p'\}_{p=1}^{+\infty }\subset AP(\mathbb {R},\mathbb {R})\) are Cauchy sequence, so that \(\{\varphi _p\}_{p=1}^{+\infty }, \{\varphi _p'\}_{p=1}^{+\infty }\subset PAP_0(\mathbb {R},\mathbb {R})\) are also Cauchy sequence.
Firstly, we show that there exists \(g\in AP(\mathbb {R},\mathbb {R})\) such that \(g_p \) uniformly converges to \( g \), as \(p\rightarrow +\infty \).
Note that \(\{g_p\}\) is Cauchy sequence in \(AP(\mathbb {R},\mathbb {R})\). \( \forall ~ \varepsilon >0\), \(\exists ~N(\varepsilon )\), such that \(\forall ~ p,q\ge N(\varepsilon )\)
So for fixed \(t\in \mathbb {R}\), it is easy to see \(\{g_p(t)\}_{p=1}^{+\infty }\) is Cauchy number sequence. Thus, the limits of \(g_p(t)\) exists as \(p\rightarrow +\infty \) and let \(g(t)=\lim \limits _{p\rightarrow +\infty }g_p(t) \). In (2.2), let \(q\rightarrow +\infty \), we have
Thus, \(g_p \) uniformly converges to \( g \), as \(p\rightarrow +\infty \). Moreover, from the theorem 1.9 [20, p. 5], we obtain \(g\in AP(\mathbb {R},\mathbb {R})\). Similarly, we also obtain that there exist \(g ^{*}\in AP (\mathbb {R},\mathbb {R})\) and \(\varphi ,\varphi ^{*}\in BC(\mathbb {R},\mathbb {R})\), such that
which imply that
where \(p\rightarrow +\infty \) and \(``\Rightarrow ''\) means uniform convergence.
Next, we claim that \(\varphi ,\varphi ^{*}\in PAP_0(R)\). Together with (2.4) and the facts that
and
we have
Hence \(\varphi ,\varphi ^{*}\in PAP_0(R)\). Let \(f=g+\varphi \), \( f^{*}=g^{*}+\varphi ^{*}\), then \(f=g+\varphi \in PAP(R)\), \(f^{*}=g^{*}+\varphi ^{*}\in PAP(R)\) and \(f_p \Rightarrow f \), \(f_p' \Rightarrow f^{*}\) as \(p\rightarrow +\infty \).
Finally, we reveal \(f' = f^{*}\). For \(t,\Delta t\in R\), it follows that
In view of the uniform convergence of \(f_p\) and \(f_p'\), let \(p\rightarrow +\infty \) for (2.5), we get
which implies that
In summary, in view of (2.3), (2.4) and (2.6), we obtain that the Cauch sequence \(\{f_p\}_{p=1}^{+\infty }\subset B^{*}\) satisfies
and \(f\in B^{*}\). This yields that \(B^{*} \) is a Banach space. The proof is completed.\(\square \)
Remark 2.1
Let \(B=\{f|f,f'\in PAP(\mathbb {R},\mathbb {R}^{n}) \}\) equipped with the induced norm defined by \(\Vert f\Vert _{B}=\max \{\Vert f\Vert _{\infty }, \Vert f'\Vert _{\infty }\}=\max \{\sup \limits _{t\in \mathbb {R}}\Vert f(t)\Vert , \sup \limits _{t\in \mathbb {R}}\Vert f'(t)\Vert \}\). It follows from Lemma 2.3 that \(B\) is a Banach space.
Definition 2.2
(see [19, 20]) Let \(x\in \mathbb {R}^{n}\) and \(Q(t)\) be a \(n\times n\) continuous matrix defined on \(\mathbb {R}\). The linear system
is said to admit an exponential dichotomy on \(\mathbb {R}\) if there exist positive constants \(k, \alpha \), projection \(P\) and the fundamental solution matrix \(X(t)\) of (2.7) satisfying
Lemma 2.4
(see [19]) Assume that \(Q(t)\) is an almost periodic matrix function and \(g (t) \in PAP(\mathbb {R},\mathbb {R}^{n}) \). If the linear system (2.7) admits an exponential dichotomy, then pseudo almost periodic system
has a unique pseudo almost periodic solution \(x(t)\), and
Lemma 2.5
(see [19, 20]) Let \(c_{i}(t)\) be an almost periodic function on \(\mathbb {R}\) and
Then the linear system
admits an exponential dichotomy on \(R\).
3 Existence and Uniqueness of Pseudo Almost Periodic Solutions
In this section, we establish sufficient conditions on the existence of pseudo almost periodic solutions of (1.1).
