一类分数阶广义KdV方程精确解
Exact Solution of a Type of Fractional Order Generalized KdV Equation
DOI: 10.12677/AAM.2018.76086, PDF, 下载: 1,113  浏览: 2,526  科研立项经费支持
作者: 李国放:北方工业大学理学院,北京
关键词: 分数阶子方程法分数阶广义KdV方程精确解Fractional Sub-Equation Method Fractional Generalized KdV Equation Exact Solutions
摘要: 分数阶子方程法是探寻分数阶偏微分方程精确解的一种行之有效的方法。本文利用该方法获得了一类分数阶广义KdV方程的精确解,并描绘精确解的图像。这些解能够对该方程在数学、物理等领域的广泛应用提供潜在的理论支持。
Abstract: Fractional sub-equation method is an effective method to explore the exact solution of fractional partial differential equations. In this paper, we use this method to obtain new exact solutions of a type of fractional generalized KdV equation. Moreover, we show the graphs of these exact solutions. Such solutions may provide potential support for the wide applications in mathematics, physics and related fields.
文章引用:李国放. 一类分数阶广义KdV方程精确解[J]. 应用数学进展, 2018, 7(6): 714-722. https://doi.org/10.12677/AAM.2018.76086

1. 引言

随着对微分方程研究的不断深入,人们对微分方程的研究逐渐从整数阶发展到分数阶。分数阶微分方程是对整数阶微分方程的推广,由它构造的系统广泛应用于物理、工程、系统识别、控制理论、信号处理、分数动力学等各个方面,其最大的优点是这种方程构造的系统的下一个状态不仅取决于它的当前状态,还取决于它的所有历史形态。近几十年间,许多数学、物理、生物等领域的学者积极投入到了对分数阶微分方程的研究中,为获得分数阶微分方程近似解和精确解提出了很多行之有效的方法,如变分迭代法 [1] [2] 、分数阶微分变换法 [3] [4] 、分数阶子方程法 [5] [6] [7] 、Adomian分解法 [8] [9] [10] 、同伦扰动法 [11] 、分数阶G'/G展开法 [12] [13] 、指数函数法 [14] 、有限差分法 [15] 等多种方法。然而,截至目前,尚未找到一种能够求解所有分数阶微分方程的方法,因此,寻找合适的方法求解微分方程仍然是一项十分有意义的研究。

本文主要研究如下方程的精确解

D t α u + u x x x x x u u x x x u x u x x u 2 u x = 0. (1.1)

该方程衍生于广义五阶KdV方程,其具体形式如下所示 [16] :

u t + ω u x x x x x + α u u x x x + β u x u x x + γ u 2 u x = 0. (1.2)

方程(1.2)既可以表示在重力作用下浅水中的长波运动,同时也是研究量子力学和非线性光学的一个重要模型。本文利用分数阶子方程法研究方程(1.1)的精确解,并利用数学软件mathematica求解计算过程中产生的超定方程组,最后获得了方程(1.1)的新的精确解。

本文的内容安排如下:第二节先介绍了修正Riemann-Liouville导数定义以及其具有的重要性质,然后介绍了本文所使用的分数阶子方程法;第三节则使用分数阶子方程法求解方程(1.1)的精确解,并给出了方程的具体图像。第四节为对本文内容的总结。

2. 方法概述

2.1. 修正Riemann-Liouville导数

修正Riemann-Liouville导数具有如下形式 [17] :

D t α f ( t ) = { 1 Γ ( α ) 0 t ( t s ) α 1 ( f ( s ) f ( 0 ) ) d s , α < 0 , 1 Γ ( 1 α ) d d t 0 t ( t s ) α ( f ( s ) f ( 0 ) ) d s , 0 < α < 1 , ( f ( n ) ( t ) ) ( α n ) , n < α < n + 1 , n 1. (2.1)

其具有的部分重要性质如下:

D t α t m = Γ ( 1 + m ) Γ ( 1 + m α ) t m α , m > 0 , (2.2)

D t α ( u ( t ) v ( t ) ) = v ( t ) D t α u ( t ) + u ( t ) D t α v ( t ) , (2.3)

