Use of Modified Variational Iteration Method In

Solving Ordinary Differential Equations

A Project Submitted in Partial Fulfilment for the Completion of

Research Based Learning-2 (RBL-2)
in
Master of Science
in
Mathematics
by
Dahiru Abdurrahman
Roll Number: 2307120009
System ID: 2023000665

UNDER THE SUPERVISION

of
Dr. Maheshwar Pathak

Department of Mathematics
Sharda School of Basic Sciences & Research
Sharda University, Greater Noida-201306, U.P., India
April, 2024
 DECLARATION

I, Dahiru Abdurrahman, hereby declare that this project entitled ”Use of

modified variational iteration method in solving ordinary differ-
ential equations” submitted as a partial fulfilment of the requirements
for the award of the degree of Master of Science in Mathematics at Sharda
University, Greater Noida, Under the supervision of Dr. Maheshwar Pathak
at the Department of Mathematics, Sharda University, Greater Noida.

I further affirm that we have not submitted this work for consideration for
any other degree or certification from this or any other University.

Date: 2nd April, 2024 Student Name: Dahiru Abdurrahman

Place: Greater Noida System Id: 2023000665

i
 Abstract

This study presents and applied the MVIM to solve some ordinary differ-
ential equations of first and second order. A series of approximations that
lead to the actual solution of an ordinary differential equation are produced
by the MVIM. Furthermore, all the repeated calculations involved in the
variational iteration method (VIM) are eliminated by the MVIM and this
simplifies the amount of computational labor involved.

Keywords: Restricted variation, Lagrange multiplier, differential equation,

correctional functional, variational iteration method

ii
 Table Of Contents
S. No. Contents Page No.
Project proposal form iv
Certificate of similarity index/ Plagarism check v
1 Introduction 1
2 Objectives of the study 1-2
3 Iterative variational approach 2
4 Literature survey 2-3
5 Modified variational iteration method 3-6
6 Applications 6-14
7 Conclusion 14
8 Recommendation 15
9 References 16
Plagarism report (annexure)

iii
 Project - Proposal form
Name of the Department: Mathematics
Name of the School: Sharda School of Basic Sciences &
Research
Academic Session: 2023-24

Details of Students ( IN BLOCK LETTERS):

S. No. Name Roll No. System ID Section Email-id Phone no
1 DAHIRU ABDURRAHMAN 2307120009 2023000665 Mathematics [email protected] 9821327536

Title of the Project: Use of Modified Variational Iteration Method In

Solving Ordinary Differential Equations.

i. Area/Concept involved: Numerical Method

ii. Language/tools that may be used: Matlab

iii. Category of the project (Comparative analysis/Simulation/Hardware/Software

Development/Experimental Research, Automation/other (SPECIFY)):
Research survey

iv. Prerequisite Knowledge Required: Differential Equations, Numerical

Method

Sign. of Student

Consent of Dissertation/ Project supervisor: Yes/No

Name of the Supervisor:

Approved/ Approved with changes/ Not approved

Signature of the Supervisor:

HOD’s Signature

Date:

iv
 Certificate of Similarity Index/ Plagiarism check

Name and Roll Number of Student Dahiru Abdurrahman

2307120009
Name of Program M.Sc Mathematics
Academic Session 2023/2024
Semester 2
Title of Project Use of modified variational
iteration method in solving
ordinary differential equations
Name of Supervisor Dr. Maheshwar Pathak
Software Name Turnitin
Date of Similarity Index/ Plagiarism Check 02-April-2024 10:49AM
Similarity Index 8%
Acceptable maximum limit (%) of similarity 10%

I hereby certify that the project entitled ’Use of modified variational it-
eration method in solving ordinary differential equations’ has been
evaluated using ’Turnitin’. I have analyzed the report produced by the
system and reported in the table.

I further certify that if similarity index is found above the acceptable, my

project may be cancelled.

Signature of student

Signature of the Supervisor

Name of Student:Dahiru Abdurrahman Name of the Supervisor:

Roll No.: 2307120009.
System ID.: 2023000665.

Date: 4th April, 2024.