For \(i, j=1, \ 2,\ldots , n,\) it will be assumed that \(h_{i} :\mathbb {R}\rightarrow [0, \ +\infty )\) is a continuous function with \(0 <\int _0^\infty {h_i }(v) dv <+\infty \) and \( \int _0^\infty v{h_i }(v)e^{\kappa v} dv <+\infty \) for a certain positive constant \(\kappa ,\, c_{i} :\mathbb {R}\rightarrow (0, \ +\infty )\) is an almost periodic function, \( \tau _{ij}:\mathbb {R}\rightarrow [0, \ +\infty )\) and \( I_{i}, \ a_{ij}, \ b_{ij} :\mathbb {R}\rightarrow \mathbb {R}\) are uniformly continuous pseudo almost periodic functions. We also make the following assumptions which will be used later.
- \((A_1)\) :
-
for each \(j\in \{1,2, \ \ldots , n \}\), there exist nonnegative constants \( L ^{f}_{j}\) and \(L^{g}_{j}\) such that
$$\begin{aligned} |f_{j}(u )-f_{j}(v )|\le L ^{f}_{j}|u -v |, |g_{j}(u )-g_{j}(v )| \le L^{g}_{j}|u -v |, \quad \hbox {for all} \quad u, \ v \in \mathbb {R}. \end{aligned}$$ - \((A_{2})\) :
-
for \(i, j\in \{1,2, \ldots , n \}\), the delay kernels \(K_{ij}:[0, \infty )\rightarrow \mathbb {R}\) are continuous, \(|K_{ij}(t)|e^{\kappa t}\) are integrable on \([0, \infty )\).
- \((A_{3})\) :
-
for each \(i\in \{1, \ 2,\ldots , n \}\), there exist constants \(\alpha _{i}>0 \) and \(\xi _{i}>0, \) such that
$$\begin{aligned}&-\,c_{i} ^{-}\int _0^\infty {h_i }(v) dv +c_{i} ^{+} \int _0^\infty v{h_i }(v) dv +\xi _{i}^{-1}\sum ^n_{j=1}a_{ij} ^{+} L^{f}_{j}\xi _{j} \\&\quad +\,\xi _{i}^{-1}\sum ^n_{j=1}b_{ij} ^{+}\int _{0}^{\infty }|K_{ij}(u)| du L^{g}_{j}\xi _{j} \le - \alpha _{i}, \end{aligned}$$and
$$\begin{aligned} c_{i}^{-}\int _0^\infty {h_i }(v) dv- \alpha _{i}+c_{i}^{+}\int _0^\infty {h_i }(v) dv(1-\frac{\alpha _{i}}{c_{i}^{-}\int _0^\infty {h_i }(v) dv})<1. \end{aligned}$$
Lemma 3.1
Assume that assumptions \((A_{1})\) and \((A_{2})\) hold. Then, for \(\varphi (\cdot )\in PAP(\mathbb {R},\mathbb {R} ) \), the function \(\int _{0}^{\infty }K_{ij}(u) g_{j}(\varphi (t-u))du\) belongs to \(PAP(\mathbb {R},\mathbb {R} ) \), where \(i,j=1, 2, \ldots , n.\)
Proof
Let \(\varphi \in PAP(\mathbb {R},\mathbb {R} ) \). Obviously, \((A_{1})\) implies that \( g_{j}\) is a uniformly continuous function on \(\mathbb {R}\). By using Corollary 5.4 in [19, p. 58], we immediately obtain the following,
where \(\chi _{ j1}\in AP(\mathbb {R},\mathbb {R} ), \chi _{ j2}\in PAP_{0}(\mathbb {R},\mathbb {R} ),\, j=1, 2, \ldots , n.\) Then, for any \(\varepsilon >0,\) it is possible to find a real number \(l=l( \varepsilon ))>0,\) for any interval with length \(l\), there exists a number \(\tau =\tau (\varepsilon )\) in this interval such that
and
It follows that
and
Thus,
which yield
The proof of Lemma 3.1 is completed.\(\square \)
Theorem 3.1
Let \((A_{1}),\, (A_{2})\) and \((A_{3})\) hold. Then, there exists a unique continuously differentiable pseudo almost periodic solution of system (1.1).
