Robustness analysis of exponential synchronization in complex dynamic networks with random perturbations

Qike Zhang; Wenxiang Fang; Tao Xie; Qike Zhang; Wenxiang Fang; Tao Xie

doi:10.3934/math.20231044

AIMS Mathematics

2023, Volume 8, Issue 9: 20487-20509. doi: 10.3934/math.20231044

Previous Article Next Article

Research article

Robustness analysis of exponential synchronization in complex dynamic networks with random perturbations

School of Mathematics and Statistics, Hubei Normal University, Huangshi 435002, Hubei, China

Received: 08 May 2023 Revised: 02 June 2023 Accepted: 12 June 2023 Published: 25 June 2023
MSC : 93B35, 93C73

This article discusses the robustness of exponential synchronization (ESy) of complex dynamic networks (CDNs) with random perturbations. Using the Gronwall-Bellman lemma and partial inequality techniques, by solving the transcendental equation, the maximum perturbation intensity of the CDN is estimated. This implies that the disturbed system achieves ESy if the disturbance intensity is within the range of our estimation. We illustrate the theoretical results with two numerical examples.

Keywords:

Citation: Qike Zhang, Wenxiang Fang, Tao Xie. Robustness analysis of exponential synchronization in complex dynamic networks with random perturbations[J]. AIMS Mathematics, 2023, 8(9): 20487-20509. doi: 10.3934/math.20231044

Related Papers:

[1]	Zui-Cha Deng, Fan-Li Liu, Liu Yang . Numerical simulations for initial value inversion problem in a two-dimensional degenerate parabolic equation. AIMS Mathematics, 2021, 6(4): 3080-3104. doi: 10.3934/math.2021187
[2]	Jia Li, Zhipeng Tong . Local Hölder continuity of inverse variation-inequality problem constructed by non-Newtonian polytropic operators in finance. AIMS Mathematics, 2023, 8(12): 28753-28765. doi: 10.3934/math.20231472
[3]	Dun-Gang Li, Fan Yang, Ping Fan, Xiao-Xiao Li, Can-Yun Huang . Landweber iterative regularization method for reconstructing the unknown source of the modified Helmholtz equation. AIMS Mathematics, 2021, 6(9): 10327-10342. doi: 10.3934/math.2021598
[4]	Jia Li, Changchun Bi . Study of weak solutions of variational inequality systems with degenerate parabolic operators and quasilinear terms arising Americian option pricing problems. AIMS Mathematics, 2022, 7(11): 19758-19769. doi: 10.3934/math.20221083
[5]	Yu Xu, Youjun Deng, Dong Wei . Numerical solution of forward and inverse problems of heat conduction in multi-layered media. AIMS Mathematics, 2025, 10(3): 6144-6167. doi: 10.3934/math.2025280
[6]	Zuliang Lu, Fei Cai, Ruixiang Xu, Chunjuan Hou, Xiankui Wu, Yin Yang . A posteriori error estimates of hp spectral element method for parabolic optimal control problems. AIMS Mathematics, 2022, 7(4): 5220-5240. doi: 10.3934/math.2022291
[7]	Batirkhan Turmetov, Valery Karachik . On solvability of some inverse problems for a nonlocal fourth-order parabolic equation with multiple involution. AIMS Mathematics, 2024, 9(3): 6832-6849. doi: 10.3934/math.2024333
[8]	Yashar Mehraliyev, Seriye Allahverdiyeva, Aysel Ramazanova . On one coefficient inverse boundary value problem for a linear pseudoparabolic equation of the fourth order. AIMS Mathematics, 2023, 8(2): 2622-2633. doi: 10.3934/math.2023136
[9]	Guojie Zheng, Baolin Ma . Observability estimate for the parabolic equations with inverse square potential. AIMS Mathematics, 2021, 6(12): 13525-13532. doi: 10.3934/math.2021785
[10]	W. Y. Chan . Blow-up for degenerate nonlinear parabolic problem. AIMS Mathematics, 2019, 4(5): 1488-1498. doi: 10.3934/math.2019.5.1488

Abstract

1. Introduction

The boundary value problems (BVPs) for differential equations have important applications in space science and engineering technology. A large number of mathematical models in the fields of engineering, astronomy, mechanics, economics, etc, are often described by differential BVPs ^[1,2,3]. Except for a few special types, the exact solution of the BVPs is difficult to express in analytical form. It is especially important to find an approximate solution to obtain its numerical solution. In ^[4], Sinc collocation method provided an exponential convergence rate for two-point BVPs. ^[5] constructed a simple collocation method by the Haar wavelets for the numerical solution of linear and nonlinear second-order BVPs with periodic boundary conditions. Erge ^[6] studied the quadratic/linear rational spline collocation method for linear BVPs. In ^[7], based on B-spline wavelets, the numerical solutions of nonlinear BVPs were derived. Pradip et al. used B-spline to Bratuis problem which is an important nonlinear BVPs in ^[8,9,10]. ^{[11,12,13,14,15,16]} solved BVPs by the reproducing kernel method. Based on the idea of least squares, Xu et al. ^[17,18,19] gave an effective algorithm in reproducing kernel space for solving fractional differential integral equations and interface problems.

It is a common technique to use orthogonal polynomials to solve differential equations. In ^{[20,21,22,23]}, the authors used Chebyshev-Galerkin scheme for the time-fractional diffusion equation. In ^[24], the authors developed Jacobi rational operational approach for time-fractional sub-diffusion equation on a semi-infinite domain. ^{[25,26,27,28]} developed multiscale orthonormal basis to solve BVPs with various boundary conditions, and the stability and convergence order were also discussed. Legendre wavelet is widely used in various fields, such as signal system, because of its good properties. In this paper, a multiscale function is constructed by using Legendre polynomials to solve the approximate solution of differential equations. We use the multiscale fine ability of Legendre wavelet to construct multiwavelet, which has better approximation than single wavelet. In addition, we improve Legendre wavelet for specific problems, and the improved one still has compact support. We know that for functions with compact support, the better the tight support, the more concentrated the energy. Moreover, in the calculation process, the calculation speed can be enhanced, and the error accumulation is low.

The purpose of this paper is to construct a set of multiscale orthonormal basis with compact support based on Legendre wavelet to find the approximate solution of the boundary value problems:

$\begin{equation} \left\{ \begin{array}{l} u^{\prime\prime}(x)+p(x)u^{\prime}(x)+q(x)u(x) = F(x, u), \quad x \in (0, 1), \\ a_1u(0)+b_1u(1)+c_1u^{\prime}(0)+d_1u^{\prime}(1) = \alpha_{1}, \\ a_2u(0)+b_2u(1)+c_2u^{\prime}(0)+d_2u^{\prime}(1) = \alpha_{2}, \end{array} \right. \end{equation}$

(1.1)

where $p(x)$ and $q(x)$ are both smooth. $a_i, b_i, c_i, d_i, \; i = 1, 2$ are constants. When $F$ is just about the function of $x$ , $F(x, u) = f(x)$ , Eq (1.1) is linear boundary value problem. According to ^[21], the nonlinear boundary value problem can be transformed into a linear boundary value problem by using Quasi-Newton method. So this paper mainly studies the case of $F(x, u) = f(x)$ , that is, the linear boundary value problem.

As we all know, if the basis function has good properties, the approximate solution of the boundary value problem has good convergence, stability and so on. In ^[25], the orthonormal basis on [0, 1] was constructed by the compact support function to obtain the numerical solution of the boundary value problem. But the basis function is not compactly supported at [0, 1], and the approximating solution is linearly convergent. In this paper, based on the idea of wavelet, a set of orthonormal bases with compact support is constructed by using Legendre polynomials, and the approximate solution of the boundary value problem is obtained by using these bases. Based on the constructed orthonormal basis, the proposed algorithm has convergence and stability, and the convergence order of the algorithm is more than 2 orders.

