An Alternative to Domain-general or Domain-specific Frameworks for Theorizing about Human Evolution and Ontogenesis

Annette Karmiloff-Smith; Annette Karmiloff-Smith

doi:10.3934/Neuroscience.2015.2.91

AIMS Neuroscience

2015, Volume 2, Issue 2: 91-104. doi: 10.3934/Neuroscience.2015.2.91

Previous Article Next Article

Review

An Alternative to Domain-general or Domain-specific Frameworks for Theorizing about Human Evolution and Ontogenesis

Annette Karmiloff-Smith ^,

Birkbeck Centre for Brain & Cognitive Development, University of London, London WC1N 7HX, UK

Received: 01 January 2015 Accepted: 11 May 2015 Published: 19 June 2015

This paper maintains that neither a domain-general nor a domain-specific framework is appropriate for furthering our understanding of human evolution and ontogenesis. Rather, as we learn increasingly more about the dynamics of gene-environment interaction and gene expression, theorists should consider a third alternative: a domain-relevant approach, which argues that the infant brain comes equipped with biases that are relevant to, but not initially specific to, processing different kinds of input. The hypothesis developed here is that domain-specific core knowledge/specialized functions do not constitute the start state; rather, functional specialization emerges progressively through neuronal competition over developmental time. Thus, the existence of category-specific deficits in brain-damaged adults cannot be used to bolster claims that category-specific or domain-specific modules underpin early development, because neural specificity in the adult brain is likely to have been the emergent property over time of a developing, self-structuring system in interaction with the environment.

Keywords:

Citation: Annette Karmiloff-Smith. An Alternative to Domain-general or Domain-specific Frameworks for Theorizing about Human Evolution and Ontogenesis[J]. AIMS Neuroscience, 2015, 2(2): 91-104. doi: 10.3934/Neuroscience.2015.2.91

Related Papers:

[1]	Fahad Alsharari, Ahmed O. M. Abubaker, Islam M. Taha . On $ r $-fuzzy soft $ \gamma $-open sets and fuzzy soft $ \gamma $-continuous functions with some applications. AIMS Mathematics, 2025, 10(3): 5285-5306. doi: 10.3934/math.2025244
[2]	Jinguo Zhang, Dengyun Yang, Yadong Wu . Existence results for a Kirchhoff-type equation involving fractional $ p(x) $-Laplacian. AIMS Mathematics, 2021, 6(8): 8390-8403. doi: 10.3934/math.2021486
[3]	Cenap Ozel, M. A. Al Shumrani, Aynur Keskin Kaymakci, Choonkil Park, Dong Yun Shin . On $ \delta b $-open continuous functions. AIMS Mathematics, 2021, 6(3): 2947-2955. doi: 10.3934/math.2021178
[4]	Ibtesam Alshammari, Islam M. Taha . On fuzzy soft $ \beta $-continuity and $ \beta $-irresoluteness: some new results. AIMS Mathematics, 2024, 9(5): 11304-11319. doi: 10.3934/math.2024554
[5]	Sang-Eon Han . Semi-topological properties of the Marcus-Wyse topological spaces. AIMS Mathematics, 2022, 7(7): 12742-12759. doi: 10.3934/math.2022705
[6]	Samirah Alzahrani, A. A. Nasef, N. Youns, A. I. EL-Maghrabi, M. S. Badr . Soft topological approaches via soft γ-open sets. AIMS Mathematics, 2022, 7(7): 12144-12153. doi: 10.3934/math.2022675
[7]	Nicky K. Tumalun, Philotheus E. A. Tuerah, Marvel G. Maukar, Anetha L. F. Tilaar, Patricia V. J. Runtu . An application of generalized Morrey spaces to unique continuation property of the quasilinear elliptic equations. AIMS Mathematics, 2023, 8(11): 26007-26020. doi: 10.3934/math.20231325
[8]	Dina Abuzaid, Samer Al-Ghour . Supra soft Omega-open sets and supra soft Omega-regularity. AIMS Mathematics, 2025, 10(3): 6636-6651. doi: 10.3934/math.2025303
[9]	Syafrizal Sy, Rinovia Simanjuntak, Tamaro Nadeak, Kiki Ariyanti Sugeng, Tulus Tulus . Distance antimagic labeling of circulant graphs. AIMS Mathematics, 2024, 9(8): 21177-21188. doi: 10.3934/math.20241028
[10]	Mesfer H. Alqahtani, Alaa M. Abd El-latif . Separation axioms via novel operators in the frame of topological spaces and applications. AIMS Mathematics, 2024, 9(6): 14213-14227. doi: 10.3934/math.2024690

Abstract

1. Introduction

Since shunting inhibitory cellular neural networks were proposed by Bouzerdoum and Pinter ^[1] as a new type of neural networks, they have received more and more attention and have been widely applied in optimisation, psychophysics, speech and other fields. At the same time, since time delays are ubiquitous, many research results have been obtained on the dynamics of shunting inhibitory cellular neural networks with time delays ^[2,3,4,5,6].

On the one hand, the quaternion is a generalization of real and complex numbers ^[7]. The skew field of quaternions is defined by

$ \mathbb{H}: = \{q = q^{R}+iq^{I}+jq^{J}+kq^{K}\}, $

where $ q^{R}, q^{I}, q^{J}, q^{K}\in\mathbb{R} $ and the elements $ i, j $ and $ k $ obey the Hamilton's multiplication rules:

$ ij = -jk = k, \quad jk = -kj = i, \quad ki = -ik = j, \quad i^{2} = j^{2} = k^{2} = -1. $

For $ q = q^{R}+iq^{I}+jq^{J}+kq^{K} $, we denote $ \check{q} = iq^{I}+jq^{J}+kq^{K} $ and $ q^R = q-\check{q} $. The norm of $ q $ is defined by $ \|q\|_{\mathbb{H}} = \sqrt{(q^{R})^2+(q^{I})^2+(q^{J})^2+(q^{K})^2} $. For $ y = (y_1, y_2, \cdots, y_n)^T\in\mathbb{H}^n $, we define $ \|y\|_{\mathbb{H}^n} = \max_{1\leq p\leq n}\{\|y\|_{\mathbb{H}}\} $, then $ (\mathbb{H}^n, \|\cdot\|_{\mathbb{H}^n}) $ is a Banach space. As we all know, quaternion-valued neural networks include real-valued neural networks and complex-valued neural networks as their special cases. Compared with complex-valued neural networks, quaternion-valued neural networks only needs half of the connection weight parameters of complex-valued neural networks when dealing with multi-level information ^[8]. In recent years, quaternion-valued neural networks have attracted the attention of many researchers, and their various dynamic behaviors, including fractional-order and stochastic quaternion-valued neural networks, have been extensively studied ^{[9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24]}.

On the other hand, because periodic and almost periodic oscillations are important dynamics of neural networks, the periodic and almost periodic oscillations of neural networks have been studied a lot in the past few decades ^{[25,26,27,28,29,30,31,32,33]}. Weyl almost periodicity is a generalization of Bohr almost periodicity and Stepanov almost periodicity ^{[34,35,36,37]}. It is a more complex recurrent oscillation. Because the spaces composed of Bohr almost periodic functions and Stepanov almost periodic functions are Banach spaces, it brings some convenience to study the existence of almost periodic solutions in these two senses of differential equations. Therefore, many results have been obtained on the Bohr almost periodic oscillation and Stepanov almost periodic oscillation of neural networks. However, the space composed of Weyl almost periodic functions is incomplete ^[38]. Therefore, the results of Weyl almost periodic solutions of neural networks are still very rare. Therefore, it is a meaningful and challenging work to study the existence of Weyl almost periodic solutions of neural networks.

Motivated by the above, in this paper, we consider the following shunting inhibitory cellular neural networks with time-varying delays:

$\begin{align} \dot{x}_{ij}(t) = &-a_{ij}(t)x_{ij}(t)-\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl}(t)f_{ij}(x_{kl}(t))x_{ij}(t)\\ &-\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl}(t)g_{ij}(x_{kl}(t-\tau_{kl}(t)))x_{ij}(t)+I_{ij}(t), \end{align}$ $

(1.1)

where $ ij\in\{11, 12, \dots, 1n, \dots, m1, m2, \dots, mn\}: = \Lambda $, $ C_{ij} $ denotes the cell at the $ (i, j) $ position of the lattice. The $ r $-neighborhood $ N_{r}(i, j) $ of $ C_{ij} $ is given as

$ N_{r}(i, j) = \{C_{kl}:\max(|k-i|, |l-j|)\leq r, 1\leq k\leq m, 1\leq l\leq n\}, $

and $ N_{s}(i, j) $ is similarly specified; $ x_{ij}(t)\in\mathbb{H} $ denotes the activity of the cell of $ C_{ij} $, $ I_{ij}(t)\in\mathbb{H} $ is the external input to $ C_{ij} $, $ a_{ij}(t)\in\mathbb{H} $ is the coefficient of the leakage term, which represents the passive decay rate of the activity of the cell $ C_{ij} $, $ B_{ij}^{kl}(t)\geq0 $ and $ C_{ij}^{kl}(t)\geq0 $ represent the connection or coupling strength of postsynaptic of activity of the cell transmitted to the cell $ C_{ij} $, the activity functions $ f_{ij}, g_{ij}:\mathbb{H}\rightarrow \mathbb{H} $ are continuous functions representing the output or firing rate of the cell $ C_{ij} $, $ \tau_{kl}(t) $ corresponds to the transmission delay and satisfies $ 0 \leq \tau_{kl}(t) \leq \tau $.

The purpose of this paper is to use the fixed point theorem and a variant of Gronwall inequality to establish the existence and global exponential stability of Weyl almost periodic solutions for a class of quaternion-valued shunting inhibitory cellular neural networks whose coefficients of the leakage terms are quaternions. This is the first paper to study the existence and global exponential stability of Weyl almost periodic solutions of system (1.1) by using the fixed point theorem and a variant of Gronwall inequality. Our result of this paper is new, and our method can be used to study other types of quaternion-valued neural networks.

For convenience, we introduce the following notations:

$ a^m = \min\limits_{ij\in\Lambda}\left\{\inf\limits_{t\in\mathbb{R}}\{a_{ij}(t)\}\right\}, \, \, a^M = \max\limits_{ij\in\Lambda}\left\{\sup\limits_{t\in\mathbb{R}}\{a_{ij}(t)\}\right\}, \, \, \check{a}_{ij}^M = \sup\limits_{t\in\mathbb{R}}\|\check{a}_{ij}(t)\|_{\mathbb{H}}, $

$ B_{ij}^{kl^M} = \sup\limits_{t\in\mathbb{R}}\{B_{ij}^{kl}(t)\}, \, \, C_{ij}^{kl^M} = \sup\limits_{t\in\mathbb{R}}\{C_{ij}^{kl}(t)\}, \, \, \tau_{ij}^M = \sup\limits_{t\in\mathbb{R}}\{\tau_{ij}(t)\}, \, \, \tau' = \max\limits_{kl\in\Lambda}\left\{\sup\limits_{t\in\mathbb{R}}\{\tau'_{kl}(t)\}\right\}. $

The initial condition of system (1.1) is given by

$ x_{ij}(s) = \varphi_{ij}(s), \quad s\in[-\tau, 0], $

where $ \varphi_{ij}\in C(\mathbb{R}, \mathbb{H}), ij\in \Lambda $.

Throughout this paper, we assume that:

$ (H_{1}) $ For $ ij, kl\in\Lambda $, functions $ a^R_{ij}, B_{ij}^{kl}, C_{ij}^{kl}\in AP(\mathbb{R}, \mathbb{R}^+) $, $ \check{a}_{ij}\in AP(\mathbb{R}, \mathbb{H}) $, $ I_{ij} \in APW^p(\mathbb{R}, \mathbb{H}) $, $ \tau_{kl}\in AP(\mathbb{R}, \mathbb{R}^{+})\cap C^1(\mathbb{R}, \mathbb{R}) $ and $ \tau' < 1 $.

$ (H_{2}) $ For $ ij\in\Lambda $, there exist positive constants $ L_{ij}^{f} $ and $ L_{ij}^{g} $ such that for all $ x, y\in\mathbb{H} $,

$\begin{eqnarray*} \|f_{ij}(x)-f_{ij}(y)\|_{\mathbb{H}}\leq L_{ij}^{f}\|x-y\|_{\mathbb{H}}, \quad \|g_{ij}(x)-g_{ij}(y)\|_{\mathbb{H}}\leq L_{ij}^{g}\|x-y\|_{\mathbb{H}}, \end{eqnarray*}$ $

and $ f_{ij}(\mathbf{0}) = g_{ij}(\mathbf{0}) = 0 $.

The rest of this paper is arranged as follows. In Section 2, we introduce some definitions and lemmas. In Section 3, we study the existence and global exponential stability of Weyl almost periodic solutions of (1.1). In Section 4, an example is given to verify the theoretical results. This paper ends with a brief conclusion in Section 5.

2. Preliminaries

Let $ (\mathbb{X}, \|\cdot\|_{\mathbb{X}}) $ be a Banach space and $ BC(\mathbb{R}, \mathbb{X}) $ be the set of all bounded continuous functions from $ \mathbb{R} $ to $ \mathbb{X} $.

Definition 2.1. ^[38] A function $ f\in BC(\mathbb{R}, \mathbb{X}) $ is said to be almost periodic, if for every $ \epsilon > 0 $, there exists a constant $ l = l(\epsilon) > 0 $ such that in every interval of length $ l(\epsilon) $ contains at least one $ \sigma $ such that

$ \|f(t+\sigma)-f(t)\|_{\mathbb{X}} < \epsilon, t\in \mathbb{R}. $

Denote by $ AP(\mathbb{R}, \mathbb{X}) $ the set of all such functions.

For $ p\in[1, \infty) $, we denote by $ L^p_{loc}(\mathbb{R}, \mathbb{X}) $ the space of all functions from $ \mathbb{R} $ into $ \mathbb{X} $ which are locally $ p $-integrable. For $ f\in L^p_{loc}(\mathbb{R}, \mathbb{X}) $, we define the following seminorm:

$ \|f\|_{W^p} = \lim\limits_{r\rightarrow +\infty}\sup\limits_{\beta\in\mathbb{R}}\bigg(\frac{1}{r}\int_{\beta}^{\beta+r} \|f(t)\|_{\mathbb{X}}^pd t\bigg)^{\frac{1}{p}}. $

Definition 2.2. ^[38] A function $ f\in L^p_{loc}(\mathbb{R}, \mathbb{X}) $ is said to be $ p $-th Weyl almost periodic ($ W^p $-almost periodic for short), if for every $ \epsilon > 0 $, there exists a constant $ l = l(\epsilon) > 0 $ such that in every interval of length $ l(\epsilon) $ contains at least one $ \sigma $ such that

$ \|f(t+\sigma)-f(t)\|_{W^p} < \epsilon. $

This $ \sigma $ is called on $ \epsilon $-translation number of $ f $. The set of all such functions will be denoted by $ APW^p(\mathbb{R}, \mathbb{X}) $.

Remark 2.1. By Definitions 2.1 and 2.2, it is easy to see that if $ f\in AP(\mathbb{R}, \mathbb{X}) $, then $ f\in APW^p(\mathbb{R}, \mathbb{X}) $.