Proof
Set
then we can transform (1.1) into the following system
Let \( \varphi \in B\). Obviously, the boundedness of \(\varphi '\) and \((A_{1})\) imply that \(f_{j} \) and \(\varphi _{j}\) are uniformly continuous functions on \(\mathbb {R}\) for \(j=1, 2, \ldots , n\). Set \(f(t,z)=\varphi _{j}(t-z) (j\in \{1, 2, \ldots , n \}).\) By Theorem 5.3 in [19, p. 58] and Definition 5.7 in [19, p. 59], we can obtain that \(f\in PAP(\mathbb {R}\times \Omega )\) and \(f\) is continuous in \(z\in K\) and uniformly in \(t\in \mathbb {R}\) for all compact subset \(K\) of \(\Omega \). This, together with \(\tau _{ij}\in PAP(\mathbb {R},\mathbb {R})\) and Theorem 5.11 in [19, p. 60], implies that
Again from Corollary 5.4 in [19, p. 58], we have
In view of \( \int _0^\infty {h_i }(v) dv <+\infty \) and \( \int _0^\infty v{h_i }(v) dv <+\infty \), by using a similar argument as in proof of Lemma 3.1, we can show
for all \( i =1,2,\ldots , n. \)
By combining (3.1), (3.2) with Lemma 3.1, we obtain
For any \( \varphi \in B\), we consider the pseudo almost periodic solution \(x^{\varphi }(t)\) of the pseudo almost periodic differential equations
Then, notice that \( M[c_{i}(t)\int _0^\infty {h_i }(s) ds]>0, \; i=1, \ 2,\ldots , n \), it follows from Lemma 2.5 that the linear system
admits an exponential dichotomy on \(\mathbb {R}\). Thus, by Lemma 2.4, we obtain that the system (3.3) has exactly one pseudo almost periodic solution:
In view of the uniform continuity of coefficients and delays, from \((A_{1}),\, (A_{2})\) and the Corollary 5.6 in [19, p. 59], we get
is a pseudo almost periodic function, where \(i= 1,2,\ldots ,n\).
Now, we define a mapping \(T:B \rightarrow B\) by setting
We next prove that the mapping \(T\) is a contraction mapping of the \(B\). In fact, in view of (3.5), \((A_{1})\) and \((A_{3})\), for \(\varphi , \psi \in B \), we have \(\ |T(\varphi (t))-T(\psi (t))|\)
and \(\ |T'(\varphi (t))-T'(\psi (t))|\)
From \((A_{3})\), we have
and
which, together with (3.6) and (3.7), yield
which implies that the mapping \(T:B \longrightarrow B \) is a contraction mapping. Therefore, the mapping \(T\) possesses a unique fixed point
By (3.3) and (3.5), \(x^{** } \) satisfies (3.3). So (1.1) has a unique continuously differentiable almost periodic solution \(x^{* }=(\xi _{1}x_{1}^{**}(t), \xi _{2}x_{2}^{**}(t), \ldots , \xi _{n}x_{n}^{**}(t))^{T}\). The proof of Theorem 3.1 is now completed.\(\square \)
4 Global Exponential Stability of the Pseudo Almost Periodic Solution
In this section, we will discuss the global exponential stability of the pseudo almost periodic solution of system (1.1).
Definition 4.1
Let \(x^{*}(t)=(x^{*}_{1}(t), x^{*}_{2}(t),\ldots ,x^{*}_{n}(t))^{T} \) be the pseudo almost periodic solution of system (1.1). If there exist constants \(\alpha >0\) and \(M>1\) such that for every solution \( x(t)=(x_{1}(t), x_{2}(t),\ldots ,x_{n}(t))^{T} \) of system (1.1) with any initial value \( \varphi (t)=(\varphi _{1}(t), \varphi _{2}(t),\ldots ,\varphi _{n}(t))^{T} \) satisfying \( ( 1.2), \)
where \( \Vert \varphi -x^{*}\Vert _{0}=\max \{\sup \limits _{t\le 0 }\max \limits _{1\le i\le n }|\varphi _{i} (t) -x^{*}_{i}(t)|, \ \sup \limits _{t\le 0}\max \limits _{1\le i\le n }|\varphi _{i}' (t)-x^{*} _{i} \ '(t) | \}\). Then \(x^{*}(t)\) is said to be globally exponentially stable.
Theorem 4.2
Suppose that all conditions in Theorem (3.1) are satisfied. Moreover, assume that
Then system (1.1) has at least one pseudo almost periodic solution \(x^{*}(t)\). Moreover, \(x^{*}(t)\) is globally exponentially stable.