The purpose of this work is to deduce the numerical solutions of Eq (1.1). In Section 2, using wavelet theory, a set of multiscale orthonormal basis is presented by Legendre polynomials in $W_2^3[0, 1]$ . The constructed basis is compactly supported. It is well known that the compact support performance generates sparse matrices during calculation, thus improving the convergence rate. The numerical method of ${\varepsilon}$ -approximate solution is presented in Section 3. And Section 4 proves the convergence order of ${\varepsilon}$ -approximate solution and stability. In Section 5, the proposed algorithm has been applied to some numerical experiments. Finally, we end with some conclusions in Section 6.

2. Basis functions in $W_2^3[0, 1]$

Wu and Lin introduced the reproducing kernel space $W_2^1[0, 1]$ and $W_2^3[0, 1]$ ^[29]. Let

$W_{2, 0}^3[0, 1] = \{u|u(0) = u^{\prime}(0) = u^{\prime\prime}(0) = 0, \ \ \ u\in W_2^3[0, 1]\}.$

Clearly, $W_{2, 0}^3[0, 1]$ is the closed subspace of $W_2^3[0, 1]$ .

Legendre polynomials are mathematically important functions. This section constructs the orthonormal basis in $W_2^3[0, 1]$ by Legendre polynomials. Legendre polynomials are known to be orthogonal on $L^2[-1, 1]$ . For convenience, we first compress Legendre's polynomials to $[0, 1]$ , and get the following four functions:

$\begin{align*} \varphi^0(x)& = 1; \quad\quad\quad\quad\quad\quad\quad\ \ \varphi^1(x) = \sqrt{3}(-1+2x);\\ \varphi^2(x)& = \sqrt{5}(1-6x+6x^2); \quad \varphi^3(x) = \sqrt{7}(-1+12x-30x^2+20x^3). \end{align*}$

By translating and weighting the above four functions, we can construct

$\begin{equation} \psi^l(x) = \sum\limits_{j = 0}^3(a_{lj}\varphi^j(2x)+b_{lj}\varphi^j(2x-1)), \quad l = 0, 1, 2, 3. \end{equation}$

(2.1)

In application, we hope $\psi^l(x)$ has good properties, for example, as many coefficients as zero and orthogonality, so $\psi^l(x)$ needs to meet the following conditions

$\begin{equation} \int_0^1x^j\psi^l(x)dx = 0, \; \; \; j = 0, 1, 2, \cdots, l+3, \end{equation}$

(2.2)

$\begin{equation} \int_0^1 \psi^i(x)\psi^j(x)dx = \delta_{ij}, \; \; \; i, j = 0, 1, 2, 3. \end{equation}$

(2.3)

The coefficients $a_{lj}, b_{lj}$ can be get by Eqs (2.2) and (2.3), immediately $\psi^l(x)$ is as follows:

$\begin{align} & \psi^0(x) = \sqrt{\frac{15}{17}}\left\{ \begin{array}{l} -3+56x-216x^2+224x^3, \quad x\in [0, \frac{1}{2}], \\ 61-296x+456x^2-224x^3, \quad x\in [\frac{1}{2}, 1]. \end{array} \right. \end{align}$

(2.4)

$\begin{align} &\psi^1(x) = \sqrt{\frac{1}{21}}\left\{ \begin{array}{l} -11+270x-1320x^2+1680x^3, \quad \; \ x\in [0, \frac{1}{2}], \\ -619+2670x-3720x^2+1680x^3, \quad x\in [\frac{1}{2}, 1]. \end{array} \right. \end{align}$

(2.5)

$\begin{align} & \psi^2(x) = \sqrt{\frac{35}{17}}\left\{ \begin{array}{l} -1+30x-174x^2+256x^3, \quad x\in [0, \frac{1}{2}], \\ 111-450x+594x^2-256x^3, \quad x\in [\frac{1}{2}, 1]. \end{array} \right. \end{align}$

(2.6)

$\begin{align} &\psi^3(x) = \sqrt{\frac{5}{21}}\left\{ \begin{array}{l} 1-36x+246x^2-420x^3, \quad x\in [0, \frac{1}{2}], \\ 209-804x+1014x^2-420x^3, \quad x\in [\frac{1}{2}, 1]. \end{array} \right. \end{align}$

(2.7)

Through the ideas of the wavelet, scale transformation of the functions $\psi^l(x)$ gets Legendre wavelet

$\psi^l_{ik}(x) = 2^{\frac{i-1}{2}}\psi^l(2^ix-k), \; \; l = 0, 1, 2, 3;\; \; i = 1, 2, \cdots;\; \; k = 0, 1, \cdots, 2^{i-1}-1.$

Clearly, $\psi_{ik}^{l}(x)$ has compactly support in $[\frac{k}{2^{i-1}}, \frac{k+1}{2^{i-1}}]$ . Let

$W_i = span\{\psi^l_{ik}(x)\}_{l = 0}^3, \; \; i = 1, 2, \cdots;\; \; k = 0, 1, \cdots, 2^{i-1}-1.$

Then,

$L^2[0, 1] = V_0\bigoplus\limits_{i = 1}^{\infty}W_i,$

where

$V_0 = \{\varphi^0(x), \varphi^1(x), \varphi^2(x), \varphi^3(x)\}.$

According to the above analysis, we can get the following theorem.

Theorem 2.1.

$\begin{align*} \begin{aligned} \{\rho_j(x)\}_{j = 1}^{\infty} = \{\varphi^0(x), \varphi^1(x), \varphi^2(x), \varphi^3(x), \psi^0_{10}(x), \psi^1_{10}(x), \psi^2_{10}(x), \psi^3_{10}(x), \cdots, \psi^0_{ik}(x), \psi^1_{ik}(x), \psi^2_{ik}(x), \psi^3_{ik}(x), \cdots \}\end{aligned} \end{align*}$

is the orthonormal basis in $L^2[0, 1]$ .

Now we generate the orthonormal basis in $W_{2, 0}^3[0, 1]$ from the basis in $L^2[0, 1]$ . Note

$\begin{equation} J^{3}u(x) = \frac{1}{2}\int_0^x (x-t)^2u(t)dt. \end{equation}$

(2.8)

Theorem 2.2. $\{J^3\rho_j(x)\}_{j = 1}^{\infty}$ is the orthonormal basis in $W_{2, 0}^3[0, 1]$ .

Proof. Only need to prove completeness and orthogonality. For $u \in W_{2, 0}^3[0, 1]$ , if

$< u, J^3\rho_j > _{W_{2, 0}^3} = 0,$

you can deduce $u \equiv 0$ , then $\{J^3\rho_j(x)\}_{j = 1}^{\infty}$ are complete. In fact,

$\begin{equation} < u, J^3\rho_j > _{W_{2, 0}^3} = < u^{\prime\prime\prime}, \rho_j > _{L^2} = \int_0^1 u^{\prime\prime\prime}\rho_j dx = 0. \end{equation}$

(2.9)

From Theorem 2.1, $u^{\prime\prime\prime}\equiv 0$ . Due to $u \in W_{2, 0}^3[0, 1]$ , $u(0) = u^{\prime}(0) = u^{\prime\prime}(0) = 0$ , then, $u\equiv 0$ .

According to Theorem 2.1 and Eq (2.9), orthonormal is obvious. □

Because of $W^3_{2, 0}[0, 1]\subset W^3_2[0, 1]$ and three more conditions for $W_{2}^3[0, 1]$ than $W_{2, 0}^3[0, 1]$ . So the orthonormal basis for $W_{2}^3[0, 1]$ as follows:

Theorem 2.3.

$\{J^3g_j(x)\}_{j = 1}^{\infty} = \{1, x, \frac{x^2}{2}\} \cup \{J^3\rho_j(x)\}_{j = 1}^{\infty}$

are the orthonormal basis in $W_{2}^3[0, 1]$ .