Similar to the proofs of the lemma on page 83 and the lemma on page 84 of ^[39], it is not difficult to prove the following two lemmas.

Lemma 2.1. If $ f\in APW^p(\mathbb{R}, \mathbb{X}) $, then $ f $ is bounded and uniformly continuous on $ \mathbb{R} $ withrespect to the seminorn $ \|\cdot\|_{W^p} $.

Using the argumentation contained in the proof of Proposition 3.21 in ^[38], one can easily prove the following.

Lemma 2.2. If $ f_{k}\in APW^p(\mathbb{R}, \mathbb{X}) $, $ k = 1, 2, \ldots, n $. Then, for every $ \epsilon > 0 $, there exist common $ \epsilon $-translation numbers for these functions.

Lemma 2.3. ^[40]Let $ g:\mathbb{R}\rightarrow \mathbb{R} $ be a continuous function such that, for every $ t\in\mathbb{R} $,

$\begin{eqnarray} 0\leq g(t)\leq\rho(t)+\gamma_1\int_{-\infty}^te^{-\eta_1(t-s)}g(s)ds+\cdots +\gamma_n\int_{-\infty}^te^{-\eta_n(t-s)}g(s)ds \end{eqnarray}$ $

(2.1)

for some locally integrable function $ \rho:\mathbb{R}\rightarrow \mathbb{R} $, and for some constants$ \gamma_1, \ldots, \gamma_n\geq0 $, and some constants $ \eta_1, \ldots, \eta_n\geq\gamma $, where$ \gamma = \sum\limits_{p = 1}^n\gamma_p $. We assume that the integrals in the right hand side of (2.1)are convergent. Let $ \eta = \min\{\eta_1, \ldots, \eta_n\} $. Then, for every $ \xi\in(0, \eta-\gamma] $such that $ \int_{-\infty}^0e^{\xi s}\rho(s)ds $ converges, we have, for every $ t\in\mathbb{R} $,

$\begin{eqnarray*} g(t)\leq\rho(t)+\gamma\int_{-\infty}^te^{-\xi(t-s)}\rho(s)ds. \end{eqnarray*}$ $

In particular, if $ \rho(t) $ is constant, we have

$\begin{eqnarray*} g(t)\leq\rho\frac{\eta}{\eta-\gamma}. \end{eqnarray*}$ $

3. Main results

Let $ BUC(\mathbb{R}, \mathbb{H}^{m\times n}) $ be a collection of bounded and uniformly continuous functions from $ \mathbb{R} $ to $ \mathbb{H}^{m\times n} $, then, the space $ BUC(\mathbb{R}, \mathbb{H}^{m\times n}) $ with the norm $ \|x\|_{\infty} = \sup\limits_{t\in\mathbb{R}}\|x(t)\|_{\mathbb{H}^{m\times n}} $ is a Banach space, where $ x\in BUC(\mathbb{R}, \mathbb{H}^{m\times n}) $.

Denote $ \phi^0 = (\phi_{11}^0, \cdots, \phi_{1n}^0, \phi_{21}^0, \cdots, \phi_{2n}^0, \cdots, \phi_{m1}^0, \cdots, \phi_{mn}^0)^T $, where

$ \phi_{ij}^0(t) = \int_{-\infty}^{t}e^{-\int_{s}^{t}{a_{ij}^R (v)}dv}I_{ij}(s)ds, \quad ij\in\Lambda. $

We will show that $ \phi^0 $ is well defined under assumption $ (H_1) $. In fact, by $ I_{ij}\in APW^p(\mathbb{R}, \mathbb{H}) $ and Lemma 2.1, there exists a constant $ M > 0 $ such that $ \|I_{ij}\|_{W^p}\leq M $ for all $ ij\in \Lambda $. According to the Hölder inequality, one has

$\begin{align} \|\phi_{ij}^0(t)\|_{\mathbb{H}} \leq&\bigg\|\int_{-\infty}^{t}e^{-a^m(t-s)}I_{ij}(s)ds\bigg\|_{\mathbb{H}}\\ \leq&\sum\limits_{r = 0}^{\infty}\bigg(\int_{t-(r+1)}^{t-r}e^{-a^mq(t-s)}ds\bigg)^{\frac{1}{q}} \bigg(\int_{t-(r+1)}^{t-r}\|I_{ij}(s)\|^p_{\mathbb{H}}ds\bigg)^{\frac{1}{p}}\\ \leq&\sum\limits_{r = 0}^{\infty} e^{-a^mr}M < +\infty, \end{align}$ $

(3.1)

where $ \frac{1}{p}+\frac{1}{q} = 1 $, which means that $ \phi^0 $ is well defined.

Take a positive constant $ \alpha\geq\|\phi^0\|_{\infty} $. Let

$ \Omega = \big\{\phi\in BUC(\mathbb{R}, \mathbb{H}^{m\times n})\Big| \|\phi-\phi^0\|_{\infty}\leq\alpha\big\}. $

Then, for every $ \phi\in\Omega $, one has

$ \|\phi\|_{\infty}\leq\|\phi-\phi^0\|_{\infty}+\|\phi^0\|_{\infty}\leq2\alpha . $

Theorem 3.1. Assume $ (H_{1}) $–$ (H_{2}) $ hold. Furthermore, suppose that

$ (H_{3}) $

$\begin{eqnarray*} \kappa = \max\limits_{ij\in\Lambda}\left\{\frac{2}{a^m}\bigg[\check{a}_{ij}^M+\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L_{ij}^f\alpha +\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L_{ij}^g\alpha\bigg]\right\} < 1, \end{eqnarray*}$ $

$ (H_{4}) $for $ p > 2 $,

$\begin{eqnarray*} &&\max\limits_{ij\in\Lambda}\Bigg\{ 24\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\bigg(\frac{4}{a^mp}\bigg)^2\bigg[2(\check{a}^M_{ij})^p +2\bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L^f_{ij}\alpha\bigg)^{p} \\ &&+\bigg(1+\frac{2e^{\frac{p}{4}a^m\tau}}{1-\tau'}\bigg) \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\bigg)^p\bigg]\Bigg\} < 1, \end{eqnarray*}$ $

and, for $ p = 2 $,

$\begin{eqnarray*} &&\max\limits_{ij\in\Lambda}\Bigg\{24 \bigg(\frac{2}{a^m}\bigg)^2\bigg[2(\check{a}^M_{ij})^2+ 2\bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L^f_{ij}\alpha\bigg)^{2} \\ &&+\bigg(1+\frac{2e^{\frac{1}{2}a^m\tau}}{1-\tau'}\bigg) \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\bigg)^2\bigg]\Bigg\} < 1, \end{eqnarray*}$ $

then system (1.1) has a unique $ W^p $-almost periodic solution in $ \Omega $.

Proof. It is easy to check that if $ x = (x_{11}, \cdots, x_{1n}, x_{21}, \cdots, x_{2n}, \cdots, x_{m1}, \cdots, x_{mn})^T\in \Omega $ is a solution of the integral equation

$\begin{align} x_{ij}(t) = &\int_{-\infty}^{t}e^{-\int_{s}^{t}{a_{ij}^R (v)}dv}\bigg[-\check{a}_{ij}(s)x_{ij}(s)-\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl}(s)f_{ij}(x_{kl}(s))x_{ij}(s)\\ &-\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl}(s)g_{ij}(x_{kl}(s-\tau_{kl}(s))) x_{ij}(s)+I_{ij}(s)\bigg]ds, \quad ij\in\Lambda, \end{align}$ $

(3.2)

then $ x $ is a solution of system (1.1).

Define an operator $ T:\Omega\rightarrow \mathbb{H}^{m\times n} $ by

$ (T\phi)(t) = ((T_{11}\phi)(t), \cdots, (T_{1n}\phi)(t), (T_{21}\phi)(t), \cdots, (T_{2n}\phi)(t), \cdots, (T_{m1}\phi)(t) , \cdots, (T_{mn}\phi)(t))^{T}, $

where

$\begin{align*} (T_{ij}\phi)(t) = &\int_{-\infty}^{t}e^{-\int_{s}^{t}{a_{ij}^R (v)}dv}\bigg[-\check{a}_{ij}(s)\phi_{ij}(s)-\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl}(s)f_{ij}(\phi_{kl}(s))\phi_{ij}(s)\nonumber\\ &-\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl}(s)g_{ij}(\phi_{kl}(s-\tau_{kl}(s))) \phi_{ij}(s)+I_{ij}(s)\bigg]ds, \quad ij\in\Lambda. \end{align*}$ $

Now, we will prove that $ T\phi $ is well defined. Actually, by $ (H_1) $–$ (H_3) $ and (3.1), for $ ij\in \Lambda $, one deduces that

$\begin{align} \|(T_{ij}\phi)(t)\|_{\mathbb{H}} \leq& \int_{-\infty}^{t}e^{-\int_{s}^{t}{a_{ij}^R (v)}dv}\bigg\|-\check{a}_{ij}(s)\phi_{ij}(s)-\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl}(s)(f_{ij}(\phi_{kl}(s))-f_{ij}(\mathbf{0}))\phi_{ij}(s)\\ &-\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl}(s)(g_{ij}(\phi_{kl}(s-\tau_{kl}(s)))-g_{ij}(\mathbf{0})) \phi_{ij}(s)\bigg\|_{\mathbb{H}}ds +\int_{-\infty}^{t}e^{-\int_{s}^{t}{a_{ij}^R(v)}dv}\|I_{ij}(s)\|_{\mathbb{H}}ds\\ \leq&\frac{1}{a^m}\bigg(\check{a}_{ij}^M+\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl^M}L_{ij}^f \|\phi\|_{\infty} +\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl^M}L_{ij}^g\|\phi\|_{\infty}\bigg)\|\phi\|_{\infty}\\ &+\int_{-\infty}^{t}e^{-\int_{s}^{t}{a_{ij}^R (v)}dv}\|I_{ij}(s)\|_{\mathbb{H}}ds\\ < &+\infty. \end{align}$ $

(3.3)

That is, $ T\phi $ is well defined.

We will divide the rest of the proof into four steps.

$ Step $ 1, we will prove that $ T\phi\in BUC(\mathbb{R}, \mathbb{H}^{m\times n}) $, for every $ \phi\in \Omega $.

In fact, by (3.3), we see that $ T\phi $ is bounded on $ \mathbb{R} $. So, we only need to show that $ T\phi $ is uniformly continuous on $ \mathbb{R} $. Based on the Hölder inequality for $ 0\leq h\leq 1 $ and $ q\geq 1 $ with $ \frac{1}{p}+\frac{1}{q} = 1 $, one has

$\begin{align*} &\|(T_{ij}\phi)(t+h)-(T_{ij}\phi)(t)\|_{\mathbb{H}}\\ = &\bigg\|\int_{-\infty}^{t+h}e^{-\int_{s}^{t+h}{a_{ij}^R (v)}dv}\bigg(-\check{a}_{ij}(s)\phi_{ij}(s)-\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl}(s)f_{ij}(\phi_{kl}(s))\phi_{ij}(s)-\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl}(s)\nonumber\\ &\times g_{ij}(\phi_{kl}(s-\tau_{kl}(s))) \phi_{ij}(s)+I_{ij}(s)\bigg)ds -\int_{-\infty}^{t}e^{-\int_{s}^{t}{a_{ij}^R (v)}dv}\bigg(-\check{a}_{ij}(s)\phi_{ij}(s)-\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl}(s)\nonumber\\ &\times f_{ij}(\phi_{kl}(s))\phi_{ij}(s)-\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl}(s)g_{ij}(\phi_{kl}(s-\tau_{kl}(s))) \phi_{ij}(s)+I_{ij}(s)\bigg)ds\bigg\|_{\mathbb{H}}\\ \leq&\int_{-\infty}^{t}\bigg|e^{-\int_{s}^{t+h}{a_{ij}^R(v)}dv}-e^{-\int_{s}^{t}{a_{ij}^R(v)}dv}\bigg| \bigg\|\check{a}_{ij}(s)\phi_{ij}(s)+\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl}(s)f_{ij}(\phi_{kl}(s))\phi_{ij}(s)+\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl}(s)\nonumber\\ &\times g_{ij}(\phi_{kl}(s-\tau_{kl}(s))) \phi_{ij}(s)\bigg\|_{\mathbb{H}}ds +\int_{-\infty}^{t}\bigg|e^{-\int_{s}^{t+h}{a_{ij}^R(v)}dv}-e^{-\int_{s}^{t}{a_{ij}^R(v)}dv} \bigg|\|I_{ij}(s)\|_{\mathbb{H}}ds\\ &+\int_{t}^{t+h}e^{-\int_{s}^{t+h}{a_{ij}^R (v)}dv}\bigg\|\check{a}_{ij}(s)\phi_{ij}(s)+\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl}(s)f_{ij}(\phi_{kl}(s))\phi_{ij}(s)+\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl}(s)\nonumber\\ &\times g_{ij}(\phi_{kl}(s-\tau_{kl}(s))) \phi_{ij}(s)\bigg\|_{\mathbb{H}}ds+\int_{t}^{t+h}e^{-\int_{s}^{t+h}{a_{ij}^R (v)}dv}\|I_{ij}(s)\|_{\mathbb{H}}ds\\ \leq&\bigg(\check{a}_{ij}^M+\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl^M}L_{ij}^f\|\phi\|_{\infty} +\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl^M}L_{ij}^g\|\phi\|_{\infty}\bigg)\|\phi\|_{\infty} \int_{-\infty}^{t}\bigg(e^{-\int_{s}^{t}{a_{ij}^R(v)}dv}\int_t^{t+h}a^R_{ij}(v)dv\bigg) ds\\ &+\int_{-\infty}^{t}\bigg(e^{-\int_{s}^{t}{a_{ij}^R(v)}dv}\int_t^{t+h} a_{ij}^R(v)dv\bigg)\|I_{ij}(s)\|_{\mathbb{H}}ds+\bigg(\check{a}_{ij}^M+\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl^M}L_{ij}^f\|\phi\|_{\infty}\\ &+\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl^M}L_{ij}^g\|\phi\|_{\infty}\bigg) \|\phi\|_{\infty}\int_{t}^{t+h}e^{-\int_{s}^{t}{a_{ij}^R (v)}dv} ds+\int_{t}^{t+h}e^{-\int_{s}^{t}{a_{ij}^R (v)}dv}\|I_{ij}(s)\|_{\mathbb{H}}ds\\ \leq&a^Mh\bigg(\check{a}_{ij}^M+\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl^M}L_{ij}^f\|\phi\|_{\infty} +\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl^M}L_{ij}^g\|\phi\|_{\infty}\bigg)\|\phi\|_{\infty}\int_{-\infty}^{t}e^{-a^m(t-s)}ds\\ &+a^Mh\int_{-\infty}^{t}e^{-\int_{s}^{t}{a_{ij}^R(v)}dv}\|I_{ij}(s)\|_{\mathbb{H}}ds+\bigg(\check{a}_{ij}^M +\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl^M}L_{ij}^f\|\phi\|_{\infty}+\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl^M}L_{ij}^g\|\phi\|_{\infty}\bigg)\|\phi\|_{\infty} \\ &\times\int_{t}^{t+h}e^{-a^m(t-s)}ds +\bigg(\int_{t}^{t+h}e^{-qa^m(t-s)}ds\bigg)^{\frac{1}{q}}\bigg(\int_{t}^{t+h} \|I_{ij}(s)\|^p_{\mathbb{H}}ds\bigg)^{\frac{1}{p}}\\ \leq&\frac{a^M}{a^m}\bigg(\check{a}_{ij}^M+\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl^M}L_{ij}^f\|\phi\|_{\infty} +\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl^M}L_{ij}^g\|\phi\|_{\infty}\bigg)\|\phi\|_{\infty}h+ a^Mh\int_{-\infty}^{t}e^{-\int_{s}^{t}{a_{ij}^R(v)}dv}\|I_{ij}(s)\|_{\mathbb{H}}ds\\ &+ e^{a^mh}\bigg(\check{a}_{ij}^M+\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl^M}L_{ij}^f\|\phi\|_{\infty} +\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl^M}L_{ij}^g\|\phi\|_{\infty}\bigg)\|\phi\|_{\infty}h+e^{a^mh}\|I_{ij}\|_{W^p}h, \end{align*}$ $

where $ ij\in\Lambda $. Hence, letting $ h\rightarrow0^+ $, by (3.1), we have

$ \|(T_{ij}\phi)(t+h)-(T_{ij}\phi)(t)\|_{\mathbb{H}}\rightarrow0, $

which means that $ (T_{ij}\phi) $ is uniformly continuous on $ \mathbb{R} $, $ ij\in\Lambda $. Therefore, $ T\phi\in BUC(\mathbb{R}, \mathbb{H}^{m\times n}) $.