Place: Greater Noida.

v
1. Introduction
Differential equations can be used to simulate many real-
world issues and solving these equations are necessary in
order to understand the problems in their physical context.
Most of the time, finding the analytical solution to the dif-
ferential equation is extremely difficult or perhaps impossi-
ble. Because of this, we must devise an additional method
for resolving these equation such as the semi-analytical
method and numerical method. Several numerical or semi-
analytical techniques are available for solving these kinds of
differential equations, but in this work we will focus on one
of the semi-analytical method known as the revised vari-
ational iteration technique. However, first we will explain
the variational iteration technique.
Chinese mathematician Ji-Huan He [4] was the first to
devise the VIM. It is a strong semi-analytical method that
provides a series of approximations that converge to the
exact solution of a differential equation. However, this
method contains a weakness that necessitates computing
duplicate terms once more for each iteration [7]. Conse-
quently, Abassy et al [1] devised the MVIM, that makes the
process simple, reliable, and efficient by doing away with
all of the repeated computations. Through the consider-
ation of a few numerical instances of ordinary differential
equations, we will test the robustness of the modified vari-
ational iteration method in this study.

2. Objectives of the study

The objectives of the study are to:
ˆ revise the variational iteration method (VIM).
ˆ derive the modified variational iteration method (MVIM)
for solving ordinary differential equations of first and
second order.

1
 ˆ solve some linear and non-linear ordinary differential
equations of first and second order using the MVIM.
3. Iterative variational approach
Examine the differential equation shown below:

T [y(x)] + U [y(x)] = v(x) (1)

with v being a continuous function, U being a non-linear operator and

T being a linear operator. Either the initial condition(s) or boundary
condition(s) must be specified in order to calculate the series of subse-
quent approximations that will yield the solution of equation (1). The
correctional functional of equation (1) is
Z x
yk+1 (x) = yk (x) + λ(s, x){T [yk (s)] + U [ỹk (s)] − v(s)}ds
x0

where ỹk (x) is the restricted variation, i.e., δ ỹk (x) = 0, and yk (x) is
the kth approximation. λ(s, x) is the Lagrange multiplier that can be
obtained optimally through variational theory [8]. The VIM yields the
iterative formula below:
Z x
yk+1 (x) = yk (x) + λ(s, x){T [y(s)] + U [y(s)] − v(s)}ds (2)
x0

y(x) = lim yk (x) is the solution of (1). The Lagrange multiplier for
k→∞
the nth order differential equation has been provided by Ji-Huan He
and Xu Hong Wu in [5] as
(−1)n
λ(s, x) = (s − x)n−1 (3)
(n − 1)!

4. Literature survey
The VIM has been utilized by numerous researchers in diverse scientific
and technical domains to get solutions for various kinds of differential
equations. The abelian differential equation is a non-linear differential
equation, and its solution was determined by M. Matinfar and S. Jafar
[8] using the VIM. The method involves building a series of approx-
imations that converges toward the actual solution of the equation.
The VIM has been used in [4] to tackle a variety of problems. Using
the VIM, they were able to find consecutive approximations for solu-
tions to isoperimetric problems, differential equations of second order
with two constraints, differential equations of first order with only one

2
constraint and volterra integral equations of second kind. They were
also able to compare the actual and approximate solutions graphically.
Additionally, by taking into consideration a few numerical instances
they employed the VIM to solve a linked first-order system of linear
ordinary differential equations [3]. It was demonstrated in that work
that each problem’s iterative approximations converge to the exact so-
lution. They have also included graphic representations of the exact
solution and a few approximations (the third, fourth, fifth, and sixth).
Ji-Huan He [6] has applied non-linear techniques of variational iter-
ation method to solved Duffing equation, the equation of motion for
human eardrum, the equation controlling the free movement of a sim-
ple pendulum, and some partial differential equations. He also makes a
comparison between Adomian’s approach and the variational iteration
approach.
Abdul-Majid Wazwaz [2] has solved a number of linear and non-
linear ordinary differential equations using the VIM, including hybrid
selection model, Thomas-Fermi equation, Riccati equation and Kidder
equation for the unsteady movement of gas through a porous media.
Additionally, he has solved numerous ordinary differential equations of
varying orders using the VIM. In addition to solving non-linear ordinary
equations like the Riccati, Lane-Emden, and Emden-Fowler equations,
variety of linear ordinary differential equations of various orders were
solved in [7] using the MVIM. They have demonstrated that the MVIM
outperforms the VIM because it eliminates some of the repetitive cal-
culations that the VIM’s iterative process involves.