3. A multiscale algorithm for Eq (1.1)

Put $L$ : $W_2^3[0, 1]\to L^2[0, 1]$ ,

$Lu = u^{\prime\prime}(x)+p(x)u^{\prime}(x)+q(x)u(x).$

$L$ is a linear bounded operator in ^[27]. Let $B_i$ : $W_2^3[0, 1]\to \mathbb{R}$ , and

$B_iu = a_iu(0)+b_iu(1)+c_iu^{\prime}(0)+d_iu^{\prime}(1), \quad i = 1, 2.$

The {Quasi-Newton} method is used to transform Eq (1.1) into a linear boundary value problem, and its operator equation is as follows:

$\begin{equation} \left\{ \begin{array}{l} \quad Lu = f(x), \\ B_1u = \alpha_{1}, \ \ B_2u = \alpha_{2}. \end{array} \right. \end{equation}$

(3.1)

Definition 3.1. $u^{{\varepsilon}}$ is named ${\varepsilon}$ -approximate solution for Eq (3.1), $\forall {\varepsilon} > 0$ , if

$\|Lu^{{\varepsilon}}-f\|_{L^2}^2+\sum\limits_{i = 1}^{2}(B_iu^{{\varepsilon}}-\alpha_i)^2 < {\varepsilon}^2.$

In ^[27], it is shown that ${\varepsilon}$ -approximate solution for Eq (3.1) exists by the following theorem.

Theorem 3.1. Equation (3.1) exists ${\varepsilon}$ -approximate solution

$u_n^{{\varepsilon}}(x) = \sum\limits_{k = 1}^{n} c^{*}_k J^3g_{k}(x) ,$

where $n$ is a natural number determined by ${\varepsilon}$ , and $c^{*}_i$ satisfies

$\|\sum\limits_{k = 1}^{n}c^*_{k}LJ^3g_k -Lu \|_{L^2}^{2}+\sum\limits_{l = 1}^{2}(\sum\limits_{k = 1}^{n}c^*_{k}J^3g_k -B_lu)^2 = \min\limits_{c_k}\{\|\sum\limits_{k = 1}^{n}c_{k}LJ^3g_k -Lu \|_{L^2}^{2}+\sum\limits_{l = 1}^{2}(\sum\limits_{k = 1}^{n}c_{k}J^3g_k -B_lu)^2\}.$

To seek the ${\varepsilon}$ -approximate solution, we just need $c_k^*$ . Let $G$ be quadratic form about

$\mathit{\boldsymbol{c}} = (c_1, \cdots, c_n)^T,$

$\begin{eqnarray} G(c_1, \cdots, c_{n}) = \|\sum\limits_{k = 1}^{n}c_{k}LJ^3g_k -Lu \|_{L^2}^{2}+\sum\limits_{l = 1}^{2}(\sum\limits_{k = 1}^{n}c_{k}J^3g_k -B_lu)^2. \end{eqnarray}$

(3.2)

From Theorem 3.1,

$\mathit{\boldsymbol{c}}^* = (c_1^*, \cdots, c_n^*)^T$

is the minimum point of $G(c_1, \cdots, c_n)$ . If $L$ is reversible, the minimum point of $G$ exists and is unique.

In fact, the partial derivative of $G(c_1, \cdots, c_n)$ with respect to $c_j$ :

$\frac{\partial G}{\partial c_j} = 2 \sum\limits_{k = 1}^{n}c_k\langle LJ^3g_k, LJ^3g_j \rangle_{L^2}-2\langle LJ^3g_j, Lu\rangle_{L^2}+2\sum\limits_{l = 1}^{2}(\sum\limits_{k = 1}^{n}c_kJ^3g_kJ^3g_j-J^3g_jB_lu).$

Let

$\frac{\partial}{\partial c_j}G(c_1, \cdots, c_{n}) = 0,$

$\begin{eqnarray} \sum\limits_{k = 1}^{n}c_k\langle LJ^3g_k, LJ^3g_j \rangle_{L^2}+2\sum\limits_{k = 1}^{n}c_{k}J^3g_k J^3g_j = \langle LJ^3g_j, Lu \rangle_{L^2}+\sum\limits_{l = 1}^{2}J^3g_jB_lu. \end{eqnarray}$

(3.3)

Let $\mathit{\boldsymbol{A_{n}}}$ be the n-order matrix and $\mathit{\boldsymbol{b_{n}}}$ be the n-dimensional vector, i.e.,

$\begin{align*} \begin{array}{c} \mathit{\boldsymbol{A_{n}}} \end{array}& = \left( \begin{array}{c} \langle LJ^3g_k, LJ^3g_j \rangle_{L^2}+2J^3g_k J^3g_j \end{array} \right)_{n\times n}, \\ \begin{array}{c} \mathit{\boldsymbol{b_{n}}} \end{array}& = \left( \begin{array}{c} \langle LJ^3g_k, Lu \rangle_{L^2}+\sum\limits_{l = 1}^{2}J^3g_jB_lu \end{array} \right)_{n}. \end{align*}$

Then Eq (3.3) changes to

$\begin{eqnarray} \mathit{\boldsymbol{A_nc = b_n}}. \end{eqnarray}$

(3.4)

If $L$ is invertible, Eq (3.4) has only one solution $\mathit{\boldsymbol{c}}^*$ , and $\mathit{\boldsymbol{c}}^*$ is minimum point of $G$ . Equation (3.4) has an unique solution is proved as follows.

Theorem 3.2. If $L$ is invertible, Eq (3.3) has only one solution.

Proof. The homogeneous linear equation of Eq (3.4) is

$\begin{eqnarray*} \sum\limits_{k = 1}^{n}c_k\langle LJ^3g_k, LJ^3g_j \rangle_{L^2}+2\sum\limits_{k = 1}^{n}c_{k}J^3g_k J^3g_j = 0. \end{eqnarray*}$

Just prove that the above equation has an unique solution. Let $c_j (j = 1, 2, \cdots, n)$ multiply to both sides of the equation, and add all equations together so that

$\begin{eqnarray*} \langle \sum\limits_{k = 1}^{n}c_k LJ^3g_k, \sum\limits_{j = 1}^{n}c_jLJ^3g_j \rangle_{L^2}+2\sum\limits_{k = 1}^{n}c_{k}J^3g_k \sum\limits_{j = 1}^{n}c_{j}J^3g_j = 0. \end{eqnarray*}$

That is

$\begin{eqnarray*} \|\sum\limits_{k = 1}^{n}c_k LJ^3g_k\|_{L^2}+2(\sum\limits_{k = 1}^{n}c_{k}J^3g_k )^2 = 0. \end{eqnarray*}$

Clearly,

$\|\sum\limits_{k = 1}^{n}c_k LJ^3g_k\|_{L^2}^2 = 0, \qquad (\sum\limits_{k = 1}^{n}c_{k}J^3g_k )^2 = 0.$

Because $J^3g_k$ is orthonormal basis, if $L$ is invertible, $c_k = 0$ . So Eq (3.3) has only one solution. □

4. Analysis of convergence and stability

Convergence and stability are important properties of algorithms. This section deals with the convergence and stability.

4.1. Analysis of convergence

In order to discuss the convergence, Theorem 4.1 is given as follows:

Theorem 4.1. $J^3\psi_{ik}^{l}(x)$ is compactly supported in $[\frac{k}{2^{i-1}}, \frac{k+1}{2^{i-1}}]$ .

Proof. When

$x < \frac{k}{2^{i-1}}, \ \ \ J^3\psi_{ik}^{l}(x) = 0.$

When $x > \frac{k+1}{2^{i-1}}$ , because of $\psi_{ik}^{l}(x)$ with compact support, then,

$\begin{align} J^3\psi_{ik}^{l}(x)& = \frac{1}{2}\int_0^x (x-t)^2\psi_{ik}^{l}(t)dt = \frac{1}{2} \int_{\frac{k}{2^{i-1}}}^{\frac{k+1}{2^{i-1}}}(x-t)^2\psi_{ik}^{l}(t)dt\\ & = 2^{\frac{i-3}{2}} \int_{\frac{k}{2^{i-1}}}^{\frac{k+1}{2^{i-1}}}(x-t)^2\psi^{l}(2^{i-1}t-k)dt\\ & = 2^{-\frac{5(i-1)}{2}} \int_0^1(s-2^{i-1}x-k)^2\psi^{l}(s)ds, \ \; \; s = 2^{i-1}t-k. \end{align}$

(4.1)

According to Eq (2.2), $J^3\psi_{ik}^{l}(x) = 0$ . So $J^3\psi_{ik}^{l}(x)$ has compactly support in $[\frac{k}{2^{i-1}}, \frac{k+1}{2^{i-1}}]$ . □

Note

$(J^3\psi_{i, k}^{l}(x))^{\prime} = J^2\psi_{i, k}^{l}(x), \ \; \; (J^3\psi_{i, k}^{l}(x))^{\prime\prime} = J^1\psi_{i, k}^{l}(x).$

By referring to the proof of Theorem 4.1, $J^1\psi_{ik}^{l}(x)$ and $J^2\psi_{ik}^{l}(x)$ are compactly supported in $[\frac{k}{2^{i-1}}, \frac{k+1}{2^{i-1}}]$ .