$ Step $ 2, we will prove that $ T $ is a self-mapping from $ \Omega $ to $ \Omega $.

Actually, for arbitrary $ \phi\in\Omega $, from $ (H_2) $–$ (H_3) $, we have

$\begin{align*} \|T\phi-\phi^0\|_{\infty} \leq&\sup\limits_{t\in\mathbb{R}}\Bigg\{\max\limits_{ij\in\Lambda}\bigg[\int_{-\infty}^{t} e^{-\int_{s}^{t}{a_{ij}^R (v)}dv}\bigg\|\check{a}_{ij}(s)\phi_{ij}(s)+\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl}(s)(f_{ij}(\phi_{kl}(s))-f(\mathbf{0}))\phi_{ij}(s)\nonumber\\ &+\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl}(s)(g_{ij}(\phi_{kl}(s-\tau_{kl}(s)))-g_{ij}(\mathbf{0})) \phi_{ij}(s)\bigg\|_{\mathbb{H}}ds\bigg]\Bigg\}\\ \leq&\max\limits_{ij\in\Lambda}\Bigg\{\frac{1}{a^m}\bigg[\check{a}_{ij}^M+\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L_{ij}^f\alpha +\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L_{ij}^g\alpha\bigg]\Bigg\}\|\phi\|_{\infty}\\ &\leq\kappa\alpha\leq\alpha, \end{align*}$ $

which implies that $ T\phi\in\Omega $. Consequently, $ T $ is a self-mapping from $ \Omega $ to $ \Omega $.

$ Step $ 3, we will prove $ T $ is a contraction mapping.

As a matter of fact, in view of $ (H_{1}) $–$ (H_{2}) $, for any $ \phi, \nu\in \Omega $, we can get

$\begin{align*} &\|T\phi-T\nu\|_{\infty}\\ \leq&\sup\limits_{t\in\mathbb{R}}\Bigg\{\max\limits_{ij\in\Lambda} \Bigg[\int_{-\infty}^{t}e^{-a^m(t-s)}\bigg(\check{a}^M_{ij}\|\phi_{ij}(s)-\nu_{ij}(s)\|_{\mathbb{H}}+\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl^M}\|f_{ij}(\phi_{kl}(s))(\phi_{ij}(s)-\nu_{ij}(s))\nonumber\\ &+(f_{ij}(\phi_{kl}(s))-f_{ij}(\nu_{kl}(s)))\nu_{ij}(s)\|_{\mathbb{H}} +\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl^M}\|g_{ij}(\phi_{kl}(s-\tau_{kl}(s)))(\phi_{ij}(s)-\nu_{ij}(s)) \\ &+(g_{ij}(\phi_{kl}(s-\tau_{kl}(s))) -g_{ij}(\nu_{kl}(s-\tau_{kl}(s))))\nu_{ij}(s)\|_{\mathbb{H}}\bigg)ds\Bigg]\Bigg\}\nonumber\\ \leq&\sup\limits_{t\in\mathbb{R}}\Bigg\{\max\limits_{ij\in\Lambda} \Bigg[\int_{-\infty}^{t}e^{-a^m(t-s)}\bigg[\check{a}^M_{ij}\|\phi_{ij}(s)-\nu_{ij}(s)\|_{\mathbb{H}}+\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl^M}\Big(2L_{ij}^f\alpha\|\phi_{ij}(s)-\nu_{ij}(s)\|_{\mathbb{H}}\nonumber\\ &+2L_{ij}^f\alpha\|\phi_{kl}(s)-\nu_{kl}(s)\|_{\mathbb{H}}\Big) +\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl^M}\Big(2L_{ij}^g\alpha\|\phi_{ij}(s)-\nu_{ij}(s)\|_{\mathbb{H}} \\ &2L_{ij}^g\alpha\|\phi_{kl}(s-\tau_{kl}(s))-\nu_{kl}(s-\tau_{kl}(s))\|_{\mathbb{H}}\Big)\bigg]ds\Bigg]\Bigg\}\nonumber\\ \leq&\max\limits_{ij\in\Lambda}\Bigg\{\frac{1}{a^m}\bigg[\check{a}_{ij}^M+\sum\limits_{C_{kl}\in N_r(i, j)}4B_{ij}^{kl^M}L_{ij}^f\alpha+\sum\limits_{C_{kl}\in N_s(i, j)}4C_{ij}^{kl^M} L_{ij}^g\alpha\bigg]\Bigg\}\|\phi-\nu\|_{\infty}. \end{align*}$ $

From this and $ (H_3) $, one has

$\begin{eqnarray*} ||T\phi-T\nu||_{\infty}\leq\kappa\|\phi-\nu\|_{\infty}. \end{eqnarray*}$ $

Noticing that $ \kappa < 1 $, $ T $ is a contraction mapping. Consequently, system $ (1.1) $ has a unique solution $ x $ in $ \Omega $.

$ Step $ 4, we will prove that the unique solution $ x\in\Omega $ is $ W^p $-almost periodic.

Indeed, since $ x = (x_{11}, \cdots, x_{1n}, x_{21}, \cdots, x_{2n}, \cdots, x_{m1}, \cdots, x_{mn})^T\in \Omega $, $ x $ is bounded and uniformly continuous. Hence, for every $ \epsilon > 0 $, there exists a $ \delta\in (0, \epsilon) $ such that for any $ t_1, t_2\in \mathbb{R} $ with $ |t_1-t_2| < \delta $ and $ ij\in \Lambda $, we have

$\begin{align} \|x_{ij}(t_1)-x_{ij}(t_2)\|_{\mathbb{H}} < \epsilon. \end{align}$ $

(3.4)

Also, for this $ \delta $, in view of $ (H_1) $ and Lemma 2.2, we see that there exists a common $ \delta $-translation number $ \sigma $ such that

$\begin{eqnarray} && \lim\limits_{r\rightarrow +\infty}\sup\limits_{\beta\in\mathbb{R}}\bigg(\frac{1}{r}\int_{\beta}^{\beta+r} \|I_{ij}(t+\sigma)-I_{ij}(t)\|_{\mathbb{H}}^pd t\bigg)^{\frac{1}{p}} < \delta < \epsilon, \end{eqnarray}$ $

(3.5)

$\begin{eqnarray} &&|B_{ij}^{kl}(t+\sigma)-B_{ij}^{kl}(t)| < \epsilon, \end{eqnarray}$ $

(3.6)

$\begin{eqnarray} && |C_{ij}^{kl}(t+\sigma)-C_{ij}^{kl}(t)| < \epsilon, \end{eqnarray}$ $

(3.7)

$\begin{eqnarray} &&|a_{ij}^R(t+\sigma)-a_{ij}^R(t)| < \epsilon, \|\check{a}_{ij}(t+\sigma)-\check{a}_{ij}(t)\|_{\mathbb{H}} < \epsilon \end{eqnarray}$ $

(3.8)

and

$\begin{align} |\tau_{ij}(t+\sigma)-\tau_{ij}(t)| < \delta, \end{align}$ $

(3.9)

where $ ij, kl\in\Lambda $. Consequently, from (3.4) and (3.9), we get

$\begin{eqnarray} \|x_{ij}(t-\tau_{ij}(t+\sigma))-x_{ij}(t-\tau_{ij}(t))\|_{\mathbb{H}} < \epsilon. \end{eqnarray}$ $

(3.10)

Since $ x $ is a solution of system (1.1), by (3.2), for $ ij\in\Lambda $, we have

$\begin{align} &\|x_{ij}(t+\sigma)-x_{ij}(t)\|_{\mathbb{H}}\\ \leq&\Bigg\|\int_{-\infty}^{t}e^{-\int_{s}^{t}{a_{ij}^R (v+\sigma)}dv}\Big(\check{a}_{ij}(s+\sigma)x_{ij}(s+\sigma)-\check{a}_{ij}(s)x_{ij}(s)\Big)ds\bigg\|_{\mathbb{H}}\\ &+\Bigg\|\int_{-\infty}^{t}e^{-\int_{s}^{t}{a_{ij}^R (v+\sigma)}dv}\sum\limits_{C_{kl}\in N_r(i, j)}\Big[B_{ij}^{kl}(s+\sigma)f_{ij}(x_{kl}(s+\sigma))x_{ij}(s+\sigma)\\ &-B_{ij}^{kl}(s)f_{ij}(x_{kl}(s))x_{ij}(s)\Big]ds\Bigg\|_{\mathbb{H}} +\Bigg\|\int_{-\infty}^{t}e^{-\int_{s}^{t}{a_{ij}^R (v+\sigma)}dv}\sum\limits_{C_{kl}\in N_s(i, j)}\Big[C_{ij}^{kl}(s+\sigma)\\ &\times g_{ij}(x_{kl}(s+\sigma-\tau_{kl}(s+\sigma))) x_{ij}(s+\sigma)-C_{ij}^{kl}(s)g_{ij}(x_{kl}(s-\tau_{kl}(s))) x_{ij}(s)\Big]ds\Bigg\|_{\mathbb{H}}\\ & +\Bigg\|\int_{-\infty}^{t}e^{-\int_{s}^{t}{a_{ij}^R (v+\sigma)}dv}\big[I_{ij}(s+\sigma)-I_{ij}(s)\big]ds\Bigg\|_{\mathbb{H}}\\ &+\Bigg\|\int_{-\infty}^{t}\bigg(e^{-\int_{s}^{t}{a_{ij}^R (v+\sigma)}dv}-e^{-\int_{s}^{t}{a_{ij}^R (v)}dv}\bigg)\check{a}_{ij}(s)x_{ij}(s)ds\bigg\|_{\mathbb{H}}\\ &+\Bigg\|\int_{-\infty}^{t}\bigg(e^{-\int_{s}^{t}{a_{ij}^R (v+\sigma)}dv}-e^{-\int_{s}^{t}{a_{ij}^R (v)}dv}\bigg)\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl}(s)f_{ij}(x_{kl}(s))x_{ij}(s)ds\Bigg\|_{\mathbb{H}}\\ &+\Bigg\|\int_{-\infty}^{t}\bigg(e^{-\int_{s}^{t}{a_{ij}^R (v+\sigma)}dv}-e^{-\int_{s}^{t}{a_{ij}^R (v)}dv}\bigg)\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl}(s)g_{ij}(x_{kl}(s-\tau_{kl}(s))) x_{ij}(s)ds\Bigg\|_{\mathbb{H}} \\ &+\Bigg\|\int_{-\infty}^{t}\bigg(e^{-\int_{s}^{t}{a_{ij}^R (v+\sigma)}dv}-e^{-\int_{s}^{t}{a_{ij}^R (v)}dv}\bigg)I_{ij}(s)ds\Bigg\|_{\mathbb{H}}\\ : = &\sum\limits_{l = 1}^8B_{lij}(t). \end{align}$ $

(3.11)

When $ p > 2 $, it follows from Hölder's inequality $ \left(\frac{2}{p}, \frac{p-2}{p}\right) $, Hölder's inequality $ \left(\frac{1}{2}, \frac{1}{2}\right) $ and $ (H_2) $ that

$\begin{align} B_{2ij}(t) \leq&\int_{-\infty}^{t}e^{-a^m(t-s)}\sum\limits_{C_{kl}\in N_r(i, j)}\Big\|B_{ij}^{kl}(s+\sigma)f_{ij}(x_{kl}(s+\sigma))(x_{ij}(s+\sigma)-x_{ij}(s))\Big\|_{\mathbb{H}}ds\\ &+\int_{-\infty}^{t}e^{-a^m(t-s)}\sum\limits_{C_{kl}\in N_r(i, j)}\Big\|B_{ij}^{kl}(s+\sigma)(f_{ij}(x_{kl}(s+\sigma))-f_{ij}(x_{kl}(s)))x_{ij}(s)\Big\|_{\mathbb{H}}ds\\ &+\int_{-\infty}^{t}e^{-a^m(t-s)}\sum\limits_{C_{kl}\in N_r(i, j)}\Big\|(B_{ij}^{kl}(s+\sigma)-B_{ij}^{kl}(s))f_{ij}(x_{kl}(s))x_{ij}(s)\Big\|_{\mathbb{H}}ds\\ \leq&\bigg(\int_{-\infty}^{t} e^{-\frac{p}{2p-4}a^m(t-s)}ds\bigg)^{\frac{p-2}{p}}\bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L_{ij}^f\alpha\|x_{ij}(s+\sigma) -x_{ij}(s))\|_{\mathbb{H}}\bigg)^{\frac{p}{2}}ds\bigg]^{\frac{2}{p}}\\ & +\bigg(\int_{-\infty}^{t} e^{-\frac{p}{2p-4}a^m(t-s)}ds\bigg)^{\frac{p-2}{p}} \bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L_{ij}^f\alpha\|x_{kl}(s+\sigma)-x_{kl}(s)\|_{\mathbb{H}} \bigg)^{\frac{p}{2}}ds\bigg]^{\frac{2}{p}}\\ &+\bigg(\int_{-\infty}^{t} e^{-\frac{p}{2p-4}a^m(t-s)}ds\bigg)^{\frac{p-2}{p}} \bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_r(i, j)}4L_{ij}^f\alpha^2 |B_{ij}^{kl}(s+\sigma)-B_{ij}^{kl}(s)|ds\bigg) ^{\frac{p}{2}}ds\bigg]^{\frac{2}{p}} \\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{\frac{p-2}{p}}\Bigg\{\bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L_{ij}^f\alpha\|x_{ij}(s+\sigma)-x_{ij}(s))\|_{\mathbb{H}}\bigg) ^{\frac{p}{2}}ds\bigg]^{\frac{2}{p}}\\ &+ \bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L^f_{ij}\alpha\|x_{kl}(s+\sigma))-x_{kl}(s)\|_{\mathbb{H}} \bigg)^{\frac{p}{2}}ds\bigg]^{\frac{2}{p}}\\ &+\bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_r(i, j)}4L_{ij}^f\alpha^2 |B_{ij}^{kl}(s+\sigma)-B_{ij}^{kl}(s)|ds\bigg) ^{\frac{p}{2}}ds\bigg]^{\frac{2}{p}}\Bigg\}\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{\frac{p-2}{p}}\Bigg\{ \bigg[\bigg(\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}ds\bigg) ^{\frac{1}{2}}\bigg(\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L_{ij}^f\alpha\|x_{ij}(s+\sigma)\\ &-x_{ij}(s))\|_{\mathbb{H}}\bigg) ^{p}ds\bigg)^{\frac{1}{2}}\bigg]^{\frac{2}{p}}+\bigg[\bigg(\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}ds\bigg) ^{\frac{1}{2}} \bigg(\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L^f_{ij}\alpha\\ &\times\|x_{kl}(s+\sigma))-x_{kl}(s)\|_{\mathbb{H}}\bigg) ^{p}ds\bigg)^{\frac{1}{2}}\bigg]^{\frac{2}{p}}+\bigg[\bigg(\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}ds\bigg) ^{\frac{1}{2}}\bigg(\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\\ &\times\bigg(\sum\limits_{C_{kl}\in N_r(i, j)}4L_{ij}^f\alpha^2 |B_{ij}^{kl}(s+\sigma)-B_{ij}^{kl}(s)|\bigg) ^{p}ds\bigg)^{\frac{1}{2}}\bigg]^{\frac{2}{p}}\Bigg\}\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{\frac{p-2}{p}}\bigg(\frac{4}{a^mp}\bigg)^{\frac{1}{p}} \Bigg\{\bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L_{ij}^f\alpha\|x_{ij}(s+\sigma)-x_{ij}(s))\|_{\mathbb{H}}\bigg) ^{p}ds\bigg]^{\frac{1}{p}}\\ &+\bigg[ \int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L^f_{ij}\alpha\|x_{kl}(s+\sigma))-x_{kl}(s)\|_{\mathbb{H}}\bigg) ^{p}ds\bigg]^{\frac{1}{p}}\\ &+\bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_r(i, j)}4L_{ij}^f\alpha^2 |B_{ij}^{kl}(s+\sigma)-B_{ij}^{kl}(s)|\bigg)^{p}ds\bigg]^{\frac{1}{p}}\Bigg\}, \end{align}$ $

