Fuzhang Wang , Kehong Zheng , Imtiaz Ahmad and Hijaz Ahmad

Gaussian radial basis functions method for linear and nonlinear convection–diffusion models in physical phenomena

De Gruyter | Published online: March 17, 2021

Abstract

In this study, we propose a simple direct meshless scheme based on the Gaussian radial basis function for the one-dimensional linear and nonlinear convection–diffusion problems, which frequently occur in physical phenomena. This is fulfilled by constructing a simple ‘anisotropic’ space–time Gaussian radial basis function. According to the proposed scheme, there is no need to remove time-dependent variables during the whole solution process, which leads it to a really meshless method. The suggested meshless method is implemented to the challenging convection–diffusion problems in a direct way with ease. Numerical results show that the proposed meshless method is simple, accurate, stable, easy-to-program and efficient for both linear and nonlinear convection–diffusion equation with different values of Péclet number. To assess the accuracy absolute error, average absolute error and root-mean-square error are used.

1 Introduction

The convection–diffusion, advection–diffusion, or drift-diffusion equations have been playing a significant role in many engineering applications. The energy can be transformed inside a physical system due to the convection and diffusion processes to describe physical phenomena.

For a large variety of problems in every subjects, it is almost impossible to get the analytical solutions for changing in time and transport processes [23,24]. Numerical approximations are alternative to the analytical solutions for convection–diffusion equations. In literatures, there are several numerical techniques for solving the convection–diffusion equations [33,41]. These numerical techniques are based on the finite-difference approximations [25,32], integral transform methods [28,29], Monte Carlo simulation [4,5], variational iteration methods [6,7, 8,9,10], and meshless method using radial basis functions [11, 12,13,14, 15,16,17, 18,19,20].

The behavior of high-order time-stepping methods combined with mesh-free methods is studied for the transient convection–diffusion equation [31]. A Petrov–Galerkin method and Green’s functions are used to solve convection–diffusion problems [21]. A new approach to construct a stable RKPM method for convection-dominated problems is presented in ref. [30]. The space–time least-squares finite element methods are constructed for the advection–diffusion equation by using both linear shape functions and quadratic B-spline shape functions [22]. The particle transport method is developed for solving linear convection problems [44]. Several literatures focus on the investigation of different splines as interpolating function for solving one-dimensional advection–diffusion equations [34,37]. These literatures focus on the one-dimensional convection–diffusion–reaction equations with constant diffusion coefficient. Recently, a high-accuracy adaptive difference strategy is investigated by Zhu and Rui [47] on 1D convection–diffusion–reaction equation with convection item. It can explain the quenching phenomena of nonlinear singular degenerate problems. Pourgholi et al. [43] proposed a meshless method using radial basis functions method based on the finite-difference method to solve a nonlinear inverse convection–reaction–diffusion problem with an unknown source function. These numerical techniques are based on two-level finite difference approximations.

It is well-known that the radial basis function methods are attractive in numerical simulation due to their simple, flexible, and truly meshfree features. In this study, we propose a direct meshless method with one-level approximation, based on the radial basis functions, for the one-dimensional linear and nonlinear convection–diffusion problems. This is fulfilled by considering the time variable as a normal space variable. There is no need to remove the time-dependent variable during the whole solution process. Under this scheme, we can solve the convection–diffusion problems in a direct way.

The structure of this paper is organized as follows. Followed by Section 2, we introduce the formulation of the direct radial basis function (DMM) with space–time distance. Section 3 presents the methodology for convection–diffusion problems under initial conditions and boundary conditions. Section 4 examines several linear with different Péclet numbers and nonlinear problems. Several numerical examples are presented to validate the accuracy and stability of the proposed algorithm for one-dimensional linear and nonlinear convection–diffusion problems. Some conclusions are given in Section 5 with some additional remarks.