5. Modified variational iteration method

5.1 First order differential equations
Consider the first order IVP below:
y ′ + g(x)y = h(x) y(0) = a (4)
As per the VIM [7], the correctional functional of (4) is utilized as
Z x
yk+1 (x) = yk (x) + λ(s, x)[yk′ (s) + g(s)ỹk (s) − h(s)]ds (5)
0

With λ(s, x) = −1 being a Lagrange multiplier obtained using equa-

tion (3), ỹk is the restricted variation i.e., δ ỹk = 0. Now, the iterative
equation below is obtained by substituting the computed lagrange mul-
tiplier into equation (5):
Z x
 ′ 
yk+1 (x) = yk (x) − yk (s) + g(s)ỹk (s) − h(s) ds (6)
0

3
Equation (4) has a solution y(x) = lim yk (x).
k→∞
However, the recursive formula (6) can be rewritten as follows to
avoid the duplication that occurs in each step of this method:
Z x
 
yk+1 (x) = yk (x) − yk (x) + yk (0) − g(s)yk (s) − h(s) ds
0
Z x
 
= yk (0) − g(s)yk (s) − h(s) ds (7)
0

Now, we can set yk (0) = y0 so that equation (7) becomes

Z x
 
yk+1 (x) = y0 − g(s)yk (s) − h(s) ds (8)
0

Clearly we have seen that through iteration formula (7), some repeated
calculations are eliminated.
In order to remove all the repeated computations, we can re-write
equation (8) as:
Z x
 
yk+1 (x) = y0 − g(s)yk (s) + g(s)yk−1 (s) − g(s)yk−1 (s) − h(s) ds
Z0 x Z x
   
= y0 − g(s)yk−1 (s) − h(s) ds − g(s) yk (s) − yk−1 (s) ds
0 0

From equation (8) we have

Z x
 
yk (x) = y0 − g(s)yk−1 (s) − h(s) ds (10)
0

Putting equation (10) into equation (9) implies

Z x
 
yk+1 (x) = yk (x)− g(s) yk (s)−yk−1 (s) ds, k > 0 (11)
0

with y−1 (x) = 0, y(0) = y0 and

Z x


y1 (x) = y0 − g(s) y0 (s) − y−1 (s) − h(s) ds
0

Remark: Regarding the non-linear ordinary differential equation of

the first order, the MVIM does not require any other treatment i.e.,
the approach is the same.
5.2 Differential equation of second order
We are now going to extend the idea of MVIM to include constant

4
coefficient second order linear differential equations.
Examine the following IVP:

y ′′ +uy ′ +vy = w y(0) = a, y ′ (0) = b (12)

with u, v being constants and w being a function of x or constant.

We first create the correctional functional of equation (13), which
is provided in accordance with the VIM concept as
Z x h i
yk+1 (x) = yk (x)+ λ(s, x) yk′′ (s)+uy˜k ′ (s)+v ỹk (s) k>0 (13)
0

The value of the Lagrange multiplier is found by using equation (3) as

λ(s, x) = s − x.
Putting this Lagrange multiplier value into equation (13) suggests
Z x h i
yk+1 (x) = yk (x)+ (s−x) yk′′ (s)+uyk′ (s)+vyk (s)−w(s) ds k>0 (14)
0

It was realised that when applying iterative formula (14), there are
repeated evaluations in each iteration. To avoid such repetitions, we
can modify the formula as
Z x Z x h i
′′ ′
yk+1 (x) = yk (x) + (s − x)yk (s)ds + (s − x) uyk (s) + vyk (s) − w(s) ds
0 0
Z x h i
′ ′
= yk (0) + xyk (0) + (s − x) uyk (s) + vyk (s) − w(s) ds (15)
0