(3.12)

for $ ij\in\Lambda $. Similarly, we have

$\begin{align} B_{1ij}(t) \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{\frac{p-2}{p}}\bigg(\frac{4}{a^mp}\bigg)^{\frac{1}{p}} \Bigg\{\check{a}_{ij}^M\bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \|x_{ij}(s+\sigma)-x_{ij}(s)\|^p_{\mathbb{H}}ds\bigg]^{\frac{1}{p}}\\ &+2\alpha\bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \|\check{a}_{ij}(s+\sigma)-\breve{a}_{ij}(s)\|^p_{\mathbb{H}}ds\bigg]^{\frac{1}{p}}\Bigg\}, \end{align}$ $

(3.13)

$\begin{align} B_{3ij}(t) \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{\frac{p-2}{p}}\bigg(\frac{4}{a^mp}\bigg)^{\frac{1}{p}} \Bigg\{\bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L_{ij}^g\alpha\|x_{ij}(s+\sigma)-x_{ij}(s))\|_{\mathbb{H}}\bigg) ^{p}ds\bigg]^{\frac{1}{p}}\\ &+\bigg[ \int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\|x_{kl}(s+\sigma-\tau_{kl}(s+\sigma)) -x_{kl}(s-\tau_{kl}(s))\|_{\mathbb{H}}\bigg) ^{p}ds\bigg]^{\frac{1}{p}}\\ &+\bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}4L_{ij}^g\alpha^2 |C_{ij}^{kl}(s+\sigma)-C_{ij}^{kl}(s))|\bigg) ^{p}ds\bigg]^{\frac{1}{p}}\Bigg\} \end{align}$ $

(3.14)

and

$\begin{align} B_{4ij}(t) \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{\frac{p-2}{p}}\bigg(\frac{4}{a^mp}\bigg)^{\frac{1}{p}} \bigg(\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \|I_{ij}(s+\sigma)-I_{ij}(s)\|^{p}_{\mathbb{H}}ds\bigg)^{\frac{1}{p}}, \end{align}$ $

(3.15)

for $ ij\in\Lambda $.

Besides, combining with Hölder's inequality $ \left(\frac{2}{p}, \frac{p-2}{p}\right) $, Hölder's inequality $ \left(\frac{1}{2}, \frac{1}{2}\right) $ and $ (H_2) $ that

$\begin{align} B_{5ij}(t) \leq&2\check{a}_{ij}^M\alpha\int_{-\infty}^{t}e^{-a^m(t-s)} \bigg(\int_s^t|a_{ij}^R(v+\sigma)-a_{ij}^R(v)|dv\bigg) ds\\ \leq&2\check{a}_{ij}^M\alpha \bigg(\int_{-\infty}^{t} e^{-\frac{p}{2p-4}a^m(t-s)}ds\bigg)^{\frac{p-2}{p}} \bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\int_s^t|a_{ij}^R(v+\sigma) -a_{ij}^R(v)|dv\bigg)^{\frac{p}{2}} ds\bigg]^{\frac{2}{p}}\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{\frac{p-2}{p}}2\check{a}_{ij}^M\alpha \bigg[\bigg(\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}ds\bigg)^{\frac{1}{2}} \bigg(\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\int_s^t|a_{ij}^R(v+\sigma)\\ &-a_{ij}^R(v)|dv\bigg)^{p} ds\bigg)^{\frac{1}{2}}\bigg]^{\frac{2}{p}}\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{\frac{p-2}{p}}\bigg(\frac{4}{a^mp}\bigg)^{\frac{1}{p}} 2\check{a}_{ij}^M\alpha \bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\int_s^t|a_{ij}^R(v+\sigma) -a_{ij}^R(v)|dv\bigg)^{p} ds\bigg]^{\frac{1}{p}}, \end{align}$ $

(3.16)

for $ ij\in\Lambda $. In a similar way, one can get that

$\begin{align} B_{6ij}(t) \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{\frac{p-2}{p}}\bigg(\frac{4}{a^mp}\bigg)^{\frac{1}{p}} \sum\limits_{C_{kl}\in N_r(i, j)}4B_{ij}^{kl^M}L_{ij}^f\alpha^2 \bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\int_s^t|a_{ij}^R(v+\sigma) -a_{ij}^R(v)|dv\bigg)^{p} ds\bigg]^{\frac{1}{p}}, \end{align}$ $

(3.17)

$\begin{align} B_{7ij}(t) \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{\frac{p-2}{p}}\bigg(\frac{4}{a^mp}\bigg)^{\frac{1}{p}} \sum\limits_{C_{kl}\in N_s(i, j)}4C_{ij}^{kl^M}L_{ij}^g\alpha^2\bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\int_s^t |a_{ij}^R(v+\sigma) -a_{ij}^R(v)|dv\bigg)^{p} ds\bigg]^{\frac{1}{p}} \end{align}$ $

(3.18)

and

$\begin{align} B_{8ij}(t) \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{\frac{p-2}{p}}\bigg(\frac{4}{a^mp}\bigg)^{\frac{1}{p}} \bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\int_s^t|a_{ij}^R(v+\sigma)-a_{ij}^R(v)|dv\bigg)^{p} \|I_{ij}(s)\|_{\mathbb{H}}^{p}ds\bigg]^{\frac{1}{p}}, \end{align}$ $

(3.19)

for $ ij\in\Lambda $. Hence, together with a change of variables, Fubini's theorem, Hölder's inequality, (3.8) and (3.13), we derive that

$\begin{align*} &\frac{1}{r}\int_\beta^{\beta+r}B^p_{1i{j}}(t)dt\nonumber\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{4}{a^mp}\Bigg\{\frac{1}{r}\int_\beta^{\beta+r} \Bigg[\bigg[\check{a}_{ij}^M\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \|x_{ij}(s+\sigma)-x_{ij}(s)\|^p_{\mathbb{H}}ds\bigg]^{\frac{1}{p}}\nonumber\\ &+2\alpha\bigg[\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \|\check{a}_{ij}(s+\sigma)-\breve{a}_{ij}(s)\|^p_{\mathbb{H}}ds\bigg]^{\frac{1}{p}}\Bigg]^pdt\Bigg\}\nonumber\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{8}{a^mp}\Bigg\{ (\check{a}_{ij}^M)^p\frac{1}{r}\int_\beta^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \|x_{ij}(s+\sigma)-x_{ij}(s)\|^p_{\mathbb{H}}dsdt\nonumber\\ &+(2\alpha)^p\frac{1}{r}\int_\beta^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \|\check{a}_{ij}(s+\sigma)-\breve{a}_{ij}(s)\|^p_{\mathbb{H}}dsdt\Bigg\}\nonumber\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{8}{a^mp} \Bigg\{(\check{a}_{ij}^M)^p\int_{-\infty}^{\beta}e^{-\frac{p}{4}a^m(\beta-s)}\bigg(\frac{1}{r}\int_s^{s+r} \|x_{ij}(t+\sigma)-x_{ij}(t)\|^p_{\mathbb{H}}dt\bigg)ds\nonumber\\ &+(2\alpha)^p\int_{-\infty}^{\beta}e^{-\frac{p}{4}a^m(\beta-s)}\bigg(\frac{1}{r}\int_s^{s+r} \|\check{a}_{ij}(t+\sigma)-\breve{a}_{ij}(t)\|^p_{\mathbb{H}}dt\bigg)ds\Bigg\}\nonumber\\ = &(\check{a}_{ij}^M)^p\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{8}{a^mp} \int_{-\infty}^{\beta}e^{-\frac{p}{4}a^m(\beta-s)}\Theta^{\sigma, r}(s)ds+\rho_{1ij}, \end{align*}$ $

where

$\begin{align*} \Theta^{\sigma, r}(s): = &\frac{1}{r}\int_s^{s+r}\|x(t+\sigma)-x(t)\|^p_{\mathbb{H}^n}dt \end{align*}$ $

and

$\begin{align} \rho_{1ij}: = &2\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\bigg(\frac{4}{a^mp}\bigg)^2(2\alpha\epsilon)^p, \end{align}$ $

(3.20)

and, together with a change of variables, Fubini's theorem, Hölder's inequality, (3.6) and (3.12), we derive that

$\begin{align*} &\frac{1}{r}\int_\beta^{\beta+r}B^p_{2ij}(t)dt\nonumber\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{4}{a^mp} \Bigg\{\frac{1}{r}\int_\beta^{\beta+r}\bigg[ \bigg(\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L_{ij}^f\alpha\|x_{ij}(s+\sigma)-x_{ij}(s))\|_{\mathbb{H}}\bigg) ^{p}ds\bigg)^{\frac{1}{p}}\nonumber\\ &+\bigg( \int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L^f_{ij}\alpha\|x_{kl}(s+\sigma)-x_{kl}(s)\|_{\mathbb{H}}\bigg) ^{p}ds\bigg)^{\frac{1}{p}}\nonumber\\ &+\bigg(\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_r(i, j)}4L_{ij}^f\alpha^2 |B_{ij}^{kl}(s+\sigma) -B_{ij}^{kl}(s)|\bigg) ^{p}ds\bigg)^{\frac{1}{p}}\bigg]^pdt\Bigg\}\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{12}{a^mp}\Bigg\{ \frac{1}{r}\int_\beta^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L_{ij}^f\alpha\|x_{ij}(s+\sigma)-x_{ij}(s)\|_{\mathbb{H}}\bigg) ^{p}dsdt\nonumber\\ & +\frac{1}{r}\int_\beta^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L^f_{ij}\alpha\|x_{kl}(s+\sigma)-x_{kl}(s)\|_{\mathbb{H}}\bigg) ^{p}dsdt\nonumber\\ &+\frac{1}{r}\int_\beta^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_r(i, j)}4L_{ij}^f\alpha^2 |B_{ij}^{kl}(s+\sigma)-B_{ij}^{kl}(s)|\bigg) ^{p}dsdt\Bigg\}\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{12}{a^mp}\Bigg[\bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L_{ij}^f\alpha\bigg)^{p} \frac{1}{r}\int_\beta^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \|x(s+\sigma)-x(s)\|^p_{\mathbb{H}^{m\times n}}dsdt\nonumber\\ &+\bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L^f_{ij}\alpha\bigg)^{p} \frac{1}{r}\int_\beta^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \|x(s+\sigma)-x(s)\|_{\mathbb{H}^{m\times n}} dsdt\nonumber\\ &+\frac{1}{r}\int_\beta^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(4mnL_{ij}^f\alpha^2\epsilon\bigg)^pdsdt\Bigg]\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{12}{a^mp}\Bigg[\bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L_{ij}^f\alpha\bigg)^{p} \int_{-\infty}^{\beta}e^{-\frac{p}{4}a^m(\beta-s)}\bigg(\frac{1}{r}\int_s^{s+r} \|x(t+\sigma)-x(t)\|^p_{\mathbb{H}^{m\times n}}dt\bigg)ds\nonumber\\ &+\bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L^f_{ij}\alpha\bigg)^{p} \int_{-\infty}^{\beta}e^{-\frac{p}{4}a^m(\beta-s)} \bigg(\frac{1}{r}\int_s^{s+r}\|x(t+\sigma)-x(t)\|_{\mathbb{H}^{m\times n}} dt\bigg)ds +\frac{4}{a^mp}\big(4mnL_{ij}^f\alpha^2\epsilon\big) ^{p}\Bigg]\\ = &\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{24}{a^mp}\bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L_{ij}^f\alpha\bigg)^{p} \int_{-\infty}^{\beta}e^{-\frac{p}{4}a^m(\beta-s)}\Theta^{\sigma, r}(s)ds +\rho_{2ij}, \end{align*}$ $

where

$\begin{align} \rho_{2ij}: = &3\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\bigg(\frac{4}{a^mp}\bigg)^2\big(4mnL_{ij}^f\alpha^2\epsilon\big) ^{p}. \end{align}$ $

(3.21)

Moreover, based on a change of variables, Fubini's theorem, Hölder's inequality, (3.7), (3.10) and (3.14), we deduce that

