I nternational J ournal of Mathematics and Statistics I nvention (I J MSI )

E-ISSN: 2321 4767 P-ISSN: 2321 - 4759

www.ijmsi.org Volume 2 Issue 1 || January. 2014 || PP-49-55
www.ijmsi.org 49 | P a g e
Variational Iteration and Homotopy Perturbation Methods for
Solving Fredholm-Volterra Integro-Differential Equations
1,
O.A.Taiwo,
2,
A. K. ADIO
1,
Department of Mathematics; University of Ilorin, Ilorin, Nigeria.
2,
Department of Basic Sciences; Babcock University, Illisan, Nigeria.


ABSTRACT: In this article, Hes variational iteration and Homotopy Perturbation methods are modified by
adding a perturbation term to the resulting system of first integro-differential equation obtained after reducing
higher order Fredholm-Volterra integro-differential equations.
The resulting revealed that both methods, are very effective and simple. We also observed that the higher the
values of n (the degree of approximant), the closer the approximate solutions obtained to the exact solutions.
Numerical examples are given to illustrate the applications of the methods.

KEYWORDS: Integro-differential equations, Variational Iteration and Homotopy Methods, Perturbed, Error.

1 INTRODUCTION
Mathematical modeling of real life problems usually results in functional equations, like ordinary or
partial differential equations, integral and integro-differential equations, stochastic equations.

Many Mathematical formulation of physical phenomena contain integro-differential equations, these
equations arise in many fields like Fluid dynamics, Biological models and chemical kinetics.
Integro-differential equations are usually difficult to solve analytically, so it is required to obtain an efficient
approximate solution (Sweilam, 2007).

Variational iteration method (He, 1997, 199, 2007) is a powerful device for solving various kinds of
equations, linear and nonlinear. The method has successfully been applied to many situations. for example, He
(2007) used the method to solve some integro differential equations where he chose initial approximate solution
in the form of exact solution with unknown constants.

Abbasbandy and Shivanion (2009) used VIM to solve systems of nonlinear Volterras Integro-
differential equations. Biazar et al (2010) employed VIM to solve linear and nonlinear system of IDEs.
A new perturbation method called Homotopy Perturbation Method (HPM) was proposed by He in 1997 and
description in 2000, which is, in fact, coupling of the traditional perturbation method and homotopy in topology.

The method has equally been applied successfully to many situations. For example, Mirzaei (2011)
employed HPM and VIM to solve Volterra integral equations.
Biazar and Eslami (2010) applied HPM to solve nonlinear Volterra-Fredholm integro-differential equations.

In this paper, we considered the Tau reduction of nth order Fredholm-Volterra integro-differential
equation into systems of first order Fredholm-Volterra integro differential equations. The resulting systems are
then perturbed and solved by Variational Iteration and Homotopy Perturbation Methods. The basic motivation
here is to get a better approximation.

In case of nonlinear Fredholm-Volterra Integro differential equations; Newtons linearization scheme
of appropriate order is used to linearize and hence leads to iterative procedure.
2. VARIATIONAL ITERATION METHOD:
Variational Iteration Method (VIM) is based on the general Langrangess multiplier method (Inokuti et
al, 1978). The main feature of the method is that the solution of a Mathematical problem with linearization
assumption is used as initial approximation. Then a more highly precise approximation at some special point can
be obtained.

Variational Iteration and Homotopy Perturbation Methods for
www.ijmsi.org 50 | P a g e
To illustrate the basic concepts of VIM, we consider the following nonlinear differential equation
) ( = x g Nu Lu + (1)

where L is a linear operator, N is a nonlinear operator and ) (x g is an inhomogeneous term.



