IOSR Journal of Mathematics (IOSR-JM)
e-ISSN: 2278-5728. Volume 5, Issue 5 (Mar. - Apr. 2013), PP 21-26
www.iosrjournals.org
A Modified Predictor-Corrector Formula For Solving Ordinary
Differential Equation Of First Order And First Degree
Mahtab Uddin And M. A. Ullah
Department of Mathematics, University of Chittagong, Chittagong-4331, Bangladesh.
Abstract: We are proposing a modified form of the Milne’s Predictor-Corrector formula for solving ordinary
differential equation of first order and first degree. Here we are approximating the value of the dependent
variable under five initial conditions (where Milne takes four initial conditions) and then improving this value
(closer to the exact value) by proper substitution in the formulae. This process is an iterative way to obtain the
values and the process continuing until we get a proper level of accuracy.
Keywords: ODE, Milne’s modified Predictor-Corrector, quantitative comparison, accuracy
I. Introduction
In the methods so far described to solve an ordinary differential equation over an interval, only the
value of y at the beginning of the interval was required. Now in the Predictor-Corrector methods, four prior
values are needed for finding the value of y at given value of x [2,4]. These methods though slightly complex,
have the advantage of giving an estimate of error from successive approximations of , with .
Then from Euler’s formula, we have
Also from modified Euler’s formula [8, 9], we have
The value of is first estimate by (1) and then by using (2) gets a better approximation of .
This value of is again substituted in (2) to find still a better approximation of . This procedure will
repeat until two consecutive iterated values of agree.
This technique of refining an initial crude estimation of by means of a more accurate formula is
known as Predictor-Corrector method [1]. The equation (1) is taken as the predictor, while (2) serves as a
corrector of .
To solve the differential equation by this method, first we are to approximate the value of
by predictor formula at , then improve this values of by using the corrector formula
after proper substitution. These formulae will be derived from the Newton’s formula of forward interpolation
[5,6].
II. Derivation Of Milne’s (Modified) Predictor Formula
We know that Newton’s formula of forward interpolation in terms of and u is given by
Author to whom all correspondences should be made (e-mail: [email protected]) .
www.iosrjournals.org 21 | Page
A Modified Predictor-Corrector Formula For Solving Ordinary Differential Equation Of First Order
Where
or,
So that
Now, integrating (3) over the interval to . i.e., to we obtain
Neglecting the terms containing and higher orders and substituting , from (4) we get
www.iosrjournals.org 22 | Page
A Modified Predictor-Corrector Formula For Solving Ordinary Differential Equation Of First Order
which is required Milne’s (modified) predictor formula.
III. Derivation Of Milne’s (Modified) Corrector Formula
To obtain the corrector formula, we integrate (3) over the interval to . i.e., to
then we get
www.iosrjournals.org 23 | Page
A Modified Predictor-Corrector Formula For Solving Ordinary Differential Equation Of First Order
Neglecting the terms containing and higher orders and substituting , from (6) we get
which is required Milne’s (modified) corrector formula.
IV. Generalization Of Milne’s (Modified) Predictor-Corrector Formula
We can write the general form [7] of Milne’s (modified) predictor and corrector formulae according to (5)
and (7) as follows
Here the index p & c indicates the predicted and corrected values of y respectively.
4.1. EXAMPLES
Problem-1:
Where initial values [2] are y(0.0) = 2.00000000, y(0.5) = 2.63610167, y(1.0) = 3.59488508, y(1.5) =
4.96800007, y(2.0) = 6.87312731
The analytical solution is
www.iosrjournals.org 24 | Page
A Modified Predictor-Corrector Formula For Solving Ordinary Differential Equation Of First Order
Problem-2:
Where initial values [2,5] are y(0.0) = 2.00000000, y(0.1) = 2.01000834, y(0.2) = 2.04013351, y(0.3) =
2.09067703, y(0.4) = 2.16214474
The analytical solution is
Problem-3:
Where initial values [5] are y(1.0) = 1.00000000, y(1.1) = 0.99573653, y(1.2) = 0.98526796, y(1.3) =
0.97104943, y(1.4) = 0.95462303
The analytical solution is
Problem-04:
Where initial values [2,5] are y(0.0) = 0.00000000, y(0.1) = 0.34869552, y(0.2) = 0.81126582,y(0.3) =
1.41677998, y(0.4) = 2.20114869
The analytical solution is y 3(e ex ) .