$\begin{align*} &\frac{1}{r}\int_\beta^{\beta+r}B^p_{3ij}(t)dt\nonumber\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{4}{a^mp} \Bigg\{\frac{1}{r}\int_\beta^{\beta+r}\bigg[ \bigg(\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L_{ij}^g\alpha\|x_{ij}(s+\sigma)-x_{ij}(s))\|_{\mathbb{H}}\bigg) ^{p}ds\bigg)^{\frac{1}{p}}\nonumber\\ &+\bigg(\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\|x_{kl}(s+\sigma-\tau_{kl}(s+\sigma))-x_{kl}(s-\tau_{kl}(s))\|_{\mathbb{H}}\bigg) ^{p}ds\bigg)^{\frac{1}{p}}\nonumber\\ &+\bigg(\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}4L_{ij}^g\alpha^2 \|(C_{ij}^{kl}(s+\sigma)-C_{ij}^{kl}(s))\|_{\mathbb{H}}\bigg) ^{p}ds\bigg)^{\frac{1}{p}}\bigg]^pdt\Bigg\}\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{12}{a^mp}\Bigg\{ \frac{1}{r}\int_\beta^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L_{ij}^g\alpha\|x_{ij}(s+\sigma)-x_{ij}(s))\|_{\mathbb{H}}\bigg) ^{p}dsdt\nonumber\\ &+\frac{1}{r}\int_\beta^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\|x_{kl}(s+\sigma-\tau_{kl}(s+\sigma))-x_{kl}(s-\tau_{kl}(s))\|_{\mathbb{H}}\bigg) ^{p}dsdt\nonumber\\ &+\frac{1}{r}\int_\beta^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}4L_{ij}^g\alpha^2 |C_{ij}^{kl}(s+\sigma)-C_{ij}^{kl}(s))|\bigg) ^{p}dsdt\Bigg\}\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{12}{a^mp}\Bigg\{ \frac{1}{r}\int_\beta^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L_{ij}^g\alpha\|x_{ij}(s+\sigma)-x_{ij}(s))\|_{\mathbb{H}}\bigg) ^{p}dsdt\nonumber\\ &+\frac{1}{r}\int_\beta^{\beta+r}\bigg[ 2\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\|x_{kl}(s+\sigma-\tau_{kl}(s+\sigma))-x_{kl}(s-\tau_{kl}(s+\sigma))\|_{\mathbb{H}}\bigg)^pd s\nonumber\\ &+2\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\sum\limits_{C_{kl}\in N_s(i, j)}4C_{ij}^{kl^M}L^g_{ij}\alpha^2\|x_{kl}(s-\tau_{kl}(s+\sigma)) -x_{kl}(s-\tau_{kl}(s))\|_{\mathbb{H}}\bigg)^pd s\bigg]dt\nonumber\\ &+ \frac{1}{r}\int_\beta^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(4mnL_{ij}^g\alpha^2\epsilon\bigg)^{p}dsdt\Bigg\}\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{12}{a^mp}\Bigg\{ \frac{1}{r}\int_\beta^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L_{ij}^g\alpha\|x_{ij}(s+\sigma)-x_{ij}(s))\|_{\mathbb{H}}\bigg) ^{p}dsdt\nonumber\\ &+\frac{1}{r}\int_\beta^{\beta+r}\bigg[\frac{2}{1-\tau'}\int_{-\infty}^{t-\tau_{kl}(t+\sigma)} e^{-\frac{p}{4}a^m(t-u-\tau)}\bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\|x_{kl}(u+\sigma)-x_{kl}(u)\|_{\mathbb{H}}\bigg)^p d u\nonumber\\ &+2\bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\epsilon\bigg)^p \int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}ds\bigg]dt +\frac{4}{a^mp}\bigg(4mnL_{ij}^g\alpha^2\epsilon\bigg)^{p} \Bigg\}\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{12}{a^mp}\Bigg\{\frac{1}{r}\int_\beta^{\beta+r} \int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L_{ij}^g\alpha\|x_{ij}(s+\sigma)-x_{ij}(s))\|_{\mathbb{H}}\bigg)^{p}dsdt\nonumber\\ &+\frac{1}{r}\int_\beta^{\beta+r}\bigg[\frac{2e^{\frac{p}{4}a^m\tau}}{1-\tau'}\int_{-\infty}^{t} e^{-\frac{p}{4}a^m(t-s)}\bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha \|x_{kl}(s+\sigma)-x_{kl}(s)\|_{\mathbb{H}}\bigg)^pds\bigg]dt\\ &+\frac{8}{a^mp}\bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\epsilon\bigg)^p +\frac{4}{a^mp}\bigg(4mnL_{ij}^g\alpha^2\epsilon\bigg)^{p} \Bigg\}\nonumber\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{12}{a^mp}\bigg[\bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L_{ij}^g\alpha\bigg)^{p} \int_{-\infty}^{\beta}e^{-\frac{p}{4}a^m(\beta-s)}\bigg( \frac{1}{r}\int_s^{s+r}\|x(t+\sigma)-x(t))\|_{\mathbb{H}^{m\times n}}dt\bigg)ds\nonumber\\ &+\frac{2e^{\frac{p}{4}a^m\tau}}{1-\tau'} \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\bigg)^p \int_{-\infty}^{\beta}e^{-\frac{p}{4}a^m(\beta-s)}\bigg(\frac{1}{r}\int_s^{s+r}\|x(t+\sigma)-x(t)\| ^{p}_{\mathbb{H}^{m\times n}}dt\bigg)ds \nonumber\\ &+\frac{8}{a^mp} \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\epsilon\bigg)^p+\frac{4}{a^mp} \bigg(4mnL_{ij}^g\alpha^2\epsilon\bigg) ^{p}\bigg]\\ = &\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{12}{a^mp} \bigg(1+\frac{2e^{\frac{p}{4}a^m\tau}}{1-\tau'}\bigg) \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\bigg)^p\int_{-\infty}^{\beta} e^{-\frac{p}{4}a^m(\beta-s)}\Theta^{\sigma, r}(s)ds +\rho_{3ij}, \end{align*}$ $

where

$\begin{align} \rho_{3ij}: = &3\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\bigg(\frac{4}{a^mp}\bigg)^2 \bigg[2\bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\bigg)^p +\bigg(4mnL_{ij}^g\alpha^2\bigg)^p\bigg]\epsilon^p. \end{align}$ $

(3.22)

In view of (3.5), (3.8) and (3.15)–(3.19), we can easily obtain that

$\begin{align} &\frac{1}{r}\int_\beta^{\beta+r}B^p_{4ij}(t)dt\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{4}{a^mp} \int_{-\infty}^{\beta}e^{-\frac{p}{4}a^m(\beta-s)}\bigg(\frac{1}{r}\int_s^{s+r} \|I_{ij}(t+\sigma)-I_{ij}(t)\|^{p}_{\mathbb{H}}dt\bigg)ds\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\bigg(\frac{4}{a^mp}\bigg)^{2}\epsilon^p : = \rho_{4ij}, \end{align}$ $

(3.23)

$\begin{align} &\frac{1}{r}\int_\beta^{\beta+r}B^p_{5ij}(t)dt\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{4}{a^mp} \bigg(2\check{a}_{ij}^M\alpha\bigg)^p \bigg[\frac{1}{r}\int_{\beta}^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\bigg(\int_s^t |a_{ij}^R(v+\sigma)\\ &-a_{ij}^R(v)|dv\bigg)^{p}dsdt\bigg]\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{4}{a^mp} \bigg(2\check{a}_{ij}^M\alpha\epsilon\bigg)^p \int_{0}^{\infty}e^{-\frac{p}{4}a^ms}s^pds : = \rho_{5ij}, \end{align}$ $

(3.24)

$\begin{align} &\frac{1}{r}\int_\beta^{\beta+r}B^p_{6ij}(t)dt\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{4}{a^mp} \bigg(\sum\limits_{C_{kl}\in N_r(i, j)}4B_{ij}^{kl^M}L_{ij}^f\alpha^2\bigg)^p \bigg[\frac{1}{r}\int_{\beta}^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\\ &\times\bigg(\int_s^t |a_{ij}^R(v+\sigma)-a_{ij}^R(v)|dv\bigg)^{p}dsdt\bigg]\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{4}{a^mp} \bigg(\sum\limits_{C_{kl}\in N_r(i, j)}4B_{ij}^{kl^M}L_{ij}^f\alpha^2\epsilon\bigg)^p \int_{0}^{\infty}e^{-\frac{p}{4}a^ms}s^pds : = \rho_{6ij}, \end{align}$ $

(3.25)

$\begin{align} &\frac{1}{r}\int_\beta^{\beta+r}B^p_{7ij}(t)dt\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{4}{a^mp} \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}4C_{ij}^{kl^M}L_{ij}^g\alpha^2\bigg)^p\bigg[\frac{1}{r}\int_{\beta}^{\beta+r} \int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}\\ &\times\bigg(\int_s^t|a_{ij}^R(v+\sigma)-a_{ij}^R(v)|dv\bigg)^{p} dsdt\bigg]\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{4}{a^mp} \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}4C_{ij}^{kl^M}L_{ij}^g\alpha^2\epsilon\bigg)^p \int_{0}^{\infty}e^{-\frac{p}{4}a^ms}s^pds : = \rho_{7ij} \end{align}$ $

(3.26)

and

$\begin{align} &\frac{1}{r}\int_\beta^{\beta+r}B^p_{8ij}(t)dt\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{4}{a^mp}\bigg[ \frac{1}{r}\int_\beta^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)} \bigg(\int_s^t|a_{ij}^R(v+\sigma)-a_{ij}^R(v)|dv\bigg)^{p}\|I_{ij}(s)\|^p_{\mathbb{H}}dsdt\bigg]\\ \leq&\epsilon^p\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{4}{a^mp}\bigg[ \frac{1}{r}\int_\beta^{\beta+r}\int_{-\infty}^{t}e^{-\frac{p}{4}a^m(t-s)}(t-s)^p \|I_{ij}(s)\|^p_{\mathbb{H}}dsdt\bigg]\\ \leq&\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{4}{a^mp}\|I_{ij}\|^p_{W^p}\epsilon^p \int_{0}^{\infty}e^{-\frac{p}{4}a^ms}s^pds : = \rho_{8ij}, \end{align}$ $

(3.27)

for $ ij\in\Lambda $. Consequently, combining with (3.11) and (3.20)–(3.27), we obtain

$\begin{align*} \Theta^{\sigma, r}(\beta) \leq&\max\limits_{ij\in\Lambda}\Bigg\{8\sum\limits_{l = 1}^8\frac{1}{r}\int_\beta^{\beta+r}B^p_{lij}(t)dt\Bigg\}\\ \leq&\rho+\gamma\int_{-\infty}^{\beta}e^{-\eta(\beta-s)}\Theta^{\sigma, r}(s)ds, \end{align*}$ $

where $ \eta = \frac{p}{4}a^m $,

$\begin{align*} \rho = &8\sum\limits_{l = 1}^6\max\limits_{ij\in\Lambda}\{\rho_{lij}\}\nonumber\\ = &8\max\limits_{ij\in\Lambda}\Bigg\{\bigg(\frac{2p-4}{a^mp}\bigg)^{p-2} \frac{4}{a^mp}\Bigg[\frac{8}{a^mp}(2\alpha)^p+\frac{12}{a^mp}(4mnL_{ij}^f\alpha^2) ^{p}\\ &+\frac{24}{a^mp}\bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\bigg)^P+\frac{12}{a^mp}(4mnL_{ij}^g\alpha^2)^p+\frac{4}{a^mp} +\bigg[\bigg(2\check{a}_{ij}^M\alpha\bigg)^p \\ &+\bigg(\sum\limits_{C_{kl}\in N_r(i, j)}4B_{ij}^{kl^M}L_{ij}^f\alpha^2\bigg)^p+\bigg(\sum\limits_{C_{kl}\in N_s(i, j)}4C_{ij}^{kl^M}L_{ij}^g\alpha^2\bigg)^p\\ &+\|I_{ij}\|^p_{W^p}\bigg]\int_{0}^{\infty}e^{-\frac{p}{4}a^ms}s^pds\Bigg] \Bigg\}\epsilon^p \end{align*}$ $

and

$\begin{align*} \gamma = &8\max\limits_{ij\in\Lambda}\Bigg\{ \bigg(\frac{2p-4}{a^mp}\bigg)^{p-2}\frac{12}{a^mp}\bigg[\frac{2}{3}(\check{a}^M_{ij})^p+ 2\bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L^f_{ij}\alpha\bigg)^{p} \\ &+\bigg(1+\frac{2e^{\frac{p}{4}a^m\tau}}{1-\tau'}\bigg) \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\bigg)^p\bigg]\Bigg\}. \end{align*}$ $

By $ (H_4) $, we have $ \gamma < \eta $. Thus, it follows from Lemma 2.3 that

$\begin{eqnarray*} \frac{1}{r}\int_{\beta}^{\beta+r} \|x(t+\sigma)-x(t)\| ^{p}_{\mathbb{H}^{m\times n}}dt\leq \rho\frac{\eta}{\eta-\gamma}. \end{eqnarray*}$ $

Hence, $ x\in APW^p(\mathbb{R}, \mathbb{H}^{m\times n}) $.

When $ p = 2 $, similar to the proof of the case of $ p > 2 $, one can obtain

$\begin{eqnarray*} \Theta^{\sigma, r}(\beta) \leq\tilde{\rho}+\tilde{\gamma}\int_{-\infty}^{\beta}e^{-\tilde{\eta}(\beta-s)}\Theta^{\sigma, r}(s)ds, \end{eqnarray*}$ $

where $ \tilde{\eta} = \frac{a^m}{2} $,

$\begin{align*} \tilde{\rho} = & 8\sum\limits_{l = 1}^8\max\limits_{ij\in\Lambda}\{\tilde{\rho}_{lij}\}\nonumber\\ = &\max\limits_{ij\in\Lambda}\Bigg\{ \frac{16}{a^m}\Bigg[\frac{4}{a^m}(2\alpha)^2+\frac{6}{a^m}(4mnL_{ij}^f\alpha^2\epsilon) ^{2}+\frac{12}{a^m}\bigg(2\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl^M}L^g_{ij}\alpha\bigg)^2\\ &+\frac{6}{a^m}(4mnL_{ij}^g\alpha^2)^2+\frac{2}{a^m} +\bigg[\bigg(2\check{a}_{ij}^M\alpha\bigg)^2+\bigg(\sum\limits_{C_{kl}\in N_r(i, j)}4B_{ij}^{kl^M}L_{ij}^f\alpha^2\bigg)^2 \\ &+\bigg(\sum\limits_{C_{kl}\in N_s(i, j)}4C_{ij}^{kl^M}L_{ij}^g\alpha^2\bigg)^2 +\|I_{ij}\|^2_{W^2}\bigg]\int_{0}^{\infty}e^{-\frac{1}{2}a^ms}s^2ds\Bigg] \Bigg\}\epsilon^2 \end{align*}$ $

and

$\begin{align*} \tilde{\gamma} = &\max\limits_{ij\in\Lambda}\Bigg\{ \frac{48}{a^m}\bigg[\frac{2}{3}(\check{a}^M_{ij})^2+2 \bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L^f_{ij}\alpha\bigg)^{2}\\ &+\bigg(1+\frac{2e^{\frac{1}{2}a^m\tau}}{1-\tau'}\bigg) \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\bigg)^2\bigg]\Bigg\}. \end{align*}$ $

By $ (H_4) $, we have $ \tilde{\gamma} < \tilde{\eta} $. Thus, it follows from Lemma 2.3 that

$\begin{eqnarray*} \frac{1}{r}\int_{\beta}^{\beta+r} \|x(t+\sigma)-x(t)\| ^{p}_{\mathbb{H}^{m\times n}}dt\leq \tilde{\rho}\frac{\tilde{\eta}}{\tilde{\eta}-\tilde{\gamma}}, \end{eqnarray*}$ $

which means that $ x\in APW^2(\mathbb{R}, \mathbb{H}^{m\times n}) $. The proof is complete.