According to VIM (He 1999, 2000, 2006), we can construct a correction functional as follows

{ }dT T g T u N T Lu x u x u
n n
x
n n
) ( ) ( ) ( ) ( = ) (
1
+ +
}
+

(2)
is a general lagrangian multiplier (Inokuti et al, 1978) which can be identified optimally via Variational
theory. Subscript n denotes the nth-order approximation,
n
u is considered as a restricted variation (He, 1999,
2000) i.e 0 =
n
u o
3. Homotopy Perturbation Method:

Consider the nonlinear algebraic equation
R x x f e 0, = ) ( (3)
The basic idea of the homotopy perturbation method is to construct a homotopy R R [0,1] : ) , ( p v H
which satisfies
[0,1] ; 0, = )) ( ) ( )( (1 ) ( = ) , ( e e + p R v x f v f p v pf p v H

(4)
or

[0,1] ; 0, = ) ( ) ( ) ( = ) , ( e e + p R v x pf x f v f p v H


where p is an embedding parameter, and

x is an initial approximation of equation (3). (Usually

x is an
initial guess close to o ).
Obviously from equation (4), we have

0, = ) ( ) ( = ,0) (

x f v f v H
(5)

0, = ) ( = ,1) (
and
v f v H
(6)
The embedding parameter p increases from 0 to 1 monotonically as trivial problem
0 = ) ( ) ( = ,0) (

x f v f v H is continuously transformed to the original problem 0 = ) ( = ,1) ( v f v H .
The HPM uses the embedding parameter p as a "small parameter and writes the solution of equation (4) as a
power series of p i.e

+ + +
2
2 1
~ ~
= p x p x x v (7)
setting 1 = p results in the approximate solution of equation (3)

+ + +

2 1
1
~ ~
=
lim
= x x x x
p
(8)
where

,
~
,
~
,
2 1
x x x are the coefficients of the power series (7).
If ) ( = p v v in equation (7), then (0), =
~
(0), =
1
v x v x '

(0),
!
1
=
~
, (0),
2!
1
=
~ ) (
2
n
n
v
n
x v x ' ' (9)
and (0)
) (n
v can be determined from ) ( , ), ( ), (
) (

x f x f x f
n
' by equation (4) or by the equation
0 = ) ( ) ( )) ( (

x pf x f p v F +
recursively.

Variational Iteration and Homotopy Perturbation Methods for
www.ijmsi.org 51 | P a g e
4. Solution Techniques:

Consider the nonlinear Fredholm-Volterra Integro differential equation of the form
) ( = ) ( ) , ( ) ( ) , ( ) (
2
1 =
2
2
1 =
1
0 =
x g dt t y t x k dt t y t x f x y p
j
j
j
x
a
i
i
i
b
a
i
i
m
i

}

+ + (10)
unde the mixed conditions
| | 1 1 , 1 , 0,1, = , = ) ( (1) 1) (
1
0 =
s s + +

c m i i c y c y b y a
j
ij
j
ij j ij
m
j
(11)


where ) (x y is an unknown function, the functions ) , ( ), ( ), ( t x f x p x g
i i
and ) , ( t x k
j
are defined on the
interval b t x a s s , and
i ij ij ij
c b a , , , , ,
2 1
are constants.
Now the general nth order Fredholm-Volterra integro differential equation is of the form :

) 12 ( , , ) ( = ) ( ) , (
) ( ) , ( ) ( ) ( ) ( ) (
2
1 =
2
2
1 =
1 2 1
b t x a x g dt t y t x k
dt t y t x f x y p x y p x y p x y p
j
j
j
x
a
i
i
i
b
a
m
m
s s +
+ + + ' ' + ' +

under the mixed conditions stated in equation (11) using the transformation
) ( = ) ( = ) ( ) ( = ) (
2 1 1
x y x y x y x y x y ' '

) ( = ) ( = ) ( ) ( = ) (
3 2 2
x y x y x y x y x y ' ' ' ' (13)



) ( = ) ( = ) ( ) ( = ) (
1
1
x y x y x y x y x y
m m
m
m
m
+

'
Equation (13)is written as a system of differential equations as
) ( =
2
1
x y
dx
dy


) ( =
3
2
x y
dx
dy
(14)