2x
Problem-5:
Where initial values [3] are y(0.0) = 0.00000000, y(0.1) = 0.10030157, y(0.2) = 0.20218090, y(0.3) =
0.30664524, y(0.4) = 0.41419888
The analytical solution is
V. Quantitative Comparison Of Numerical Results
Problem In Milne’s (modified) In Adam-Moulton’s
Exact value In Milne’s method.
No. method. method.
9.46150093 9.46137662 9.46166121
9.46137183
01 7th iteration 6th iteration 7th iteration
│ER│= 0.00137% │ER│= 0.00005% │ER│= 0.00306%
2.25525212
2.25525202 2.25525194
3rd iteration
02 2.25525193 3rd iteration 3rd iteration
│ER│=
│ER│= 0.0000040% │ER│= 0.0000004%
0.0000084%
0.93697840 0.93697747 0.93698138
03 0.93697674 4th iteration 4th iteration 4th iteration
│ER│= 0.00018% │ER│= 0.00008% │ER│= 0.00050%
www.iosrjournals.org 25 | Page
A Modified Predictor-Corrector Formula For Solving Ordinary Differential Equation Of First Order
3.20870662 3.20868230 3.20873816
04 3.20868167 6th iteration 5th iteration 6th iteration
│ER│= 0.00078% │ER│= 0.00002% │ER│= 0.00176%
0.52495807 0.52496060
0.52495910
4th iteration 4th iteration
05 0.52495805 4th iteration
│ER│= 0.000004% │ER│= 0.000486%
│ER│= 0.000200%
VI. Conclusions
From above comparison table we have shown that Milne’s (modified) predictor-corrector formulae
gives better accuracy and it also can minimize the calculating time as it takes less number of iterations. But, it is
yet to implement proposed formulae to the real world problems. Though Milne’s (modified) predictor-corrector
formula seems to be lengthy process of solving ordinary differential equations, it has following advantages over
previous methods. Such as, (i) the previous methods estimates the value of y respecting a given value of x by
means of four initial conditions whereas the Milne’s (modified) predictor-corrector formulae estimate the value
of y respecting a given value of x by means of five initial conditions, which is more logical, (ii) to obtain value
of y at any value of x, previous methods are need to be calculating up to fourth difference of Newton’s formula
of forward interpolation but Milne’s (modified) predictor-corrector formulae need to be calculating up to fifth
difference of Newton’s formula of forward interpolation, which will give better accuracy, (iii) at Milne’s
(modified) corrector formula the co-efficient of is zero, then the truncation error converging to zero, this will
upgrade the level of accuracy of the method. Next, we are to construct a generalized formula for predictor-
corrector method for solving ordinary differential equations of first order and first degree.
Acknowledgement
Authors are highly grateful to Professor Dr. Munshi Nazrul Islam, Department of Mathematics,
University of Chittagong, Bangladesh for his valuable suggestions of this work.
References
[1] A. RALSTON and P. RABINOWITZ, “A first course in numerical analysis”, McGraw-Hill Book Company, Auckland, 1988, PP.
193.
[2] A. R. VASISTHA, V. VASISTHA, “Numerical analysis”, KedarNath-Ram Nath, Meerut, 1999, PP.284,288.
[3] B. D. SHARMA, “Differential equations”, KedarNath-Ram Nath, Meerut, 2006, PP.06
[4] B. S. GOEL and S. K. MITTAL, “Numerical analysis”, Pragati-Prakashan, India, 1995, PP.518
[5] V. N. VEDAMURTHY and N. Ch. S. N. IYENGAR, “Numerical methods”, Vikas Publishing house Private Ltd., New Delhi, 2002,
PP.11.49, 11.21, 11.58
[6] FRANCIS SCHELD, “Numerical analysis”, Schaum’s Outline Series, McGraw-Hill, 1988, PP.471.
[7] JAMES B. SCARBOROUGH, “ Numerical mathematical analysis” Oxford and IBM Publishing Co. Pvt. Ltd., 1966, PP.354.
[8] S. BALACHANDRA RAO and C. K. SHANTHA, “Numerical methods”, Universities Press India Ltd., Hyderabad, 2000, PP.386.
[9] S. S. SASTRY, 2002. “Introductory methods of numerical analysis”, Third edition, Prentic-Hall of India Private Ltd., New Delhi,
2002, PP.303.
www.iosrjournals.org 26 | Page