D t α ( u ( v ( t ) ) ) = u v ( v ( t ) ) D t α ( v ( t ) ) = D v α ( u ( v ( t ) ) ) ( v ( t ) ) α . (2.4)

2.2. 分数阶子方程法

本文以如下分数阶偏微分方程

F ( u , u t , u x , D t α u , D x α u , D t 2 α u , D x 2 α u , ) = 0 , 0 < α 1. (2.5)

为例来简要介绍分数阶子方程法。其中, u = u ( t , x ) 是一个未知函数, D t α u , D x α u , D t 2 α u , D x 2 α u 是关于u的改进Riemann-Liouville导数,F是一个包含有非线性项和最高阶导数项的多项式。该方法主要分为以下几步:

1) 对方程(2.5)进行行波变换

u ( t , x ) = U ( θ ) , θ = m t + n x . (2.6)

其中 m , n 为待定常数,则方程(2.5)可转化为

G ( U ( θ ) , m U ( θ ) , n U ( θ ) , m α D θ α U ( θ ) , n α D θ α U ( θ ) , ) = 0. (2.7)

2) 假设方程具有如下形式的解:

U ( θ ) = i = 0 n a i ( Φ ( θ ) ) i , (2.8)

其中, a i 为待定常数, Φ = Φ ( θ ) 满足如下的分数阶黎卡提微分方程

D θ α Φ = σ + Φ 2 , 0 < α 1. (2.9)

其中, σ 为待定常数,方程(2.9)的一些解如下所示:

Φ ( θ ) = { σ tanh α ( σ θ ) , σ < 0 , σ coth α ( σ θ ) , σ < 0 , σ tan α ( σ θ ) , σ > 0 , σ cot α ( σ θ ) , σ > 0 , Γ ( 1 + α ) θ α + ω , ω , σ = 0. (2.10)

广义的三角函数和双曲函数如下所示 [17] :

sin α ( θ ) = E α ( i θ α ) E α ( i θ α ) 2 i ,

cos α ( θ ) = E α ( i θ α ) + E α ( i θ α ) 2 i ,

tan α ( θ ) = sin α ( θ ) cos α ( θ ) ,

cot α ( θ ) = cos α ( θ ) sin α ( θ ) ,

sinh α ( θ ) = E α ( θ α ) E α ( θ α ) 2 ,

cosh α ( θ ) = E α ( θ α ) + E α ( θ α ) 2 ,

tanh α ( θ ) = sinh α ( θ ) cosh α ( θ ) ,

coth α ( θ ) = cosh α ( θ ) sinh α ( θ ) . (2.11)

其中, E α ( θ α ) 是Mittag-Leffler方程,具体表现形式为:

E α ( θ α ) = k = 0 θ k Γ ( 1 + k α ) , α > 0. (2.12)

特例:当 α = 1 2 时,Mittag-Leffler方程可转化为:

E 1 2 ( θ 1 2 ) = e θ E r f c ( θ 1 2 ) , Erfc为误差函数. (2.13)

3) 将方程(2.8)和(2.9)代入方程(2.7)中,之后平衡方程的非线性项和最高阶导数项,即可确定方程(2.8)中的n。

4) 将已经确定n的方程(2.8)以及方程(2.9)代入方程(2.7)中,可以得到关于 Φ i 的方程。之后,令这个方程组的系数为零,能够得到一系列关于 a i , m , n , σ , α 的方程组,求解这些方程组,即可求得方程(2.5)的解。

3. 方程求解

本节将采用上述分数阶子方程法求解方程(1.1),对于该方程,我们令

u ( t , x ) = u ( θ ) , θ = x + c t . (3.1)

则方程(1.1)转化为

c α D t α u + u ( 5 ) u u ( 3 ) u u u 2 u = 0. (3.2)

假设方程具有(2.8)形式的解,即

u ( θ ) = i = 0 a i ( Φ ( θ ) ) i . (3.3)

将方程(3.3)代入方程(3.2)中,平衡非线性项 u 2 u 和最高阶导数项 u ( 5 ) 可得 n = 2 ,即

u ( θ ) = a 0 + a 1 Φ ( θ ) + a 2 ( Φ ( θ ) ) 2 . (3.4)

