A Meshless Method for Burgers’ Equation Using
Multiquadric Radial Basis Functions With
a Lie-Group Integrator
Department of Mathematics, Faculty of Art and Science, Mu¸s Alparslan University, 49100 Mu¸s, Turkey;
Received: 27 November 2018; Accepted: 21 January 2019; Published: 22 January 2019
An efﬁcient technique is proposed to solve the one-dimensional Burgers’ equation based on
multiquadric radial basis function (MQ-RBF) for space approximation and a Lie-Group scheme for
time integration. The comparisons of the numerical results obtained for different values of kinematic
viscosity are made with the exact solutions and the reported results to demonstrate the efﬁciency and
accuracy of the algorithm. It is shown that the numerical solutions concur with existing results and
the proposed algorithm is efﬁcient and can be easily implemented.
Burgers’ equation; meshless method; multiquadric radial basis function (MQ-RBF);
We consider the nonlinear Burgers’ equation in the one-dimensional case
∂xF(u(x,t) = v∂2u(x,t)
with initial condition
and boundary conditions
u(a,t)=h1(t),u(b,t)=h2(t), 0 ≤t(2)
0 is interpreted as kinematic viscosity number and
are known sufﬁciently smooth functions of the time and space variables, respectively. The steady
solutions of the Burgers’ equation, for the ﬁrst time, are presented by Bateman [
]. Burger [
this equation as a model of turbulence. The Burgers equation is considered to be a simpliﬁed model of
the Navier-Stokes equation because they share common nonlinear and viscosity terms. Additionally,
it may be helpful to understand the physical mechanisms of the wave motion from advection and
dissipation terms. The analytical solutions in the terms of inﬁnite series can be computed by using the
Hopf-Cole transformation [
] and this gives the opportunity for a comparison between numerical
algorithms. Throughout history, there have been many investigations of Burgers’ equation in a variety
of different areas including nonlinear wave, shock waves, trafﬁc ﬂows, gas dynamics, elasticity, etc.
] and references therein). For all these reasons, Burgers’ equation has been dealt with by many
researchers. The authors of [
] examined different exact solutions of the one-dimensional Burgers’
equation. Many researchers approximate the solutions of the Burgers’ equation by various numerical
Mathematics 2019,7, 113; doi:10.3390/math7020113 www.mdpi.com/journal/mathematics
Mathematics 2019,7, 113 2 of 11
schemes such as ﬁnite difference, ﬁnite element, exponentially ﬁtted, Haar wavelet, differential
A kind of univariate multiquadric (MQ) quasi-interpolation technique to approximate a solution
of the Burgers’ equation is tailored by Chen and Zu [
]. A meshless method of lines based on
radial basis functions (RBFs) is proposed for the numerical solution of Burgers’-type equations in [
Xie et al. 
presented the method of particular solutions based on RBFs and ﬁnite difference scheme
to solve one-dimensional time-dependent inhomogeneous Burgers’ equations. Two mesh-free methods,
in which the MQ quasi-interpolation method is applied in direct and indirect forms for the numerical
solution of the Burgers’ equation, are proposed in
. Fan et al. 
developed a mesh-free numerical
method based on a combination of the multiquadric RBFs (LRBFCM) and the ﬁctitious time integration
method (FTIM) to solve the two-dimensional Burgers’ equations. Bouhamidi et al. [
RBFs interpolation technique for spatial discretization and implicit Runge–Kutta (IRK) schemes for
temporal discretization of the unsteady coupled Burgers’-type equations. Xie et al. [
the solution of the Burgers’ equation spatially by the multiquadric MQ-RBF and used C-N ﬁnite
difference scheme as temporal discretization technique.
On the other hand, the goal of Geometric Numerical Integration (GNI) is to preserve geometric
structure of the systems when approximating their solutions. Thus, the discretized solution of the
differential equation has the same qualitative properties of the system [
]. Lie-group methods are
a class of geometric numerical integrators which preserve qualititative structure of the exact solution.