2 Formulation of the direct radial basis function

To describe the interaction between convection effects and diffusion transports, we can get the general mathematical formulation of convection–diffusion–reaction problem

(1) C ( x , t ) t + α C ( x , t ) x = β 2 C ( x , t ) x 2 + f ( x , t ) , 0 < x < L , t > 0
where C is the unknown temperature or scalar quantity, α is the convection velocity, β is the diffusion, and f ( x , t ) is the prescribed source term. For linear convection–diffusion equations, α and β are positive numbers. For nonlinear convection–diffusion equations, one can change β into a nonlinear convection term β = β ( C ) with a nonlinear source term f ( C ; x , t ) . To seek for the solutions of convection–diffusion–reaction problem equation ( 1), one should consider the additional initial condition
(2) C ( x , 0 ) = g 1 ( x ) , 0 x L
and boundary conditions
(3) C ( x , t ) = g 2 ( x , t ) , x = { 0 , L } , t > 0 .

For traditional numerical techniques, equation (1) should be discretized using the finite difference method or integral transform method, which leads to a steady-state equation. Then, the other numerical techniques can be used to get the numerical solutions. This provides a two-level procedure. To obtain a one-level procedure, we propose a direct collocation scheme by using the Gaussian radial basis function (GRBF).

For direct RBF-based collocation methods, the approximate solution can be written as a linear combination of RBFs for the approximation space under consideration for 2D or more higher-dimensional problems. We take the following GRBF for 2D problems as an example ref. [27]

(4) ϕ ( r j ) = e ( ε r j ) 2 ,
where r j = X X j = ( x x j ) 2 + ( y y j ) 2 is the Euclidean distance between two points X = ( x , y ) and X j = ( x j , y j ) , and ε is the RBF shape parameter.

However, there is only one space variable x for the one-dimensional convection–diffusion problems, the traditional RBF can not be used directly. Here, we construct a simple space–time radial basis function by combining the space variable x and time variable t as a point ( x , t ) for physical domains [ 0 , L ] × T . More specifically, the interval [ 0 , L ] is uniformly divided into segments firstly 0 = x 0 < x 1 < < x n = L with the corresponding finess h = L / n . For non-uniform cases, the corresponding finess is h = max { Δ x j } = max { x j x j 1 } j = 1 n . The time variable is evenly chosen from the initial time t 0 = 0 to a final time t n = T as 0 = t 0 < t 1 < < t n = T with time-step Δ t = T / n . The corresponding configuration of the space–time coordinate is shown in Figure 1, where stands for the value of space variable x , stands for the value of time-variable t , and stands for the point ( x , t ) . Then, we have the simple anisotropic space–time radial basis function

(5) φ ( r j ) = e ε 2 ( x x j ) 2 + ( t t j ) 2 .
This can be easily extended to two-dimensional or high-dimensional cases.

Figure 1 
               Configuration of the space–time coordinate.

Figure 1

Configuration of the space–time coordinate.

In the literature, there is a product model of a space–time radial basis function [42], which was introduced by Myers et al. [39,40],

(6) φ ( r j ) = 1 e ε 2 ( x x j ) 2 1 e ε 2 ( t t j ) 2 .

The other types of definitions of radial or non radial space–time radial basis functions can be found in ref. [35,36].

3 Methodology for DMM

Based on the definition of space–time radial basis functions, the above-mentioned equations (1)–(3) can be solved directly in a one-level approximation. Thus, the approximate solution of the function C ( x , t ) has the form

(7) C N ( ) j = 1 N λ j φ j ( ) .
To seek for the unknown coefficients λ j , we can collocate the convection–diffusion equation ( 1) at N I internal points, while the initial condition equation ( 2) and boundary condition equation ( 3) are collocated at N 1 initial collocation points and N 2 boundary points, respectively. This procedure yields the following equations:
(8) j = 1 N λ j L φ j ( P i , P j ) = f ( P i , P j ) , i = 1 , , N I ,
(9) j = 1 N λ j φ j ( P i , P j ) = g 1 ( P i , P j ) , i = N I + 1 , , N I + N 1 ,
(10) j = 1 N λ j φ j ( P i , P j ) = g 2 ( P i , P j ) , i = N I + N 1 + 1 , , N ,
with
(11) L φ j = φ j t + α φ j x β 2 φ j x 2 ,
N 1 is the total number of x j and N 2 is the total time step number. Equations ( 8)–( 10) have the matrix form as
(12) A λ = b ,
where
(13) A = A 11 A 12 A 13 A 21 A 22 A 23 A 31 A 32 A 33
is N × N known matrix with submatrices

