See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/332555930

Solving Linear Boundary Value Problem Using Shooting Continuous Explicit

Runge-Kutta Method

Article · April 2019

CITATIONS READS

0 1,872

2 authors:

Bushra E. Kashem Madeha Shaltagh

University of Technology, Iraq University of technology
14 PUBLICATIONS   5 CITATIONS    6 PUBLICATIONS   1 CITATION   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Generalizing of Finite Difference Method for Certain Fractional Order Parabolic PDE's View project

All content following this page was uploaded by Bushra E. Kashem on 22 April 2019.

The user has requested enhancement of the downloaded file.

2013@‚b«@H3@ÖÜ»€a@I@26@Ü‹1a@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ÚÓ‘Ój�n€a@Î@Úœäñ€a@‚Ï‹»‹€@·rÓ:a@Âig@Ú‹©
Ibn Al-Haitham Jour. for Pure & Appl. Sci. Vol. 26 (3) 2013

Solving Linear Boundary Value Problem Using Shooting

Continuous Explicit Runge-Kutta Method
Madeha Sh. Yousif
Bushra E. Kashiem
University of Technology

Received in : 20 May 2012 , Accepted in : 8 April 2013

Abstract
In this paper we shall generalize fifth explicit Runge-Kutta Feldberg(ERKF(5)) and
Continuous explicit Runge-Kutta (CERK) method using shooting method to solve second
order boundary value problem which can be reduced to order one.These methods we shall
call them as shooting Continuous Explicit Runge-Kutta method, the results are computed
using matlab program.
Keyword: Boundary Value problems, Shooting Method ,Continuous Explicit Runge-
Kutta Method

324 | Mathematics
2013@‚b«@H3@ÖÜ»€a@I@26@Ü‹1a@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ÚÓ‘Ój�n€a@Î@Úœäñ€a@‚Ï‹»‹€@·rÓ:a@Âig@Ú‹©
Ibn Al-Haitham Jour. for Pure & Appl. Sci. Vol. 26 (3) 2013

Introduction
The boundary value problem with second order describe many phenomena in the applied
sciences, found in the theory of nonlinear diffusion generated by nonlinear sources, thermal
ignition of gases and concentration in chemical or biological problems.[1]
Consider second order linear two-point boundary value problem of the form: -[2]

y ′′( x) + f ( x) y ′ + g ( x) y ( x) = h( x) (1)

with the following boundary conditions:-

y (a) = α , y (b) = β

where a < x < b and f ( x), g ( x), h( x) are given continues functions.

The boundary value problem of second order have been studied by a number of researchers,
for example, Min and Salvador using spline collocation methods to solve fractional second
order Boundary Value Problems [3], while Guoju and Xiuli investigate the existence of
continues solutions for the second order boundary value problems with integral boundary
conditions[4].
In this paper, we generalize the Shooting Continuous Runge-Kutta method to solve the
boundary value problem of second order.

Explicit Runge-Kutta Method (ERK)[5,6,7,8]

Explicit Rung-Kutta method is one of the methods to solve initial value problems,Bucher
put on the basic rules of using the explicit Rung-Kutta method to solve ODE'S and DDE's,
which is completed by Jeltsch and Torrilhon.
The general form of s-stage explicit Runge-Kutta formula, to compute the numerical
solution of ordinary differential equations:-
y ′( x) = f ( x, y ) (2)
With initial value condition y ( x 0 ) = y0 at ( xn + hn ) is defined by:-
s
y n+1 = y n + hn ∑ bi ki (3)
i =1
i −1 i −1
where k i = f ( xn + ci hn , y n + hn ∑ aij k j ) , ci = ∑ aij (4)
j =1 j =1
for i=1,2,…,s and aij , bi and ci are the coefficients of Runge-Kutta formula and y n is
the associated approximation to y ( xn ) .