Proof
By Theorem 3.1, (1.1) has a unique continuously differentiable pseudo almost periodic solution \(x^{*}(t)=(x^{*}_{1}(t), x^{*}_{2}(t),\ldots ,x^{*}_{n}(t))^{T} \). Suppose that \( x(t)=(x_{1}(t), x_{2}(t),\ldots ,x_{n}(t))^{T} \) is an arbitrary solution of (1.1) associated with initial value \( \varphi (t)=(\varphi _{1}(t), \varphi _{2}(t),\ldots ,\varphi _{n}(t))^{T} \) satisfying (1.2).
Let
Then
where \( i=1, 2,\ldots , n.\)
Define continuous functions \(\Gamma _{i}(\omega )\) and \(\Pi _{i}(\omega )\) by setting
and
\(\text{ where } \ t>0, \omega \in [0, \ \kappa ], \; i=1,2,\ldots , n.\) Then, from \((A_{3})\) and (4.1), we have
and
which, together with the continuity of \(\Gamma _{i} (\omega )\) and \(\Pi _{i} (\omega )\), implies that we can choose a constants \(\lambda \in (0, \ \min \{\kappa , \min \limits _{i=1, 2,\cdots , n}c_{i} ^{-}\}) \) such that
and
where
and \(i=1, 2,\ldots , n.\)
Let
and \(M \) be a constant such that
which, together with (4.3), yields
Consequently, for any \(\varepsilon >0\), it is obvious that
In the following, we will show
Otherwise, there must exist \(i \in \{ 1, 2,\ldots , n \}\) and \(\theta >0 \) such that
Note that
Multiplying both sides of (4.10) by \(e ^{ \int _{0}^{s}c_{i}(u) \int _0^\infty {h_i }(v) dvdu} \), and integrating on \( [0, t]\), we get
Thus, with the help of (4.9), we have
which, together with (4.7) and (4.9), implies that
and
From (4.4) and (4.7), (4.10) and (4.11) yield
which contradicts (4.12). Hence, (4.8) holds. Letting \(\varepsilon \longrightarrow 0^{+}\), we have from (4.8) that
which implies
This completes the proof.\(\square \)
5 An Example
In this section, we give an example to demonstrate the results obtained in previous sections.
Example 1
Consider the following CNNs with continuously distributed leakage delays:
where \(f_{1}(x)\!=\!f_{2}(x)\!=\!g_{1}(x)\!=\!g_{2}(x)\!=\! |x|,\,I_{1}(t)\!=\!I_{2}(t)\!=\!\frac{1}{1000}\left( 4\!+\!\sin t\!+\!\sin \sqrt{2}t\!+\!\frac{1}{ 1+t^{2} }\right) \).
Obviously,
Let \(\xi _{1}= \xi _{2}=1\). Then,
and
Hence, from Theorem (4.2), system (5.1) has exactly one pseudo almost periodic solution \(x^{*}(t)\). Moreover, all solutions and their derivatives of solutions for (5.1) with initial conditions (1.2) converge exponentially to \(x^{*}(t)\) and \(x^{*} \ '(t)\), respectively. The exponential convergent rate is about \( 0.001\). The fact is verified by the numerical simulation in Fig. 1.
Remark 5.1
To the best of our knowledge, there is no research on the globally exponential convergence of the pseudo almost periodic solution of DCNNs with continuously distributed leakage delays. We also mention that all results in the reference [26] cannot be applied to imply that all solutions and their derivatives of solutions for (5.1) with initial conditions (1.4) converge exponentially to \(x^{*}(t)\) and \(x^{*} \ '(t)\), respectively. Here we employ a novel proof to establish some criteria to guarantee the existence and exponential stability of pseudo almost periodic solutions for DCNNs with continuously distributed leakage delays. This implies that the results of this paper are essentially new.