) ( = ) ( ) , ( ) ( ) , ( ) ( = ) (
2
1 =
2
2
1 =
1
0 =
x g dt t y t x k dt t y t x f x g x y p
dx
d
j
j
j
x
a
i
i
i
b
a
i
i
m
i

}

Thus, equation (14) is expressible in matrix form as


|
|
|
|
|
|
.
|

\
|
+
|
|
|
|
|
|
.
|

\
|
|
|
|
|
|
|
.
|

\
|

|
|
|
|
|
|
.
|

\
|

) (
0
0
0
0 0 0 1 0 0
0 0 0 0 1 0
0 0 0 0 0 1
=
3
2
1
1 5 4 3 2 1
3
2
1
x f y
y
y
y
p p p p p p y
y
y
y
m m m

(15)
where
dt t y t x k dt t y t x f y p x g x f
j
j
j
x
a
i
i
i
b
a
) ( ) , ( ) ( ) , ( ) ( = ) (
2
1 =
2
2
1 =
1 1
}

}

Variational Iteration and Homotopy Perturbation Methods for
www.ijmsi.org 52 | P a g e
In view of VIM, a correction functional is of the form
{ } c c c d f y y y
j n j
x
n j n j
' + '
}
+
) ( ) ( =
, , 1 ,

(16)
where

1 2 1 1 2 1
) ( = y p y p y p y p x g y
j j j j j
'




N
j
j
j
x
a
i
i
i
b
a
T dt t y t x k dt t y t x f
1
2
1 =
2
2
1 =
1
) ( ) , ( ) ( ) , ( t +

}

}
(4.8)
m j , 1,2, =

j
f is considered as a restricted variation i.e 0 =
j
f o ,
1
t is a constant parameter to be determined and ) (x T
N

is a chebyshev polynomial of degree 1,2) (= N valid in the interval 1 1 s s x given by
)
cos
( cos = ) (
1
x N x T
N

(17)
which satisfy the recurrence relation given by
1 ), ( ) ( 2 = ) (
1
>

N x T x xT x T
N N N

For nonlinear (FVIDE) (10) and (11), in order to solve the equations by Tau-variational iteration method,
equation (10) is linearized using Newtons linearization scheme of the form

) 18 ( ) , ( ) , ( ) ( ) (
) (
2
1 =
2
) (
2
1 =
1
0 =
(

c
c
A +
(

c
c
A +

}

t x K
y
G
y dt t x F
y
G
y x y x p
j
j
n
j
n
j
b
a
j
j
n
j
n
j
b
a
k
k
m
k


where

) ( ) (
1
) (
=
j
n
j
n
j
n
y y y V
+
(19)

The basic idea of Tau-HPM is the addition of perturbation terms to equation (10) and then writing the
equation as a system of IDEs using the transformation in (13). Thus, we have a system of equations of the form
) ( = ) (
) (
2 1 1
dt t y p x y
k
z
}
+

o

dt t y p x y
k
z
) ( = ) (
) (
3 2 2
}
+

o (20)



{ }dt t y p t y p t y p t y p t g p x y
m m m m
z
m m
) ( ) ( ) ( ) ( ) ( = ) (
1 2 1 1 2 1

+

}
o

dt t H p dw t y t w k p dw t y t w f p
n
x
j
j
j
w
a
x
i
i
j
b
a
x
) ( ) ( ) , ( ) ( ) , (
2
1 =
2
2
1 =
1
}

} }

} }
+
)
`

)
`

>
+
n x T x H
k n k
m
k
n
), ( = ) (
2
1 =
t (21)
is the perturbation term.
Thus, by HPM, equation (20) is written in expansion form and coefficients of like powers of
0) ( > i p
i
are compared to give the values of the constants s b s a
i i ' '
, and s c
i'
. The unknown function
function ) (x y is expressed as summation of s a
i
'
'
i.e
Variational Iteration and Homotopy Perturbation Methods for
www.ijmsi.org 53 | P a g e

+ + +

2 1
0 =
1
= = ) ( = ) ( a a a a x y x y
i
i
(22)

using the initial/boundary conditions, the unknown constants can be evaluated, the values of which are then
substituted back into the approximate solutions earlier obtained.