将方程(3.4)代入(3.2)中,令 Φ i ( θ ) = 0 ,可以得到一系列关于 a 0 , a 1 , a 2 , c , α , σ 代数方程,具体方程如下所示:

( Φ ( θ ) ) 0 : c α σ a 1 + 16 σ 3 a 1 2 σ 2 a 0 a 1 σ a 0 2 a 1 2 σ 3 a 1 a 2 = 0 ,

( Φ ( θ ) ) 1 : 4 σ 2 a 1 2 2 σ a 0 a 1 2 + 2 c α σ a 2 + 272 σ 3 a 2 16 σ 2 a 0 a 2 2 σ a 0 2 a 2 4 σ 3 a 2 2 = 0 ,

( Φ ( θ ) ) 2 : c α a 1 + 136 σ 2 a 1 8 σ a 0 a 1 a 0 2 a 1 σ a 1 3 32 σ 2 a 1 a 2 6 σ a 0 a 1 a 2 = 0 ,

( Φ ( θ ) ) 3 : 12 σ a 1 2 2 a 0 a 1 2 + 2 c α a 2 + 1232 σ 2 a 2 40 σ a 0 a 2 2 a 0 2 a 2 4 σ a 1 2 a 2 36 σ 2 a 2 2 4 σ a 0 a 2 2 = 0 ,

( Φ ( θ ) ) 4 : 240 σ a 1 6 a 0 a 1 a 1 3 70 σ a 1 a 2 6 a 0 a 1 a 2 5 σ a 1 a 2 2 = 0

( Φ ( θ ) ) 5 : 8 a 1 2 + 1680 σ a 2 24 a 0 a 2 4 a 1 2 a 2 68 σ a 2 2 4 a 0 a 2 2 2 σ a 2 3 = 0 ,

( Φ ( θ ) ) 6 : 120 a 1 40 a 1 a 2 5 a 1 a 2 2 = 0 ,

( Φ ( θ ) ) 7 : 720 a 2 36 a 2 2 2 a 2 3 = 0. (3.5)

之后,借助mathematica软件求解这些方程,可以得到如下几种形式的解。

情形一:此时 a 0 = a 0 , a 1 = 0 , a 2 = 30 , c = ( 11 a 0 2 100 ) 1 α , σ = a 0 20 。通过这个解,我们可以得到如下精确解:

u 1 ( t , x ) = a 0 30 ( a 0 20 tanh α ( a 0 20 ( x + ( 11 a 0 2 100 ) 1 α t ) ) ) 2 , a 0 > 0 ,

u 2 ( t , x ) = a 0 30 ( a 0 20 coth α ( a 0 20 ( x + ( 11 a 0 2 100 ) 1 α t ) ) ) 2 , a 0 > 0 ,

u 3 ( t , x ) = a 0 30 ( a 0 20 tan α ( a 0 20 ( x + ( 11 a 0 2 100 ) 1 α t ) ) ) 2 , a 0 < 0 ,

u 4 ( t , x ) = a 0 30 ( a 0 20 cot α ( a 0 20 ( x + ( 11 a 0 2 100 ) 1 α t ) ) ) 2 , a 0 < 0. (3.6)

在精确解 u 1 ( t , x ) 中,我们令 a 0 = 10 , α = 2 3 , c = 11 11 , σ = 1 2 ,并将公式(2.11)和(2.12)代入 u 1 ( t , x )

中,借助于mathematica软件,可得

u 1 ( t , x ) = 10 15 ( k = 0 ( x + 11 11 t 2 ) k Γ ( 1 + 2 3 k ) k = 0 ( x + 11 11 t 2 ) k Γ ( 1 + 2 3 k ) k = 0 ( x + 11 11 t 2 ) k Γ ( 1 + 2 3 k ) + k = 0 ( x + 11 11 t 2 ) k Γ ( 1 + 2 3 k ) ) 2 (3.7)

其图像如图1所示。

Figure 1. Figure of solution (3.7)

图1. 解(3.7)的图

在精确解 u 3 ( t , x ) 中,我们令 a 0 = 10 , α = 4 5 , c = 11 5 4 , σ = 1 2 ,并将公式(2.11)和(2.12)代入 u 3 ( t , x ) 中,

