A second-order scheme with nonuniform time grids for Caputo–Hadamard fractional sub-diffusion equations☆
Introduction
In the last few decades, a great deal of efforts have been done in the study of fractional calculus and fractional differential equations (FDEs) [1], [2], [3]. Until now, many researchers have carried on the thorough study to several kinds of fractional integrals and derivatives, such as Riemann–Liouville calculus, Caputo calculus, etc. Although Hadamard fractional derivative was first proposed by Hadamard in 1892 [4], there are few studies on this kind of fractional problem besides [5], [6], [7], [8]. In practice, the Hadamard derivative is worth being further investigated as well. It has been found that the Hadamard derivative and Hadamard-type fractional differential equations are of value in mechanics, engineering and so on for the past few years [8]. In addition, compared with the Riemann–Liouville derivative, there are two characteristics of Hadamard derivative [9]. For one thing, in Hadamard derivatives, the integral is kernel to the power of , while that in the Riemann–Liouville one is . For another, the Hadamard derivative is regarded as a generalization of the operator , while the Riemann–Liouville one is considered as an extension of the classical operator . Furthermore, the Hadamard derivative begins at the initial moment that is greater than zero, whereas the Riemann–Liouville one tends to start at the origin or any other real number.
Particularly, Gohar et al. [10] first investigated the smoothness properties of the solution to the initial value problems with Caputo–Hadamard derivative. In [11], the existence and uniqueness of the solution to the fractional ordinary differential equations with Caputo–Hadamard derivative were studied. Li et al. [9] studied the regularity and logarithmic decay for Caputo–Hadamard fractional sub-diffusion equations. In addition, the finite difference scheme with nonuniform meshes was employed to approximate the time fractional derivative and the local discontinuous Galerkin method was used to approximate the spacial derivative. However, the temporary accuracy of its derived scheme is , where denotes the numbers of grids in temporal direction and is the fractional order. In this paper, we study the high order finite difference scheme for the following Caputo–Hadamard fractional sub-diffusion equations: where is the Laplacian, are given functions, and the symbol is Caputo–Hadamard derivative with order : where , .
In the literature, several numerical Caputo formulas have a high-order accuracy, such as the 1-2 formula [12] and the formula [13]. But they achieved second-order accuracy for adequately smooth solutions. It is worth mentioning that the solution of time-fractional differential equations typically displays a weak singularity near the initial time, that is, for more general given data, the high-order time partial derivatives of the solution may not be bounded in the whole closed time domain especially at the initial point, which leads to the loss of time accuracy for many related high-order numerical methods. Recently, more and more researchers have begun to take the initial singularity into account in time fractional models, see [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33] and their references. The nonuniform grids technique [14], [17], [19], [22], [27], [28], [29], [30], [31], [32] works efficiently for time fractional partial/ordinary differential equations with non-smooth solutions. This motivates us to select the nonuniform time meshes , where the grading parameter is adopted to represent the strength of the singularity. And its efficiency is demonstrated by the theoretical and numerical results that will be provided later.
Liao et al. [14] derived a second-order Crank–Nicolson-like scheme with nonuniform meshes for linear reaction–subdiffusion problems with non-smooth solutions by approximation formula. By virtue of the ideas derived in [14], we consider the interpolation for the Caputo–Hadamard derivative. The proposed method in this paper is analyzed under the following regularity assumptions [9]: for all and is a positive constant. Throughout this paper, the generic constant is a positive constant and may take different values at different places.
The structure of the paper is as follows. In Section 2, the approximation to the Caputo–Hadamard fractional derivative and the Crank–Nicolson-like scheme are proposed. In Section 3, the properties of discrete kernels and global consistency error are investigated. The stability and convergence of the resulting discrete scheme are analyzed by the energy method in Section 4. Eventually, the numerical example presenting the validity of theoretical results is given in Section 5 and some conclusions are drawn in Section 6.
Section snippets
Preliminaries
Besides, we divide time interval and space interval as follows. On space interval, for a positive integer , let . On time interval, we adopt nonuniform meshes. For a positive integer , the interval is divided into with Correspondingly, we also divide the interval into with and then we denote . We define a fractional time
The properties of discrete kernels
Before studying the kernels , we introduce the following lemma which is crucial to study the properties of coefficients and error estimate.
Lemma 3.1 Suppose that has a continuous -derivative of order in field of point . A Taylor-like formula with integral remainder will be introduced as follows:
Proof We consider the following formula
Stability and convergence analysis
Before verifying the stability and convergence, a significant lemma is given as follows.