Definition 3.1. ^[14] Let $ x $ be a solution of system (1.1) with the initial value $ \varphi $ and $ y $ be an arbitrary solution of system (1.1) with the initial value $ \psi $, respectively. If there exist positive constants $ \lambda $ and $ M $ such that

$\begin{eqnarray*} \|x(t)-y(t)\|_{\mathbb{H}^{m\times n}}\leq M e^{-\lambda t}\|\varphi-\psi\|_{\tau}, \, \, t\in\mathbb{R}^{+}, \end{eqnarray*}$ $

where $ \|\varphi-\psi\|_{\tau} = \sup\limits_{t\in[-\tau, 0]}\|\varphi(t)-\psi(t)\|_{\mathbb{H}^{m\times n}}. $ Then the solution $ x $ of system (1.1) is said to be globally exponentially stable.

Theorem 3.2. Assume that $ (H_{1}) $–$ (H_{3}) $ hold, then system (1.1) has a unique $ W^p $-almost periodic solution that is globallyexponentially stable.

Proof. Let $ x(t) $ be the $ W^p $-almost periodic solution with the initial value $ \varphi(t) $ and $ y(t) $ be an arbitrary solution with the initial value $ \psi(t) $. Taking

$ z_{ij}(t) = x_i(t)-y_{ij}(t), \quad \phi_{ij}(t) = \varphi_{ij}(t)-\psi_{ij}(t), \quad ij\in\Lambda, $

we have

$\begin{align} \dot{z}_{ij}(t) = &-a_{ij}(t)z_{ij}(t)-\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl}(t)(f_{ij}(x_{kl}(t))x_{ij}(t)\\ &-f_{ij}(y_{kl}(t))y_{ij}(t)-\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl}(t)(g_{ij}(x_{kl}(t-\tau_{kl}(t)))x_{ij}(t)\\ &-g_{ij}(y_{kl}(t-\tau_{kl}(t)))y_{ij}(t)), \quad ij\in\Lambda. \end{align}$ $

(3.28)

For $ ij\in\Lambda $, we define the following functions:

$ \Pi_{ij}(u) = a^m-u-\bigg(\check{a}_{ij}^M+\sum\limits_{C_{kl}\in N_r(i, j)}4B_{ij}^{kl^M}L_{ij}^{f}\alpha +\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl}L_{ij}^{g}\alpha e^{u\tau^M_{ij}} +\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl}L_{ij}^{g}\alpha\bigg). $

From $ (H_3) $, we get

$ \Pi_{ij}(0) = a^m-\bigg(\check{a}_{ij}^M+\sum\limits_{C_{kl}\in N_r(i, j)}4B_{ij}^{kl^M}L_{ij}^f\alpha+\sum\limits_{C_{kl}\in N_s(i, j)}4C_{ij}^{kl^M} L_{ij}^g\alpha\bigg) > 0, \quad ij\in\Lambda. $

Since $ \Pi_{ij}(u) $ is continuous on $ [0, +\infty) $ and $ \Pi_{ij}(u)\rightarrow -\infty $ as $ u\rightarrow \infty $, there exists $ \zeta_{ij} > 0 $ such that $ \Pi_{ij}(\zeta_{ij}) = 0 $ and $ \Pi_{ij}(u) > 0 $, for $ u\in(0, \zeta_{ij}) $, $ ij\in\Lambda $. Let $ \varsigma = \min\limits_{ij\in\Lambda}\{\zeta_{ij}\} $, then we have $ \Pi_{ij}(\varsigma)\geq0, ij\in\Lambda $. Hence, we can choose a positive constant $ \lambda $ such that $ 0 < \lambda < \min\{\varsigma, a^m\} $ and $ \Pi_{ij}(\lambda) > 0 $. Thus, one has

$\begin{eqnarray} &&\frac{1}{a^m-\lambda}\bigg(\check{a}_{ij}^M+\sum\limits_{C_{kl}\in N_r(i, j)}4B_{ij}^{kl^M}L_{ij}^{f}\alpha +\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl}L_{ij}^{g}\alpha e^{\lambda\tau^M_{ij}}\\ &&+\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl}L_{ij}^{g}\alpha\bigg) < 1, \end{eqnarray}$ $

(3.29)

where $ ij\in\Lambda $. Take a constant $ M = \max\limits_{i\in\Lambda}\bigg\{\frac{a^m}{\check{a}_{ij}^M+\sum\limits_{C_{kl}\in N_r(i, j)}4B_{ij}^{kl^M}L_{ij}^f\alpha+\sum\limits_{C_{kl}\in N_s(i, j)}4C_{ij}^{kl^M} L_{ij}^g\alpha}\bigg\} $, then by $ (H_3) $, we have $ M > 1 $. Thus,

$\begin{eqnarray} &&\frac{1}{M}-\frac{1}{a^m-\lambda}\bigg(\check{a}_{ij}^M+\sum\limits_{C_{kl}\in N_r(i, j)}4B_{ij}^{kl^M}L_{ij}^{f}\alpha +\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl}L_{ij}^{g}\alpha e^{\lambda\tau^M_{ij}}\\ &&+\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl}L_{ij}^{g}\alpha\bigg) < 0, \quad ij\in\Lambda. \end{eqnarray}$ $

(3.30)

From (3.28), we have

$\begin{align} \dot{z}_{ij}(t)+a^R_{ij}(t)z_{ij}(t) = &-\check{a}_{ij}(t)z_{ij}(t)-\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl}(t)\Big(f_{ij}(x_{kl}(t))x_{ij}(t)-f_{ij}(y_{kl}(t))y_{ij}(t)\Big)\\ &-\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl}(t)\Big(g_{ij}(x_{kl}(t-\tau_{kl}(t)))x_{ij}(t)\\ &-g_{ij}(y_{kl}(t-\tau_{kl}(t)))y_{ij}(t)\Big), \quad ij\in\Lambda. \end{align}$ $

(3.31)

Multiplying both sides of (3.31) by $ e^{\int_{0}^{t}a_{ij}^R(v)dv} $ and integrating over $ [0, t] $, we have

$\begin{align*} z_{ij}(t) = &\phi_{ij}(0)e^{-\int_{0}^{t}a_{ij}^R(v)dv}+\int_{0}^{t}e^{-\int_{s}^{t}a_{ij}^R(v)dv} \bigg[-\check{a}_{ij}(t)z_{ij}(s)-\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl}(t)\Big(f_{ij}(x_{kl}(t))x_{ij}(t)\nonumber\\ &-f_{ij}(y_{kl}(t))y_{ij}(t)\Big)-\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl}(t)\Big(g_{ij}(x_{kl}(t-\tau_{kl}(t)))x_{ij}(t)\nonumber\\ &-g_{ij}(y_{kl}(t-\tau_{kl}(t)))y_{ij}(t)\Big)\bigg]ds, \quad ij\in\Lambda. \end{align*}$ $

Hence, for any $ \epsilon > 0 $, it is easy to see that

$ \|z(t)\|_{\mathbb{H}^{m\times n}} < (\|\phi\|_\tau+\epsilon)e^{-\lambda t} < M(\|\phi\|_\tau+\epsilon)e^{-\lambda t}, \quad \forall t\in(-\tau, 0]. $

We claim that

$\begin{eqnarray} \|z(t)\|_{\mathbb{H}^{m\times n}} < M(\|\phi\|_\tau+\epsilon)e^{-\lambda t}, \quad \forall t\in[0, +\infty). \end{eqnarray}$ $

(3.32)

Otherwise, there exists $ t^* > 0 $ such that

$\begin{eqnarray} \|z(t^*)\|_{\mathbb{H}^{m\times n}} = M(\|\phi\|_\tau+\epsilon)e^{-\lambda t^*} \end{eqnarray}$ $

(3.33)

and

$\begin{eqnarray} \|z(t)\|_{\mathbb{H}^{m\times n}} < M(\|\phi\|_\tau+\epsilon)e^{-\lambda t}, \quad t < t^*. \end{eqnarray}$ $

(3.34)

From (3.29), (3.30) and (3.34), we get

$\begin{align*} \|z_{ij}(t^*)\|_{\mathbb{H}}\leq&\|\phi_{ij}(0)\|_{\mathbb{H}} e^{-t^*a^m}+\int_{0}^{t^*}e^{-t^*a^m} \bigg[\check{a}_{ij}^M\|z_{ij}(s)\|_{\mathbb{H}}+\sum\limits_{C_{kl}\in N_r(i, j)}B_{ij}^{kl}(t)\nonumber\\ &\times\Big(\|f_{ij}(x_{kl}(t))(x_{ij}(t)-y_{ij}(t))\|_{\mathbb{H}}+\|(f_{ij}(x_{kl}(t)) -f_{ij}(y_{kl}(t)))y_{ij}(t)\|_{\mathbb{H}}\Big)\nonumber\\ &+\sum\limits_{C_{kl}\in N_s(i, j)}C_{ij}^{kl}(t)\Big(\|g_{ij}(x_{kl}(t-\tau_{kl}(t)))(x_{ij}(t)-y_{ij}(t))\|_{\mathbb{H}}\\ &+\|(g_{ij}(x_{kl}(t-\tau_{kl}(t)))-g_{ij}(y_{kl}(t-\tau_{kl}(t))))y_{ij}(t)\|_{\mathbb{H}}\Big)\bigg]ds\\ \leq&(\|\phi\|_\tau+\epsilon)e^{-a^mt^*}+M(\|\phi\|_\tau+\epsilon)\int_{0}^{t^*}e^{-(t^*-s)a^m} \bigg(\check{a}_{ij}^M+\sum\limits_{C_{kl}\in N_r(i, j)}4B_{ij}^{kl^M}L_{ij}^{f}\alpha\\ &+\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl}L_{ij}^{g}\alpha e^{\lambda\tau^M_{ij}} +\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl}L_{ij}^{g}\alpha\bigg)e^{-\lambda s}ds\\ \leq&(\|\phi\|_\tau+\epsilon)e^{-\lambda t^*}e^{(\lambda-a^m)t^*}+M(\|\phi\|_\tau+\epsilon)\frac{e^{-\lambda t^*}(1-e^{(\lambda-a^m)t^*})}{a^m-\lambda}\bigg(\check{a}_{ij}^M\\ &+\sum\limits_{C_{kl}\in N_r(i, j)}4B_{ij}^{kl^M}L_{ij}^{f}\alpha +\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl}L_{ij}^{g}\alpha e^{\lambda\tau^M_{ij}} +\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl}L_{ij}^{g}\alpha\bigg)\\ = &M(\|\phi\|_\tau+\epsilon)e^{-\lambda t^*}\bigg[\frac{e^{(\lambda-a^m)t^*}}{M}+\frac{1-e^{(\lambda-a^m)t^*}}{a^m-\lambda} \bigg(\check{a}_{ij}^M+\sum\limits_{C_{kl}\in N_r(i, j)}4B_{ij}^{kl^M}L_{ij}^{f}\alpha\\ &+\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl}L_{ij}^{g}\alpha e^{\lambda\tau^M_{ij}} +\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl}L_{ij}^{g}\alpha\bigg)\bigg]\\ = &M(\|\phi\|_\tau+\epsilon)e^{-\lambda t^*}\bigg[\bigg(\frac{1}{M}-\frac{1}{a^m-\lambda}\bigg(\check{a}_{ij}^M+\sum\limits_{C_{kl}\in N_r(i, j)}4B_{ij}^{kl^M}L_{ij}^{f}\alpha \\ &+\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl}L_{ij}^{g}\alpha e^{\lambda\tau^M_{ij}}+\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl}L_{ij}^{g}\alpha\bigg) e^{(\lambda-a^m)t^*}+\frac{1}{a^m-\lambda}\bigg(\check{a}_{ij}^M\\ &+\sum\limits_{C_{kl}\in N_r(i, j)}4B_{ij}^{kl^M}L_{ij}^{f}\alpha+\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl}L_{ij}^{g}\alpha e^{\lambda\tau^M_{ij}} +\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl}L_{ij}^{g}\alpha\bigg)\bigg]\\ < &M(\|\phi\|_\tau+\epsilon)e^{-\lambda t^*}, \quad ij\in\Lambda, \end{align*}$ $

which contradicts the Eq (3.33). Hence, (3.32) holds. Letting $ \epsilon\rightarrow0^+ $, from (3.32), we have

$\begin{eqnarray*} \|z(t)\|_{\mathbb{H}^{m\times n}}\leq M\|\phi\|_\tau e^{-\lambda t}, \quad \forall t > 0. \end{eqnarray*}$ $

Therefore, the $ W^p $-almost periodic solution of system (1.1) is globally exponentially stable. This completes the proof.

4. A numerical example

Example 4.1. In system (1.1), let $ i, j = 1, 2 $, $ r = s = 1 $ and take

$ x_{ij}(t) = x_{ij}^{R}(t)+ix_{ij}^{I}(t)+jx_{ij}^{J}(t)+kx_{ij}^{K}(t)\in\mathbb{H}, $

$ a_{11}(t) = 32\sin^2(\sqrt{3}t)+i\frac{\sqrt{2}}{5}\cos(2t)-j\frac{2}{5}\sin(\sqrt{5}t)+k\frac{1}{100}\cos^2(3t), $

$ a_{21}(t) = 39\left|\cos(\sqrt{2}t)\right|-i\frac{2}{10}\sin(2t)+j\frac{3}{10}\sin^2(\sqrt{7}t)+k\frac{\sqrt{3}}{100}\cos(t), $

$ a_{12}(t) = 33\cos^2(t)-i\frac{\sqrt{3}}{6}\sin(5t)-j\frac{1}{4}\cos^3(\sqrt{5}t)+k\frac{1}{6}\cos^2(3t) $

$ a_{22}(t) = 37\sin^4(\sqrt{3}t)+i\frac{1}{4}\cos(7t)-j\frac{\sqrt{29}}{16}\left|\sin(\sqrt{11}t)\right| +k\frac{1}{8}\cos^2(5t), $

$ \tau_{11}(t) = \frac{\sqrt{2}}{32}\sin^{8} (\sqrt{2}t), \tau_{12}(t) = \frac{1}{16}\cos^{2}(\sqrt{3}t), \tau_{21}(t) = \frac{1}{41}\sin^{2}(\frac{\sqrt{2}}{5}t), \tau_{22}(t) = \frac{1}{29}\sin^{4} (\frac{\sqrt{3}}{7}t), $

$ B_{11}(t) = \frac{1}{3}|\cos(\sqrt{2}t)|, B_{12}(t) = \frac{1}{6}\sin(4t)+\frac{1}{2} B_{21}(t) = \frac{3}{4}\sin(2t)+1, B_{22}(t) = \frac{1}{4}\sin^2(\sqrt{5}t), $

$ C_{11}(t) = \cos(\sqrt{3}t)+3, C_{12}(t) = |\sin(\sqrt{5}t)|, C_{21}(t) = 2\sin^2(\sqrt{7}t), C_{22}(t) = \frac{1}{4}\sin(2t)+\frac{7}{4}, $

$ f_{11}(x) = f_{12}(x) = \frac{1}{5}\sin(\frac{1}{4}x^{R}+\frac{\sqrt{3}}{6}x^{K}) -i\frac{1}{5}\bigg|\frac{3\sqrt{2}}{4}x^{I}+\frac{\sqrt{3}}{5}x^{J}\bigg| +k\frac{1}{4}\sin(\frac{1}{5}x^{R}), $