In this paper, we use one important type of Runge-Kutta method when aij = 0 , f (i ≤ j )
and c1 = 0 which is called Explicit Runge-Kutta method.
The interval [ x0 , x N ] can be divided into a so called mesh points x0 < x1 < ... < x N = x N and
generate a discrete approximation yi ≅ y ( xi ) for each associated distribution of these mesh
points is usually determined adaptively by the method in an attempt to deliver acceptable
accuracy at minimum cost.

325 | Mathematics
2013@‚b«@H3@ÖÜ»€a@I@26@Ü‹1a@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ÚÓ‘Ój�n€a@Î@Úœäñ€a@‚Ï‹»‹€@·rÓ:a@Âig@Ú‹©
Ibn Al-Haitham Jour. for Pure & Appl. Sci. Vol. 26 (3) 2013

These methods generally accomplish this objective by keeping N as small as possible subject
to constraint at an indirect measure of max y ( xi ) − yi would be kept small (relation to
i =1, 2,..., N
an accuracy parameter TOL).

Continuous Explicit Runge-Kutta Method (CERK)[9, 10,11]

The ERK methods can be developed to be called Continuous Explicit Rung-Kutta method
to solve ordinary differential equation (ODE) and Delay differential equation (DDE). The first
of research in this subject, Enright and Hayashi.

A Continuous Explicit Runge-Kutta (CERK) method provides an approximation to an

initial value problem. Such method may be obtained by appending additional stage to a
discrete method, or alternatively by solving the appropriate order conditions directly. So, for
some application, an approximation to the solution of the ODE:-
y ′ = f ( x, y ( x)) (5)
y ( x0 ) = y 0

on the interval [ x0 , x N ] is needed, to obtain dense, output for plotting the solution of a
standard ODE, to find the roots of the function associated with the solution of a standard
ODE.
Consequently, continuous (CERK) formula, which produces continuous approximation to the
solution of an ODE on [ x0 , x N ] has been developed Enright and have the form:-
s +1
z n ( xn + rhn ) = z n−1 ( xn ) + hn ∑ bi (r )k ;
i =1
ki = f ( xn + ci hni −1 , yi ) (6)
i −1
yi = z n−1 ( xn ) + hn ∑ aij k j
j =1
for i=1,2,…, s + 1 , and r ∈ [0,1] where b1 ( r ), b2 ( r ),..., bs +1 ( r ) are polynomial with respect
to r and z n−1 ( x0 ) = y0
This class of (CERK) formula is constructed by adding extra stage such that (s+1) becomes
( s + 1) with cs +1 = 1, a s +1, j = b j for j=1,2,…,s so that the approximation can easily be
developed as a c′ interpolate. Note that since k s +1 = f ( xn+1 , z n ( xn+1 )) will be the first
stage on the next attempted stops the actual cost per successful step is s derivative evaluation.

Shooting Method[12,13,14]
Arciniega and Allen applied a shooting method procedure to numerically solve linear
systems of stratonovich boundary value problems.
To illustrate the approach, we use the following TPBVP:-
y ′′( x) + f ( x) y ′ + g ( x) y ( x) = h( x)
y (a) = α , y (b) = β
(7)
The curve that represents y between these two points is desired, we anticipate that some such
curve, such as the dotted line, its slope and curve true are interrelated to y and x the diff. eq.
(7)

326 | Mathematics
2013@‚b«@H3@ÖÜ»€a@I@26@Ü‹1a@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ÚÓ‘Ój�n€a@Î@Úœäñ€a@‚Ï‹»‹€@·rÓ:a@Âig@Ú‹©
Ibn Al-Haitham Jour. for Pure & Appl. Sci. Vol. 26 (3) 2013

If we assume the slope of the curve at y(a), say y ′( a ) = G1 then we solve the equation as an
initial value problem using this assumed value, the result of this not the desired to y (b) = β ,
we assume anther value of y ′(a ) smaller for the slope, say y ′( a ) = G2 and repeat. After
these two trials, we linearly extrapolate for a third trial as follows:-
G2 − G1
Extrapolated estimate for Initial slope = G1 + * ( D − R1 )
R2 − R1
(8)
where R1=first result at endpoint (using G1)
R2=second result at endpoint (using G2)
D= the desired value t the endpoint
so the correct value of y(b) result.