A 11 with elements L φ j ( P i , P j ) , i , j = 1 , 2 , , N I ,

A 12 with elements L φ j ( P i , P j ) , i = 1 , 2 , , N I , j = N I + 1 , , N I + N 1 ,

A 13 with elements L φ j ( P i , P j ) , i = 1 , 2 , , N I , j = N I + N 1 + 1 , , N ,

A 21 with elements φ j ( P i , P j ) , i = N I + 1 , , N I + N 1 , j = 1 , 2 , , N I ,

A 22 with elements φ j ( P i , P j ) , i = N I + 1 , , N I + N 1 , j = N I + 1 , , N I + N 1 ,

A 23 with elements φ j ( P i , P j ) , i = N I + 1 , , N I + N 1 , j = N I + N 1 + 1 , , N ,

A 31 with elements φ j ( P i , P j ) t , i = N I + N 1 + 1 , , N , j = 1 , 2 , , N I ,

A 32 with elements φ j ( P i , P j ) t , i = N I + N 1 + 1 , N , j = N I + 1 , , N ,

A 33 with elements φ j ( P i , P j ) t , i = N I + N 1 + 1 , , N , j = N I + N 1 + 1 , , N , and

(14) b = f g 1 g 2
is N × 1 vectors. This can be solved by the backslash computation in MATLAB codes.

4 Numerical experiments

To compare with the previous literatures, we consider using the absolute error, root-mean-square error (RMSE) [45,46], and average absolute error (AAE) defined as follows:

(15) RMSE = 1 N t j = 1 N t C ( x j ) C N ( x j ) 2 ,
(16) AAE = 1 N t j = 1 N t C ( x j ) C N ( x j ) ,
where C ( j ) is the analytical solution at test points x j , j = 1 , 2 , , N t , and C N ( j ) is the numerical solutions at the test points x j , j = 1 , 2 , , N t . N t is the number of test points on the physical domain. Since the DMM method is stable, the choice of RBF parameter ε = 1 is fixed for all the following cases. The optimal choice of RBF parameter is beyond the scope of our current research. For more details about this topic, we refer readers to ref. [ 3, 26] and references therein.

In the first two examples, we consider two linear cases with governing equation

(17) C t + α C x β 2 C x 2 = 0 , 0 < x < L , t > 0 .

4.1 Linear example 1

This example considers the following initial condition

(18) C ( x , 0 ) = a e c x ,
and boundary conditions
(19) C ( 0 , t ) = a e b t , C ( 1 , t ) = a e b t + c , t > 0 .

The corresponding analytical solution is given by

(20) C ( x , t ) = a e b t + c x ,
and
(21) c = α ± α 2 + 4 β b 2 β .

For fair comparison, the dimensionless Péclet number is defined as Pe = β α h [1,2]. The coefficients in the governing equation are L = 1 , α = 0.1 , β = 0.1 , a = 0.1 , b = 1.25 , and the corresponding low Péclet number is Pe = 1.0 . For h = Δ t = 1 / 15 , we compare the DMM results with compactly supported radial basis functions, thin-plate spline radial basis function, and the finite difference method [1,2], detailed results are listed in Table 1. For parameters α = 1 and β = 0.01 with the other parameters fixed as earlier, the corresponding high Péclet number is Pe = 100 . Numerical results of the DMM are listed in Table 2. It should be pointed that our time step h = Δ t = 1 / 15 , which leads to less computations, is larger than the one d t = 0.01 in ref. [1]. However, the DMM results are more accurate than all results (TPS, FD, CS) presented in ref. [1].