$ f_{21}(x) = f_{22}(x) = \frac{1}{4}\bigg|\frac{\sqrt{2}}{3}x^{I}+\frac{\sqrt{3}}{7}x^{K}\bigg| -i\frac{\sqrt{2}}{8}\sin(\frac{1}{5}x^{R}+\frac{\sqrt{2}}{5}x^{K}) +k\frac{\sqrt{2}}{3}\sin(\frac{1}{5}x_{j}^{I}), $

$ g_{11}(x) = g_{12}(x) = \frac{1}{20}\sin(\frac{9\sqrt{2}}{2}x^{R})-i\frac{\sqrt{3}}{40}\sin(\frac{7}{5}x^{K}) +j\frac{\sqrt{2}}{20}\sin(\frac{1}{3}x^{J}+\frac{3}{4}x^{I}), $

$ g_{21}(x) = g_{22}(x) = \frac{\sqrt{2}}{25}\bigg|\frac{1}{3}x^{J}+\frac{2}{3}x^{I}\bigg| -i\frac{\sqrt{3}}{40}\sin(\frac{7}{5}x^{K}+x^R) +j\frac{1}{21}\sin(4\sqrt{2}x^{R}), $

$ I_{11}(t) = I_{21}(t) = \sqrt{2}\sin t+i\frac{4}{3}e^{-|t|}+k\frac{\sqrt{2}}{3}\cos(\frac{1}{2}t), $

$ I_{12}(t) = I_{22}(t) = i\frac{\sqrt{2}}{3}\sin \sqrt{2}t-j\frac{1}{5}\cos2t+ke^{-|t|}. $

By computing, we obtain

$ L_{11}^{f} = L_{12}^{f} = \frac{3}{10}, L_{21}^{f} = L_{22}^{f} = \frac{1}{6}, L_{11}^{g} = L_{12}^{g} = \frac{9}{20}, L_{21}^{g} = L_{22}^{g} = \frac{8}{21}, $

$ a^m = 32, \check{a}_{11}^M = \frac{1}{2}, \check{a}_{12}^M = \frac{2}{5}, \check{a}_{21}^M = \frac{5}{12}, \check{a}_{22}^M = \frac{7}{16}, \tau = \frac{1}{16}, \tau' = \frac{1}{2}, $

$ \sum\limits_{C_{kl}\in N_1(1, 1)}B_{11}^{kl^M} = \sum\limits_{C_{kl}\in N_1(1, 2)}B_{12}^{kl^M} = \sum\limits_{C_{kl}\in N_1(2, 1)}B_{21}^{kl^M} = \sum\limits_{C_{kl}\in N_1(2, 2)}B_{22}^{kl^M} = 4, $

$ \sum\limits_{C_{kl}\in N_1(1, 1)}C_{11}^{kl^M} = \sum\limits_{C_{kl}\in N_1(1, 2)}C_{12}^{kl^M} = \sum\limits_{C_{kl}\in N_1(2, 1)}C_{21}^{kl^M} = \sum\limits_{C_{kl}\in N_1(2, 2)}C_{22}^{kl^M} = 8. $

Choose a constant $ \alpha = \frac{1}{16}\geq \|\phi^0\|_{\infty} $, for $ i, j = 1, 2 $, we obtain that

For $ i, j = 1, 2 $, take $ p = 3 $, it is easy to obtain that

$\begin{eqnarray*} &&\max\limits_{ij\in\Lambda}\Bigg\{ 96\bigg(\frac{2}{3a^m}\bigg)^3\bigg[\frac{2}{3}(\check{a}^M_{ij})^3+2 \bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L^f_{ij}\alpha\bigg)^{3} \\ &&+\bigg(1+\frac{2e^{\frac{3}{4}a^m\tau}}{1-\tau'}\bigg) \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\bigg)^3\bigg]\Bigg\} = 0.0016 < 1. \end{eqnarray*}$ $

Take $ p = 2 $, we have

$\begin{eqnarray*} &&\max\limits_{ij\in\Lambda}\Bigg\{24 \bigg(\frac{2}{a^m}\bigg)^2\bigg[\frac{2}{3}(\check{a}^M_{ij})^2+2 \bigg(\sum\limits_{C_{kl}\in N_r(i, j)}2B_{ij}^{kl^M}L^f_{ij}\alpha\bigg)^{2} \\ &&+\bigg(1+\frac{2e^{\frac{1}{2}a^m\tau}}{1-\tau'}\bigg) \bigg(\sum\limits_{C_{kl}\in N_s(i, j)}2C_{ij}^{kl^M}L^g_{ij}\alpha\bigg)^2\bigg]\Bigg\} = 0.2832 < 1. \end{eqnarray*}$ $

Thus, conditions $ (H_1) $–$ (H_4) $ of Theorems 3.1 and Theorems 3.2 are satisfied. Hence, system (1.1) has a unique $ W^p $-almost periodic solution that is globally exponentially stable (see Figures 1, 2).

Figure 1. Curves of $ (x_{11}^{l}(t), x_{12}^{l}(t))^T $ of system (1.1) with the initial values $ (x_{11}^{l}(0), x_{12}^{l}(0))^T = (1, -2)^T, (3, -4)^T, (5, -5)^T, (7, -7)^T, (9, -9)^T $, l = R, I, J, K.

DownLoad: Full-Size Img PowerPoint

Figure 2. Curves of $ (x_{21}^{l}(t), x_{22}^{l}(t))^T $ of system (1.1) with the initial values $ (x_{21}^{l}(0), x_{22}^{l}(0))^T = (-8, -1)^T, (-9, -3)^T, (5, -5)^T, (7, -2)^T, (9, 4)^T $, l = R, I, J, K.

DownLoad: Full-Size Img PowerPoint

Remark 4.1. No known results are available to give the results of Example 4.1.

5. Conclusions

In this paper, the existence and global exponential stability of Weyl almost periodic solutions for a class of quaternion-valued neural networks with time-varying delays are established. Even when the system we consider is a real-valued system, our results are brand-new. In addition, the method in this paper can be used to study the existence of Weyl almost periodic solutions for other types of neural networks.

Acknowledgments

This work is supported by the National Natural Science Foundation of China under Grant 11861072.

Conflict of interest

The authors declare no conflict of interest in this paper.