Runge-Kutta Shooting Method for Solving TPBVP

Consider the following second order TPBVP equations:-
y ( x ) + f ( x ) y ′ + g ( x ) y ( x ) = h( x )
′′

Subject to the boundary conditions where f, g, h are function and α , β are constant:-

y (a) = α , y (b) = β
these equation changed into a first- order differential system:-
0 1   y1  0 
y′ =     + 
 g ( x) f ( x )   y 2   − h( x ) 
(9)
β −α
with assume the initial condition y1 ( a ) = α , y1′ (a ) = = G1
b−a
applying explicit and continuous Rung-Kutta steps, the result we compare the computed
value with the given conditions at the other boundary.
Repeat the solution with assume anther value of y1′ ( a ) = G2 and with extrapolated estimate
until agreement is attained at the other boundary.

Algorthim
Example (1):-
Consider the following two-point boundary value Problem:-
y ′′ − 4 y = 0
with boundary conditions:-
y(-1)=-1.1752 , y(1)=10.0179
while the exact solution is :-
y(x)  sinh(2x+1)
Table (1) presents the results of the , Shooting fifth Explicit Runge-Kutta (SERKF(5)) and
Shooting Continuous Explicit Runge-kutta method (SCERK) are used with differential value
of r .

Example (2):-
Consider the following two-point boundary value Problem:-

327 | Mathematics
2013@‚b«@H3@ÖÜ»€a@I@26@Ü‹1a@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ÚÓ‘Ój�n€a@Î@Úœäñ€a@‚Ï‹»‹€@·rÓ:a@Âig@Ú‹©
Ibn Al-Haitham Jour. for Pure & Appl. Sci. Vol. 26 (3) 2013

y ′′ = y − x 2 − x
with boundary conditions:-
y(0)=2 , y(1)=4
while the exact solution is :-
y ( x) = x 2 + x + 2 
Table (2) presents the results of the, Shooting fifth Explicit Runge-Kutta (SERKF(5)) and
Shooting Continuous Explicit Runge-kutta method (SCERK) are used with differential value
of r .

Conclusions
The obtains results show that the SCERK Method gives better accuracy than SERKF(5) of
the same order, since the error in SCERK method is decreasing when r ∈ [0,1) is increasing.

References
1. Wenjing,S and Wenjine,G.(2011) "Positive Solutions For a Second-Order System With
Integral Boundary Conditions", Electronic Journal of Differential Equations, 13, pp. 1–9.
2. Hikmet,C and Canan,A." The numerical solution of the singular two-point boundary value
problems by using non-polynomial Spline functions" Proceedings of the 9th WSEAS Int.
Conference on Applied Computer and Applied Computational Science.
3. Min,L and Salvador,J.(2009)"Solving Two-Point Boundary Value Problems of Fractional
Differential Equations by Spline Collocation Methods" www.cc.ac.cn/ research- report/
0903.pdf.
4. Guoju,Ye and Xiuli,Li.(2010) "Existence of Solutions of Second Order Boundary
Value Problems with Integral Boundary Conditions and Singularities" Hindawi Publishing
Corporation Journal of Inequalities and Applications Volume 2010, p.p.1-14,Article ID
807178,doi:10.1155.
5. Butcher,J','(1987) " The Numerical Analysis of ordinary differential equation", Jhon Wiley
& sons, New York.
6. Baker.C.T.H., Pual.C.A.H.,(1994) " Computing Stability regions of Rung-Kutta Methods
for DDE" IMA, J. of Num. and .No. 14,p.p. 347-362.
7. Baker,C and Pual,C.(1998) "Numerical Modeling by Retarded functional Differential
Equations" Numerical analysis Report No.335, October .
8. Rolf,J and Manuel,T.(2006) " Flexible Stability Domains for Explicit Rung-Kutta
Methods",Seminar for Applied Mathematics,8092
Zurich,Switzerland,email:{jeltsch,matorril}@math.ethz.ch
9. Enright,W.(1999)" Continues Numerical Method for ODE's with Defect Control",
University of Toronto, Canada, M533, GU, December.
10. Verner.J .H and Zennaro.M.(1995) " Continuous Explicit Rung-Kutta Methods of ORD
ER5",July,Vol.64,No.211.p.p 1123-1146.
11. Hayashi,H.(1996) "Numerical solution of Retarded and Neutral Delay Differential
Equations Using Continues Rung-Kutta Methods", PH.D.Thesis of Toronto University.
12. Armando,A and Edward,A.(2004) "Shooting Methods for Numerical Solution of
Stochastic Boundary-Value Problems" Stochastic Analysis and Applications. 22. 5, 1295–
1314.
13. Armando,A.(2007)" Shooting Methods for Numerical Solution of Nonlinear Stochastic
Boundary-Value Problems" Stochastic Analysis and Applications, 25: 187–200.
14. Curtis,F.(1984)"Applied Numerical Analysis", Third Edition, Addison-Wesley Publishing
Company,.