5. Applications

We illustrate the ability and reliability of the methods with the following examples.

Example 1:
Consider the Volterra-Fredholm integro differential equations (Biazer, J and Eslami, M. (2010)).
dt t xty dt t y t x x f x xy x y x x y
x
) ( )) ( )( 2 ( ) ( = ) ( ) ( ) (
1
1
2
1
} }

+ + + ' ' '
where

3
5
15
23
2
3
1
25
2
= ) (
2 3 4 6
+ + x x x x x x f

with condition 0 = (0) 1, = (0) y y ' and exact solution 1 = ) (
2
x x y .
Now using the transformation in equation (4.4), we have

|
|
.
|

\
|
+ +
+
|
|
.
|

\
|
|
|
.
|

\
|

|
|
.
|

\
|
} }

dt t xty dt t y t x x f
y
y
x x y
y
x
) ( ) ( ) 2 ( ) (
0
1 0
=
1
1
2
1
2
1
2
1

and

2
2, 1,
3 4 5 7
2, 1 2,
15
23
2
1
3
2
4
1
15
1
175
2
3
5
= ) ( x y y x x x x x y x y
n n n n
|
.
|

\
|
+ + + + +
+


1) (2 1) (
2
1
1) (
1
2 2
1,
2 2
1,
+ + + x x x y x x y
n n
t
Example 2:

dt t y e x x e x xy x y x x y
t x
) ( 2 ) sin (1 = ) ( sin ) ( ) (
2 2
1
1

}
+ + + ' ' ' ' '

with conditions
1 = (0) = (0) = (0) y y y ' ' '
Exact solution for the problem is
x
e x y = ) (

The Newtons linearization scheme of order three is given as
0 =
k
k
k
k
k
k
k
k
y
y
G
y
y
G
y
y
G
y
y
G
G ' ' ' V
' ' ' c
c
+ ' ' V
' ' c
c
+ ' V
' c
c
+ V
c
c
+
where
i
k
i
k
i
k
y y y V
+1
=

and
dt t y e x x e x xy x y x x y G
t x
) ( 2 ) sin (1 ) ( sin ) ( ) ( =
2 2
1
1

}
+ + + ' ' ' ' '




Thus,
dt y y e x x e dt t y e y y x xy x y x x y
t x t
) 2 (1 2 ) sin (1 = ) ( 2 ) ( sin ) ( ) (
2 2
1
1
2 2
1
1
' + + + ' + ' ' ' ' '

} }
Variational Iteration and Homotopy Perturbation Methods for
www.ijmsi.org 54 | P a g e
1) 8 (8 ) 3 (4 1) (2
2 4 3
2
2
1
+ + + + x x x x x t t t

Thus following the procedure and comparing the coefficients of like powers of 0) ( > i P
i
, we have
Coefficients of

p : 1 =

a
1 =

b
1 =

c
Coefficients of
4
p :
2
2 4
= x c a
x c b
3 4
=
y y x c x x c c x c ' +
3
1
3
1 3
2
4
6 7.25372081 sin = and

{ ) sin (1 6 7.25372081 sin ) ( 1 = ) (
3
1
2 4 2 2
x x xe y y x c x x x x x x x x y
x
+ + ' + + + + +

) 8 (8 ) 3 (4 ) (2 2 ) 2 (1 6 7.25372081
3 5
3
2 4
2
3
1
2
x x x x x x x x y y + + + + ' + t t t

}
4 4
6 7.25372081 sin x y y x x ' +

Remark : The values of Tau-parameters are determined in each example using the conditions given

