Malaya Journal of Matematik, Vol. 9, No.

1, 701-708, 2021

[Link]

A study of variational iteration method for solving

various types of problems
E.Rama1 * , K.Somaiah2 and K.Sambaiah3

Abstract
This paper presents the application of He’s Variational Iteration Method for various types of problems. This
method is applied to find successive approximate solutions of first order differential equation with single condition,
second order differential equations with two conditions, an isoperimetric problem and Volterra integral equations
of second kind. It was shown that they are converging to their exact solutions. The successive approximations
and exact solution are shown graphically.
Keywords
Successive Approximate Solution, Variational Iteration Method
AMS Subject Classification
65F10.
1 Department of Mathematics, University College of Science, Saifabad, Osmania University, Hyderabad, Telangana State, India.
2 Department of Mathematics, Kakatiya University, Vidyaranyapuri, Warangal, Telangana State, India.
3 Flat No.101, Riddhi’s Pebble, Madhapur, Hyderabad, Telangana State, India.

*Corresponding author: 1 ramamathsou@[Link]

Article History: Received 14 January 2021; Accepted 27 February 2021 c 2021 MJM.

Contents ferential and partial differential equations. There are several

methods to get the approximate solutions of such equations.
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701 Adomain decomposition method, pertubation method, numer-
2 Variational Iterative Method . . . . . . . . . . . . . . . . . . . . . . . 701 ical methods can be applied to get approximate solution of
differential and partial differential equation. Numerous re-
3 Variational Iterative method for obtaining the solu-
searchers [4–7] are using variation iteration method to find the
tion of first order differential equation with single given
condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 702 solutions of such problems. The variational iteration method
is easy to compute the successive approximations. In a fewer
4 Variational Iterative method for obtaining the solu- iterations we get exact solutions of problems if solution of
tion of second order differential equations with two
those problems exist. In this paper, we find the solutions
given conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 703
of first order differential equation with one given condition,
5 Variational Iterative method for the solution of Volterra second order differential equation with two given conditions,
Integral equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 705 isoperimetric problem and Volterra integral equation using the
6 Variational Iteration Method for solving isoperimetric variational iteration method. Further it is observed that the
problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707 successive approximations converges to the exact solution in
7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708 two iterations in some problems.
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708
2. Variational Iterative Method
1. Introduction We briefly explain the concept of the Variational Iterative
Method to find the solution of the equations of the type
J.H. He [1–3] has initiated the Variational Iteration Method.
Usually modeling problems yield either differential or par- L[y(x)] + N[y(x)] = g(x) (2.1)
tial differential equations. For their physical interpretation
we need their solutions. The methods given in the theory of where L is linear operator, N is a non-linear operator and g(x)
differential equations may fail to give solutions of certain dif- is known continuous function. This method is introduced by
 A study of variational iteration method for solving various types of problems
— 702/708

[Link] [8] and the solutions of the problems are approximated Simplifying we get
by a set of functions that may contain unknown constants Z x
which are to be determined either from initial conditions or y1 (x) = 1 − {0 + 1 − 2et }dt = 2ex − 1 − x.
boundary values. He introduced two terms namely restricted 0
variation and correction functional. The correction functional Substituting n = 1 in equation (3.1), we get
of (2.1) is given by
Z x
Z x y2 (x) = y1 (x) − {y01 (t) + y1 (t) − 2et }dt
yn+1 (x) = yn (x) + λ (t, x)[L[yn (t)] + N[yn (t)] − g(t)]dt Z x 0
x0
(2.2) = y1 (x) − {2et − 1 + 2et − 1 − t − 2et }dt
0

where λ (t, x) is a general Lagrange Multiplier which can be x2

= 1+x+ .
computed using variational theory. Let λ f (t, x) be the com- 2
puted Lagrange multiplier. Further yn (x) is the nth approx- Similarly, we get
imate solution and y˜n (x) denotes restricted variation, that it
δ ˜yn (x) = 0. The successive approximations yn (x) are com- x2 x3 x2 x3 x4
y3 = 2ex − (1 + x + + ), y4 = 1 + x + + + ,
puted by the variation iteration formula 2! 3! 2! 3! 4!
x 2 x 3 x4 x5
y5 = 2ex − (1 + x + + + + ),
Z x
yn+1 (x) = yn (x) + λ f (t, x){L[yn (t)] + N[yn (t)] −g(t)}dt 2! 3! 4! 5!
x0 x2 x3 x4 x5 x6
(2.3) y6 = 1 + x + + + + +
2! 3! 4! 5! 6!
and properly chosen initial approximation y0 (x). and so on.
The solution of (2.1) is given by
The exact solution and approximate solutions of IVP are
y(x) = lim yn (x) shown in the Fig.1. It is observed that the even numbered
n→∞
approximations are below the exact solution curve and odd
Ji-Huan He and XuHong Wu [9] given that the Lagrange
numbered approximations are above of it. Further it is ob-
multiplier is
served that the subsequence y2 , y4 , y6 , y8 , ....of {yn } converge
(−1)k to exact solution ex and another subsequence y1 , y3 , y5 , ....is
λ f (t, x) = (t − x)k−1 (2.4) also converge to exact solution y = ex .
(k − 1)!
for the differential equation
y(k) + f (y, y0 , y00 , ...., y(k−1) ) = 0. (2.5)