Thus, one can preserve the structural properties of the true ﬂows [
]. The authors of [
a group preserving scheme based on Cayley transformation for Burgers’ equation, in which ﬁnite
difference method is used for spatial discretization. Furthermore, a combination of the one-step
backward group preserving scheme (BGPS) and the numerical method of line to discretize the time
and spatial variables, respectively, is presented in  to solve Burgers’ equation numerically.
In the present work, the numerical solutions of the one-dimensional Burgers’ equation is analyzed
by the MQ-RBFs and a Lie-Group version of Euler method for space and temporal discretization
technique, respectively. The goal is to take advantage of both meshless method and GNI method for
numerical approximations obtained by this combination of the Lie-Group method based on radial
basis functions (LG-RBFs). This method has been recently applied to the Heat equation [
high-dimensional generalized Benjamin-Bona-Mahony-Burgers’ equation  .
2. Numerical Scheme
In this section, we introduce a scheme which consists of the MQ-RBFs and the first-order explicit
Lie-group version of Euler method to solve Burgers’ equation numerically.
2.1. The Multiquadric Radial Basis Approximation
We approximate Burgers’ equation spatially by the MQ-RBFs to obtain a nonlinear system of
ordinary differential equations. The approximation of the function u(x,t) can be written as
where the MQ-RBFs deﬁned as
ϕ(x) = ϕrj=qr2
is the Euclidean norm and the unknown functions
’s can be determined
by the collocation technique [
]. The number
appeared in Formula
interpreted as a shape
parameter and it plays an important role in the accuracy of the schemes. One can control the shape
of RBFs with the free parameter
. Thus, the selection of its appropriate value is very important to
minimize the error of approximation. Mostly authors use "trial and error" to describe an acceptable
Mathematics 2019,7, 113 3 of 11
value of shape parameter for highly accurate results. We also follow the same technique to compute
shape parameters for experiments. However, collocating
at the M points yields the M equations,
an extra two equations are required. This is assured by considering additional two conditions for
given as 
where λjn=λj(tn). Following , one can write
where the collocation points
on the interval [a, b] are called for 2
2 as center points and
as boundary points. If one inserts Equation
2 into Equation
then obtains following nonlinear system of differential equations
dt u(t)=f(t,u(t)), (6)
f(t,u(t)) = ν
2.2. Lie-Group Method
We have used Lie-group version of Euler method as a temporal approximation scheme to solve
the system of ordinary differential Equation (6) numerically. Using Equation (6) one has
Merging Equations (6) and (9) leads to a system of differential equations
is an augmented vector,
Mathematics 2019,7, 113 4 of 11
is an augmented matrix of state variables [
]. Notice that the ﬁrst equation in
is the same as
and the second one is known as Minkowskian structure for the augmented system.
The augmented vector U veriﬁes the cone condition
UTgU =u.u−kuk2=kuk2−kuk2=0, (12)
where Minkowski metric gis deﬁned as
identity matrix given with
and the dot between two
-dimensional vectors deﬁnes the
Euclidean inner product [
]. Furthermore, the matrix
deﬁned by (11) belongs to the Lie algebra of
the Lorentz group SO0(d, 1), i.e.,
Now we have extended
-dimensional system (10) of the
-dimensional system (6), which has
the additional property that the solutions should satisfy cone condition (12). Thus, it is very important
to preserve this property by the numerical integrators. All one must do is to check the properties
with G(tn)∈SO0(d, 1)
is the zeroth component of
]. A Lie-group version of ﬁrst-order Euler
method can be applied to solve Equation (10) as follows
where Un+1∈SO0(d, 1). On the other hand, it is easy to express exp(∆tA(tn)) in closed-form
where , un≈u(tn),∆t=tn+1−tn,fn=f(un,tn)and
Inserting this expression of exp(∆tA(tn)) for G(tn)into (13) gives
ζn=(an−1)fn.un+bnkunk k fnk
kun+1k=bn(fn.un) + ankunk k fnk
Mathematics 2019,7, 113 5 of 11
The Equations (14)–(16) preserve the cone condition for each advanced time. Furthermore,
the scheme (14) unconditionally preserves the ﬁxed point and the geometric property of the true ﬂows
of the original equation (see , Theorem 1,2).