Now by setting yk (0) + xyk′ (0) = y(0) + xy ′ (0) = a + bx = y0 , equation

(15) becomes
Z x h i
yk+1 (x) = y0 + (s−x) uyk′ (s)+vyk (s)−w(s) ds (16)
0

In order to cancel out all the repeated evaluations involved in VIM,

equation (16) can be re-witten as follows:
Z x n h i h i
yk+1 (x) = y0 + (s − x) u yk′ (s) − yk−1
′
(s) + v yk (s) − yk−1 (s)
0
h io
′
+ uyk−1 (s) + vyk−1 (s) − w(s) ds
Z x h i
′
= y0 + (s − x) uyk−1 (s) + vyk−1 (s) − w(s) ds+
0
Z x n h i h io
(s − x) u yk′ (s) − yk−1
′
(s) + v yk (s) − yk−1 (s) ds (17)
0

5
From equation (16) we have
Z x h i
′
yk (x) = y0 + (s−x) uyk−1 (s)+vyk−1 (s)−w(s) ds (18)
0

Putting equation (18) into equation (17), we obtained the recursive

relation:
Z x n h i h io
yk+1 (x) = yk (x)+ (s−x) u yk′ (s)−yk−1
′
(s) +v yk (s)−yk−1 (s) ds, k>0 (19)
0

where y0 = a + bx, y−1 (x) = 0 and

Z x n h i h i o
′ ′
y1 (x) = y0 + (s−x) u y0 (s)−y−1 (s) +v y0 (s)−y−1 (s) −w(s) ds
0

Rermark: All of the repeated calculations in VIM were removed by

the final updated formula (19).

6. Applications
In order to assess the MVIM’s power, we will look at a few numerical
instances of first and second order differential equations below:
Example 1 : Use MVIM to obtain the solution of the problem below:
y ′ −2y = 0, y(0) = 5 (20)
To solve equation (20) by MVIM, we apply the iteration formulae (11)
as follows:
Z x
y1 (x) = 5 + 10 ds
0
= 5 + 10x
= 5(1 + 2x)
Z x
y2 (x) = (5 + 10x) + 2 [(5 + 10s) − 5]ds
0
Z x
= (5 + 10x) + 2 10s ds
0
= 5 + 10x + 10x2
= 5(1 + 2x + 2x2 )
" #
(2x)2
= 5 1 + 2x +
2!

6
 Z x
2
y3 (x) = (5 + 10x + 10x ) + 2 [(5 + 10s + 10s2 ) − (5 + 10s)]ds
Z0 x
= (5 + 10x + 10x2 ) + 2 10s2 ds
0
20
= 5 + 10x + 10x2 + x3
3

2 4 3

= 5 1 + 2x + 2x + x
" 3 #
2
(2x) (2x)3
= 5 1 + 2x + +
2! 3!

Z xh

2 20 3  2 20 3
2

i
y4 (x) = 5 + 10x + 10x + x + 2 5 + 10s + 10s + s − 5 + 10s + 10s ds
3 0 3
Z x
 20  20
= 5 + 10x + 10x2 + x3 + 2 s3 ds
3 0 3
20 10
= 5 + 10x + 10x2 + x3 + x4
3 3
 4 2 
= 5 1 + 2x + 2x2 + x3 + x4
" 3 3 #
2 3
(2x) (2x) (2x)4
= 5 1 + 2x + + +
2! 3! 4!
·
·
·
" #
(2x)2 (2x)3 (2x)4 (2x)k
yk (x) = 5 1 + 2x + + + + ... +
2! 3! 4! k!

Recall that y(x) = lim yk (x) provides the actual solution to the equa-
k→∞
tion.
Hence y(x) = 5e2x is the solution to the IVP.
NOTE: In MVIM, the inhomogeneous term is used once (i.e., in the
first iteration only) as we have seen in the above example.

Example 2 : Use MVIM to obtain the solution of the IVP below:

y ′ +3x2 y = x2 , y(0) = 2 (21)

7
Equation (21) can be solved using MVIM by the iteration formula (11)
as follows:
Z x
y1 (x) = 2 − 5s2 ds
0
5
= 2 − x3
3
1 5 5 3
= + − x
3 3 3
1 5
= + 1 − x3


3 3
Z x
 5 3 h 5
i
y2 (x) = 2 − x − 3s2 (2 − s3 − 2 ds
3 3
Z0 x
 5 
= 2 − x3 + 5s5 ds
3 0
5 3 5 6
=2− x + x
3 6
1 5 5 3 5 6
= + − x + x
3 3" 3 6 #
3 2
1 5 (−x )
= + 1 + (−x3 ) +
3 3 2!