Table 1

Numerical results with Pe = 1

t x = 0.1 x = 0.3 x = 0.5 x = 0.7 x = 0.9 AAE
0.1 TPS 0.0834 0.0451 0.0244 0.0132 0.0069 0.0000
FD 0.0836 0.0454 0.0244 0.0133 0.0073 0.0001
CS 0.0848 0.0452 0.0244 0.0132 0.0077 0.0003
DMM 0.0833 0.0451 0.0244 0.0132 0.0072 7.95 × 1 0 6
0.5 TPS 0.1376 0.0744 0.0402 0.0216 0.0113 0.0001
FD 0.1384 0.0757 0.0413 0.0225 0.0123 0.0008
CS 0.1407 0.0757 0.0407 0.0223 0.0131 0.0010
DMM 0.1374 0.0744 0.0402 0.0218 0.0118 5.57 × 1 0 6
1 TPS 0.2570 0.1390 0.0751 0.0404 0.0211 0.0002
FD 0.2587 0.1420 0.0778 0.0425 0.0232 0.0018
CS 0.2631 0.1420 0.0768 0.0421 0.0246 0.0023
DMM 0.2568 0.1389 0.0752 0.0407 0.0220 8.35 × 1 0 6
Table 2

Numerical results with Pe = 100

t x = 0.1 x = 0.3 x = 0.5 x = 0.7 x = 0.9 AAE
0.1 TPS 0.1001 0.0782 0.0611 0.0478 0.0373 0.0000
FD 0.1006 0.0788 0.0616 0.0481 0.0469 0.0013
CS 0.1005 0.0782 0.0611 0.0477 0.0372 0.0001
DMM 0.1000 0.0779 0.0607 0.0473 0.0368 3.46 × 1 0 7
0.5 TPS 0.1650 0.1289 0.1007 0.0787 0.0615 0.0001
FD 0.1663 0.1316 0.1038 0.0817 0.0858 0.0042
CS 0.1661 0.1297 0.1011 0.0787 0.0613 0.0004
DMM 0.1649 0.1285 0.1001 0.0780 0.0607 4.04 × 1 0 7
1 TPS 0.3083 0.2409 0.1882 0.1470 0.1149 0.0001
FD 0.3106 0.2460 0.1948 0.1545 0.1626 0.0087
CS 0.3102 0.2424 0.1893 0.1478 0.1152 0.0009
DMM 0.3081 0.2100 0.1870 0.1457 0.1135 5.33 × 1 0 7

As is known to all, when the Péclet number is low, the diffusion term dominates. In order to compare the DMM with the cubic C 1 -spline collocation method (CFD) in ref. [38], the corresponding dimensionless Péclet number is defined as Pe = β α . Numerical results of the DMM in terms of average errors are listed in Table 3. From which we can find that the CFD method performs not well for relatively large Péclet numbers even for dense discretization h = 1 / 128 . However, the DMM results are stable and accurate than the CFD [1]. This reveals that the DMM is stable for high Péclet numbers.

Table 3

Average errors for DMM with different Péclet numbers

Pe Pe = 10 Pe = 20 Pe = 100 Pe = 1,000
FD 1/16 2.5 × 1 0 5 2.0 × 1 0 4 1.3 × 1 0 2 2.9 × 1 0 1
1/32 1.5 × 1 0 6 1.2 × 1 0 5 1.2 × 1 0 3 7.6 × 1 0 2
1/64 9.4 × 1 0 8 7.5 × 1 0 7 8.8 × 1 0 5 1.5 × 1 0 2
1/128 5.8 × 1 0 9 4.7 × 1 0 8 5.7 × 1 0 6 2.4 × 1 0 3
DMM 1/16 6.5 × 1 0 8 1.3 × 1 0 6 8.9 × 1 0 8 2.3 × 1 0 7

4.2 Linear example 2

Once the Péclet number increases, i.e., the convection term completely dominates over the diffusion term. In order to investigate problems with the far higher Péclet numbers, we compare the DMM results with the compact finite-difference approximation of fourth-order and the cubic C 1 -spline collocation method [38].