借助于mathematica软件,可得

u 3 ( t , x ) = 10 15 ( k = 0 ( x + 11 5 4 t 2 i ) k Γ ( 1 + 4 5 k ) k = 0 ( x + 11 5 4 t 2 i ) k Γ ( 1 + 4 5 k ) k = 0 ( x + 11 5 4 t 2 i ) k Γ ( 1 + 4 5 k ) + k = 0 ( x + 11 5 4 t 2 i ) k Γ ( 1 + 4 5 k ) ) 2 (3.8)

其图像如图2所示。

情形二:此时 a 0 = a 0 , a 1 = 0 , a 2 = 12 , c = ( a 0 2 4 ) 1 α , σ = a 0 8 。通过这个解,我们可以得到如下精确解:

u 5 ( t , x ) = a 0 + 12 ( a 0 8 tanh α ( a 0 8 ( x + ( a 0 2 4 ) 1 α t ) ) ) 2 , a 0 < 0 ,

u 6 ( t , x ) = a 0 + 12 ( a 0 8 coth α ( a 0 8 ( x + ( a 0 2 4 ) 1 α t ) ) ) 2 , a 0 < 0 ,

u 7 ( t , x ) = a 0 + 12 ( a 0 8 tan α ( a 0 8 ( x + ( a 0 2 4 ) 1 α t ) ) ) 2 , a 0 > 0 ,

u 8 ( t , x ) = a 0 + 12 ( a 0 8 cot α ( a 0 8 ( x + ( a 0 2 4 ) 1 α t ) ) ) 2 , a 0 > 0. (3.9)

在精确解 u 5 ( t , x ) 中,我们令 a 0 = 2 2 , α = 1 2 , c = 4 , σ = 2 2 8 ,并将公式(2.11)和(2.13)代入 u 5 ( t , x )

中,借助于mathematica软件,可得

u 5 ( t , x ) = 2 2 + 3 2 ( e ( x + 4 t ) 2 2 2 E r f c [ x + 4 t 2 3 4 ] e ( x + 4 t ) 2 2 2 E r f c [ x + 4 t 2 3 4 ] e ( x + 4 t ) 2 2 2 E r f c [ x + 4 t 2 3 4 ] + e ( x + 4 t ) 2 2 2 E r f c [ x + 4 t 2 3 4 ] ) 2 (3.10)

其图像如图3所示。

在精确解 u 8 ( t , x ) 中,我们令 a 0 = 2 , α = 2 3 , c = 1 , σ = 1 4 ,并将公式(2.11)和(2.12)代入 u 8 ( t , x ) 中,借助

于mathematica软件,可得

u 8 ( t , x ) = 2 + 3 ( k = 0 ( x + t 2 i ) k Γ ( 1 + 2 3 k ) k = 0 ( x + t 2 i ) k Γ ( 1 + 2 3 k ) k = 0 ( x + t 2 i ) k Γ ( 1 + 2 3 k ) + k = 0 ( x + t 2 i ) k Γ ( 1 + 2 3 k ) ) 2 (3.11)

其图像如图4所示。

Figure 2. Figure of solution (3.8)

图2. 解(3.8)的图

Figure 3. Figure of solution (3.10)

图3. 解(3.10)的图

Figure 4. Figure of solution (3.11)

图4. 解(3.11)的图

4. 总结

本文主要应用分数阶子方程法寻找分数阶广义KdV方程(1.1)的精确解,最后给出了方程新的广义双曲函数和三角函数形式的解,画出了解的图像。这些解能够加深人们对分数阶KdV方程的理解,从而促进该方程在非线性系统和孤子理论等领域的应用。