Z x
 5 3 5 6 h 5 5   5 i
y3 (x) = 2 − x + x − 3s2 2 − s3 + s6 − 2 − s3 ds
3 6 3 6 3
Z0 x
 5 5  5 8
= 2 − x3 + x6 + s ds
3 6 0 2
5 5 5
= 2 − x 3 + x6 − x9
3 6 18
1 5 5 3 5 6 5
= + − x + x − x9
3 3" 3 6 18 #
3 2 3 3
1 5 (−x ) (−x )
= + 1 + (−x3 ) + +
3 3 2! 3!

8
 Z x
 5 3 5 6 5 9 h 5 5 5   5 5 i
y4 (x) = 2 − x + x − x − 3s2 2 − s3 + s6 − s9 − 2 − s3 + s6 ds
3 6 18 3 6 18 3 6
Z0 x
 5 5 5  5 11
= 2 − x3 + x6 − x9 + s ds
3 6 18 0 6
5 5 5 5
= 2 − x3 + x6 − x9 − x12
3 6 18 72
1 5 5 3 5 6 5 9 5
= + − x + x − x − x12
3 3" 3 6 18 72 #
1 5 1 1 1
= + 1 + (−x3 ) + (−x3 )2 + (−x3 )3 + (−x3 )4
3 3 2! 3! 4!
·
·
·
" #
3 2 3 3 3 4 3 k
1 5 (−x ) (−x ) (−x ) (−x )
yk (x) = + 1 + (−x3 ) + + + + ... +
3 3 2! 3! 4! k!

1 5 −x3
Hence y(x) = + e is the solution to the IVP.
3 3

Example 3 : Obtain the solution of the Riccati equation below:

y ′ −y 2 +6xy = 9x2 +3, y(0) = 2 (22)

To solve equation (22) by MVIM, we follow the procedure presented

above as follows:
Z xh i
2 2 2



y1 (x) = y0 − − y0 (s) − y−1 (s) + 6s y0 (s) − y−1 (s) − 9s + 3) ds
0
Z xh i
2
=2− − (4 − 0) + 6s(2 − 0) − 9s − 3 ds
Z0 x
=2− (−4 + 12s − 9s2 − 3)ds
Z0 x
=2− (−7 + 12s − 9s2 )ds
0
= 2 − (−7x + 6x2 − 3x3 )
= 2 + 7x − 6x2 + 3x3

9
 Z x h
i
− y12 (s) − y02 (s) + 6s y1 (s) − y0 (s) ds


y2 (x) = y1 (x) −
0
Z x
2 3
− (28s + 25s2 − 72s3 + 78s4 − 36s5 + 9s6 )−

= (2 + 7x − 6x + 3x ) −
0
6s(7s − 6s2 + 3s3 ) ds

Z x
2 3
= (2 + 7x − 6x + 3x ) − (−28s − 67s2 + 108s3 − 96s4 + 36s5 − 9s6 )ds
0
 67 96 9 
= (2 + 7x − 6x2 + 3x3 ) − − 14x2 − x3 + 27x4 − x5 + 6x6 − x7
3 5 7
76 96 9
= 2 + 7x + 8x2 + x3 − 27x4 + x5 − 6x6 + x7
3 5 7
Z x h
i
− y22 (s) − y12 (s) + 6s y2 (s) − y1 (s) ds


y3 (x) = y2 (x) −
0
Z x
 76 3 2 4 96 5 6 9 7 856 3 698 4
= 2 + 7x + 8x + x − 27x + x − 6x + x − − 56s2 − s − s
3 5 7 0 3 3
2102 5 20051 6 39888 7 8119 8 46208 9 132612 10 10494 11
− s − s + s − s + s − s + s −
15 45 35 5 35 175 35
2988 12 108 13 81 14 576 5 54 
s + s − s + 84s2 + 128s3 − 162s4 + s − 36s6 + s7 ds
35 7 49 5 7
 76 96 9  Z x 472 3 1184 4
= 2 + 7x + 8x2 + x3 − 27x4 + x5 − 6x6 + x7 − 28s2 − s − s−
3 5 7 0 3 3
374 5 21671 6 40158 7 8119 8 46208 9 132612 10 10494 11 2988 12
s − s + s − s + s − s + s − s +
15 45 35 5 35 175 35 35
108 13 81 14 
s − s ds
7 49
 76 96 9   28 3 118 4 1184 5
= 2 + 7x + 8x2 + x3 − 27x4 + x5 − 6x6 + x7 − x − x − x−
3 5 7 3 3 15
187 6 21671 7 20079 8 8119 9 23104 10 132612 11 1749 12 2988 13
x − x + x − x + x − x + x − x
45 315 140 45 175 1925 70 455
54 27 15 
+ x14 − x
49 245
 37 1472 5 83 6 22076 20079 8 8119 9
= 2 + 7x + 8x2 + 16x3 + x4 + x − x + + x − x+
3 15 45 315 140 45
23104 10 132612 11 1749 12 2988 13 54 14 27 15 
x − x + x − x + x − x
175 1925 70 455 49 245