We consider the convection–diffusion equation (1) with the following initial condition:

(22) C ( x , 0 ) = sin x ,
and boundary conditions
(23) C ( 0 , t ) = e a t sin ( b t ) , C ( 1 , t ) = e a t sin ( 1 b t ) , t > 0 .

The analytical solution is given by

(24) C ( x , t ) = e a t sin ( x b t ) .
For a fair comparison, we use the definition of dimensionless Péclet number Pe = β α in ref. [ 38]. We choose L = 2 , α = 1 , b = 1 . Numerical results of the DMM are listed in Table 4 with various Péclet numbers for L = 2 at final time T = 1 and h = 1 / 20 . We note that our time step Δ t = 1 / 20 , which leads to less computations, is larger than the one 0.001 in ref. [ 38]. For different values of x < 1.75 , we can find that the DMM results are more accurate than the FD results presented in ref. [ 38] for Péclet number Pe = 1,000 . For x = 1.75 , the results are similar. For Péclet number Pe = 10,000 and Pe = 20,000 , the DMM results are more accurate than the FD results presented in ref. [ 38]. This reveals that the DMM is more stable than the FD method for various Péclet numbers.

Table 4

Numerical results with different Péclet numbers Pe = β α

Pe DMM FD DMM FD DMM FD
x 1,000 1,000 10,000 10,000 20,000 20,000
0.25 9.3 × 1 0 8 1.4 × 1 0 6 2.5 × 1 0 7 1.1 × 1 0 6 1.8 × 1 0 7 1.0 × 1 0 4
0.50 1.7 × 1 0 8 1.8 × 1 0 6 1.9 × 1 0 7 4.9 × 1 0 7 1.4 × 1 0 7 2.2 × 1 0 4
0.75 2.3 × 1 0 8 7.9 × 1 0 6 6.2 × 1 0 7 6.9 × 1 0 6 1.7 × 1 0 7 3.5 × 1 0 4
1.00 6.7 × 1 0 8 9.8 × 1 0 7 0 2.6 × 1 0 5 3.6 × 1 0 7 4.7 × 1 0 4
1.25 7.0 × 1 0 10 1.1 × 1 0 7 3.9 × 1 0 7 7.0 × 1 0 5 3.5 × 1 0 7 6.5 × 1 0 4
1.50 9.6 × 1 0 9 2.1 × 1 0 7 1.5 × 1 0 7 1.6 × 1 0 4 1.2 × 1 0 7 8.0 × 1 0 4
1.75 2.1 × 1 0 7 2.9 × 1 0 7 3.1 × 1 0 7 2.9 × 1 0 4 2.5 × 1 0 8 7.4 × 1 0 4

4.3 Nonlinear example 3

In this example, we consider a typical nonlinear convection–diffusion equation

(25) C t = a C x x x C x + F ( x , t ) , 1 < x < 1 , t > 0 ,
with the exact solution
(26) C ( x , t ) = tanh x e t 4 a ,
to measure the performance of the DMM by comparing with the method in ref. [ 47]. The source term F ( x , t ) can be derived from the exact solution. The following initial condition and boundary conditions are considered:
(27) C ( x , 0 ) = tanh x 4 a , C ( 1 , t ) = tanh e t 4 a , C ( 1 , t ) = tanh e t 4 a .

We note that for smaller values of a , the exact solution varies greatly at x = 0 . For fair comparison with the method in ref. [47], we consider T = 1 and a = 0.1 , for our DMM method, respectively. Detailed results are listed in Table 5.

Table 5

Numerical results for a = 0.1

h DMM Uniform Non-uniform
8 4.42 × 1 0 5
16 2.79 × 1 0 5
32 2.30 × 1 0 6
64 5.15 × 1 0 5 1.01 × 1 0 2 6.65 × 1 0 3
128 1.02 × 1 0 2 6.73 × 1 0 3
256 1.02 × 1 0 2 6.78 × 1 0 3

For a = 0.1 , the non-uniform scheme in ref. [47] performances better than the uniform scheme with dense mesh k = 64 or more. Obviously, the DMM method performs better than these two schemes with sparse mesh k = 32 or less. It can be seen that the computation accuracy of our DMM is superior to those of the others from the above-mentioned results.

