Solution of KDV Equation by Computer Algebra: He Ping, Chen Zheng, Fu Jun

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 5

Applied Mathematics and Computation 136 (2003) 511–515

www.elsevier.com/locate/amc

Solution of KdV equation


by computer algebra
a,*
He Ping , Chen Zheng b, Fu Jun c

a
Physics Department, Guangxi Teachers University, 19 MingXiu Road, Nanning,
Guangxi 530001, China
b
Nanning New & High-Tech Industrial Development Zone, Nanning, Guangxi 530002, China
c
Physics Department, Hainan Teachers University, Haikou 571158, China

Abstract
In this paper we use the real exponential approach by draw up Mathematica pro-
gram to calculate the KdV (Korteweg–de Vries) equation. We get the same soliton
solutions as conventional analytical methods of nonlinear equation.
Ó 2002 Elsevier Science Inc. All rights reserved.

Keywords: Computer algebra; KdV equation; Real exponential approach; Soliton solutions

1. Introduction

Nonlinear reaction of physical quantities in most dynamic models can be


turned to solve nonlinear evolution equation. For this reaction, solving non-
linear evolution equation and making researches on solving methods is an
active task in branches of physics. Among many of nonlinear equations in
physics, KdV equation is a typical, relatively simple and classical one. Many
literatures of nonlinear equation theory [1–5] usually take solving soliton so-
lutions of KdV equation for example to introduce nonlinear theory, method
and character of soliton solutions. In this paper we use the real exponential
approach by making Mathematica program to calculate the KdV equation. We
get the same soliton solutions as conventional analytical methods of nonlinear

*
Corresponding author.
E-mail addresses: hp8827@263.net, heping@webmail.gxtc.edu.cn (H. Ping).

0096-3003/02/$ - see front matter Ó 2002 Elsevier Science Inc. All rights reserved.
PII: S 0 0 9 6 - 3 0 0 3 ( 0 2 ) 0 0 0 7 6 - 0
512 H. Ping et al. / Appl. Math. Comput. 136 (2003) 511–515

equations. Using computer algebra method to solve nonlinear equation is a


new attempt.

2. Real exponential approach of nonlinear equation and computer algebra


calculate

Real exponential approach was put forward by Korpel in 1978. Later this
method was developed to solve the nonlinear equation by Hereman et al. [6],
etc. In this method series solution of linear equation was used to express so-
lution of nonlinear equation. Putting the solution of equation into nonlinear
equation, then decompound nonlinear equation to compare the coefficients gn .
We can get the expression of series coefficients an . It is difficult to directly use
real exponential approach to calculate nonlinear equation because it is not easy
to get a general term of the series coefficient. We use the real exponential ap-
proach by making Mathematica program to solve series coefficient of equation
and the general term of an is convenient to obtain. We adopt LogicalExpand
function to realize this process in making program. This method can reduce
calculating quantity and raise calculating velocity.

3. Solution of KdV equation by computer algebra

Taking an example, we make Mathematica program to calculate soliton


solutions of KdV equation. KdV equation is a dispersion equation. Its general
form is [1–3]
ou ou o3 u
þ au þ 3 ¼ 0: ð1Þ
ot ox ox
Using real exponential approach, we decompound u into series form:
X
1
u ¼ a0 þ an g n ðg < 1Þ; ð2Þ
n¼1

X
1
u ¼ a0 þ an gn ðg > 1Þ; ð3Þ
n¼1

an is the decomposition coefficient, and


g ¼ eðkxþxtÞ : ð4Þ
The key to solve equation by real exponential approach lies in finding out the
general term solution of an . Now we solve equations with zero boundary
condition (a0 ¼ 0) and non-zero boundary condition (a0 6¼ 0), respectively.
H. Ping et al. / Appl. Math. Comput. 136 (2003) 511–515 513

3.1. Zero boundary condition (a0 ¼ 0)

a1 is arbitrary constant when a0 ¼ 0. Draw up program by Mathemat-


ica:
deg ¼ 10; n0 ¼ 1;
v ¼ Sum[a[n] g^n, n, n0, deg}];
vt ¼ Sum[a[n](n x) g^n, {n, n0, deg}];
vx ¼ Sum[a[n] (-nk) g^n, {n, n0, deg}];
vxxx ¼ Sum[a[n] (n x)3 g^n, {n, n0, deg}];
u ¼ Series[v, {g,0, deg}];
ut ¼ Series[vt , {g, 0, deg}];
ux ¼ Series[vx , {g, 0, deg}];
uxxx ¼ Series[vxxx , {g, 0, deg}];
f ¼ LogicalExpand[ut þ a ux u þ uxxx ¼¼ 0]
After calculating, we get
 k 3 a1 þ xa1 ¼ 0; ð5Þ
3
 8k a2 þ 2xa2  kaa21 ¼ 0; ð6Þ
3
 27k a3 þ 3xa3  3kaa1 a2 ¼ 0; ð7Þ
..
.
X
n1
ðnkÞ3 an þ nxan  ak am ðn  mÞanm ¼ 0; ð8Þ
m¼1