10
y4 (x) = 2 + 7x + 8x2 + 16x3 + 32x4 + 64x5 + 128x6 + 256x7 + 512x8 + ...

= 3x + 2 + 4x + 8x2 + 16x3 + 32x4 + 64x5 + 128x6 + 256x7 + 512x8 + ...

·
·
·
h i
yk (x) = 3x + 2 1 + 2x + (2x)2 + (2x)3 + (2x)4 + (2x)5 + (2x)6 + (2x)7 + ... + (2x)k−1
Therefore, the solution to the equation is
2 1
y(x) = 3x + , for |x| <
1 − 2x 2

Example 4 : Utilizing the MVIM, solve the following IVP:

y ′′ +y = 1, y(0) = 2 y ′ (0) = 7 (23)
To solve equation (23) by MVIM, we use the iteration formula (19) as
follows:
Z x
y1 (x) = (2 + 7x) + (s − x)(1 + 7s)ds
Z0 x
s + 7s2 − x − 7xs ds


= (2 + 7x) +
0
 s2 7 3 7 2 x
= (2 + 7x) + + s − xs − xs
2 3 2 0
1 7 3 7 3
2 2
= (2 + 7x) + x + x −x − x
2 3 2
1 2 7 3
= 2 + 7x − x − x
2 6
 1 2  7 3
= 1 + 1 − x + 7x − x
2 6
 x2   x3 
=1+ 1− +7 x−
2! 3!

Z x
 1 7 3 h 1 2 7 3  i
y2 (x) = 2 + 7x − x2 − x + (s − x) 2 + 7s − s − s − 2 + 7s ds
2 6 0 2 6
Z x
 1 7 3   1 7 
= 2 + 7x − x2 − x − (s − x) s2 + s3 ds
2 6 2 6
Z0 x 
 1 2 7 3 1 3 7 4 1 2 7 3
= 2 + 7x − x − x − s + s − xs − xs ds
2 6 0 2 6 2 6

11
  1 7  1 4 7 1 7 
= 2 + 7x − x2 − x3 − x + x5 − x4 − x5
2 6 8 30 6 24
1 2 7 3 1 4 7 5
= 2 + 7x − x − x + x + x
2 6 24 120
 1 1   7 7 5
= 1 + 1 − x2 + x4 + 7x − x3 + x
2 24 6 120
 1 1   1 1 
= 1 + 1 − x2 + x4 + 7 x − x5 + x 5
2! 4! 3! 5!

Z x
 1 2 7 3 1 4 7 5 h 1 7 1
y3 (x) = 2 + 7x − x − x + x + x + (s − x) 2 + 7s − s2 − s3 + s4 +
2 6 24 120 0 2 6 24 1
 1 2 7 3 i
2 + 7s − s − s
2 6 Z x
 1 2 7 3 1 4 7 5 1 7 5
= 2 + 7x − x − x + x + x + (s − x) s4 + s ds
2 6 24 120 0 24 120
Z x
 1 2 7 3 1 4 7 5 1 5 7 6 1 7 
= 2 + 7x − x − x + x + x + s + s − xs4 − xs5 ds
2 6 24 120 0 24 120 24 120
 1 7 1 7 5   1 6 1 7 1 6 7 7
= 2 + 7x − x2 − x3 + x4 + x + x + x − x − x
2 6 24 120 144 120 120 720
1 7 1 7 5 1 6 1 7
= 2 + 7x − x2 − x3 + x4 + x − x − x
2 6 24 120 720 720
 1 1 1 6  7 7 5 1 7
= 1 + 1 − x2 + x4 − x + 7x − x3 + x − x
2 24 720 ! 6 120 120!
1 1 1 1 1 1
= 1 + 1 − x2 + x4 − x6 + 7 x − x3 + x5 − x7
2! 4! 6! 3! 5! 7!
·
·
·
" # #
x2 x4 x 6 (−1)k x2k h x3 x 5 x7 (−1)k x2k+1
yk (x) = 1 + 1 − + − + ... + +7 x− + − + ... +
2! 4! 6! (2k)! 3! 5! 7! (2k + 1)!