The order of convergence will proceed below. Assume

$\begin{equation} u(x) = \sum\limits_{j = 0}^2 c_j\frac{x^j}{j!}+\sum\limits_{j = 0}^3 d_j J^3\varphi^j(x)+\sum\limits_{i = 1}^{\infty}\sum\limits_{k = 0}^{2^{i-1}-1}\sum\limits_{l = 0}^3\left(c_{i, k}^{(l)}J^3\psi_{i, k}^{l}\right), \end{equation}$

(4.2)

where

$c_j = < u, \frac{x^j}{j!} > _{W_2^3}, \ \ \ d_j = < u, \varphi^j(x) > _{W_2^3},$

and

$c_{i, k}^{(l)} = < u, \ \ \ J^3\psi_{i, k}^{l}(x) > _{W_2^3}.$

And

$u_n(x) = \sum\limits_{j = 0}^2 c_j\frac{x^j}{j!}+\sum\limits_{j = 0}^3 d_j J^3\varphi^j(x)+\sum\limits_{i = 1}^{n}\sum\limits_{k = 0}^{2^{i-1}-1}\sum\limits_{l = 0}^3\left(c_{i, k}^{(l)}J^3\psi_{i, k}^{l}\right).$

Theorem 4.2. Assume $u_n^{{\varepsilon}}(x)$ is the ${\varepsilon}$ -approximate solution of Eq (3.1). If $u^{(m)}(x)$ is bounded in $[0, 1]$ , $m \in \mathbb{N}, 3\le m\le7$ , then,

$|u(x)-u_n^{{\varepsilon}}(x)|\le 2^{-(m-2)n}M,$

here $M$ is a constant.

Proof. From Definition 3.1 and Theorem 3.1, we get

$\begin{align} |u(x)-u_n^{{\varepsilon}}(x)|&\le M_0\|u-u_n^{{\varepsilon}}\|_{W_2^3}\le M_0\|L^{-1}\|\|L(u-u_n^{{\varepsilon}})\|_{L^2}\\ &\le M_0\|L^{-1}\|(\|L(u-u_n^{{\varepsilon}})\|_{L^2}+|B_1(u-u_n^{{\varepsilon}})|+|B_2(u-u_n^{{\varepsilon}})|)\\ &\le M_0\|L^{-1}\|(\|L(u-u_n)\|_{L^2}+\|B_1(u-u_n)\|+\|B_2(u-u_n)\|). \end{align}$

Obviously,

$\|B_1(u-u_n)\| = 0, \; \; \; \|B_2(u-u_n)\| = 0.$

That is

$\begin{align} |u(x)-u_n^{{\varepsilon}}(x)| &\le M_0\|L^{-1}\|\|L(u-u_n)\|_{L^2}\le M_0\|L^{-1}\|(\int_0^1 (L(u-u_n))^2dx)^{\frac{1}{2}}\\ &\le M_0\|L^{-1}(\|\max\limits_{x\in [0, 1]}\{|L(u-u_n)|^2\})^{\frac{1}{2}}\\ &\le 3M_0\|L^{-1}\|M_1 \max\limits_{x\in[0, 1]}\{|u-u_n|, |u^{\prime}-u_n^{\prime}|, |u^{\prime\prime}-u_n^{\prime\prime}|\}, \end{align}$

where

$M_1 = \max\limits_{x\in [0, 1]}\{1, |p(x)|, |q(x)|\}.$

We know

$|u-u_n| = |\sum\limits_{i = n+1}^{\infty}\sum\limits_{k = 0}^{2^{i-1}-1}\sum\limits_{l = 0}^{3}c_{i, k}^{(l)}J^3\psi_{i, k}^{l}(x)| \le \sum\limits_{i = n+1}^{\infty}\sum\limits_{k = 0}^{2^{i-1}-1}\sum\limits_{l = 0}^{3}|c_{i, k}^{(l)}||J^3\psi_{i, k}^{l}(x)|.$

By the compactly support of $J^p\psi_{ik}^{l}(x), \; p = 1, 2, 3$ , fixed $i$ , then $J^p\psi_{ik}^{l}(x)\neq 0$ only in $[\frac{k}{2^{i-1}}, \frac{k+1}{2^{i-1}}]$ ,

$|u-u_n| \le \sum\limits_{i = n+1}^{\infty}\sum\limits_{l = 0}^{3}|c_{i, k}^{(l)}||J^3\psi_{i, k}^{l}(x)|.$

Similarly,

$|u^{\prime}-u_n^{\prime}| \le \sum\limits_{i = n+1}^{\infty}\sum\limits_{l = 0}^{3}|c_{i, k}^{(l)}||J^2\psi_{i, k}^{l}(x)|$

and

$|u^{\prime\prime}-u_n^{\prime\prime}| \le \sum\limits_{i = n+1}^{\infty}\sum\limits_{l = 0}^{3}|c_{i, k}^{(l)}||J^1\psi_{i, k}^{l}(x)|.$

Through $J^1\psi_{i, k}^{l}(x), J^2\psi_{i, k}^{l}(x)$ and $J^3\psi_{i, k}^{l}(x)$ , you can get

$|u(x)-u_n^{{\varepsilon}}(x)|\le 3M_0 M_1\|L^{-1}\| |u^{\prime\prime}-u_n^{\prime\prime}|.$

As $|u^{\prime\prime}-u_n^{\prime\prime}|$ , $|c_{i, k}^{(l)}|$ and $|J^1\psi_{i, k}^{l}(x)|$ will be discussed below. We can get that $|c_{i, k}^{(l)}|$ is related to $u^{(m)}(x)$ . In fact,

$\begin{equation} |c_{i, k}^{(l)}| = | < u, J^3\psi_{i, k}^{l}(x) > _{W_2^3}| = |\int_0^1(u^{\prime\prime\prime}(x))\psi_{i, k}^{l}(x)dx| = |\int_{\frac{k}{2^{i-1}}}^{\frac{k+1}{2^{i-1}}}u^{\prime\prime\prime}(x)\psi_{i, k}^{l}(x)dx|. \end{equation}$

(4.3)

Taylor's expansion of $u^{\prime\prime\prime}(x)$ at $\frac{k}{2^{i-1}}$ is

$u^{\prime\prime\prime}(x) = \sum\limits_{j = 3}^{m-1}\frac{u^{(j)}(\frac{k}{2^{i-1}})}{(j-3)!}(x-\frac{k}{2^{i-1}})^{j-3}+\frac{u^{(m)}(\xi)}{(m-3)!}(x-\frac{k}{2^{i-1}})^{m-3}, \ \; \; \; \xi\in[\frac{k}{2^{i-1}}, \frac{k+1}{2^{i-1}}].$