Now we apply scheme (14) to solve Equation (6) given with Equations (7)–(8). Let
tn=n∆t,n=0, . . . , N
be discrete time points with steplength
. If one considers scheme (14) for
0 at initial
time t0=0 then has
M+2=u0(xi) + ζ0f0,i=2, . . . , M−1,
can be computed from Equation (15). For 1
i=2, . . . , M−1 one has
can be computed from Equation (15). One can write Equations (5), (17), (18) and boundary
conditions (2) in the matrix form as follows
ϕ(r11)ϕ(r12 ). . . ϕ(r1M)x11
ϕ(r21)ϕ(r22 ). . . ϕ(r2M)x21
ϕ(rM1)ϕ(rM2). . . ϕ(rMM )xM1
x1x2. . . xM0 0
1 1 . . . 1 0 0
2, . . . , λn+1
2, . . . , bn+1
M, 0, 0iT.
Each elements of vector Bcan be computed for i=2, . . . , M−1 as
i=u0(xi) + ζ0f0,
M+2+ζnfn,n=1, . . . , N−1,
, respectively. The coefﬁcients
are computed by
solving the resulting system of (19) and then the numerical solutions are obtained using them in (3).
can be shown to be invertible and it is often ill-conditioned. One can determine how
Mathematics 2019,7, 113 6 of 11
accurately to solve system of (19) by considering the conditioning of
. The condition number of this
system related to small perturbation in Dis deﬁned for any norm k.kas
κ(D) = kDk
It is possible to lose
digits through the computational process of the system (19) including
]. To solve ill-posed system (19) we have used Gaussian elimination with
partial pivoting. We obtained the solutions by using the standard ﬂoating-point arithmetic in our
computational algorithm. The shape parameter also plays an important role for the conditioning of the
. One can have high accuracy with smaller value of
for a ﬁxed number of points
this case one often faces with the ill-conditioned matrix
. On the other hand, the condition number
κ(D)increases with the number of collocation points Mfor ﬁxed the values of c.
One can investigate the numerical stability of the method of line by Rule of Thumb [
]. If the
eigenvalues of the discretized spatial operator, scaled by steplength ∆t, are inside the stability region
of Lie-Group version of Euler integrator than the proposed algorithm is convergent.
3. Numerical Results
We consider Burgers’ equation on the interval [0, 1]with the following initial condition
u(x, 0) = sin(πx),
and the following boundary conditions
u(0, t) = u(1, t) = 0, t>0.
Its analytical solution is given with the Hopf-Cole transformation as [3,4]
u(x,t) = 2νπ ∑∞
(i=1, 2, 3 . . .).
We compute the analytical solutions from (20) up to 500 series terms for all comparisons in
experiments. The shape parameters are computed by trial and error. We compare the errors for
different values of shape parameters which are leading the lower conditioned number of matrix
in (19). We select the shape parameter c for the lowest error values in each experiment. Accuracy of
these methods is examined by computing error in L2and L∞norm.