Hence y(x) = 1 + cos(x) + sin(x) is the actual solution to the IVP.

Example 5 : Utilizing the MVIM, find the solution to the following IVP:

y ′′ +y ′ −2y 2 = x, y(0) = 1 y ′ (0) = 2 (24)

12
To solve equation (24) by MVIM, we use the iteration formula (19) as
follows:
Z x h i
y1 (x) = y0 + ′
(s − x) (y0′ (s) − y−1 2
(s)) − 2(y02 (s) − y−1 (s)) − x ds
0
Z x h i
(s − x) 2 − 0) − 2 1 + 4s + 4s2 − 0 − x ds


= (1 + 2x) +
Z0 x
= (1 + 2x) + (s − x)(−8s − 8s2 − x)ds
Z0 x
= (1 + 2x) − (s − x)(8s + 8s2 + x)ds
Z0 x
= (1 + 2x) − (8s2 + 8s3 − 7xs − 8xs2 − x2 )ds
0
8 8 7 8
= 1 + 2x − x3 − x4 + x3 + x4 + x3
3 4 2 3
11 3 2 4
= 1 + 2x + x + x
6 3
Z x h
i
(s − x) y1′ (s) − y0′ (s) − 2 y12 (s) − y02 (s) ds


y2 (x) = y1 (x) +
0
Z x
 11 3 2 4  h 11 8 11
(s − x) 2 + s2 + s3 − 2 − 2 1 + 4s + 4s2 + s3


= 1 + 2x + x + x +
6 3 0 2 3 3
26 8 121 6 22 7 4 8 i
+ s4 + s5 + s + s + s − 1 − 4s − 4s2 ds


3 3 36 9 Z 9
x
 11 3 2 4   11 14 52 16 121 6 44 7
= 1 + 2x + x + x + (s − x) s2 − s3 − s4 − s5 − s − s−
6 3 0 2 3 3 3 18 9
8 8 
s ds
9 Z x
 11 3 2 4  8 8 44 44 121 7 121 6
= 1 + 2x + x + x + − s9 + s8 x − s8 + s7 x − s + sx
6 3 0 9 9 9 9 18 18
16 6 16 5 52 5 52 4 14 4 14 3 11 3 11 2 
− s + s x − s + s x − s + s x + s − s x ds
3 3 3 3 3 3 2 2
 11 3 2 4   4 10 8 9 44 9 11 8 121 8 121 7
= 1 + 2x + x + x + − s + s x − s + s x − s + s x−
6 3 45 81 81 18 144 126
16 7 8 6 26 52 14 7 11 11  x
s + s x − s6 + s5 x − s5 + s4 x + s4 − s3 x
21 9 9 15 15 6 8 6 0
 11 3 2 4   4 10 8 10 44 9 11 9 121 8 121 8
= 1 + 2x + x + x + − x + x − x + x − x + x−
6 3 45 81 81 18 144 126
16 7 8 7 26 6 52 6 14 5 7 5 11 4 11 4 
x + x − x + x − x + x + x − x
21 9 9 15 15 6 8 6
11 3 5 4 7 5 26 6 8 7 121 8 11 9 4 10
= 1 + 2x + x + x + x + x + x + x + x + x
6 24 30 45 63 1008 162 405