References

[1]	Baillargeon R, Carey S (2012) Core cognition and beyond: The acquisition of physical and numerical knowledge. In S. Pauen (Ed.) Early childhood development and later outcome. Cambridge, England: Cambridge University Press, 33-65.
[2]	Carey S (2011) The Origin of Concepts: A Precis. Behav Brain Sci 34: 113-162
[3]	Spelke ES, Kinzler KD (2007) Core knowledge. Dev Sci 10: 89-96. doi: 10.1111/j.1467-7687.2007.00569.x
[4]	van der Lely HKJ (2005) Domain-specific cognitive systems: Insight from grammatical specific language impairment. Trends Cog Sci 9: 53-59. doi: 10.1016/j.tics.2004.12.002
[5]	Van der Lely JKJ, Pinker S (2014) The biological basis of language: insights from developmental grammatical impairments. Trends Cog Sci 18: 586-595. doi: 10.1016/j.tics.2014.07.001
[6]	Elman JL, Bates E, Johnson MH, et al. (1996) Rethinking Innateness: A Connectionist Perspective on Development. Cambridge, Mass: MIT Press.
[7]	Kirkham NZ, Slemmer JA, Johnson SP (2002) Visual statistic learning in infancy: evidence for a domain-general learning mechanism. Cognition, 83: B35-B42. doi: 10.1016/S0010-0277(02)00004-5
[8]	Lany J, Saffran JR (2013) Statistical learning mechanisms in infancy. In: Comprehensive Developmental Neuroscience: Neural Circuit Development and Function in the Brain, Volume 3 (Rubenstein JLR, Rakic P, eds). Amsterdam, Elsevier: 231-248.
[9]	Pinker S (2002) The Blank Slate. The modern denial of human nature. New York, NY: Penguin Group.
[10]	Cosmides L, Barrett HC, Tooby J (2010) Adaptative specializations, social exchange, and the evolution of human intelligence. Proc Natl Acad Sci 107: 9007-9014. doi: 10.1073/pnas.0914623107
[11]	Agrillo C, Ranpura A, Butterworth B (2010) Time and numerosity estimation are independent: Behavioral evidence for two different systems using a conflict paradigm. Cog Neurosci 1: 96-101. doi: 10.1080/17588921003632537
[12]	Clahsen H, Temple C (2003) Words and rules in children with Williams syndrome. In: Language competence across populations (Levy Y, Schaeffer J eds). Erlbaum: Mahwah, NJ, 323-352.
[13]	Cohen Kadosh R, Bahrami B, Walsh V, et al. (2011) Specialization in the human brain: The case of numbers. Front Hum Neurosci 5: 62.
[14]	Bates E, Elman J, Johnson MH, Karmiloff-Smith A, et al. (1998) Innateness and Emergentism. In A Companion to Cognitive Science (Bechtel W, Graham G, eds). Oxford: Basil Blackwell, 590-601.
[15]	Karmiloff-Smith A, Plunkett K, Johnson M, et al. (1998) What does it mean to claim that something is ‘innate'? Mind Lang 13: 588-597. doi: 10.1111/1468-0017.00095
[16]	Hyde DC, Spelke ES (2011) Neural signatures of number processing in human infants: Evidence for two core systems underlying numerical cognition. Dev Sci 14: 360-371. doi: 10.1111/j.1467-7687.2010.00987.x
[17]	Lee SA, Spelke ES (2010) A modular geometric mechanism for reorientation in children. Cognitive Psychology 61: 152-176. doi: 10.1016/j.cogpsych.2010.04.002
[18]	Spelke ES, Kinzler KD (2007) Core Knowledge. Dev Sci 10: 89-96. doi: 10.1111/j.1467-7687.2007.00569.x
[19]	Xu F, Spelke ES, Goddard S (2005) Number sense in human infants. Dev Sci 8: 88-101. doi: 10.1111/j.1467-7687.2005.00395.x
[20]	Dehaene S (1997) The number sense: how the mind creates mathematics. Oxford: Oxford University Press.
[21]	Butterworth B (1999) The mathematical brain. Macmillan: London.
[22]	Gelman R, Gallistel R (1986) The child's understanding of number. Cambridge, MA: Harvard University Press.
[23]	Duchaine B (2000) Developmental propagnosia with normal configural processing. NeuroReport 11: 79-83. doi: 10.1097/00001756-200001170-00016
[24]	Duchaine B, Nakayama K (2005) Dissociations of face and object recognition in developmental prosopagnosia. J Cog Neurosci 17: 249-261. doi: 10.1162/0898929053124857
[25]	Duchaine B, Nakayama K (2006) Developmental prosopagnosia: a window to content-specific face processing. Cur Op Neurobio 16: 166-173. doi: 10.1016/j.conb.2006.03.003
[26]	Duchaine B, Yovel G, Butterworth E, et al. (2006) Prosopagnosia as an impairment to face-specific mechanisms: elimination of the alternative explanations in a developmental case. Cognitive Neuropsychology 23: 714-747. doi: 10.1080/02643290500441296
[27]	Landau B, Hoffman JE, Kurz N (2005) Object definitions with severe spatial deficits in Williams syndrome: sparing and breadkdown. Cognition 100: 483-510.
[28]	Piattelli-Palmarini M (2001) Speaking of learning: How do we acquire our marvellous facility for expressing ourselves in words? Nature 411: 887-888. doi: 10.1038/35082123
[29]	Spelke ES, Kinzler KD (2009) Innateness, learning and rationality. Child Dev Pers 3: 96-98. doi: 10.1111/j.1750-8606.2009.00085.x
[30]	Duchaine B, Cosmides L, Tooby J (2001) Evolutionary psychology and the brain. Curr Op Neurobio 11: 225-230. doi: 10.1016/S0959-4388(00)00201-4
[31]	Hauser MD, Spelke ES (2004). Evolutionary and developmental foundations of human knowledge: A case study of mathematics. In M Gazzaniga (Ed.) The Cognitive Neurosciences, Vol. 3. Cambridge: MIT Press, 853-864.
[32]	Bolhuis JJ, Brown GR, Richardson RC, et al. (2011) Darwin in Mind: New Opportunities for Evolutionary Psychology. PLoS Biology 9: 1-8.
[33]	Baron-Cohen S, Leslie AM, Frith U (1985) Does the autistic child have a “theory of mind”? Cognition 21: 37-46. doi: 10.1016/0010-0277(85)90022-8
[34]	Leslie AM (1992) Pretence, autism, and the theory-of-mind-module. Cur Dir Psych Sci 1: 18-21. doi: 10.1111/1467-8721.ep10767818
[35]	Baron-Cohen S, Ring HA, Wheelwright S, et al. (1999) Social intelligence in the normal and autistic brain: an fMRI study. Eur J Neurosci 11: 1891-8. doi: 10.1046/j.1460-9568.1999.00621.x
[36]	Temple CM (1997) Cognitive neuropsychology and its application to children. J Child Psychol Psychiatry 38: 27-52. doi: 10.1111/j.1469-7610.1997.tb01504.x
[37]	Butterworth B (2008) State-of-science review: Dyscalculia. In Goswami UC (Ed.) Foresight Mental Capital and Mental Wellbeing. Office of Science and Innovation, London.
[38]	Duchaine B, Nieminen-von Wendt T, New J, et al. (2003) Dissociations of visual recognition in a developmental agnosic: evidence for separate developmental processes. Neurocase 9: 380-389. doi: 10.1076/neur.9.5.380.16556
[39]	Molko N, Cachia A, Rivière D, et al. (2003) Functional and Structural Alterations of the Intraparietal Sulcus in a Developmental Dyscalculia of Genetic Origin. Neuron 40: 847-858. doi: 10.1016/S0896-6273(03)00670-6
[40]	Rice M (1999) Specific grammatical limitations in children with Specific Language Impairment. In: Neurodevelopmental disorders (Tager-Flusberg H, ed). Cambridge, Mass: MIT Press, 331-360.
[41]	Shalev RS, Manor O, Gross-Tsur V (2005) Developmental dyscalculia: a prospective six-year follow-up. Dev Med Child Neurol 2: 121-125.
[42]	Castle A, Coltheart M (1993) Varieties of developmental dyslexia. Cognition 47: 149-180. doi: 10.1016/0010-0277(93)90003-E
[43]	Gopnik M (1990) Genetic basis of grammar defect. Nature 347: 26-26.
[44]	Karmiloff-Smith A (1992) Beyond Modularity: A Developmental Perspective on Cognitive Science. Cambridge, Mass: MIT Press/Bradford Books.
[45]	Karmiloff-Smith A (1998) Development itself is the key to understanding developmental disorders. Trends Cog Sci 2: 389-398. doi: 10.1016/S1364-6613(98)01230-3
[46]	Mareschal D, Johnson MH, Sirios S, et al. (2007) Neuroconstructivism: Vol. I. How the brain constructs cognition. Oxford, England: Oxford University Press.
[47]	Westerman G, Marescal D, Johnson MH, et al. (2007) Neuroconstructivism. Dev Sci 10: 75-83. doi: 10.1111/j.1467-7687.2007.00567.x
[48]	Westermann G, Thomas MSC, Karmiloff-Smith A (2010) Neuroconstructivism. In: Handbook of Childhood Development (Goswami U, ed). Oxford: Wiley-Blackwell, 723-748.
[49]	Karmiloff-Smith A (2013) Challenging the use of adult neuropsychological models for explaining neurodevelopmental disorders: Developed versus developing brains. Quart J Exp Psych 66: 1-14. doi: 10.1080/17470218.2012.744424
[50]	Paterson SJ, Brown JH, Gsödl MK, et al. (1999) Cognitive Modularity and Genetic Disorders, Science 286: 2355-2358.
[51]	Johnson MH (2001) Functional brain development in humans. Nat Rev Neurosci 2: 475-483. doi: 10.1038/35081509
[52]	Stiles J (2009) The Fundamentals of Brain Development. Cambridge MA: Havard University Press.
[53]	Thomas MSC, Knowland VC, Karmiloff-Smith A (2011) Mechanisms of developmental regression in autism and the broader phenotype: a neural network modeling approach. Psych Rev 118: 637-654. doi: 10.1037/a0025234
[54]	Casey BJ, Giedd JN, Thomas KM (2000) Structural and functional brain development and its relation to cognitive development. Biol Psychol 54: 241-257. doi: 10.1016/S0301-0511(00)00058-2
[55]	Huttenlocher PR, de Courten C (1987) The development of synapses in striate cortex of man. Hum Neurobio 6: 1-9.
[56]	Huttenlocher PR, Dabholkar AS (1997) Regional Differences in Synaptogenesis in Human. J Comp Neurol 387: 167-178. doi: 10.1002/(SICI)1096-9861(19971020)387:2<167::AID-CNE1>3.0.CO;2-Z
[57]	Giedd J, Blumenthal J, Jeffries N, et al. (1999) Brain development during childhood and adolescence: A longitudinal MRI study. Nat Neurosci 2: 861-863. doi: 10.1038/13158
[58]	Giedd J, Rumsey J, Castellanos F, et al. (1996) A quantitative MRI study of the corpus callosum in children and adolescents. Dev Brain Res 91: 274-280. doi: 10.1016/0165-3806(95)00193-X
[59]	Cohen Kadosh K., Henson RN, Cohen Kadosh R, et al. (2009) Task-dependent activation of face-sensitive cortex: an fMRI adaptation study. J Cog Neurosci, 22: 903-917.
[60]	De Haan M, Humphreys K, Johnson MH (2002) Developing a brain specializes for face processing: A converging methods approach. Dev Psychobiology 40: 200-212. doi: 10.1002/dev.10027
[61]	Krishnan S, Leech R, Mercure E, et al. (2014) Convergent and divergent fMRI responses in children and adults to increasing language production demands. Cereb Cortex pii: bhu120.
[62]	Mills DL, Coffy-Corins S, Neville H (1997) Language comprehension and cerebral specialisation from 13-20 months. Dev Psych 13: 397-445.
[63]	Minagawa-Kawai Y, Mori K, Naoi N, et al. (2007) Neural attunement processes in infants during the acquisition of language-specific phonemic contrasts. J Neurosci 3: 315-321.
[64]	Neville H, Mills D, Bellugi U (1994) Effects of altered auditory sensitivity and age of language acquisition on the development of language-relevant neural systems: Preliminary studies of William syndrome. In: Atypical cognitive deficits in developmental disorders: Implications for brain function (Broman S, Grafman J, eds). Hillsdale, NJ: Erlbaum, 67-83.
[65]	Stiles J (2012) Neural plasticity and cognitive development: Insights from children with perinatal brain injury. Oxford UK: Oxford University Press.
[66]	Bates E, Roe K (2001) Language development in children with unilateral brain injury. In: Handbook of Developmental Cognitive Neuroscience (Nelson CA, Luciana M, eds). Cambridge, MA: MIT Press, 281-307.
[67]	Casey BJ, Tottenham N, Liston C, et al. (2005) Imaging the developing brain. What have we learned about cognitive development? Trends Cog Sci 9: 104-110.
[68]	Durston S, Davidson MC, Tottenham N, et al. (2006). A shift from diffuse to focal cortical activity with development. Dev Sci 9: 1-8.
[69]	Paterson SJ, Heim S, Friedman JT, et al. (2006) Development of structure and function in the infant brain: Implications for cognition, language and social behaviour. Neurosci Biobehav Rev 30: 1087-1105. doi: 10.1016/j.neubiorev.2006.05.001
[70]	Smith LB, Thelen E (2003) Development as a dynamic system. Trends Cog Sci 7: 343-348. doi: 10.1016/S1364-6613(03)00156-6
[71]	Tyler LK, Shafto MA, Randall B, et al. (2010) Preserving Syntactic Processing across the Adult Life Span: The Modulation of the Frontotemporal Language System in the Context of Age-Related Atrophy. Cereb Cortex 20: 352-364. doi: 10.1093/cercor/bhp105
[72]	Dehaene S, Cohen L (2007) Cultural recycling of cortical maps. Neuron 56: 384-398. doi: 10.1016/j.neuron.2007.10.004
[73]	Fodor J (1983) Modularity of Mind. Cambridge, MA: MIT Press.
[74]	Dehaene S, Charles L, King JR, et al. (2014) Toward a computational theory of conscious processing. Curr Op Neurobio 25: 76-84. doi: 10.1016/j.conb.2013.12.005
[75]	Anderson ML (2010) Neural reuse: A fundamental organizational principle of the brain. Behav Brain Sci 33: 245-313. doi: 10.1017/S0140525X10000853
[76]	Dehaene S (2009) Reading in the brain: The science and evolution of a human invention. New York, NY: Penguin Group
[77]	Goldman-Rakic PS (1987) Circuitry of the pre- frontal cortex and the regulation of behavior by representational knowledge. In: Handbook of Physiology 5 (Plum F, Mountcastle V, eds). Bethesda, MD: American Physiological Society, 373-417.
[78]	Johnson MH (2004) Plasticity and functional brain development: The case of face processing. In: Attention & Performance XX: Functional neuroimaging of visual cognition (Kanwisher N, Duncan J eds). Oxford: Oxford University Press, 257- 263.
[79]	Kuhl P (2004) Early language acquisition: cracking the speech code. Nat Rev Neurosci 5: 831-843. doi: 10.1038/nrn1533
[80]	Majdan M, Schatz CJ (2006) Effects of visual experience on activity-dependent gene regulation in cortex. Nat Neurosci 9: 650-659. doi: 10.1038/nn1674
[81]	Meaney MJ, Szyf M (2005) Maternal care as a model for experience-dependent chromatin plasticity? Trends Neurosci 28: 456-463. doi: 10.1016/j.tins.2005.07.006
[82]	Bhattacharya J, Petsche H (2005) Drawing on mind's canvas: Differences in cortical integration patterns between artists and non-artists. Hum Brain Mapping 26: 1-14. doi: 10.1002/hbm.20104
[83]	Crone EA, Elzinga B (2014) Changing brains: How longitudinal fMRI studies can inform us about cognitive and social-affective growth trajectories. WIREs Cogn Sci 6: 53-63.
[84]	Brown TT, Petersen SE, Schlaggar BL (2006) Does human functional brain organization shift from different to focal with development? Dev Sci 9: 9-10. doi: 10.1111/j.1467-7687.2005.00455.x
[85]	Karmiloff-Smith A (1997) Promissory notes, genetic clocks or epigenetic outcomes? Behav Brain Sci 20: 359-377.
[86]	Karmiloff-Smith A (2007) Atypical epigenesis. Dev Sci 10: 84-88. doi: 10.1111/j.1467-7687.2007.00568.x
[87]	Kingsbury MA, Finlay LF (2001) The cortex in multidimensional space: where do cortical areas come from? Developmental Science 4: 125-142. doi: 10.1111/1467-7687.00158
[88]	Bishop KM, Goudreau G, O'Leary DDM (2000) Regulation of area identity in the mammalian neocortex by Emx2 and Pax6. Science 288: 344-349. doi: 10.1126/science.288.5464.344
[89]	Kaffman A, Meaney MJ (2007) Neurodevelopmental sequelae of postnatal maternal care in rodents: clinical and research implications of molecular insights. J Child Psychol. Psychiatry 48: 224-244. doi: 10.1111/j.1469-7610.2007.01730.x
[90]	Gotlieb, G (2007) Probabilistic epigenesis. Dev Sci 10: 1-11. doi: 10.1111/j.1467-7687.2007.00556.x
[91]	Scerif G, Karmiloff-Smith A (2001) Genes and environment: What does interaction really mean? Trends Gen 17: 418-419. doi: 10.1016/S0168-9525(01)02337-X
[92]	Quartz S, Sejnowski T (1997) The neural basis of cognitive development: A constructivist manifesto. Behav Brain Sci 20: 537-596.
[93]	Finlay BL (2007) E pluribus unum: Too many unique human capacities and too many theories. In: The evolution of mind: Fundamental questions and controversies (Gangestad S, Simpson J, eds). New York: Guilford Press, 294-304.
[94]	Karmiloff-Smith A (2010) Neuroimaging of the developing brain: Taking “developing” seriously. Hum Brain Map 31: 934-941. doi: 10.1002/hbm.21074
[95]	Karmiloff-Smith A (2012) Brain: The Neuroconstructivist Approach. In: Neuro- developmental disorders across the lifespan: A neuroconstructivist approach (Farran EK, Karmiloff-Smith A, eds) Oxford: Oxford University Press, 37-58.
[96]	Karmiloff-Smith A, Thomas MSC (2004) Can developmental disorders be used to bolster claims from Evolutionary Psychology? A neuroconstructivist approach. In: Biology and Knowledge Revisited: From Neurogenesis to Psychogenesis (Langer J, Taylor Parker S, Milbrath C, eds) Mahwah, NJ: Lawrence Erlbaum Associates, 307-322.
[97]	Karmiloff-Smith A (2009) Nativism versus Neuroconstructivism: Rethinking the Study of Developmental Disorders. Interplay of Biology and Environment, Dev Psych 45: 56-63.
[98]	Pinker S (1994) The language instinct: How the mind creates language. New York: W.Morrow.
[99]	Pinker S (1999) Words and rules: The ingredients of language. New York: Basic Books.
[100]	Karmiloff-Smith A, Thomas MSC, Annaz D, et al. (2004) Exploring the Williams Syndrome Face Processing Debate: The importance of building developmental trajectories. J Child Psych Psychiatry 45: 1258-1274. doi: 10.1111/j.1469-7610.2004.00322.x
[101]	Karmiloff-Smith A, D'Souza D, Dekker TM, et al. (2012) Genetic and environmental vulnerabilities in children with neurodevelopmental disorders. Proc Natl Acad Sci 109: 17261-5. doi: 10.1073/pnas.1121087109
[102]	D'Souza D, Karmiloff-Smith A (2011) When modularization fails to occur: a developmental perspective. Cog Neuropsych 28: 276-287. doi: 10.1080/02643294.2011.614939
[103]	Mackinlay R, Charman T, Karmiloff-Smith A (2006) High functioning children with autism spectrum disorder: A novel test of multitasking. Brain Cog 61: 14-24. doi: 10.1016/j.bandc.2005.12.006
[104]	Comery TA, Harris JB, Willems PJ, et al. (1997) Abnormal dendritic spines in fragile X knockout mice: Maturation and pruning deficits. Proc Natl Acad Sci 94: 5401-5404. doi: 10.1073/pnas.94.10.5401
[105]	Oliver A, Johnson MH, Karmiloff-Smith A, et al. (2000) Deviations in the emergence of representations: A neuroconstructivist framework for analysing developmental disorders. Dev Sci 3: 1-23. doi: 10.1111/1467-7687.00094
[106]	Karmiloff-Smith A, Scerif G, Ansari D (2003) Double dissociations in developmental disorders? Theoretically misconceived, empirically dubious. Cortex 39: 161-163.
[107]	Johnson MH, Halit H, Grice S, et al. (2002) Neuroimaging of typical and atypical development: A perspective from multiple levels of analysis. Dev Psychopath 14: 521-536.
[108]	Cicchetti D, & Tucker D (1994) Development and self-regulatory structures of the mind. Dev Psychopath 6: 533-549. doi: 10.1017/S0954579400004673
[109]	Sur M, Pallas SL, Roe AW (1990) Crossmodal plasticity in cortical development: Differentiation and specification of sensory neocortex. TINS 13: 227-233.
[110]	Webster MJ, Bachevalier J, Ungerleider LG (1995) Development and plasticity of visual memory circuits. In: Maturational windows and adult cortical plasticity in human development: Is there reason for an optimistic view? (Julesz B, Kovacs I, eds). Reading, MA: Addison-Wesley.
[111]	Bavelier D, Levi DM, Li RW, et al. (2010) Removing brakes on adult plasticity: from molecular to behavioural interventions. J Neurosci 30: 14964-71. doi: 10.1523/JNEUROSCI.4812-10.2010
[112]	Hensch TK (2005) Critical period plasticity in local cortical circuits. Nat. Rev. Neurosci 6: 877-888.
[113]	Karmiloff-Smith A (1994) Transforming a partially structured brain into a creative mind. Behav Brain Sci 17: 732-745. doi: 10.1017/S0140525X00036906

This article has been cited by:

1.	Adnène Arbi, Najeh Tahri, New results on time scales of pseudo Weyl almost periodic solution of delayed QVSICNNs, 2022, 41, 2238-3603, 10.1007/s40314-022-02003-0
2.	Marko Kostić, Weyl ρ-Almost Periodic Functions in General Metric, 2023, 73, 1337-2211, 465, 10.1515/ms-2023-0035
3.	Qi Shao, Yongkun Li, Almost periodic solutions for Clifford-valued stochastic shunting inhibitory cellular neural networks with mixed delays, 2024, 9, 2473-6988, 13439, 10.3934/math.2024655
4.	Adnène Arbi, Najeh Tahri, Synchronization analysis of novel delayed dynamical Clifford-valued neural networks on timescales, 2024, 18, 1748-3018, 10.1177/17483026241241492

Reader Comments

Your name:*

Email:*
© 2015 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)

通讯作者: 陈斌, bchen63@163.com

1.
沈阳化工大学材料科学与工程学院沈阳 110142

AIMS Neuroscience

3.1 4.2

Metrics

Article views(11168) PDF downloads(2026) Cited by(53)

Preview PDF

Download XML

Export Citation

Article outline

Show full outline

AIMS Neuroscience

An Alternative to Domain-general or Domain-specific Frameworks for Theorizing about Human Evolution and Ontogenesis

Related Papers:

Abstract

1. Introduction

2. Preliminaries

3. Main results

4. A numerical example

5. Conclusions

Acknowledgments

Conflict of interest

References

This article has been cited by:

Reader Comments

通讯作者: 陈斌, bchen63@163.com

Metrics

Other Articles By Authors

Catalog

AIMS Neuroscience

An Alternative to Domain-general or Domain-specific Frameworks for Theorizing about Human Evolution and Ontogenesis

Related Papers:

Abstract

1. Introduction

2. Preliminaries

3. Main results

4. A numerical example

5. Conclusions

Acknowledgments

Conflict of interest

References

This article has been cited by:

Reader Comments

通讯作者: 陈斌, bchen63@163.com

Metrics

Other Articles By Authors

Related pages

Tools

Export File

Citation

Format

Content

Catalog