We compute the results of the LG-RBFs for
1 at different ﬁnal times
0.1, 0.15, 0.2, 0.25
0.299, 0.297, 0.299, 0.299 and at points
0.25, 0.5, 0.75. We compare the
numerical solutions of the LG-RBFs with GPS [
and with [
Mathematics 2019,7, 113 7 of 11
.The all results tabulated in Table 1demonstrate that the LG-RBFs method is highly accurate
compared to that of other methods. Now we compute the solutions with LG-RBFs for
at different times
0.4, 0.6, 0.8, 1, 3 with respectively
0.128, 0.312, 0.312, 0.312 and at points
0.25, 0.5, 0.75. Numerical solutions obtained by LG-RBFs with
, have been
tabulated and compared in Table 2with exact solution and the numerical solutions presented in [
. It is clear from table that LG-RBFs provides
accurate results than are obtained by [
]. Finally, we compute the solutions with LG-RBFs for
0.01 at different times
0.4, 0.6, 0.8, 1, 3 with respectively
0.121, 0.034, 0.036, 0.114, 0.111 and
0.25, 0.5, 0.75. The results compared in Table 3with exact solutions and the numerical
solutions presented in [
for all methods. Table 3indicate that the
accuracy of LG-RBFs solutions is compatible with the numerical solutions obtained in [
]. It is
easy to see that the all results tabulated in Tables for the LG-RBFs are in good agreement with the
Comparison of the approximate solution at different final times with
x t Galerkin  GPS  LG-RBFs Exact
0.25 0.10 0.25469 0.25376 0.25364 0.25364
0.15 0.15672 0.15672 0.15660 0.15660
0.20 0.09619 0.09654 0.09644 0.09644
0.25 0.05924 0.05929 0.05922 0.05922
0.50 0.10 0.37134 0.37177 0.37157 0.37158
0.15 0.22674 0.22700 0.22682 0.22682
0.20 0.13829 0.13862 0.13847 0.13847
0.25 0.08457 0.08464 0.08454 0.08454
0.75 0.10 0.27102 0.27273 0.27258 0.27258
0.15 0.16411 0.16450 0.16437 0.16437
0.20 0.09929 0.09954 0.09943 0.09944
0.25 0.06036 0.06042 0.06035 0.06035
Comparison of the approximate solutions at different final times with
x t Galerkin  GPS  LG-RBFs Exact
0.25 0.4 0.31429 0.30889 0.30864 0.30889
0.6 0.24373 0.24077 0.24061 0.24074
0.8 0.19758 0.19573 0.19559 0.19568
1.0 0.16391 0.16264 0.16251 0.16256
3.0 0.02743 0.02725 0.02720 0.02720
0.50 0.4 0.57636 0.56988 0.56944 0.56963
0.6 0.45169 0.44745 0.44707 0.44721
0.8 0.36245 0.35948 0.35914 0.35924
1.0 0.29437 0.29215 0.29184 0.29192
3.0 0.04057 0.04028 0.04020 0.04020
0.75 0.4 0.62952 0.62605 0.62564 0.62544
0.6 0.49034 0.48778 0.48727 0.48721
0.8 0.37713 0.37438 0.37389 0.37392
1.0 0.29016 0.28784 0.28741 0.28747
3.0 0.01334 0.02983 0.02977 0.02977
Mathematics 2019,7, 113 8 of 11
Comparison of the approximate solutions at different ﬁnal times with
x t Exact-Explicit  RK4  GPS  LG-RBFs Exact
0.25 0.4 0.34164 0.34197 0.34193 0.34189 0.34191
0.6 0.26890 0.26900 0.26897 0.26890 0.26896
0.8 0.22150 0.22151 0.22149 0.22140 0.22148
1.0 0.18825 0.18821 0.18820 0.18816 0.18819
3.0 0.07515 0.07512 0.07511 0.07512 0.07511
0.50 0.4 0.65606 0.66083 0.66079 0.66068 0.66071
0.6 0.52658 0.52950 0.52946 0.52938 0.52942
0.8 0.43743 0.43919 0.43916 0.43910 0.43914
1.0 0.37336 0.37446 0.37443 0.37439 0.37442
3.0 0.15015 0.15019 0.15018 0.15018 0.15018
0.75 0.4 0.90111 0.91053 0.91058 0.91031 0.91026
0.6 0.75862 0.76741 0.76739 0.76723 0.76724
0.8 0.64129 0.64750 0.64747 0.64737 0.64740
1.0 0.55187 0.55620 0.55609 0.55603 0.55605
3.0 0.22454 0.22484 0.22483 0.22481 0.22481
error norms of the LG-RBFs for
0.588, 0.587 and for
3, 3.5 with respectively
0.6, 0.611. We compare these with the error norms of the backward differentiation formula of order
one (BDF1) reported in [
]. The results are tabulated for the both ﬁrst-order schemes (in time) in the
Tables 4and 5
. The tables indicate that the accuracy of LG-RBFs is much better than BDF1 method [
It can be seen from Tables 4and 5that the error of LG-RBFs gets smaller than the error of BDF1
method when the ﬁnal time increases, as we expect. In Figure 1we show the numerical solutions of
the LG-RBFs which are computed for
0.1, 0.01 with
different ﬁnal times, and at
1. The presence of the steep descent can be seen from the
Figure 1for small values of viscosity. Thus, proposed scheme can exhibit the correct physical behavior
of the problem.