328 | Mathematics
2013@‚b«@H3@ÖÜ»€a@I@26@Ü‹1a@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ÚÓ‘Ój�n€a@Î@Úœäñ€a@‚Ï‹»‹€@·rÓ:a@Âig@Ú‹©
Ibn Al-Haitham Jour. for Pure & Appl. Sci. Vol. 26 (3) 2013

5. Anna,N.(2010)" Economical Rung-Kutta Methods with weak second order for Stochastic
differential equations", Int,J. contempt.Math.Sciences,Vol.5,No.24,1151-1160.

Table (1) results the solution

SERKF(5) SCERK
H x
r =1 r = 0 . 25 |error|* r = 0.5 |error|*
-0.5 0.0070096 0.0070096 0.0067226 0.0069 0.0067526 0.0069
0.1 0.5 3.6859483 3.6859483 3.6853062 0.0584 3.6854109 0.0583
1 10.1787000 10.1787000 10.1787000 0.0608 10.1787000 0.0608
-0.5 0.0070095 0.0070095 0.0067941 0.0069 0.0068918 0.0069
0.05 0.5 3.6859281 3.6859281 3.6853237 0.0584 3.6856672 0.0583
1 10.1787000 10.1787000 10.1787000 0.0608 10.1787000 0.0608

Table (2) results the solution

SCERK
h x SERKF(5)
r =1 r = 0.25 |error|* r = 0.5 |error|*
0.1 2.1100000 2.1100000 2.1093142 7.0000e-004 2.1098533 1.0000e-004
0.1 0.5 2.7500011 2.7500011 2.7477002 0.0023 2.7497009 3.0000e-004
1 4.0000000 4.0000000 4.0000000 0.0000 4.0000000 0.0000
0.1 2.1100000 2.1100000 2.1093142 7.0000e-004 2.1098533 1.0000e-004
0.05 0.5 2.7500011 2.7500011 2.7477002 0.0023 2.7497009 3.0000e-004
1 4.0000000 4.0000000 4.0000000 0.0000 4.0000000 0.0000
* Where |error| represent the difference between the exact solution and (SCERK) for
r=0.25, r=0.5.