Table 1: Results obtained for example 1 and error
x Exact Solution Approx. Solution Error
-1 0.000000 0.000041 4.00E-6
-0.8 -3.600000 -0.359997 3.00E-6
-0.6 -0.640000 -0.639902 9.80E-6
-0.4 -0.840000 -0.839004 9.96E-4
-0.2 -0.960000 -0.959793 2.07E-4
0.0 -1.000000 -0.999782 2.18E-4
0.2 -0.960000 -0.959793 2.07E-4
0.4 -0.840000 -0.839004 9.96E-4
0.6 -0.640000 -0.639902 9.80E-5
0.8 -0.360000 -0.359997 3.00E-6
1.0 0.000000 0.000004 4.00E-6

Table 2: Results obtained for example 2 and error
x Exact Solution Approx. Solution Error
-1 0.3678794412 0.3678735712 5.87E-6
-0.8 0.4493289641 0.4493223041 6.66E-6
-0.6 0.5448811636 0.5448805126 6.51E-6
-0.4 0.6703200460 0.6903186060 1.44E-6
-0.2 0.8187307531 0.8187296431 1.11E-6
0.0 1.0000000000 0.9999968900 3.12E-6
0.2 1.2214027580 1.2213625580 4.02E-5
0.4 1.4918246980 1.4917423980 8.23E-5
0.6 1.8221188000 1.8221129000 7.51E-5
0.8 2.2255409280 2.2255084280 3.25E-5
1.0 2.7182818280 2.7182716280 1.02E-5

6. CONCLUSION
In this paper, Tau-VIM and Tau-HPM have been successfully applied to find solutions of Fredholm-
Volterra integro differential equations.

The solutions are expressed as polynomials and the errors are also obtained. Both methods performed
creditably well for the examples considered. It is observed that the higher the values of n (the degree of
approximant), the closer the approximate solutions obtained to the exact solutions.

Variational Iteration and Homotopy Perturbation Methods for
www.ijmsi.org 55 | P a g e
REFERENCES
Method Iteration l Variationa the of cation E..``Appli Shivamoin, & S., , Abbasbandy [1]
' ' calandCom .Mathemati s equation al Differenti - ntegro VolterraI Nonlinear of System for
158. 147 14(2), (2009), ns, Applicatio putational
Nonlinear the of Solutions ical I..``Numer shim, M.S.M.,&Ha Noorani, B., Batiha, [2]
1(1). (2008), Compt., problems n .Int.J.Ope n Equatio al Differenti - Integro ' '
dfor ationMetho tionalIter H..``Varia himi, M.G.,&Ebra i, Porshokouh J., Biazar, [3]
forum, matical .Int.Mathe n Equatio al Differenti - Integro of Systems Nonlinear and Linear ' '
. 3191 4),3185 (2010),5(6
integro Fredholm Volterra Nonlinear for Solutions M..``Exact Eslami, J.,& Biazar, [4]
289 285 9(3), (2010), Sciences, Nonlinear .Int.J.of M HP s e H by Equation al Differenti ' ' '
Value Boundary initial order Eight for Method Iteration iational J.H..``Var He, [5]
682 680 (2007),76, .Phys.Sc., m Proble ' '
e onTechniqu Perturbati a and echnique Homotopy t a of Method coupling J.H..``A He, [6]
43. 37 (2000),35, Mechanics. Nonlinear .Int.J.of s problem Nonlinear for ' '
Technique, Analytical Nonlinear of kind A : ethod IterationM iational J.H..``Var He, [7]
708. 4),699 (1999),34( Mech, inear Int.J.Nonl , s example Some ' '
Nonlinear in Multiplier Lagrange the of use l T.``Genera H.,&Mura, M.Sekine, Inokuti, [8]
anics in theMech Method l Variationa ed., Nassed, In.S.Nemat Physics al Mathematic
162 (1978),156 NewYork, Press, .Pergamon d Soli of ' '
Method Iteration l Variationa and Method on Perturbati otopy S.M..``Hom Mirzaei, [9]
(2011) tics, Bioinforma and Maths Applied of .Journal Equations" ntegral VolterraI for
113 1(1),105
l Variationa using Equations al Differenti Integro Order rth N.H..``Fou Sweilam, [10]
1091. 8),1086 7 (2007),54( pl., MathwithAp & mp Method".Co Iteration