Table 4. Comparison of error norms for different values of ﬁnal times with M=10 and ∆t=0.001.
BDF-1, M=80  1.8457 ×10−62.6102 ×10−61.9600 ×10−10 2.7713 ×10−10
LG-RBFs 1.1727 ×10−10 1.5587 ×10−10 4.8120 ×10−14 6.7895×10−14
Table 5. Comparison of error norms for different values of ﬁnal times with M=10 and ∆t=0.01.
BDF-1  4.2130 ×10−45.9753 ×10−43.0024 ×10−44.2509 ×10−4
LG-RBFs 5.0581 ×10−57.1572 ×10−53.0291 ×10−54.3053 ×10−5
Mathematics 2019,7, 113 9 of 11
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Physical behavior of numerical solutions
40 (top left panel) and
80 (top right panel) and
80 (bottom panel) at different ﬁnal times for
The ﬁrst-order Lie-Group version of Euler method combined with MQ-RBFs method is proposed
to approximate solutions of the one-dimensional Burgers’ equation. The obtained numerical results
are compared with analytical and reported results in the literature for a test example at different
values of viscosity. It is concluded that the numerical solutions of the proposed scheme are effective in
comparison with the analytical and numerical solutions presented in the previous studies. The results
show that the presented scheme can capture the nonlinear steep behavior of the Burgers’ equation in
the case of smaller values of viscosity. The proposed algorithm can also be used for solving other more
Funding: This research received no external funding.
Conﬂicts of Interest: The author declares no conﬂict of interest.
Bateman, H. Some recent researches on the motion of ﬂuids. Mon. Weather Rev.
Burger, J.M. A mathematical model illustrating the theory of turbulence. Adv. Appl. Math.
3. Cole, J.D. On a quasi-linear parabolic equation in aerodynamics. Q. Appl. Math. 1948,9, 225–236.
Hopf, E. The partial differential equation
.Commun. Pure Appl. Math.
Bonkile, M.P.; Awasthi, A.; Lakshmi, C.; Mukundan, V.; Aswin, V.S. A systematic literature review of Burgers’
equation with recent advances. Pramana 2018,90, 69, doi:10.1007/s12043-018-1559-4.
Caldwell, J.; Smith, P. Solution of Burgers’ equation with a large Reynolds numbe. Appl. Math. Model.
6, 381–385, doi:10.1016/S0307-904X(82)80102-9.
Mathematics 2019,7, 113 10 of 11
Kutluay, S.; Bahadir, A.R.; Ozdes, A. Numerical solution of one-dimensional Burgers’ equation:
Explicit and exact-explicit ﬁnite difference methods. J. Comput. Appl. Math.
Acedo, L. On the gravitational instability of a set of random walkers. Europhys. Lett.
Jiwari, R.; Mittal, R.C.; Sharma, K.K. A umerical scheme based on weighted average differential quadrature
method for the numerical solution of Burgers’ equation. Appl. Math. Comput.
Cordero, A.; Franques, A.; Torregrosa, J.R. Numerical solution of turbulence problems by solving Burgers’
equation, Algorithms 2015,8, 224–233, doi:10.3390/a8020224.
Jiwari, R. A hybrid numerical scheme for the numerical solution of the Burgers’ equation. Comput. Phys. Commun.
2015 188, 59–67, doi:10.1016/j.cpc.2014.11.004.
Chen, R.; Wu, Z. Applying multiquadric quasi-interpolation to solve Burgers’ equation. Appl. Math. Comput.
2006 172, 472–484, doi:10.1016/j.amc.2005.02.027.
Haq, S.; Hussain, A.; Uddin, M. On the numerical solution of nonlinear Burgers’-type equations using
meshless method of lines. Appl. Math. Comput. 2012,218, 6280–6290, doi:10.1016/j.amc. 2011.11.106.