Lemma 4.1 From the Lemma 3.3(II)–(III), the discrete Caputo–Hadamard formula (2.5) with the discrete kernels satisfies The proof of this lemma is similar to [14], [33].
Theorem 4.1 The stability equation is as follows: Then, it holds that
Proof Taking the inner product
Numerical experiments
In this section, we carry out numerical experiments to illustrate our theoretical statements and all our tests are done in MATLAB with a laptop. The -norm errors between the exact and numerical solutions are shown in the following tables. Furthermore, the temporal convergence order denoted by is reported.
The theoretical analysis are confirmed by practical numerical experiments. The current model displays a weak singularity near the initial time,
Conclusions
In this paper, we deal with the Caputo–Hadamard fractional sub-diffusion equations with initial singularity. In order to study the discrete kernels and truncation error, a Taylor-like expansion with integral remainder is proposed. Adopting piecewise interpolation on nonuniform meshes, the ECS analysis that includes an ECS bound and a global consistency error, is proposed for investigating the numerical approximations to the Caputo–Hadamard fractional derivative. As a result, a
Acknowledgments
The authors would like to thank the editor and reviewers for their constructive comments and suggestions, which helped the authors to improve the quality of the paper significantly.
References (34)
- et al.
Mathematical analysis and the local discontinuous Galerkin method for Caputo-Hadamard fractional partial differential equation
J. Sci. Comput.
(2020) - et al.
A new fractional differentiation formula to approximate the Caputo fractional derivative and its applications
J. Comput. Phys.
(2014) A new difference scheme for the time fractional diffusion equation
J. Comput. Phys.
(2015)- et al.
A second-order accurate numerical method with graded meshes for an evolution equation with a weakly singular kernel
J. Comput. Appl. Math.
(2019) - et al.
A fast linearized finite difference method for the nonlinear multi-term time-fractional wave equation
Appl. Numer. Math.
(2020) - et al.
An alternating direction implicit orthogonal spline collocation method for the two-dimensional multi-term time fractional integro-differential equation
Appl. Numer. Math.
(2020) - et al.
Sharp error estimate of a compact L1-ADI scheme for the two-dimensional time-fractional integro-differential equation with singular kernels
Appl. Numer. Math.
(2021) - et al.
Second order difference schemes for time-fractional KdV-Burgers equation with initial singularity
Appl. Math. Lett.
(2021) - et al.
Time two-grid technique combined with temporal second order difference method for two-dimensional semilinear fractional sub-diffusion equations
Appl. Math. Lett.
(2022) - et al.
Mathematical analysis and numerical methods for Caputo-Hadamard fractional diffusion-wave equations
Appl. Numer. Math.
(2022)
A numerical method for distributed order time fractional diffusion equation with weakly singular solutions
Appl. Math. Lett.
A second-order and nonuniform time-stepping maximum-principle preserving scheme for time-fractional Allen-Cahn equations
J. Comput. Phys.
Fractional Integrals and Derivatives: Theory and Applications
Fractional Differential Equations
Theory and Applications of Fractional Differential Equations
Essai sur létude des fonctions données par leur développement de Taylor
J. Math. Pures Appl.
Hadamard-type fractional calculus
J. Korean Math. Soc.
Cited by (20)
Legendre-finite difference method for solving fractional nonlinear Sobolev equation with Caputo derivative
2023, Journal of Computational ScienceA Galerkin finite element method for the space Hadamard fractional partial differential equation
2023, Mathematics and Computers in SimulationA linear Galerkin numerical method for a quasilinear subdiffusion equation
2023, Applied Numerical MathematicsMathematical Analysis and a Second-Order Compact Scheme for Nonlinear Caputo–Hadamard Fractional Sub-diffusion Equations
2024, Mediterranean Journal of MathematicsA new nonlinear compact difference scheme for a fourth-order nonlinear Burgers type equation with a weakly singular kernel
2024, Journal of Applied Mathematics and Computing
- ☆
This research was partly supported by the National Natural Science Foundation of China (No. 11701103), Young Top-notch Talent Program of Guangdong Province (No. 2017GC010379), Natural Science Foundation of Guangdong Province (Nos. 2022A1515012147, 2019A1515010876), the Project of Science and Technology of Guangzhou (No. 201904010341, 202102020704), and the Opening Project of Guangdong Province Key Laboratory of Computational Science at the Sun Yat-sen University (2021023), the Science and Technology Development Fund, Macau SAR (File No. 0005/2019/A) and University of Macau (File Nos. MYRG2020-00035-FST, MYRG2018-00047-FST).