5 Conclusions

In this study, a new direct meshless scheme is proposed for the one-dimensional linear and nonlinear convection–diffusion problems. The present numerical procedure, in which the time variable is considered as normal space variable, is based on the Gaussian radial basis function. There is no need to remove time-dependent variable during the whole solution process. Numerical results for several typical examples show that the proposed method is better than some other numerical methods given in the recent literature in terms of solution accuracy, stability and efficiency for the linear convection–diffusion equation with different values of Péclet number. These results lead us that the proposed method can successfully be used to nonlinear problems with accurate numerical results.

Acknowledgments

This work was supported by the Natural Science Foundation of Anhui Province (Project No. 1908085QA09) and the University Natural Science Research Project of Anhui Province (Project No. KJ2019A0591 & KJ2020B06).

    Conflict of interest: The authors declare that there is no conflict of interests regarding the publication of this paper.

    Data availability statement: The data used to support the findings of this study are available from the corresponding author upon request.

References

[1] Boztosun I , Charafi A , Boztosun D . On the numerical solution of linear advection–diffusion equation using compactly supported radial basis functions. In: Griebel M , Schweitzer MA (eds) Meshfree methods for partial differential equations. Lecture notesin computational science and engineering, vol 26. Berlin, Heidelberg: Springer; 2003. Search in Google Scholar

[2] Boztosun I , Charafi A , Zerroukat M , Djidjeli K . Thin-plate spline radial basis function scheme for advection–diffusion problems. Electron J Bound Elements. 2002;BETEQ 2001:267–82. Search in Google Scholar

[3] Chen W , Hong YX , Lin J . The sample solution approach for determination of the optimal shape parameter in the Multiquadric function of the Kansa method. Comput Math Appl. 2018;75:2942–54. Search in Google Scholar

[4] Fulger D , Scalas E , Germano G . Monte Carlo simulation of uncoupled continuous-time random walks yielding a stochastic solution of the space–time fractional diffusion equation. Phys Rev E. 2008;77(2):021122. Search in Google Scholar

[5] Koley U , Risebro NH , Schwab C , Weber F . A multilevel Monte Carlo finite difference method for random scalar degenerate convection–diffusion equations. J Hyperbolic Differ Equ. 2017;14(3):415–54. Search in Google Scholar

[6] Ahmad H , Khan TA , Ahmad I , Stanimirović PS , Chu Y-M . A new analyzing technique for nonlinear time fractional Cauchy reaction-diffusion model equations. Results Phys. 2020;103462. 10.1016/j.rinp.2020.103462. Search in Google Scholar

[7] Ahmad H , Akgül A , Khan TA , Stanimirović PS , Chu Y-M . New perspective on the conventional solutions of the nonlinear time-fractional partial differential equations. Complexity. 2020;2020:8829017. 10.1155/2020/8829017. Search in Google Scholar

[8] Ahmad H , Khan TA , Stanimirović PS , Chu Y-M , Ahmad I . Modified variational iteration algorithm-II: convergence and applications to diffusion models. Complexity. 2020;2020:8841718. 10.1155/2020/8841718. Search in Google Scholar

[9] Ahmad H , Khan TA , Stanimirovic PS , Ahmad I Modified variational iteration technique for the numerical solution of fifth order KdV type equations. J Appl Comput Mech. 2020;6(SI):1220–7. Search in Google Scholar

[10] Ahmad H , Seadawy AR , Khan TA , Thounthong P . Analytic approximate solutions for some nonlinear Parabolic dynamical wave equations. J Taibah Univ Sci. 2020;14(1):346–58. Search in Google Scholar

[11] Ahmad I , Ahmad H , Inc M , Yao S-W , Almohsen B . Application of local meshless method for the solution of two term time fractional-order multi-dimensional PDE arising in heat and mass transfer. Thermal Sci. 2020;24:95–105. 10.2298/TSCI20S1095A. Search in Google Scholar