3. Variational Iterative method for

obtaining the solution of first order
differential equation with single given
condition
Example 3.1. Suppose y0 (x)+y = 2ex and the given condition
is y(0) = 1.
Here Ly = y0 , Ny = y and g(x) = 2ex . Order of linear
operator is one and given equation is in the form of equation
(2.1). For the problem which is under consideration, the
(n + 1)th approximation is given by
Z x
yn+1 (x) = yn (x) − {y0n (t) + yn (t) − 2et }dt (3.1)
0

Choose y0 (x) = y(0). Then y0 (x) = 1 since y(0) = 1. Substi-

tuting n = 0 in equation (3.1) we get Fig.1
Z x
y1 (x) = y0 (x) − {y00 (t) + y0 (t) − 2et }dt
0

702
 A study of variational iteration method for solving various types of problems
— 703/708

2k+1
Example 3.2. Suppose y0 (x) − 2x = cos x and given condition x
and nth approximation is given by yn (x) = − ∑nk=1 (2k+1)!
is y(0) = 0. −x x
Further y(x) = limn→∞ yn (x) = e 2 − e2 + x which is exact
The exact solution of this first order differential equation is solution of the problem considered. Curves representing exact
y = x2 + sin x. The variation iteration formula for the problem solution, first, second, third and fourth approximations are
considered is shown graphically in Fig.2 to Fig.5 . It is observed that higher
Z x numbered approximations are very very close to the curve
yn+1 (x) = yn (x) − {y0n (t) − 2t − cost}dt (3.2) representing the exact solution of the problem.
0

Substituting n = 0 in (3.2) we obtain

Z x
y1 (x) = y0 (x) − {y00 (t) − 2t − cost}dt (3.3)
0

Choose y0 (x) = 0. Then the equation (3.3) reduces to

Z x
y1 (x) = − {−2t − cost}dt = x2 + sin x (3.4)
0

Substituting n = 1 in equation (3.2) we get

Z x
y2 (x) = y1 (x) − {y01 (t) − 2t − cost}dt
0
Z x
= y1 (x) − {2t + cost − 2t − cost}dt = y1 (x)
0

Therefore, yk (x) = x2 + sin x for k = 1, 2, 3, ......

We can observe that in the first iteration itself we got the exact
solution.
Fig.2
4. Variational Iterative method for
obtaining the solution of second order
differential equations with two given
conditions
Example 4.1. Suppose the second order differential equation
is y00 − y + x = 0 and the given two conditions are y(0) = 0
and y0 (0) = 0.
−x x
The exact solution of it is y(x) = e 2 − e2 + x . The varia-
tion iteration formula for the problem under consideration is

Z x
yn+1 (x) = yn (x) + (t − x){y00n (t) − yn (t) + t}dt (4.1)
0

Choose y0 (x) = 0. Substituting n = 0 in the equation (4.1) we

get

x3
Z x
y1 (x) = y0 (x) + (t − x){y000 (t) − y0 (t) + t}dt = −
0 3!
Similarly substituting n = 1, 2, . . . in (4.1) we get

Fig.3
x3 x5 x3 x5 x7
y2 (x) = − − , y3 (x) = − − − ,
3! 5! 3! 5! 7!
x 3 x 5 x7 x9
y4 (x) = − − − − , ....
3! 5! 7! 9!

703
 A study of variational iteration method for solving various types of problems
— 704/708

Substituting n = 1 in the equation (4.2) we get

Z x
y2 (x) = y1 (x) + (t − x){y001 (t) + 16y1 (t)}dt
0
Z x
2
= 1 − 8x + (t − x){−16 + 16(1 − 8t 2 )}dt
0
32 4 (4x)2 (4x)4
y2 (x) = y1 (x) + x = 1− +
3 2! 4!

Continuing this process we get

(4x)2 (4x)4 (4x)6 (4x)2n

yn (x) = 1 − + − + ....... + (−1)n
2! 4! 6! (2n)!