13
 Z x h
i
(s − x) y2′ (s) − y1′ (s) − 2 y22 (s) − y12 (s) ds


y3 (x) = y2 (x) +
0
 11 5 7 26 8 121 8 11 9 4 10 
= 1 + 2x + x3 + x4 + x5 + x6 + x7 + x + x + x
Z x 6 24 30 45 63 1008 162 405
h 11 5 7 52 8 121 7 11 8 8 11
+ (s − x) 2 + s2 + s3 + s4 + s5 + s6 + s + s + s9 − 2 − s2
0 2 6 6 15 9 126 18 81 2
8   11 31 13 109 6 8389 7 33203 8
− s3 − 2 1 + 4s + 4s2 + s3 + s4 + s5 + s + s + s+
3 3 4 10 20 2520 20160
64223 9 59657 10 25991 11 1883549 12 449 13 15133 14 101411 15
s + s + s + s + s + s + s +
22680 56700 32400 2721600 1680 79380 893025
9845777 16 1097 17 6413 18 44 19 16 11
s + s + s + s + s20 − 1 − 4s − 4s2 − s3
228614400 58320 918540 32805 164025 3
26 4 8 5 121 6 22 7 4 8 i
− s − s − s − s − s ds
3 3 36 9 9
 11 3 5 4 7 5 26 6 8 7 121 8 11 9 4 10 
= 1 + 2x + x + x + x + x + x + x + x + x +
6 24 30 45 63 1008 162 405
 11 1 31 7 37 8 1019 9 18083 10 63103 11 59657 12
x5 − x6 − x + x + x + x + x + x
120 10 210 630 90720 907200 1247400 3742200
25991 13 1883549 14
+ x + x +···
2527200 247665600
11 5 13 43 13 7 901 8 2393 9
= 1 + 2x + x3 + x4 + x5 + x6 − x + x + x+
6 24 40 90 630 5040 30240
2190483 10 63103 11 59657 12 25991 13 1883549 14
x + x + x + x + x + · · ·,
73483200 1247400 3742200 2527200 247665600
Similarly, other approximations can be obtained by using (19).

7. Conclusion
In this work we have presented the MVIM by deriving its iterative
procedure and also seen the strength of the method after solving some
various ordinary differential equations. The presented analysis of the
method involves two cases starting with first order via second order
linear differential equations. The displayed examples we looked at ear-
lier, that we looked at earlier demonstrates that MVIM is more efficient
than the current method. We have seen that for non-linear differen-
tial equation where the analytical solution is difficult to obtain, a few
number of approximations can be obtain using the method to get the
solution, as it cancels out some repeated calculations in VIM and so
reduce the amount of computational work and also save time.

14
8. Recommendation
Our research was limited to applying the MVIM to solve ordinary dif-
ferential equations. Therefore, I recommend the researchers to take
partial differential equations into consideration in order to expand the
application of MVIM beyond the realm of ordinary differential equa-
tions.

15
References

[1] Abassy, T.A., M.A. El-Tawil and H. El-Zoheiry, Toward a modified

variational iteration method, Journal of computational and Applied
Mathematics, 207(1): 137-147 (2011).
[2] Abdul-Majid Wazwaz, The variational iteration method for solv-
ing linear and non-linear ODEs and scientific models with variable
coefficients, Cent. Euro. Jr. Eng., 4(1), 64-71 (2014).
[3] E.Rama and K.Sambaiah, Application of variational iteration
method to find the solution of coupled system of first order differ-
ential equations, J. Math. Comput. Sci, No.4, 11 (2021), 4118-4128.
[4] E.Rama, K.Somaiah and K.Sambaiah, A study of variational itera-
tion method for solving various types of problems, Malaya Journal
of Matematik, vol.9, No.1, 701-708, (2021).
[5] Ji-Huan He and Xu-Hong Wu, Variational iteration method: new
development and applications, Science Direct, Compu and Mathe-
matics Applications, 54, 881-894, (2007).
[6] Ji-Huan He, Variational iteration method - a kind of non-linear
analytical technique: some examples, International Journal of Non-
linear Mechanics, 34 (1999), 699-708.
[7] M.A. Fariborzi Araghi, S. Gholizadeh Dogaheh and Z. Sayadi, The
modified variational iteration method for solving linear and non-
linear ordinary differential equations, Australian Journal of Basic
and Applied Sciences, 5 (10): 406-416, (2011).
[8] M. Matinfar and S. Jafar-Nodeh, Application of He’s variational
iteration method to abelian differential equation, Journal of Applied
Mathematics, Islamic Azad University of Lahijan, vol.7, No.4 (27),
71-75, (2011).

16