Equation (4.3) is changed to

$\begin{align} |c_{i, k}^{(l)}|& = \left|\int_{\frac{k}{2^{i-1}}}^{\frac{k+1}{2^{i-1}}}\left(\sum\limits_{j = 3}^{m-1}\frac{u^{(j)}(\frac{k}{2^{i-1}})}{(j-3)!}(x-\frac{k}{2^{i-1}})^{j-3}+\frac{u^{(m)}(\xi)}{(m-3)!}(x-\frac{k}{2^{i-1}})^{m-3}\right)\psi_{i, k}^{l}(x)dx\right|\\ & = \left|\sum\limits_{j = 3}^{m-1}\frac{u^{(j)}(\frac{k}{2^{i-1}})}{(j-3)!}\int_{\frac{k}{2^{i-1}}}^{\frac{k+1}{2^{i-1}}}(x-\frac{k}{2^{i-1}})^{j-3}\psi_{ik}^l(x)dx\right|+\left|\int_{\frac{k}{2^{i-1}}}^{\frac{k+1}{2^{i-1}}}\frac{u^{(m)}(\xi)}{(m-3)!}(x-\frac{k}{2^{i-1}})^{m-3})\psi_{i, k}^{l}(x)dx\right|, \end{align}$

where

$\begin{align*} \begin{aligned} \int_{\frac{k}{2^{i-1}}}^{\frac{k+1}{2^{i-1}}}(x-\frac{k}{2^{i-1}})^{j-3}\psi_{ik}^l(x)dx = 2^{\frac{i-1}{2}}\int_{\frac{k}{2^{i-1}}}^{\frac{k+1}{2^{i-1}}}(x-\frac{k}{2^{i-1}})^{j-3}\psi^l(2^{i-1}x-k)dx \mathop = \limits^{t = 2^{i-1}x-k}2^{\frac{(-3-2j)(i-1)}{2}}\int_0^1(t)^{j-3}\psi^l(t)dt.\end{aligned} \end{align*}$

According to Eq (2.2),

$\sum\limits_{j = 3}^{m-1}\frac{u^{(j)}(\frac{k}{2^{i-1}})}{(j-3)!}\int_{\frac{k}{2^{i-1}}}^{\frac{k+1}{2^{i-1}}}(x-\frac{k}{2^{i-1}})^{j-3}\psi_{ik}^l(x)dx = 0,$

$\begin{align} |c_{i, k}^{(l)}|& = \left|\int_{\frac{k}{2^{i-1}}}^{\frac{k+1}{2^{i-1}}}\frac{u^{(m)}(\xi)}{(m-3)!}(x-\frac{k}{2^{i-1}})^{m-3}\psi_{i, k}^{l}(x)dx\right|\\ &\le \left|\frac{u^{(m)}(\xi)}{(m-3)!}\right|\int_{\frac{k}{2^{i-1}}}^{\frac{k+1}{2^{i-1}}}|x-\frac{k}{2^{i-1}}|^{m-3}|\psi_{i, k}^{l}(x)|dx\\ &\le \left|\frac{u^{(m)}(\xi)}{(m-3)!}\right| 2^{-(m-3)(i-1)}\int_{\frac{k}{2^{i-1}}}^{\frac{k+1}{2^{i-1}}}|\psi_{i, k}^{l}(x)|dx\\ &\le \left|\frac{u^{(m)}(\xi)}{(m-3)!}\right| 2^{-(m-3)(i-1)} 2^{-\frac{i-1}{2}}. \end{align}$

Because $u^{(m)}(x)$ is bounded,

$|\frac{u^{(m)}(\xi)}{(m-3)!}|\le M_3,$

then,

$\begin{equation} |c_{i, k}^{(l)}|\le 2^{-\frac{(2m-5)(i-1)}{2}}M_3. \end{equation}$

(4.4)

By the compactly support of $J^1\psi_{ik}^{l}(x)$ ,

$|J^1\psi_{i, k}^{l}(x)| = \left|\int_0^x \psi_{i, k}^{l}(t) dt\right|\le \int_{\frac{k}{2^{i-1}}}^{\frac{k+1}{2^{i-1}}}|\psi_{i, k}^{l}(t)| dt\le 2^{-\frac{i-1}{2}}.$

According to the above analysis,

$|u^{\prime\prime}-u_n^{\prime\prime}|\le\sum\limits_{i = n+1}^{\infty}4M_32^{-\frac{(2m-5)(i-1)}{2}}2^{-\frac{(i-1)}{2}} = 4M_3 2^{-(m-2)n}.$

That is

$|u(x)-u_n^{{\varepsilon}}(x)|\le 2^{-(m-2)n}M,$

where $M$ is a constant. □

4.2. Analysis of stability

Stability analysis is conducted below. According to the third section, the stability of the algorithm is related to the stability of Eq (3.4). By the following Property 4.1, the stability of the algorithm can be discussed by the number of conditions of the matrix $\mathit{\boldsymbol{A}}$ .

Property 4.1. If the matrix $\mathit{\boldsymbol{A}}$ is symmetric and reversible, then

$cond(A) = \bigg|\frac{\lambda_{max}}{\lambda_{min}}\bigg|,$

where ${\lambda_{max}}$ and ${\lambda_{min}}$ are the largest and smallest eigenvalues of $\mathit{\boldsymbol{A}}$ respectively.

In this paper,

$\begin{array}{c} \mathit{\boldsymbol{A_{n}}} \end{array} = \left( \begin{array}{c} a_{ij} \end{array} \right)_{n\times n} = \left( \begin{array}{c} \langle LJ^3g_i, LJ^3g_j \rangle_{L^2}+2J^3g_i J^3g_j \end{array} \right)_{n\times n}.$

Clearly, $\mathit{\boldsymbol{A_{n}}}$ is symmetric. From Theorem 3.2, $\mathit{\boldsymbol{A_{n}}}$ is reversible. In order to discuss the stability of the algorithm, only the eigenvalues of matrix $\mathit{\boldsymbol{A_{n}}}$ need to be discussed.

Theorem 4.3. Assume $u\in W_2^3$ and $\|u\|_{W_2^3} = 1$ . If $L$ is an invertible differential operator, then,

$\|Lu\|_{L^2}\geq \frac{1}{\|L^{-1}\|}.$

Proof. Since $L$ is an invertible, assume $Lu = v$ , then $u = L^{-1}v$ . Moreover

$1 = \|u\|_{W_2^3} = \|L^{-1}v\|_{W_2^3}\leq \|L^{-1}\|\|v\|_{L^2}.$

Then,

$\|v\|_{L^2}\geq \frac{1}{L^{-1}}.$

That is,

$\|Lu\|_{L^2}\geq \frac{1}{\|L^{-1}\|}.$

□

Theorem 4.4. Let ${\lambda}$ {be} the eigenvalues of matrix $\mathit{\boldsymbol{A}}$ of Eq (3.4), $\mathit{\boldsymbol{x}} = (x_1, \cdots, x_n)^{T}$ is related eigenvalue of $\lambda$ and $\|\mathit{\boldsymbol{x}}\| = 1$ , then,

$\lambda\leq \|L\|^2+2.$

Proof. By $\mathit{\boldsymbol{Ax}} = \lambda \mathit{\boldsymbol{x}}$ ,

$\begin{align} \lambda x_i& = \sum\limits_{j = 1}^na_{ij}x_j = \sum\limits_{j = 1}^n(\langle LJ^3g_i, LJ^3g_j \rangle_{L^2}+2J^3g_i J^3g_j)x_j\\ & = \langle LJ^3g_i, \sum\limits_{j = 1}^nx_jLJ^3g_j \rangle_{L^2}+2J^3g_i \sum\limits_{j = 1}^n(J^3g_jx_j), \; \; \; i = 1, \cdots, n. \end{align}$

(4.5)

Let $x_i$ multiply to both sides of Eq (4.5), and then add the equations from $j = 1$ to $j = n$ together so that

$\begin{align} \lambda& = \lambda x_i^2 = \langle \sum\limits_{i = 1}^nx_iLJ^3g_i, \sum\limits_{j = 1}^nx_jLJ^3g_j \rangle_{L^2}+2\sum\limits_{i = 1}^n(J^3g_ix_i)\sum\limits_{j = 1}^n(J^3g_jx_j)\\ & = \big\|\sum\limits_{i = 1}^nx_iLJ^3g_i\big\|_{L^2}^2+2\big(\sum\limits_{i = 1}^n(J^3g_ix_i)\big)^2\\ &\leq \|L\|^2\sum\limits_{i = 1}^nx_i^2+2\sum\limits_{i = 1}^nx_i^2\\ & = (\|L\|^2+2)\|\mathit{\boldsymbol{x}}\|. \end{align}$

(4.6)

Since

$\|\mathit{\boldsymbol{x}}\| = 1, \ \ \ \lambda\leq \|L\|^2+2.$

□

From Theorem 4.3 and Eq (4.6), we can get

$\lambda \geq \big\|\sum\limits_{i = 1}^nx_iLJ^3g_i\big\|_{L^2} = \big\|L(\sum\limits_{i = 1}^nx_iJ^3g_i)\big\|_{L^2}\geq\frac{1}{L^{-1}}.$

Then,

$cond(\mathit{\boldsymbol{A}}) = \big|\frac{\lambda_{max}}{\lambda_{min}}\big|\leq\frac{\|L\|^2+2}{\frac{1}{\|L^{-1}\|}} = (\|L\|^2+2)\|L^{-1}\|.$

That is the condition number of $\mathit{\boldsymbol{A}}$ is bounded, so the presented method is stable.