References
Chua LO, Roska T (1990) Cellular neural networks with nonlinear and delay-type template elements. In: Proc. 1990 IEEE int. workshop on cellular neural networks and their applications, pp 12–25
Jiang H, Teng Z (2004) Global exponential stability of cellular neural networks with time-varying coefficients and delays. Neural Netw 17:1415–1425
Kwon O, Lee S, Park J (2010) Improved results on stability analysis of neuralnet works with time-varying delays: novel delay-dependent criteria. Mod Phys Lett B 24:775–789
Kwon O, Park J (2009) Exponential stability analysis for uncertain neural networks with interval time-varying delays. Appl Math Comput 212:530–541
Wu J (2001) Introduction to neural dynamics and signal trasmission delay. Walter de Gruyter, Belin
Park JH (2006) Further result on asymptotic stability criterion of cellular neural networks with time-varying discrete and distributed delays. Appl Math Comput 182:1661–1666
Li X, Ding C, Zhu Q (2010) Synchronization of stochastic perturbed chaotic neural networks with mixed delays. J Franklin Inst 347:1266–1280
Kwon O, Park J (2008) Delay-dependent stability for uncertain cellular neural networks with discrete and distribute time-varying delays. J Franklin Inst 345:766–778
Haykin S (1994) Neural networks. Prentice-Hall, Upper Saddle River
Kosok B (1992) Neural networks and fuzzy systems. Prentice-Hall, New Delhi
Gopalsamy K (1992) Stability and oscillations in delay differential equations of population dynamics. Kluwer Academic Publishers, Dordrecht
Gopalsamy K (2007) Leakage delays in BAM. J Math Anal Appl 325:1117–1132
Liu B (2013) Global exponential stability for BAM neural networks with time-varying delays in the leakage terms. Nonlinear Anal 14:559–566
Peng S (2010) Global attractive periodic solutions of BAM neural networks with continuously distributed delays in the leakage terms. Nonlinear Anal 11:2141–2151
Chen Z, Meng J (2012) Exponential convergence for cellular neural networks with time-varying delays in the leakage terms, Hindawi Publishing Corporation. Abstr Appl Anal 2012(941063):1–11
Chen Z, Yang M (2012) Exponential convergence for HRNNs with continuously distributed delays in the leakage terms. Neural Comput Appl. doi:10.1007/s00521-012-1172-2
Chen Z (2012) A shunting inhibitory cellular neural network with leakage delays and continuously distributed delays of neutral type. Neural Comput App. doi:10.1007/s00521-012-1200-2
Xiong W, Meng J (2013) Exponential convergence for cellular neural networks with continuously distributed delays in the leakage terms. Electron J Qual Theory Differ Equ 2013(10):1–12
Zhang C (2003) Almost periodic type functions and ergodicity. Kluwer Academic, Beijing
Fink AM (1974) Almost periodic differential equations. Lecture Notes in Mathematics, vol 377. Springer, Berlin
Chérif Farouk (2012) Existence and global exponential stability of pseudo almost periodic solution for SICNNs with mixed delays. J Appl Math Comput 39:235–251
Zhou T, Liu Y, Chen A (2006) Almost periodic solution for shunting inhibitory cellular neural networks with time-varying delays and variable coefficients. Neural Process Lett 23(3):243–255
Li Y, Wang C, Li X (2013) Existence and global exponential stability of almost periodic solution for high-order BAM neural networks with delays on time scales. Neural Process Lett. doi:10.1007/s11063-013-9302-0
Zhang H, Shao J (2013) Existence and exponential stability of almost periodic solutions for CNNs with time-varying leakage delays. Neurocomputing 121(9):226–233
Zhang H, Shao J (2013) Almost periodic solutions for cellular neural networks with time-varying delays in leakage terms. Appl Math Comput 219(24):11471–11482
Zhang H (2014) Existence and stability of almost periodic solutions for CNNs with continuously distributed leakage delays. Neural Comput Appl 2014(24):1135–1146
Liu B (2014) Global exponential stability of positive periodic solutions for a delayed Nicholson’s blowflies model. J Math Anal Appl 412:212–221
Meng J (2013) Global exponential stability of positive pseudo almost periodic solutions for a model of hematopoiesis. Abstr Appl Anal 2013(463076):1–11
Wang W, Liu B (2014) Global exponential stability of pseudo almost periodic solutions for SICNNs with time-varying leakage delays. Abstr Appl Anal 2014(967328):1–17
Li L, Fang Z, Yang Y (2012) A shunting inhibitory cellular neural network with continuously distributed delays of neutral type. Nonlinear Anal Real World Appl 13:1186–1196
Acknowledgments
The author would like to express the sincere appreciation to the editor and reviewers for their helpful comments in improving the presentation and quality of the paper. This work was supported by the construct program of the key discipline in Hunan province (Mechanical Design and Theory), the Scientific Research Fund of Hunan Provincial Natural Science Foundation of PR China (Grant No. 11JJ6006), and the Natural Scientific Research Fund of Hunan Provincial Education Department of PR China (Grants No. 11C0916, 11C0915).
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Liu, B. Pseudo Almost Periodic Solutions for CNNs with Continuously Distributed Leakage Delays. Neural Process Lett 42, 233–256 (2015). https://doi.org/10.1007/s11063-014-9354-9
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11063-014-9354-9
Keywords
- Cellular neural networks
- Pseudo almost periodic solution
- Exponential stability
- Continuously distributed leakage delay