Abstract

In this paper, we present an efficient method for the numerical investigation of three-dimensional non-integer-order convection-diffusion equation (CDE) based on radial basis functions (RBFs) in localized form and Laplace transform (LT). In our numerical scheme, first we transform the given problem into Laplace space using Laplace transform. Then, the local radial basis function (LRBF) method is employed to approximate the solution of the transformed problem. Finally, we represent the solution as an integral along a smooth curve in the complex left half plane. The integral is then evaluated to high accuracy by a quadrature rule. The Laplace transform is used to avoid the classical time marching procedure. The radial basis functions are important tools for scattered data interpolation and for solving partial differential equations (PDEs) of integer and non-integer order. The LRBF and global radial basis function (GRBF) are used to produce sparse collocation matrices which resolve the issue of the sensitivity of shape parameter and ill conditioning of system matrices and reduce the computational cost. The application of Laplace transformation often leads to the solution in complex plane which cannot be generally inverted. In this work, improved Talbot’s method is utilized which is an efficient method for the numerical inversion of Laplace transform. The stability and convergence of the method are discussed. Two test problems are considered to validate the numerical scheme. The numerical results highlight the efficiency and accuracy of the proposed method.

Highlights

  • Fractional calculus is a fast growing field and has attracted the research community due to its applications in engineering and other sciences [1,2,3]

  • Gao and Sun [38] obtained the numerical approximation of fractional sub-diffusion equation via finite difference scheme

  • Wei et al [30] obtained the numerical solution of 2D fractional-order diffusion equation based on the implicit RBF method

Read more

Summary

Introduction

Fractional calculus is a fast growing field and has attracted the research community due to its applications in engineering and other sciences [1,2,3]. Zhai et al [34] proposed a compact finite difference scheme for the numerical simulation of fractional-order 3D CDE. In [36], the numerical scheme for the approximation of 2D sub-diffusion equation of fractional order was proposed. Gao and Sun [38] obtained the numerical approximation of fractional sub-diffusion equation via finite difference scheme. E authors in [39] derived the error estimates of the numerical solution for reaction-sub-diffusion of fractional order via the meshless method. Wei et al [30] obtained the numerical solution of 2D fractional-order diffusion equation based on the implicit RBF method. Approximation of diffusion equation of variable fractional order with different boundary conditions based on local RBF method was carried out in [41].

Time Discretization via Laplace Transform
Space Discretization via LRBF Method
Stability
Numerical Inversion of Laplace Transform
Numerical Experiments
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call