5. Numerical examples

This section discusses numerical examples to reveal the accuracy of the proposed algorithm. Examples 5.1 and 5.3 are linear and nonlinear BVPs respectively. Example 5.2 shows that our method also applies to Eq (1.1) with other linear boundary value conditions. In this paper, $N$ is the number of bases, and

$N = 7+4*(2^n-1), \; \; \; n = 1, 2, \cdots.$

$e_N(x)$ is the absolute errors. $C.R.$ and $cond$ represent the convergence order and the condition number respectively. For convenience, we denote

$e_N(x) = |u(x)-u_N(x)|$

and

$C.R. = \log_2\frac{\max|e_{N}(x)|}{\max|e_{N+1}(x)|}.$

Example 5.1. Consider the test problem suggested in ^[28,30]

$\begin{equation*} \left\{ \begin{array}{l} u^{\prime\prime} = u^{\prime}+2u+4x-2e^x, \quad x \in (0, 1), \\ u(0) = 2, u(1) = e-1, \end{array} \right. \end{equation*}$

where the exact solution is $u(x) = e^x-2x+1$ . The numerical results are shown in Table 1. It is clear from Table 1 that the present method produces a converging solution for different values. In addition, the results of the proposed algorithm in Table 1 are compared with those in ^[28,30]. Obviously, the proposed algorithm is better. shows $e_N(x)$ , $C.R.$ , $cond$ and CPU time. The unit of CPU time is second, expressed as $s$ .

Table 1.

$e_N(x)$ of Example 5.1.

$x$	$e_N(x)$ of ^[30]	$e_{66}(x)$ of ^[28]	$e_{35}(x)$	$e_{67}(x)$
$0$	0	5.67e-9	9.94e-14	8.88e-16
$0.1$	1.19e-5	3.35e-9	2.04e-13	2.22e-16
$0.2$	4.18e-5	3.93e-10	2.16e-13	1.55e-15
$0.3$	4.96e-5	1.33e-9	1.42e-13	8.88e-16
$0.4$	6.04e-5	1.40e-9	1.68e-14	1.33e-15
$0.5$	6.33e-5	1.82e-9	1.82e-13	4.44e-16
$0.6$	6.23e-5	5.96e-9	1.52e-13	2.66e-15
$0.7$	5.76e-5	1.14e-8	1.66e-13	8.88e-16
$0.8$	4.23e-5	1.52e-8	4.36e-13	4.44e-15
$0.9$	2.15e-5	1.66e-8	4.93e-13	4.44e-16
$1.0$	0	1.90e-8	2.67e-13	4.44e-15

| Show Table

DownLoad: CSV

Table 2.

$e_N(x)$ ,

$C.R.$ and

$cond$ of Example 5.1.

$n$	$N$	$\max{e_N(x)}$	$C.R.$	$cond$	CPU(s)
$1$	11	1.66e-8		274.262	2.57
$2$	19	6.85e-11	7.92	274.262	7.89
$3$	35	6.55e-13	6.71	274.262	24.42
$4$	67	7.93e-15	6.40	274.262	82.73

| Show Table

DownLoad: CSV

Example 5.2. Consider the problem suggested in ^[25,28].

$\begin{equation*} \left\{ \begin{array}{l} u^{\prime\prime}+u^{\prime}+xu = f(x), \quad x \in (0, 1), \\ u(0) = 2, \quad u(1)+u(\frac{1}{2}) = \sin{\frac{1}{2}}+\sin{1}. \end{array} \right. \end{equation*}$

The exact solution is $u(x) = \sin{x}$ , and $f(x) = \cos{x}-\sin{x}+x\sin{x}$ . This problem is the boundary value problem with the multipoint boundary value conditions. shows maximum absolute error $ME_n$ , $C.R.$ and $cond$ ., which compared with the other algorithms, the results obtained demonstrate that our algorithm is remarkably effective. The numerical errors are provided in Figures 1 and 2, also show a good accuracy.

Table 3.

$ME_n$ , C.R. and

$cond$ of Example 5.2.

The present method				^[25]				^[28]
$n$	$ME_n$	C.R.	$cond$	$n$	$ME_n$	C.R.	$cond$	$n$	$ME_n$	C.R.	$cond$
11	3.73e-9		195.05	10	6.34e-6	3.96	182.06	11	1.88e-4
19	2.97e-11	6.97	195.05	18	4.04e-7	3.98	182.06	19	5.99e-5	1.65	1.49 $\times 10^6$
35	2.13e-13	7.12	195.05	34	2.54e-8	4.06	182.06	35	1.84e-5	1.70	3.74 $\times 10^8$
67	2.77e-15	6.27	195.05	66	1.59e-9	3.95	182.06	67	4.62e-6	1.99	8.87 $\times 10^{10}$

| Show Table

DownLoad: CSV

Figure 1.

$e_N(x)$ of Example 5.3 (n = 35).

DownLoad: Full-Size Img PowerPoint

Figure 2.

$e_N(x)$ of Example 5.3 (n = 67).

DownLoad: Full-Size Img PowerPoint

Example 5.3. Consider a nonlinear problem suggested in ^[7,9]

$\begin{eqnarray*} \left\{ \begin{array}{l} u^{\prime\prime}+\lambda e^{u} = 0, \quad x\in(0, 1)\\ u(0) = 0, u(1) = 0, \end{array} \right. \end{eqnarray*}$

where

$u(x) = -2ln(\cosh((x-\frac{1}{2})(\theta/2))/\cosh(\theta/4)),$

and $\theta$ satisfies

$\theta-\sqrt{2\lambda}\cosh(\theta/4) = 0.$

This is the second-order nonlinear Bratu problem. { Bratu equation is widely used in engineering fields, such as spark discharge, semiconductor manufacturing, etc. In the field of physics, the Bratu equation is used to describe the physical properties of microcrystalline silica gel solar energy. In the biological field, the Bratu equation is used to describe the kinetic model of some biochemical reactions in living organisms.} To this problem, taking $u_0(x) = x(1-x), \; k = 3$ , where $k$ is the number of iterations of the algorithm mentioned in ^[27]. when $\lambda = 1, \lambda = 2$ , $e_N(x)$ are listed in Tables 4 and 5, respectively.

Table 4.

$e_N(x)$ of Example 5.3 (

$\lambda = 1$ ).

$x$	$e_N(x)$ of ^[8]	$e_N(x)$ of ^[9]	$e_N(x)$
0	0	$0$	$4.4959e-11$
0.2	$1.4958e-9$	$2.4390e-5$	$4.1096e-11$
0.4	$2.7218e-9$	$4.2096e-5$	$7.1502e-12$
0.6	$2.7218e-9$	$4.2096e-5$	$7.1483e-12$
0.8	$1.4958e-9$	$2.4390e-5$	$4.1104e-11$

| Show Table

DownLoad: CSV

Table 5.

$e_N(x)$ of Example 5.3 (

$\lambda = 2$ ).

