Beruflich Dokumente
Kultur Dokumente
www.elsevier.com/locate/amc
Abstract
In this paper, we give a more general definition of integral equations of the second
and first kind which are called functional integral equations. Fredholm and Volterra
integral equations are special type of these equations. We derive an expansion method
to treat these equations. Finally, we introduce numerical examples and conclusion.
Ó 2003 Elsevier Inc. All rights reserved.
Keywords: Functional integral equations; Lagrange interpolation; Volterra and Fredholm integral
1. Introduction
In this paper, we introduce new definition for integral equations such that
Fredholm or Volterra integral equations are special type [1–3]. Fredholm
functional integral equation of the second kind is given by
Z b
yðxÞ þ pðxÞyðhðxÞÞ þ k kðx; tÞyðtÞ dt ¼ gðxÞ; a 6 x 6 b: ð1:1Þ
a
Also, Volterra functional integral equation of the second kind is given by
Z x
yðxÞ þ pðxÞyðhðxÞÞ þ k kðx; tÞyðtÞ dt ¼ gðxÞ; a 6 x 6 b: ð1:2Þ
a
0096-3003/$ - see front matter Ó 2003 Elsevier Inc. All rights reserved.
doi:10.1016/j.amc.2003.08.003
508 M.T. Rashed / Appl. Math. Comput. 156 (2004) 507–512
2. Suggested method
X
n
yn ðhðxÞÞ ¼ ‘n ðhðxÞ; xj Þyðxj Þ ð2:2Þ
j¼0
and
X
n
kðx; tÞyðtÞ ¼ ‘n ðt; xj Þkðx; xj Þyðxj Þ; ð2:3Þ
j¼0
where
Y x xk
‘n ðx; xj Þ ¼ :
x xk
k¼0 j
k6¼j
0 6 i 6 n; ð2:4Þ
X
n
yðxj Þf‘n ðxi ; xj Þ þ pðxi Þ‘n ðhðxi Þ; xj Þ þ kkðxi ; xj ÞBðxi ; xj Þg ¼ gðxÞ;
j¼0
0 6 i 6 n; ð2:5Þ
X
n
yðxj Þfpðzi Þ‘n ðhðzi Þ; xj Þ þ kðzi ; xj ÞBðxn ; xj Þg ¼ gðzi Þ; 0 6 i 6 n; ð2:6Þ
j¼0
M.T. Rashed / Appl. Math. Comput. 156 (2004) 507–512 509
X
n
yðxj Þfpðzi Þ‘n ðhðzi Þ; xj Þ þ kðzi ; xj ÞBðzi ; xj Þg ¼ gðzi Þ; 0 6 i 6 n; ð2:7Þ
j¼0
where
iðb aÞ
xi ¼ a þ ; i ¼ 0ð1Þn
n
and
ðb aÞ ði 0:45Þp b þ a
zi ¼ cos þ ; i ¼ 0ð1Þn:
2 n 2
The points zi , i ¼ 0ð1Þn in (2.6) and (2.7) are very useful in making a system of
equations stable, where
Z xi Z xi Y
n
ðt xk Þ
Bðxi ; xj Þ ¼ ‘n ðt; xj Þ dt ¼ ¼ dt
a a ðx j xk Þ
k¼0
k6¼j
Z Y xi a
xi a 1 n
2
t þ xi þa
2
xk
¼ dt
2 1 xj xk
k¼0
k6¼j
kp
xi a X
n Y
n xi a
2
cos n
þ xi a
2
xk
¼ Bðtk Þ ;
2 k¼0 xj xk
k¼0
k6¼j
i ¼ 0ð1Þn; j ¼ 0ð1Þn;
00 1
2dk X n
skp ; k ¼ 0; n;
Bðtk Þ ¼ Is cos ; dk ¼ 2
n s¼0 n 1; 0 < k < n;
8
< 2; s ¼ 0;
Is ¼ 0; s is odd;
: 2
1s 2 ; s is even:
A summation symbol with double primes denotes a sum with first and last
terms halved.
