Commun Nonlinear Sci Numer Simulat 14 (2009) 2592–2596
Contents lists available at ScienceDirect
Commun Nonlinear Sci Numer Simulat journal homepage: www.elsevier.com/locate/cnsns
Solving singular boundary value problems of higher-order ordinary differential equations by modified Adomian decomposition method q Yahya Qaid Hasan, Liu Ming Zhu * Department of Mathematics, Harbin Institute of Technology, 11, Siling Street, Harbin, Helongjiang 150001, PR China
a r t i c l e
i n f o
Article history: Received 28 June 2008 Received in revised form 30 September 2008 Accepted 30 September 2008 Available online 10 October 2008
a b s t r a c t In this paper, we use modified Adomian decomposition method to solving singular boundary value problems of higher-order ordinary differential equations. The proposed method can be applied to linear and nonlinear problems. The scheme is tested for some examples and the obtained results demonstrate efficiency of the proposed method. Ó 2008 Elsevier B.V. All rights reserved.
PACS: 02.60.Lj Keywords: Adomian decomposition method Singular boundary value problems Higher-order ordinary differential equation
1. Introduction The decomposition method has been shown [2–4,8–14] to solve effectively, easily and accurately a large class of linear and nonlinear, ordinary, partial, deterministic or stochastic differential equations with approximate solutions which converge rapidly to accurate solutions. The goal of this paper is to introduce a new reliable modification of Adomian decomposition method. For this reason, a new differential operator is proposed which can be used for singular boundary value problem. In this paper we find solutions without massive computations and restrictive assumptions which change the physics problem into a mathematically tractable problem. This method is easy than the traditional methods. The theoretical treatment of the Convergence of Adomian’s method has been considered in[1,5–7]. 2. Modified Adomian decomposition method Consider the singular boundary value problem of n þ 1 order ordinary differential equation in the form
m ðnÞ y þ Ny ¼ gðxÞ; x yð0Þ ¼ a0 ; y0 ð0Þ ¼ a1 ; . . . ; yðn1Þ ð0Þ ¼ an1 ; yðbÞ ¼ c; yðnþ1Þ þ
q
This paper is supported by National Natural Science Foundation of China (10471067). * Corresponding author. Tel.: +68 451 86 403055; fax: +86 451 86403055. E-mail addresses:
[email protected] (Y.Q. Hasan),
[email protected] (L.M. Zhu).
1007-5704/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.cnsns.2008.09.027
ð1Þ
2593
Y.Q. Hasan, L.M. Zhu / Commun Nonlinear Sci Numer Simulat 14 (2009) 2592–2596
where N is a nonlinear differential operator of order less than n, gðxÞ is given function and a0 ; a1 ; . . . ; an1 ; c; b are given constants. We propose the new differential operator, as below n
Lð:Þ ¼ x1
d 1þnm d mn x ð:Þ; nx dx dx
ð2Þ
where m 6 n; n P 1, so, the problem (1) can be written as
Ly ¼ gðxÞ Ny:
ð3Þ 1
The inverse operator L
L1 ð:Þ ¼ xnm
Z
is therefore considered a n þ 1-fold integral operator, as below
x
xmn1
b
Z
x
Z
Z
x
x
xð:Þdx . . . dx:
... 0
0
ð4Þ
0
By applying L1 on (3), we have
yðxÞ ¼ UðxÞ þ L1 gðxÞ L1 Ny:
ð5Þ
such that
LUðxÞ ¼ 0: The Adomian decomposition method introduce the solution yðxÞ and the nonlinear function Ny by infinite series
yðxÞ ¼
1 X
yn ðxÞ;
ð6Þ
n¼0
and
Ny ¼
1 X
An ;
ð7Þ
n¼0
where the components yn ðxÞ of the solution yðxÞ will be determined recurrently. Specific algorithms were seen in [8,12] to formulate Adomian polynomials. The following algorithm:
A0 ¼ FðuÞ; A1 ¼ F 0 ðu0 Þu1 ; 1 A2 ¼ F 0 ðu0 Þu2 þ F 00 ðu0 Þu21 ; 2 A3 ¼ F 0 ðu0 Þu3 þ F 00 ðu0 Þu1 u2 þ
ð8Þ 1 000 F ðu0 Þu31 ; 3!