[12] Shakeel M , Hussain I , Ahmad H , Ahmad I , Thounthong P , Zhang Y-F . Meshless technique for the solution of time-fractional partial differential equations having real-world applications. J Funct Spaces. 2020;8898309. 10.1155/2020/8898309. Search in Google Scholar

[13] Li JF , Ahmad I , Ahmad H , Shah D , Chu YM , Thounthong P , Ayaz M . Numerical solution of two-term time-fractional PDE models arising in mathematical physics using local meshless method. Open Phys. 2020;18(1):1063–72. Search in Google Scholar

[14] Khan MN , Ahmad I , Akgül A , Ahmad H , Thounthong P . Numerical solution of time-fractional coupled Korteweg-de Vries and Klein-Gordon equations by local meshless method. Pramana-J Phys. 2021;95(1):1–13. Search in Google Scholar

[15] Ahmad I , Siraj-ul-Islam M , Zaman S . Local meshless differential quadrature collocation method for time-fractional PDEs. Discrete & Continuous Dynamical Systems-S. 2020;13(10):2641. Search in Google Scholar

[16] Inc M , Khan MN , Ahmad I , Yao SW , Ahmad H , Thounthong P . Analysing time-fractional exotic options via efficient local meshless method. Results Phys. 2020;19:103385. 10.1016/j.rinp.2020.103385. Search in Google Scholar

[17] Ahmad I , Ahmad H , Thounthong P , Chu YM , Cesarano C . Solution of multi-term time-fractional PDE models arising in mathematical biology and physics by local meshless method. Symmetry. 2020;12(7):1195. Search in Google Scholar

[18] Ahmad I , Khan MN , Inc M , Ahmad H , Nisar KS . Numerical simulation of simulate an anomalous solute transport model via local meshless method. Alex Eng J. 2020;59(4):2827–38 Search in Google Scholar

[19] Khan MN , Ahmad I , Ahmad H . A radial basis function collocation method for space-dependent inverse heat problems. J Appl Comput Mech. 2020;6 (SI):1187–99. Search in Google Scholar

[20] Srivastava MH , Ahmad H , Ahmad I , Thounthong P , Khan NM . Numerical simulation of three-dimensional fractional-order convection–diffusion PDEs by a local meshless method. Therm Sci. 2020;210. 10.2298/TSCI200225210S Search in Google Scholar

[21] Cyron CJ , Nissen K , Gravemeier V , Wolfgang AW . Stable meshfree methods in fluid mechanics based on Greenas functions. Comput Mech. 2010;46:287–300. Search in Google Scholar

[22] Dag I , Irk D , Tombul M . Least-squares finite element method for the advection–diffusion equation. Appl Math Comput. 2006;173:554–65. Search in Google Scholar

[23] Wang FZ , Hou ER . A direct meshless method for solving two-dimensional second-order hyperbolic telegraph equations. J Math. 2020;2020:8832197. Search in Google Scholar

[24] Dehghan M . Numerical solution of the three-dimensional advection–diffusion equation. Appl Math Comput. 2004;150:5–19. Search in Google Scholar

[25] Ding HF , Zhang YX . A new difference scheme with high accuracy and absolute stability for solving convection–diffusion equations. J Comput Appl Math. 2009;230:600–6. Search in Google Scholar

[26] Fasshauer GE , Zhang JG . On choosing optimal shape parameters for RBF approximation. Numer Algorithms. 2007;45:345–68. Search in Google Scholar

[27] Fornberg B , Larsson E , Flyer N . Stable computations with Gaussian radial basis functions. SIAM J Sci Comput. 2011;33:869–92. Search in Google Scholar

[28] Glushkov EV , Glushkova NV , Chen CS . Semi-analytical solution to heat transfer problems using fourier transform technique, radial basis functions, and the method of fundamental solutions. Numer Heat Tr B. 2007;52:409–27. Search in Google Scholar

[29] Gu Y , He X , Chen W , Zhang C . Analysis of three-dimensional anisotropic heat conduction problems on thin domains using an advanced boundary element method. Comput Math Appl 2018;75:33–44. Search in Google Scholar