329 | Mathematics
 ‫©‹‪2013@‚b«@H3@ÖÜ»€a@I@26@Ü‹1a@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ÚÓ‘Ój�n€a@Î@Úœäñ€a@‚Ï‹»‹€@·rÓ:a@Âig@Ú‬‬
‫‪Ibn Al-Haitham Jour. for Pure & Appl. Sci.‬‬ ‫‪Vol. 26 (3) 2013‬‬

‫ﺣﻞ ﻣﺴﺎﺋﻞ اﻟﻘﯿﻢ اﻟﺤﺪودﯾﺔ اﻟﺨﻄﯿﺔ ﺑﺎﺳﺘﺨﺪام طﺮﯾﻘﺔ ﺗﮭﺪﯾﻒ راﻧﺞ‪ -‬ﻛﺘﺎ اﻟﺼﺮﯾﺤﺔ‬
‫اﻟﻤﺴﺘﻤﺮة‬
‫ﻣﺪﯾﺤﺔ ﺷﻠﺘﺎغ ﯾﻮﺳﻒ‬
‫ﺑﺸﺮى ﻋﯿﺴﻰ ﻏﺸﯿﻢ‬
‫ﻓﺮع اﻟﺮﯾﺎﺿﯿﺎت ‪ /‬ﻗﺴﻢ اﻟﻌﻠﻮم اﻟﺘﻄﺒﯿﻘﯿﮫ ‪ /‬اﻟﺠﺎﻣﻌﮫ اﻟﺘﻜﻨﻠﻮﺟﯿﮫ‬

‫اﺳﺘﻠﻢ اﻟﺒﺤﺚ ﻓﻲ ‪ 20 :‬اﯾﺎر ‪ ، 2012‬ﻗﺒﻞ اﻟﺒﺤﺚ ﻓﻲ ‪ 8‬ﻧﯿﺴﺎن ‪2013‬‬

‫اﻟﺨﻼﺻﺔ‬
‫ﻓﻲ ھﺬا اﻟﺒﺤﺚ ﻗﻤﻨﺎ ﺑﺘﻌﻤﯿﻢ طﺮﯾﻘﺔ راﻧﺞ‪-‬ﻛﺘﺎ اﻟﺼﺮﯾﺤﺔ ﻣﻦ اﻟﺮﺗﺒﺔ اﻟﺨﺎﻣﺴﺔ وطﺮﯾﻘﺔ راﻧﺞ‪-‬ﻛﺘﺎ اﻟﺼﺮﯾﺤﺔ اﻟﻤﺴﺘﻤﺮة ﺑﺎﺳﺘﺨﺪام‬
‫طﺮﯾﻘﺔ اﻟﺘﮭﺪﯾﻒ ﻟﺤﻞ ﻣﺴﺎﺋﻞ اﻟﻘﯿﻢ اﻟﺤﺪودﯾﺔ ﻣﻦ اﻟﺮﺗﺒﺔ اﻟﺜﺎﻧﯿﺔ وذﻟﻚ ﺑﺘﺨﻔﯿﻀﮭﺎ اﻟﻰ اﻟﺮﺗﺒﺔ اﻻوﻟﻰ ھﺬه اﻟﻄﺮاﺋﻖ دﻣﺠﺖ ﺑﺄﺳﻢ‬
‫طﺮﯾﻘﺔ اﻟﺘﮭﺪﯾﻒ ﻟﺮاﻧﺞ‪-‬ﻛﺘﺎ اﻟﺼﺮﯾﺤﺔ اﻟﻤﺴﺘﻤﺮة‪ .‬وﻗﺪ أُ ُ◌ﺳﺘﺨﺪم ﺑﺮﻧﺎﻣﺞ اﻟﻤﺎﺗﻼب ﻟﺤﺴﺎب اﻟﻨﺘﺎﺋﺞ‪.‬‬
‫اﻟﻜﻠﻤﺎت اﻟﻤﻔﺘﺎﺣﯿﺔ‪-:‬ﻣﺴﺎﺋﻞ اﻟﻘﯿﻢ اﻟﺤﺪودﯾﺔ‪،‬طﺮﯾﻘﺔ اﻟﺘﮭﺪﯾﻒ‪،‬طﺮﯾﻘﺔ راﻧﺞ‪-‬ﻛﺘﺎ اﻟﺼﺮﯾﺤﺔ اﻟﻤﺴﺘﻤﺮة‪.‬‬

‫‪330 | Mathematics‬‬

‫‪View publication stats‬‬