Consequently the solution of the differential equation is

Fig.4
y(x) = lim yn (x) = cos(4x)
n→∞

In Fig.6 to Fig.9 we have drawn the curves representing the

exact solution and first, second , third, fourth approximations
separately to know how the approximations are approaching
to the exact values of function which is the solution of
considered problem. It is observed that higher numbered
approximations are close to the exact solution.

Fig.5

Example 4.2. We know that y00 (x) + 16y(x) = 0 with the

conditions y(0) = 1 and y0 (0) = 0 represent simple harmonic
motion. The exact solution of this problem is y = cos(4x).

The variation iteration formula for y00 (x) + 16y(x) = 0 is

Z x
yn+1 (x) = yn (x) + (t − x){y00n (t) + 16yn (t)}dt (4.2)
0

Choose y0 (x) = 1. Substitute n = 0 in the equation (4.2) we

get
Z x
y1 (x) = y0 (x) + (t − x){y000 (t) + 16y0 (t)}dt
0 Fig.6
Z x
= 1+ (t − x){16}dt = 1 − 8x2
0

(4x)2
y1 (x) = 1 − (4.3)
2!

704
 A study of variational iteration method for solving various types of problems
— 705/708

5. Variational Iterative method for the

solution of Volterra Integral equation

Example 5.1. Let the Volterra integral equation be

Z x
y(x) = x − (x − t)y(t)dt (5.1)
0

Using Leibnitz rule, differentiating (5.1) with ‘x’ we obtain

Z x Z x
∂
y0 (x) = 1 − {(x −t)y(t)}dt ⇒ y0 (x) = 1 − y(t)dt
0 ∂x 0
(5.2)
Fig.7

Simplifying (5.2) we get

Z x
y0 (x) + y(t)dt − 1 = 0 (5.3)
0

Hence the iteration formula is given by

Z x Z t
yn+1 (x) = yn (x) − [y0n (t) − 1 + yn (s)ds]dt (5.4)
0 0

From equation (5.1) we have y(0) = 0. Choose y0 (x) = 0.

Substituting n = 0 in (5.4) we get y1 (x) = x.
Fig.8 Now, substituting n = 1 in the equation (5.4) we get

Z x Z t
y2 (x) = y1 (x) − [y01 (t) − 1 + y1 (s)ds]dt
0 0
x3
Z x Z t
= x− [1 − 1 + sds]dt = x −
0 0 3!
x3 x5 x3 x5 x7
y3 (x) = x − + , y4 (x) = x − + − , .....
3! 5! 3! 5! 7!

y(x) = lim yn (x) = sin x

n→∞

In this example it is observed that the fourth approximation is

Fig.9 more accurate than preceeding approximations to the exact
solution and it is shown in the Fig.10 to Fig.13 .

705
 A study of variational iteration method for solving various types of problems
— 706/708

Fig.13
Fig.10
Example 5.2. The equation given below is a Volterra
integral equation
Z x
y(x) = 1 + x + (x − t)y(t)dt (5.5)
0

Differentiating the above equation we get

Z x
∂
y0 (x) = 1 + [(x − t)y(t)]dt
0 ∂x
i.e.,
Z x
y0 (x) − y(t)dt − 1 = 0 (5.6)
0

The variational iteration formula for (5.6) is given by

Z x Z t
yn+1 (x) = yn (x) − [y0n (t) − 1 − yn (s)ds]dt (5.7)
0 0

Fig.11 From (5.5) we get y(0) = 1. Choosey0 (x) = 1. Substituting

n = 0 in equation (5.7) we obtain
Z x Z t
y1 (x) = y0 (x) − [y00 (t) − 1 − y0 (s)ds]dt
0 0
x2
Z x Z t
= 1− [0 − 1 − 1ds]dt = 1 + x +
0 0 2!
2
Thus y1 (x) = 1 + x + x2! . Substituting n = 1, 2, 3, . . . . in
equation (5.7) we obtain

x2 x3 x4
y2 (x) = 1 + x + + + ,
2! 3! 4!
x2 x3 x4 x5 x6
y3 (x) = 1 + x + + + + + ,
2! 3! 4! 5! 6!
x2 x3 x4 x5 x6 x7 x8
y4 (x) = 1 + x + + + + + + +
2! 3! 4! 5! 6! 7! 8!

Fig.12 706
 A study of variational iteration method for solving various types of problems
— 707/708

,...... and so on
The solution of the Volterra integral equation is

y(x) = lim yn (x) = ex

n→∞

The approximations and exact solution of the problem are

shown in the Fig.14 to Fig.17 . It is observed that the curve
representing the fourth approximation and the curve
Fig.16
representing the exact solutions coincide for small values of
x.