where we use CauchyÕs rule [6] to get formula (8). According to recursion re-
lations (5)–(8), we can get expression of an about a1 :
k2
an ¼ 12 ð1Þðnþ1Þ nan ; ð9Þ
a
where
a1 a
x ¼ k3; a¼ :
12k 2
Putting (9) into (2), we get
12k 2 ag
u¼ 2
: ð10Þ
að1 þ agÞ
Formula (10) was used following the formula:

2
X
1
nþ1
ð1 þ xÞ ¼ ð1Þ nxn1 : ð11Þ
n¼1
514 H. Ping et al. / Appl. Math. Comput. 136 (2003) 511–515

Fig. 1. k ¼ 1:25, t ¼ 0, a ¼ 1.

Putting
ag ¼ aeðkxþxtÞ ¼ eðkxþxtþln aÞ ð12Þ
into formula (10), we get
 
3k 2 1
u¼ sech2 ðkx  xtÞ : ð13Þ
a 2
Formula (13) is soliton solution of KdV equation under zero boundary
condition. Fig. 1 is its graph drawing up by Mathematica program.

3.2. Non-zero boundary condition (a0 6¼ 0)

Drawing up Mathematica program:


deg ¼ 10; n0 ¼ 0;
v ¼ Sum[a[n] g^n, {n, n0, deg}];
vt ¼ Sum[a[n] (n x) g^n, {n, n0, deg}];
vx ¼ Sum[a[n] (-nk) g^n, {n, n0, deg}];
vxxx ¼ Sum[a[n] (n x)3 g^n, {n, n0, deg}];
u ¼ Series[v, {g, 0, deg}];
ut ¼ Series[vt , {g, 0, deg}];
ux ¼ Series[vx , {g, 0, deg}];
uxxx ¼ Series[vxxx , {g, 0, deg}];
f ¼ LogicalExpand[ut þ a ux u þ uxxx ¼¼ 0]
after calculating, we get
 k 3 a1 þ xa1  kaa0 a1 ¼ 0; ð14Þ
3
 8k a2 þ 2xa2 þ aðka21  2ka0 a2 Þ ¼ 0; ð15Þ
..
.
3
X
n1
ðnkÞ an þ nxan  ak am ðn  mÞanm ¼ 0: ð16Þ
m¼1
H. Ping et al. / Appl. Math. Comput. 136 (2003) 511–515 515

Fig. 2. k ¼ 1:25, t ¼ 0, a ¼ 1.

According to recursion relations (14)–(16), we can get expression of an


k2
an ¼ 12 ð1Þðnþ1Þ nan ; ð17Þ
a
where
a1 a
x ¼ k 3 þ kaa0 ; a¼ :
12k 2
Recursion relation of an is same as zero boundary condition, but x was
changed. Putting formula (17) into formula (2), we can get soliton solution of
equation:
 
3k 2 1
u ¼ a0 þ sech2 ðkx  xtÞ : ð18Þ
a 2
Formulas (13) and (18) are same as in the literature [2,3]. Fig. 2 is its graph.
Using the real exponential approach by drawing up Mathematica program to
calculate the nonlinear equation, we get the same soliton solutions as con-
ventional analytical methods of nonlinear equation.

References

[1] B.-L. Guo, Nonlinear Evolution Equation [M], Shanghai Science and Technology Press,
Shanghai, 1996.
[2] S.-K. Liu, S.-D. Liu, Nonlinear Equation in Physics [M], Peking University Press, Peking, 2000.
[3] L.-Y. Zhuo et al., Nonlinear Physics Threoy and Application [M], Science Press, Peking, 2000.
[4] R.M. Miura (Ed.), Backlaund Transformation C, Springer Lecture Notes in Mathematica, vol.
515, Springer, New York, 1976.
[5] W.R. Rogerscand Shaduick, Backland Transformation and their Applications, Academic Press,
New York, 1982.
[6] W. Hereman, P.P. Banerjee, A. Korpel, G. Assanto, A. Immerzeele, A. Vanand, Meepeel,
Physica A 19 (1986) 607–628.

You might also like