.. . can be used to construct Adomian polynomials, when FðuÞ is a nonlinear function. By substituting (6) and (7) into (5), 1 X
yn ¼ UðxÞ þ L1 gðxÞ L1
n¼0
1 X
An :
ð9Þ
n¼0
Through using Adomian decomposition method, the components yn ðxÞ can be determined as
y0 ¼ UðxÞ þ L1 gðxÞ; ynþ1 ¼ L1 An ; n P 0;
ð10Þ
which gives
y0 ¼ UðxÞ þ L1 gðxÞ; y1 ¼ L1 A0 ; y2 ¼ L1 A1 ;
ð11Þ
1
y3 ¼ L A3 ; ... From (8) and (11), we can determine the components yn ðxÞ, and hence the series solution of yðxÞ in (6) can be immediately obtained. For numerical purposes, the n-term approximate
2594
Y.Q. Hasan, L.M. Zhu / Commun Nonlinear Sci Numer Simulat 14 (2009) 2592–2596
Wn ¼
n1 X
yk ;
ð12Þ
n¼0
can be used to approximate the exact solution. The approach presented above can be validated by testing it on a variety of several linear and nonlinear initial value problems. 3. Numerical examples Example 1. We consider the linear boundary value problem:
q
y0 ¼ x1q cos x ð2 qÞxq sin x; x yð0Þ ¼ 0; yð1Þ ¼ cos 1:
y00 þ
ð13Þ
We put
Lð:Þ ¼ x1
d 2q d 1þq ð:Þ; x x dx dx
so
L1 ð:Þ ¼ x1q
Z
x
x2þq
1
Z
x
xð:Þdxdx:
0
In an operator form, Eq. (13) becomes
Ly ¼ x1q cos x ð2 qÞxq sin x: 1
By applying L
ð14Þ
to both sides of (14) we have
y ¼ yð0Þ þ ðyð1Þ yð0ÞÞx1q þ L1 ðx1q cos x ð2 qÞxq sin xÞ Z x Z x ¼ x1q cos 1 þ x1q x2þq ðx1q cos x ð2 qÞxq sin xÞdxdx; 1
0
and it implies,
yðxÞ ¼ x1q cos 1 x1q cos 1 þ x1q cos x ¼ x1q cos x: so, the exact solution is easily obtained by this method. Example 2. We consider the non-linear boundary value problem:
1 4x2 y y00 y0 ¼ e; x 4 þ x2 1 1 ; yð1Þ ¼ ln : yð0Þ ¼ ln 4 5
ð15Þ
We put
Lð:Þ ¼ x1
d 3 d 2 x x ð:Þ; dx dx
so
L1 ð:Þ ¼ x2
Z
x
1
x3
Z
x
xð:Þdxdx: 0
In an operator form, Eq. (15) becomes
Ly ¼
4x2 y e: 4 þ x2
ð16Þ
Applying L1 on both sides of (16) we find
y ¼ yð0Þ þ ðyð1Þ yð0ÞÞx2 þ 4L1
x2 ey : 4 þ x2
2595
Y.Q. Hasan, L.M. Zhu / Commun Nonlinear Sci Numer Simulat 14 (2009) 2592–2596
By modified Adomian decomposition method [9] we obtain
1 ; 4 4 2 x2 x þ 4L1 A0 ; y1 ¼ ln 5 4 þ x2 2 x Ak ; k P 1: ykþ1 ¼ 4L1 4 þ x2 y0 ¼ ln
ð17Þ
The Adomian polynomials for the nonlinear term FðyÞ ¼ ey are computed as follows:
A0 ¼ ey0 ; A1 ¼ y1 ey0 ; 1 A2 ¼ y2 þ y21 ey0 ; 2 1 A3 ¼ y3 þ y1 y2 þ y31 ey0 ; 3!
ð18Þ
1 Substituting (18) into (17) and it must noted that, to compute y1 we use the Taylor series of 4þx 2 with order 10. In this case we obtain
y0 ¼ ln
1 ; 4
y1 ¼ 0:2520728x2 þ 0:03125x4 0:0026042x6 þ 0:0003255x8 0:0000488x10 þ 8:1380208 106 x12 1:4532180 106 x14 ; y2 ¼ 0:0098820x2 0:0105030x6 þ 0:0006510x8 0:0000326x10 þ 2:7126736 106 x12 2:9064360 107 x14 þ 3:6330450 108 x16 5:0458959 109 x18 ; .. . In Fig. 1 we have plotted
P4
i¼0 yi ðxÞ,
which is almost equal to the exact solution yðxÞ ¼ ln
1 4þx2
.
Example 3. Consider the non-linear boundary value problems:
2 y000 y00 y y2 ¼ gðxÞ; x yð0Þ ¼ 0;
y0 ð0Þ ¼ 0;
ð19Þ
yð1Þ ¼ 2:71828182:
Where
gðxÞ ¼ 7x2 ex þ 6xex 6ex x6 e2x : Here, we use Taylor series of g(x) with order 10,
gðxÞ 6 þ 10x2 þ 9x3 þ
17 4 41 5 2 6 325 7 5729 8 20137 9 67181 10 x þ x x x x x x : 4 30 3 168 2880 15120 100800
y
-1.3865
-1.387
-1.3875
x -0.1
-0.05
0.05
0.1
-1.3885
1 Fig. 1. The exact solution y ¼ ln 4þx 2 and the Adomian decomposition solution y ¼
P4
i¼10 yiðxÞ .