Fig.14 Fig.17

6. Variational Iteration Method for solving

isoperimetric problem
Example
R 1 02
6.1. We obtain the extremal of the functional
y (x)dx subject to the conditions y(0) = 1 ,y(1) = 1 and
R01
0 y(x)dx = 2. The exact solution of this problem is
y(x) = 1 + 6x − 6x2 .
Let F ∗ be the auxiliary function and λlmi be the Lagrange
multiplier of isoperimetric problem.
Then F ∗ = y0 (x)2 + λlmi y(x)
∗ d ∂ F∗
Then, Euler Lagrange equation ∂∂Fy − dx ∂ y0 = 0 for this
problem gives the following differential equation.

2y00 (x) − λlmi = 0 (6.1)

Fig.15 The variational iteration formula for the differential equation

707
 A study of variational iteration method for solving various types of problems
— 708/708

(6.1) we obtain as References

Z x [1] J.H. He, A new approach to nonlinear partial differential
yn+1 (x) = yn (x) + (t − x){y00n (t) − λlmi }dt (6.2)
0 equations, [Link] Nonlinear Sci. and Numer. Simul.2
(4), 230-235, (1997).
we choose [2] J.H. He, Approximate analytical solution for seepage
flow with fractional derivatives in porous media, Comp.
y0 (x) = 1 + ax − ax2 (6.3) Methods Appl. Mech. Engg. , Vol.167, 57-68, (1998).
[3] J.H. He, Approximate solution of non linear differential
So that it satisfies y(0) = 1, y(1) = 1. From the isoperimetric
equations with convolution product non-linearities,
condition
Comp. Methods Appl. Mech. Engg. , Vol.167, 69-73,
Z 1 Z 1 (1998).
y0 (x)dx = 2 ⇒ (1 + ax − ax2 )dx = 2 ⇒ a = 6 (6.4) [4] Fateme Ghomanjani, Sara Gaderi, Varitional iterative
0 0
method applied to variational problems with moving
Hence boundaries, Appl. Maths, 3, 395-402, (2012).
[5] Junfeng Lu, Variational iteration method for solving
2
y0 (x) = 1 + 6x − 6x (6.5) two-point boundary value problems, Jr. of Comp. and
Appl. Math, 207, 92-95, (2007).
Substituting n = 0 in (6.2) we get [6] Sallid Abbasbandy, Elyas Shivanian, Application of the
Z x variational iteration method for system of the non-linear
y1 (x) = y0 (x) + (t − x){y000 (t) − λlmi }dt Volterra’s integro-differential equations, Math and Comp.
0
Applications, vol. 14, No.2, 147-159, (2009).
[7] Abdal-Majid waz waz, The variational iteration method
In view of (6.4) and (6.5) we have
Z x
for solving linear and non-linear ODEs and scientific
models with variable coefficients, Cent. Euro. Jr. Engg. ,
y1 (x) = 1 + 6x − 6x2 + (t − x){−24 − λlmi }dt
0 4(1), 64-71, (2014).
[8] [Link] Variational Iteration method a kind of non-linear
analytical technique: some examples, Int. Jr. of
x2 nonlinear Mechanics ,Vol.34, 699-708, (1999).
y1 (x) = 1 + 6x − 6x2 + (24 + λlmi ) (6.6)
2 [9] Ji-Huan He,Xu-Hong Wu,Variational Iteration Method:
Now from the isoperimetric condition we have New development and applications, Science Direct,
Compu. and Mathematics Applications. ,54, 881-894,
(2007).
x2
Z 1 Z 1
y1 (x)dx = 2 ⇒ {1 + 6x − 6x2 + (24 + λlmi ) =2
0 0 2 ?????????
(6.7) ISSN(P):2319 − 3786
Malaya Journal of Matematik
ISSN(O):2321 − 5666
?????????
⇒ λlmi = −24 (6.8)

Hence, y1 (x) = 1 + 6x − 6x2 which is equal to the exact

solution. It can be shown that yk (x) = y1 (x)(k = 2, 3, 4, . . . .).
In the first iteration itself we got the exact solution of the
considered isoperimetric problem.

7. Conclusion
This paper is aimed to obtain the approximate solution of
initial value problems of first and second orders, Volterra
integral equation and isoperimetric problem. By observing
the solutions of problems that we considered one can say that
Variational Iterative Method developed by [Link] can be
applied to various types of problems. By observing the
graphs of exact and approximate solutions we can say that
solution is obtained in fewer iterations.

708