[30] Hillman M , Chen JS . An implicit gradient meshfree formulation for convection-dominated problems. In: Bazilevs Y , Takizawa K (eds), Advances in computational fluid-structure interaction and flow simulation. Birkhäuser, Cham: Modeling and Simulation in Science, Engineering and Technology; 2016. Search in Google Scholar

[31] Huerta A , Fernandez-Mendez S . Time accurate consistently stabilized mesh-free methods for convection dominated problems. Int J Numer Methods Eng. 2001;50:1–18. Search in Google Scholar

[32] Karahan H . A third-order upwind scheme for the advection diffusion equation using spreadsheets. Adv Eng Softw. 2007;38:688–97. Search in Google Scholar

[33] Karahan H . Unconditional stable explicit finite difference technique for the advection–diffusion equation using spreadsheets. Adv Eng Softw. 2007;8:80–6. Search in Google Scholar

[34] Korkmaz A , Dag I . Cubic B-spline differential quadrature methods for the advection–diffusion equation. Int J Numer Methods H. 2012;22:1021–36. Search in Google Scholar

[35] Ku CY , Liu CY , Xiao JE , Chen MR . Solving backward heat conduction problems using a novel space–time radial polynomial basis function collocation method. Appl Sci. 2020;10:3215. Search in Google Scholar

[36] Liu CY , Ku CT , Xiao JE , Yeih WC . A novel spacetime collocation meshless method for solving two-dimensional backward heat conduction problems. CMES-Comp Model Eng. 2019;118:229–52. Search in Google Scholar

[37] Mohammadi R . Exponential B-spline solution of convection–diffusion equations. Appl Math. 2013;4:933–44. Search in Google Scholar

[38] Mohebbi A , Dehghan M . High-order compact solution of the one-dimensional heat and advection–diffusion equations. Appl Math Model. 2010;34:3071–84. Search in Google Scholar

[39] Myers DE . Anisotropic radial basis functions. Int J Pure Appl Math. 2008;42:197–203. Search in Google Scholar

[40] Myers DE , Iaco SD , Posa D , Cesare LD . Space-time radial basis functions. Comput Math Appl. 2002;43:539–49. Search in Google Scholar

[41] Nazir T , Abbas M , Ismail AIM , Majid AA , Rashid A . The numerical solution of advection–diffusion problems using new cubic trigonometric B-splines approach. Appl Math Model. 2016;40:4586–611. Search in Google Scholar

[42] Parand K , Rad JA . Kansa method for the solution of a parabolic equation with an unknown space wise-dependent coefficient subject to an extra measurement. Comput Phys Commun. 2013;184:582–95. Search in Google Scholar

[43] Pourgholi R , Saeedi A , Hosseini A . Determination of nonlinear source term in an inverse convection–reaction–diffusion problem using radial basis functions method. Iran J Sci Technol A. 2019;43:2239–52. Search in Google Scholar

[44] Smolianski A , Shipilova O and Haario H . A fast high-resolution algorithm for linear convection problems: particle transport method. Int J Numer Methods Eng. 2007;70:655–84. Search in Google Scholar

[45] Wang FZ , Chen W , Jiang XR . Investigation of regularization techniques for boundary knot method. Commun Numer Meth En. 2010;26:1868–77. Search in Google Scholar

[46] Wang FZ , Chen W , Ling L . Combinations of the method of fundamental solutions for general inverse source identification problems. Appl Math Comput. 2012;219:1173–82. Search in Google Scholar

[47] Zhu XL , Rui HX . High-order compact difference scheme of 1D nonlinear degenerate convection–reaction–diffusion equation with adaptive algorithm. Numer Heat Tr B-Fund. 2019;75:43–66. Search in Google Scholar

Received: 2020-10-04
Revised: 2021-01-29
Accepted: 2021-02-11
Published Online: 2021-03-17

© 2021 Fuzhang Wang et al., published by De Gruyter

This work is licensed under the Creative Commons Attribution 4.0 International License.