致谢

本文得到北京市教委面上项目(编号:KM201710009011)和北京市教委基础科研计划项目资助。

参考文献

[1] Zhang, Y. (2014) Formulation and Solution to Time-Fractional Generalized Korteweg-de Vries Equation via Variational Methods. Advances in Difference Equations, 2014, 65.
https://doi.org/10.1186/1687-1847-2014-65
[2] Elsaid, A. (2010) The Variational Iteration Method for Solving Riesz Fractional Partial Differential Equations. Computers & Mathematics with Applications, 60, 1940-1947.
https://doi.org/10.1016/j.camwa.2010.07.027
[3] Khan, N.A., Ara, A. and Mahmood, A. (2012) Numerical Solu-tions of Time-Fractional Burger Equations: A Comparison between Generalized Transformation Technique with Homotopy Pertur-bation Method. International Journal of Numerical Methods for Heat & Fluid Flow, 24,175-193.
https://doi.org/10.1108/09615531211199818
[4] Odibat, Z. and Momani, S. (2008) A Generalized Differential Transform Method for Linear Partial Differential Equations of Fractional Order. Applied Mathematics Letters, 21, 194-199.
https://doi.org/10.1016/j.aml.2007.02.022
[5] Zhang, S. and Zhang, H.Q. (2011) Fractional Sub-Equation Method and Its Applications to Nonlinear Fractional PDEs. Physics Letters A, 375, 1069-1073.
https://doi.org/10.1016/j.physleta.2011.01.029
[6] Guo, S., Mei, L., Li, Y., et al. (2012) The Improved Fractional Sub-Equation Method and Its Applications to the Space-Time Fractional Differential Equations in Fluid Mechanics. Physics Letters A, 376, 407-411.
https://doi.org/10.1016/j.physleta.2011.10.056
[7] Sahoo, S. and Ray, S.S. (2015) Improved Fractional Sub-Equation Method for (3+1)-Dimensional Generalized Fractional KdV-Zakharov-Kuznetsov Equations. Computers & Mathematics with Applications, 70, 158-166.
https://doi.org/10.1016/j.camwa.2015.05.002
[8] El-Sayed, A.M.A. and Gaber, M. (2006) The Adomian Decomposition Method for Solving Partial Differential Equations of Fractal Order in Finite Domains. Physics Letters A, 359, 175-182.
https://doi.org/10.1016/j.physleta.2006.06.024
[9] Momani, S. and Odibat, Z. (2006) Analytical Approach to Linear Fractional Partial Differential Equations Arising in Fluid Mechanics. Physics Letters A, 355, 271-279.
https://doi.org/10.1016/j.physleta.2006.02.048
[10] Matinfar, M., Nasseri, S.H., Ghanbari, M., et al. (2009) Numerical Solution of Linear Fuzzy Fredholm Integro-Differential Equations by Adomian Decomposition Method. International Journal of Applied Mathematics, 22, 677-688.
[11] Markosian, A.A. (2010) A Numerical Algorithm for the Solution of an Intermediate Fractional Ad-vection Dispersion Equation. Communications in Nonlinear Science & Numerical Simulation, 15, 1253-1258.
https://doi.org/10.1016/j.cnsns.2009.05.072
[12] Wang, M., Li, X. and Zhang, J. (2008) The (G′/G)-Expansion Method and Travelling Wave Solutions of Nonlinear Evolution Equations in Mathematical Physics. Physics Letters A, 372, 417-423.
https://doi.org/10.1016/j.physleta.2007.07.051
[13] Zheng, B. (2012) (G'/G)-Expansion Method for Solving Fractional Partial Differential Equations in the Theory of Mathematical Physics. Communications in Theoretical Physics, 58, 623-630.
https://doi.org/10.1088/0253-6102/58/5/02
[14] Ma, W.X. and Zhu, Z. (2012) Solving the (3 + 1)-Dimensional Generalized KP and BKP Equations by the Multiple Exp-Function Algorithm. Applied Mathematics & Computation, 218, 11871-11879.
https://doi.org/10.1016/j.amc.2012.05.049
[15] Cui, M. (2009) Compact Finite Difference Method for the Fractional Diffusion Equation. Journal of Computational Physics, 228, 7792-7804.
https://doi.org/10.1016/j.jcp.2009.07.021
[16] Salas, A.H. (2008) Exact Solutions for the General Fifth KdV Equation by the Exp Function Method. Applied Mathematics & Computation, 205, 291-297.
https://doi.org/10.1016/j.amc.2008.07.013
[17] Wang, G.W. and Xu, T.Z. (2014) Invariant Analysis and Exact Solutions of Nonlinear Time Fractional Sharma-Tasso-Olver Equation by Lie Group Analysis. Nonlinear Dynamics, 76, 571-580.
https://doi.org/10.1007/s11071-013-1150-y