$x$	$e_N(x)$ of ^[7]	$e_N(x)$ of ^[9]	$e_N(x)$
0	$5.8988e-26$	$0$	$1.1801e-12$
0.2	$1.3070e-7$	$6.9297e-5$	$2.5646e-10$
0.4	$1.4681e-7$	$1.0775e-4$	$1.6666e-9$
0.6	$1.4681e-7$	$1.0775e-4$	$1.6666e-9$
0.8	$1.3070e-7$	$6.9297e-5$	$2.5646e-10$

| Show Table

DownLoad: CSV

6. Conclusions

In this paper, based on Legendre's polynomials, we construct orthonormal basis in $L^2[0, 1]$ and $W_2^3[0, 1]$ , respectively. It proves that this group of bases is orthonormal and compactly supported. According to the orthogonality of the basis, we present an algorithm to obtain the approximate solution of the boundary value problems. Using the compact support of the basis, we prove that the convergence order of the presented method related to the boundedness of $u^{(m)}(x)$ . Finally, three numerical examples show that the absolute error and convergence order of the algorithm are better than other methods.

Use of AI tools declaration

The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.

Acknowledgments

This study was supported by National Natural Science Funds of China by Grant number (12101164), Characteristic Innovative Scientific Research Project of Guangdong Province (2023KTSCX181, 2023KTSCX183) and Basic and Applied Basic Research Project Zhuhai City (ZH24017003200026PWC).

Conflict of interest

The authors have no conflicts of interest to declare.