3. Numerical examples
ex x2
yðxÞ ¼ x; gðxÞ ¼ ex hðxÞ ðe 1Þ;
2
Z 1
pðxÞyðhðxÞÞ þ kðx; tÞyðtÞ dt ¼ gðxÞ; 0 6 x 6 1; ð3:4Þ
0
where
en ðxj Þ ¼ yexact ðxj Þ yn ðxj Þ
yexact is the exact solution of (1.1) and yn ðxj Þ is the approximated solution of
(1.1) functional integral equations with using the suggested method.
In Table 1 we give the computed errors rn for example (3.1) which are given
by the suggested method for different hðxÞ and n:
x2
ð1Þ hðxÞ ¼ lnðx þ 0:5Þ; ð2Þ hðxÞ ¼ ;
2
ð3Þ hðxÞ ¼ xex ; ð4Þ hðxÞ ¼ 0:8x:
M.T. Rashed / Appl. Math. Comput. 156 (2004) 507–512 511
Table 1
Computed errors rn for example (3.3)
n hðxÞ
(1) (2) (3) (4)
2 5.2D)02 7.2D)01 7.9D)01 6.1D)01
3 3.8D)02 2.8D)01 2.9D)01 3.1D)01
4 5.7D)03 8.5D)04 4.2D)02 9.2D)02
5 5.1D)04 4.2D)02 8.5D)03 3.2D)02
6 1.6D)04 4.5D)02 1.6D)03 1.7D)02
7 7.2D)06 2.5D)03 2.1D)04 4.6D)03
8 2.0D)06 6.4D)04 2.4D)05 1.0D)03
9 1.7D)07 1.8D)04 2.4D)06 7.6D)04
10 4.9D)09 7.6D)06 7.0D)08 1.1D)04
Table 2
Computed errors rn for example (3.4)
n hðxÞ
(1) (2) (3) (4)
2 2.9D)02 5.8D)03 2.2D)03 3.9D)03
3 2.5D)03 1.9D)04 1.3D)04 3.9D)04
4 2.2D)04 6.2D)05 7.0D)06 2.7D)05
5 2.3D)05 5.6D)06 2.3D)07 1.2D)06
6 2.4D)06 1.5D)07 1.1D)08 3.1D)08
7 2.9D)07 5.8D)09 3.3D)10 6.2D)10
8 2.6D)08 1.8D)10 8.3D)12 4.3D)11
9 3.1D)09 5.9D)12 3.1D)13 1.0D)12
10 3.2D)10 2.3D)13 7.8D)15 1.7D)14
Table 3
Computed errors rn for examples (3.3) and (3.4)
n pðxÞ
ex 0.0 ex 0.0
2 2.8D)02 4.4D)02 4.7D)02 1.1D)01
3 1.5D)02 2.1D)02 9.3D)03 1.1D)02
4 1.6D)03 1.8D)03 1.2D)03 7.9D)04
5 2.9D)04 3.5D)04 1.3D)04 4.9D)05
6 7.1D)05 4.7D)05 9.6D)06 2.7D)06
7 1.4D)05 5.6D)06 4.9D)07 1.2D)07
8 5.3D)06 9.5D)07 1.9D)08 5.1D)09
9 5.9D)07 5.8D)08 1.0D)09 2.0D)10
10 5.1D)07 1.3D)08 1.3D)09 6.4D)11
Table 2 contains the computed errors rn for example (3.2) which are given by
the suggested method for different hðxÞ and n.
512 M.T. Rashed / Appl. Math. Comput. 156 (2004) 507–512
The computed errors rn for examples (3.3) and (3.4) are given in Table 3
with using hðxÞ ¼ lnðx þ 0:5Þ and pðxÞ
4. Conclusion
References
[1] C.T.H. Baker, The Numerical Treatment of Integral Equations, Oxford University Press,
Oxford, 1977.
[2] L.M. Delves, J.L. Mohamed, Computational Methods for Integral Equations, Cambridge
University Press, Cambridge, 1985.
[3] L.M. Delves, J. Walsh, Numerical Solution of Integral Equations, Oxford University Press,
Oxford, 1974.