Xie, H.; Li, D.; Li, F. A new Numerical Method of Particular Solutions for Inhomogeneous Burgers’ equation.
Math. Prob. Eng. 2013, doi:10.1155/2013/974808.
Sarboland, M.; Aminataei, A. On the Numerical Solution of One-Dimensional Nonlinear Nonhomogeneous
Burgers’ Equation. J. Appl. Math. 2014, doi:10.1155/2014/598432.
Fan, C.M.; Yang, C.H.; Gu, M.H. Applications of the local RBF collocation method and the ﬁctitious time
integration method for Burgers’ equations. Procedia Eng.
,79, 569–574, doi:10.1016/j.proeng.2014.06.380.
Bouhamidi, A.; Hached, M.; Jbilou, K. A meshless RBF method for computing a numerical solution of
unsteady Burgers’-type equations. Comput. Math. Appl.
,68, 238–256, doi:10.1016/j. camwa.2014.05.022.
Xie, H.; Zhou, J.; Jiang, Z.; Guo, X. Approximations for Burgers’ equations with C-N scheme and RBF
collocation methods. J. Nonlinear Sci. Appl. 2016,9, 3727–3734, doi:10.22436/jnsa. 009.06.23.
Hairer, E.; Lubich, C.; Wanner, G. Geometric Numerical Integration: Structure-Preserving Algorithms for Ordinary
Differential Equations; Springer Verlag: Berlin, Germany, 2006.
Blanes, S.; Casas, F. A Concise Introduction to Geometric Numerical Integration; CRC Press: New York, NY, USA, 2016.
21. Iserles, A.; Munthe-Kaas, H.Z.; Norsett, S.P.; Zanna, A. Lie-group methods. Acta Numer. 2000,9, 215–365.
Liu, C.S. A Group Preserving Scheme for Burgers Equation with Very Large Reynolds Number. CMES
Liu, C.S. An Efﬁcient Backward Group Preserving Scheme for the Backward in Time Burgers’ Equation.
CMES 2006,12, 5565.
Hajiketabi, M.; Abbasbandy, S. The combination of meshless method based on radial basis functions with a
geometric numerical integration method for solving partial differential equations: Application to the heat
equation. Eng. Anal. Bound. Elem. 2018,87, 36–46, doi:10.1016/j. enganabound.2017.11.008.
25. Hajiketabi, M.; Abbasbandy, S.; Casas, F. The Lie-group method based on radial basis functions for solving
nonlinear high dimensional generalized Benjamin-Bona-Mahony-Burgers equation in arbitrary domains.
Appl. Math. Comput. 2018,321, 223–243, doi:10.1016/j.amc.2017.10.051.
Kansa, E.J. Multiquadrics—A scattered data approximation scheme with applications to computational ﬂuid
dynamics—I surface approximations and partial derivative estimates. Comput. Math. Appl.
27. Kansa, E.J. Multiquadrics—A scattered data approximation scheme with applications to computational fluid
dynamics—II solutions to parabolic, hyperbolic and elliptic partial differential equations. Comput. Math. Appl.
1990,19, 147–161, doi:10.1016/0898-1221(90)90271-K.
Liu, C.S. Cone of non-linear dynamical system and group preserving schemes. Int. J. Nonlinear Mech.
36, 1047–1068, doi:10.1016/S0020-7462(00)00069-X.
29. Trefethen, L.N.; Bau, D., III. Numerical Linear Algebra; Siam: Philadelphia, PA, USA, 1997.
30. Trefethen, L.N. Spectral Methods in MATLAB; Siam: Philadelphia, PA, USA, 2000.
Mathematics 2019,7, 113 11 of 11
Özi¸s, T.; Aksan, E.N.; Özde¸s, A. A finite element approach for solution of Burgers’ equation. Appl. Math. Comput.
2003,139, 417–428, doi:10.1016/S0096-3003(02)00204-7.
Mukundan, V.; Awasthi, A. Efﬁcient numerical techniques for Burgers’ equation. Appl. Math. Comput.
262, 282–297, doi:10.1016/j.amc.2015.03.122.
2019 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).