2596
Y.Q. Hasan, L.M. Zhu / Commun Nonlinear Sci Numer Simulat 14 (2009) 2592–2596
We put 2
Lð:Þ ¼ x1
d
2
dx
So
L1 ð:Þ ¼ x4
Z
x5
d 4 x ð:Þ: dx
x
x5
Z
1
x 0
Z
x
xð:Þdxdxdx:
0
In an operator form, Eq. (19) becomes
Ly ¼ gðxÞ þ y þ y2 : 1
Applying L
ð20Þ
on both sides of (20) we find
y ¼ 2:71828182x4 þ L1 gðxÞ þ L1 y þ L1 y2 : Proceeding as before we obtained the recursive relationship
y0 ¼ 2:71828182x4 þ L1 gðxÞ; ynþ1 ¼ L1 yn þ L1 An ;
n P 0:
ð21Þ
Where An ’s are Adomian polynomials of nonlinear term y2 , as below
A0 ¼ y20 ; A1 ¼ 2y0 y1 ; A2 ¼ 2y2 y0 þ y21 ; .. .
ð22Þ
So, by substituting (22) into (21), we have
y0 ¼ x3 þ 1:0382041x4 þ 0:5x5 þ 0:15x6 þ 0:0337302x7 þ 0:0061012x8 þ 0:001851852x9 0:003582451x10 0:00258342x11 0:001261185x12 0:0004746x13 þ . . . ; y1 ¼ 0:0388054x4 þ 0:016666x6 þ 0:00823972x7 þ 0:00223214x8 þ 0:00319444x9 þ 0:00390766x10 þ 0:00270645x11 þ 0:00126549x12 þ 0:000445398x13 þ . . . ; y2 ¼ 0:000413381x4 0:000307979x7 þ 0:0000462963x9 0:000128465x10 0:000101745x11 2:15681 106 x12 þ 0:0000308777x13 þ . . . ; .. . This means that the solution in a series form is given by
y ¼ y0 þ y1 þ y2 ¼ x3 þ 0:999812x4 þ 0:5x5 þ 0:1666666x6 þ 0:0416619x7 þ 0:0083333x8 þ 0:00138889x9 þ 0:000196747x10 þ . . . : Note that, the Taylor series of exact solution yðxÞ ¼ xex with order 10 is as below
yðxÞ ¼ x3 þ x4 þ 0:5x5 þ 0:166666x6 þ 0:04166666x7 þ 0:008333x8 þ 0:0013888x9 þ 0:000198x10 þ . . . : References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]
Abbaoui K, Cherruault Y. New ideas for proving convergence of decomposition methods. Comput Math Appl 1995;29:103–8. Adomian G. A review of the decomposition method and some recent results for nonlinear equation. Math Comput Model 1992;13(7):17–43. Adomain G. Differential equations with singular coefficients. Appl Math Comput 1992;47:179–84. Adomian G. Solving Frontier problems of physics: the decomposition method. Boston, MA: Kluwer; 1994. Cherruault Y, Adomian G, Abbaoui K, Rach R. Further remarks on convergence of decomposition method. Bio-Med Comput 1995;38:89–93. Hosseini MM, Nasabzadeh H. On the convergence of Adomian decomposition method. Appl Math Comput 2006;182:536–43. Lesnic D. Convergence of Adomian decomposition method: Periodic temperatures. Comput Math Appl 2002;44:13–24. Wazwaz AM. A first course in integral equation. Singapore: World Scientific; 1997. Wazwaz AM. A reliable modification of Adomian decomposition method. Appl Math Comput 1999;102:77–86. Wazwaz AM. Approximate solutions to boundary value problems of higher-order by the modified decomposition method. Comput Math Appl 2000;40:679–91. Wazwaz AM. The numerical solution of fifth-order BVP by the decomposition method. J Comput Appl Math 2001;136:59–270. Wazwaz AM. A new algorithm for calculating Adomian polynomials for nonlinear operators. Appl Math Comput 2000;111(1):53–69. Wazwaz AM. A new method for solving singular initial value problems in the second-order ordinary differential equations. Appl Math Comput 2002;128:45–57. Wazwaz AM. A new algorithm for solving boundary value problems for higher-order integro-differential equations. Appl Math Comput 2001;118:327–42.