References

[1]	D. J. Watts, S. Strogatz, Collective dynamics of 'small-world' networks, Nature, 393 (1998), 440–442. https://doi.org/10.1038/30918 doi: 10.1038/30918
[2]	A. L. Barabási, R. Albert, Emergence of scaling in random networks, Science, 286 (1999), 509–512. https://doi.org/10.1126/science.286.5439.509 doi: 10.1126/science.286.5439.509
[3]	J. Lü, X. Yu, G. Chen, Chaos synchronization of general complex dynamical networks, Physica A, 334 (2004), 281–302. https://doi.org/10.1016/j.physa.2003.10.052 doi: 10.1016/j.physa.2003.10.052
[4]	S. Strogatz, Exploring complex networks, Nature, 410 (2001), 268–176. https://doi.org/10.1038/35065725 doi: 10.1038/35065725
[5]	M. E. J. Newman, The structure and function of complex networks, SIAM Review, 45 (2003), 167–256. https://doi.org/10.1137/S003614450342480 doi: 10.1137/S003614450342480
[6]	Y. Shen, J. Wang, Robustness analysis of global exponential stability of recurrent neural networks in the presence of time delays and random disturban, IEEE T. Neur. Net. Lear., 45 (2011), 87–96. https://doi.org/10.1109/TNNLS.2011.2178326 doi: 10.1109/TNNLS.2011.2178326
[7]	W. Fang, T. Xie, B. Li, Robustness analysis of BAM cellular neural network with deviating arguments of generalized type, Discrete Dyn. Nat. Soc., 2023 (2023), 9570805. https://doi.org/10.1155/2023/9570805 doi: 10.1155/2023/9570805
[8]	C. Huygens, H. Oscillatorium, The pendulum clock, Trans RJ Blackwell, Ames: The Iowa State University Press, 1986.
[9]	I. Blekhman, The problem of synchronization of dynamical systems, J. Appl. Math. Mec., 28 (1964), 239–265. https://doi.org/10.1016/0021-8928(64)90160-1 doi: 10.1016/0021-8928(64)90160-1
[10]	A. Arenas, A. Díaz-Guilera, J. Kurths, Y. Moreno, C. Zhou, Synchronization in complex networks, Phys. Rep., 469 (2004), 93–153. https://doi.org/10.1016/j.physrep.2008.09.002 doi: 10.1016/j.physrep.2008.09.002
[11]	Y. Xu, H. Yang, D. Tong, Y. Wang, Adaptive exponential synchronization in pth moment for stochastic time varying multi-delayed complex networks, Nonlinear Dynam., 73 (2013), 1423–1431. https://doi.org/10.1007/s11071-013-0873-0 doi: 10.1007/s11071-013-0873-0
[12]	Y. Li, C. Li, Complete synchronization of delayed chaotic neural networks by intermittent control with two switches in a control period, Neurocomputing, 173 (2016), 1341–1347. https://doi.org/10.1016/j.neucom.2015.09.007 doi: 10.1016/j.neucom.2015.09.007
[13]	G. Wang, J. Xiao, Y. Wang, J. Yi, Adaptive pinning cluster synchronization of fractional-order complex dynamical networks, Appl. Math. Comput., 231 (2014), 347–356. https://doi.org/10.1016/j.amc.2014.01.023 doi: 10.1016/j.amc.2014.01.023
[14]	Z. Zhang, H. Wu, Cluster synchronization in finite/fixed time for semi-markovian switching ts fuzzy complex dynamical networks with discontinuous dynamic nodes, AIMS Math., 7 (2022), 11942–11971. http://dx.doi.org/10.3934/math.2022666 doi: 10.3934/math.2022666
[15]	Y. Cao, Y. Kao, J. H. Park, H. Bao, Global mittag-leffler stability of the delayed fractional-coupled reaction-diffusion system on networks without strong connectedness, IEEE T. Neur. Net. Lear., 31 (2021), 6473–6483. https://doi.org/10.1109/TNNLS.2021.3080830 doi: 10.1109/TNNLS.2021.3080830
[16]	Y. Kao, Y. Li, J. H. Park, X. Chen, Mittag-leffler synchronization of delayed fractional memristor neural networks via adaptive control, IEEE T. Neur. Net. Lear., 32 (2020), 2279–2284. https://doi.org/10.1109/TNNLS.2020.2995718 doi: 10.1109/TNNLS.2020.2995718
[17]	H. Du, P. Shi, N. Lü, Function projective synchronization in complex dynamical networks with time delay via hybrid feedback control, Nonlinear Anal.-Real, 14 (2013), 1182–1190. https://doi.org/10.1016/j.nonrwa.2012.09.009 doi: 10.1016/j.nonrwa.2012.09.009
[18]	W. Wong, H. Li, S. Leung, Robust synchronization of fractional-order complex dynamical networks with parametric uncertainties, Com. Nonlinear Sci., 17 (2012), 4877–4890. https://doi.org/10.1016/j.cnsns.2012.05.020 doi: 10.1016/j.cnsns.2012.05.020
[19]	R. Cheng, M. Peng, W. Yu, Pinning synchronization of delayed complex dynamical networks with nonlinear coupling, Physica A, 413 (2014), 426–431. https://doi.org/10.1016/j.physa.2014.06.034 doi: 10.1016/j.physa.2014.06.034
[20]	W. Shen, Z. Zeng, S. Wen, Synchronization of complex dynamical network with piecewise constant argument of generalized type, Neurocomputing, 173 (2016), 671–675. https://doi.org/10.1016/j.neucom.2015.08.014 doi: 10.1016/j.neucom.2015.08.014
[21]	C. Shi, G. Yang, X. Li, Event-triggered output feedback synchronization control of complex dynamical networks, Neurocomputing, 275 (2018), 29–39. https://doi.org/10.1016/j.neucom.2017.05.014 doi: 10.1016/j.neucom.2017.05.014
[22]	Q. Dong, S. Shi, Y. Ma, Non-fragile synchronization of complex dynamical networks with hybrid delays and stochastic disturbance via sampled-data control, ISA T., 105 (2020), 174–189. https://doi.org/10.1016/j.neucom.2017.05.014 doi: 10.1016/j.neucom.2017.05.014
[23]	Z. Wu, J. H. Park, H. Su, B. Song, J. Chu, Exponential synchronization for complex dynamical networks with sampled-data, J. Franklin I., 349 (2012), 2735–2749. https://doi.org/10.1016/j.jfranklin.2012.09.002 doi: 10.1016/j.jfranklin.2012.09.002
[24]	H. Yang, X. Wang, S. Zhong, L. Shu, Observer-based asynchronous event-triggered control for interval type-2 fuzzy systems with cyber-attacks, Inform. Sciences, 606 (2022), 805–818. https://doi.org/10.1016/j.ins.2022.05.087 doi: 10.1016/j.ins.2022.05.087
[25]	X. Wang, J. H. Park, Z. Liu, H. Yang, Dynamic event-triggered control for GSES of memristive neural networks under multiple cyber-attacks, IEEE T. Neur. Net. Lear., 2022. https://doi.org/10.1109/TNNLS.2022.3217461 doi: 10.1109/TNNLS.2022.3217461
[26]	X. Wang, J. H. Park, H. Yang, An improved protocol to consensus of delayed MASs with UNMS and aperiodic DoS cyber-attacks, IEEE T. Netw. Sci. Eng., 8 (2021), 2506–2516. https://doi.org/10.1109/TNSE.2021.3098258 doi: 10.1109/TNSE.2021.3098258
[27]	X. Li, H. Wu, J. Cao, Prescribed-time synchronization in networks of piecewise smooth systems via a nonlinear dynamic event-triggered control strategy, Math. Comput. Simulat., 203 (2023), 647–668. https://doi.org/10.1016/j.matcom.2022.07.010 doi: 10.1016/j.matcom.2022.07.010
[28]	X. Li, H. Wu, J. Cao, A new prescribed-time stability theorem for impulsive piecewise-smooth systems and its application to synchronization in networks, Appl. Math. Model., 115 (2023), 385–397. https://doi.org/10.1016/j.apm.2022.10.051 doi: 10.1016/j.apm.2022.10.051
[29]	Q. Zhang, G. Chen, L. Wan, Exponential synchronization of discrete-time impulsive dynamical networks with time-varying delays and stochastic disturbances, Neurocomputing, 309 (2018), 62–69. https://doi.org/10.1016/j.neucom.2018.04.070 doi: 10.1016/j.neucom.2018.04.070
[30]	H. Li, Y. Kao, H. Bao, Y. Chen, Uniform stability of complex-valued neural networks of fractional order with linear impulses and fixed time delays, IEEE T. Neur. Net. Lear., 33 (2021), 5321–5331. https://doi.org/10.1109/TNNLS.2021.3070136 doi: 10.1109/TNNLS.2021.3070136
[31]	W. Zhou, T. Wang, J. Mou, J. Fang, Mean square exponential synchronization in Lagrange sense for uncertain complex dynamical networks, J. Franklin I., 349 (2012), 1267–1282. https://doi.org/10.1016/j.jfranklin.2012.01.011 doi: 10.1016/j.jfranklin.2012.01.011
[32]	Y. Zhang, D. Gu, S. Xu, Global exponential adaptive synchronization of complex dynamical networks with neutral-type neural network nodes and stochastic disturbances, IEEE T. Circuits-I, 60 (2013), 2709–2718. https://doi.org/10.1109/TCSI.2013.2249151 doi: 10.1109/TCSI.2013.2249151
[33]	Y. Wang, Y. Ma, A. Chen, Exponential synchronization of markovian jump complex dynamical networks with partially uncertain transition rates and stochastic disturbances, Neurocomputing, 304 (2018), 30–46. https://doi.org/10.1016/j.neucom.2018.03.044 doi: 10.1016/j.neucom.2018.03.044
[34]	R. Sipahi, S. I. Niculescu, C. T. Abdallah, W. Michiels, K. Q. Gu, Stability and stabilization of systems with time delay, IEEE Contr. Syst. Mag., 31 (2011), 38–65. https://doi.org/10.1109/MCS.2010.939135 doi: 10.1109/MCS.2010.939135
[35]	R. Sipahi, S. I. Niculescu, C. T. Abdallah, W. Michiels, K. Q. Gu, Stabilization of continuous-time hybrid stochastic differential equations by discrete-time feedback control, Automatica, 49 (2013), 3677–3681. https://doi.org/10.1016/j.automatica.2013.09.005 doi: 10.1016/j.automatica.2013.09.005
[36]	H. Gao, J. Lam, G. Chen, New criteria for synchronization stability of general complex dynamical networks with coupling delays, Phys., Lett., A, 360 (2006), 263–273. https://doi.org/10.1016/j.physleta.2006.08.033 doi: 10.1016/j.physleta.2006.08.033
[37]	R. Cheng, M. Peng, W. Yu, B. Sun, J. Yu, Stability analysis and synchronization in discrete-time complex networks with delayed coupling, Chaos: An Interdisciplinary J. Nonlinear Sci., 23 (2013). https://doi.org/10.1063/1.4825095 doi: 10.1063/1.4825095
[38]	X. Liu, T. Chen, Synchronization analysis for nonlinearly-coupled complex networks with an asymmetrical coupling matrix, Physica A, 387 (2008), 4429–4439. https://doi.org/10.1016/j.physa.2008.03.005 doi: 10.1016/j.physa.2008.03.005
[39]	J. Zhu, G. Yang, Robust H dynamic output feedback synchronization for complex dynamical networks with disturbances, Neurocomputing, 175 (2016), 287–292. https://doi.org/10.1016/j.neucom.2015.10.061 doi: 10.1016/j.neucom.2015.10.061
[40]	P. Selvaraj, O. Kwon, R. Sakthivel, Disturbance and uncertainty rejection performance for fractional-order complex dynamical networks, Neural Networks, 112 (2019), 73–84. https://doi.org/10.1016/j.neunet.2019.01.009 doi: 10.1016/j.neunet.2019.01.009
[41]	W. Fang, T. Xie, B. Li, Robustness analysis of fuzzy cellular neural network with deviating argument and stochastic disturbances, IEEE Access, 11 (2023), 3717–3728. https://doi.org/10.1109/ACCESS.2023.3233946 doi: 10.1109/ACCESS.2023.3233946
[42]	W. Si, T. Xie, B. Li, Robustness analysis of exponential stability of neutral-type nonlinear systems with multi-interference, IEEE Access, 9 (2021), 116015–116032. https://doi.org/10.1109/ACCESS.2021.3105521 doi: 10.1109/ACCESS.2021.3105521
[43]	W. Si, T. Xie, B. Li, Exploration on robustness of exponentially global stability of recurrent neural networks with neutral terms and generalized piecewise constant arguments, Discrete Dyn. Nat. Soc., 2021 (2021), 9941881. https://doi.org/10.1155/2021/9941881 doi: 10.1155/2021/9941881
[44]	W. Si, T. Xie, B. Li, Further results on exponentially robust stability of uncertain connection weights of neutral-type recurrent neural networks, Complexity, 2021 (2021), 6941701. https://doi.org/10.1155/2021/6941701 doi: 10.1155/2021/6941701
[45]	B. Lia, J. Huang, D. Wang, Robustness analysis of control laws in complex dynamical networks evoked by deviating argument, Discrete Dyn. Nat. Soc., 2022 (2022). https://doi.org/10.1155/2022/2033708 doi: 10.1155/2022/2033708
[46]	X. Mao, Stochastic differential equations and applications, 2 Eds., UK: Woodhead Publishing, 2008.
[47]	N. Li, H. Sun, X. Jing, Q. Zhang, Exponential synchronisation of united complex dynamical networks with multi-links via adaptive periodically intermittent control, IET Control Theory A., 7 (2013), 1725–1736. https://doi.org/10.1049/iet-cta.2013.0159 doi: 10.1049/iet-cta.2013.0159

This article has been cited by:

1.	Mohammed Elamine Beroudj, Abdelaziz Mennouni, Carlo Cattani, Hermite solution for a new fractional inverse differential problem, 2024, 0170-4214, 10.1002/mma.10516
2.	苗苗宋, Inverse Problem of Option Drift Rate Based on Degenerate Parabolic Equations, 2023, 12, 2324-7991, 3814, 10.12677/AAM.2023.129375
3.	Yilihamujiang Yimamu, Zui-Cha Deng, C. N. Sam, Y. C. Hon, Total variation regularization analysis for inverse volatility option pricing problem, 2024, 101, 0020-7160, 483, 10.1080/00207160.2024.2345660

Reader Comments

Your name:*

Email:*
© 2023 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)