Solution of First Order Initial Value Problem by Sixth Order Predictor Corrector Method

Similar documents
Continuous Block Hybrid-Predictor-Corrector method for the solution of y = f (x, y, y )

Fourth Order RK-Method

Modified Milne Simpson Method for Solving Differential Equations

Runge Kutta Collocation Method for the Solution of First Order Ordinary Differential Equations

A Modified Predictor-Corrector Formula For Solving Ordinary Differential Equation Of First Order And First Degree

Numerical Differential Equations: IVP

multistep methods Last modified: November 28, 2017 Recall that we are interested in the numerical solution of the initial value problem (IVP):

Chapter 8. Numerical Solution of Ordinary Differential Equations. Module No. 2. Predictor-Corrector Methods

9.6 Predictor-Corrector Methods

Differential Equations

One-Step Hybrid Block Method with One Generalized Off-Step Points for Direct Solution of Second Order Ordinary Differential Equations

ECE257 Numerical Methods and Scientific Computing. Ordinary Differential Equations

Numerical Methods - Initial Value Problems for ODEs

Applied Math for Engineers

One Step Continuous Hybrid Block Method for the Solution of

Z. Omar. Department of Mathematics School of Quantitative Sciences College of Art and Sciences Univeristi Utara Malaysia, Malaysia. Ra ft.

Modified block method for the direct solution of second order ordinary differential equations

Jim Lambers MAT 772 Fall Semester Lecture 21 Notes

Second Order ODEs. CSCC51H- Numerical Approx, Int and ODEs p.130/177

Integration of Ordinary Differential Equations

Solving Ordinary Differential Equations

Adebayo O. Adeniran1, Saheed O. Akindeinde2 and Babatunde S. Ogundare2 * Contents. 1. Introduction

TS Method Summary. T k (x,y j 1 ) f(x j 1,y j 1 )+ 2 f (x j 1,y j 1 ) + k 1

NORTH MAHARASHTRA UNIVERSITY JALGAON.

5.6 Multistep Methods

Mathematics for chemical engineers. Numerical solution of ordinary differential equations

8.1 Introduction. Consider the initial value problem (IVP):

16.7 Multistep, Multivalue, and Predictor-Corrector Methods

Module 6 : Solving Ordinary Differential Equations - Initial Value Problems (ODE-IVPs) Section 1 : Introduction

2.1 NUMERICAL SOLUTION OF SIMULTANEOUS FIRST ORDER ORDINARY DIFFERENTIAL EQUATIONS. differential equations with the initial values y(x 0. ; l.

Unit I (Testing of Hypothesis)

-Stable Second Derivative Block Multistep Formula for Stiff Initial Value Problems

KINGS COLLEGE OF ENGINEERING DEPARTMENT OF MATHEMATICS ACADEMIC YEAR / EVEN SEMESTER QUESTION BANK

COSC 3361 Numerical Analysis I Ordinary Differential Equations (II) - Multistep methods

Do not turn over until you are told to do so by the Invigilator.

Milne s Spreadsheet Calculator Using VBA with Excel Programming for solving Ordinary Differential Equations

HIGHER ORDER METHODS. There are two principal means to derive higher order methods. b j f(x n j,y n j )

NUMERICAL ANALYSIS SYLLABUS MATHEMATICS PAPER IV (A)

Initial value problems for ordinary differential equations

Chapter 8. Numerical Solution of Ordinary Differential Equations. Module No. 1. Runge-Kutta Methods

Two Step Hybrid Block Method with Two Generalized Off-step Points for Solving Second Ordinary Order Differential Equations Directly

Solving Delay Differential Equations (DDEs) using Nakashima s 2 Stages 4 th Order Pseudo-Runge-Kutta Method

Review Higher Order methods Multistep methods Summary HIGHER ORDER METHODS. P.V. Johnson. School of Mathematics. Semester

On Nonlinear Methods for Stiff and Singular First Order Initial Value Problems

1 Ordinary Differential Equations

Block Algorithm for General Third Order Ordinary Differential Equation

A Variable-Step Double-Integration Multi-Step Integrator

Predictor Corrector Methods of High Order for Numerical Integration of Initial Value Problems

Ordinary differential equations - Initial value problems

Solving PDEs with PGI CUDA Fortran Part 4: Initial value problems for ordinary differential equations

+ h4. + h5. 6! f (5) i. (C.3) Since., it yields

On the Computational Procedure of Solving Boundary Value Problems of Class M Using the Power Series Method

Uniform Order Legendre Approach for Continuous Hybrid Block Methods for the Solution of First Order Ordinary Differential Equations

AN OVERVIEW. Numerical Methods for ODE Initial Value Problems. 1. One-step methods (Taylor series, Runge-Kutta)

A NOTE ON EXPLICIT THREE-DERIVATIVE RUNGE-KUTTA METHODS (ThDRK)

2.29 Numerical Fluid Mechanics Fall 2011 Lecture 20

M.SC. PHYSICS - II YEAR

MTH 452/552 Homework 3

AP Calculus Testbank (Chapter 9) (Mr. Surowski)

Ordinary Differential Equations

Math 128A Spring 2003 Week 11 Solutions Burden & Faires 5.6: 1b, 3b, 7, 9, 12 Burden & Faires 5.7: 1b, 3b, 5 Burden & Faires 5.

Virtual University of Pakistan

Lecture Notes on Numerical Differential Equations: IVP

A three point formula for finding roots of equations by the method of least squares

BHARATHIAR UNIVERSITY, COIMBATORE. B.Sc. Mathematics CA (Revised papers with effect from onwards)

Chapter 5 Exercises. (a) Determine the best possible Lipschitz constant for this function over 2 u <. u (t) = log(u(t)), u(0) = 2.

PowerPoints organized by Dr. Michael R. Gustafson II, Duke University

VEER NARMAD SOUTH GUJARAT UNIVERSITY, SURAT. SYLLABUS FOR B.Sc. (MATHEMATICS) Semester: III, IV Effective from July 2015

Multistep Methods for IVPs. t 0 < t < T

1 Error Analysis for Solving IVP

Math 225 Differential Equations Notes Chapter 1

Romberg Integration and Gaussian Quadrature

Scientific Computing: An Introductory Survey

Numerical Methods in Physics and Astrophysics

Computational Methods

Hybrid Methods for Solving Differential Equations

Department of Mathematics Faculty of Natural Science, Jamia Millia Islamia, New Delhi-25

CS 450 Numerical Analysis. Chapter 9: Initial Value Problems for Ordinary Differential Equations

Some New Three Step Iterative Methods for Solving Nonlinear Equation Using Steffensen s and Halley Method

On the reliability and stability of direct explicit Runge-Kutta integrators

Previous Year Questions & Detailed Solutions

MATHEMATICAL METHODS INTERPOLATION

Module 4: Numerical Methods for ODE. Michael Bader. Winter 2007/2008

USHA RAMA COLLEGE OF ENGINEERING & TECHNOLOGY

16.7 Multistep, Multivalue, and Predictor-Corrector Methods

Lecture Notes to Accompany. Scientific Computing An Introductory Survey. by Michael T. Heath. Chapter 9

Applied Numerical Analysis

Numerical Solution of Linear Fredholm Integro-Differential Equations by Non-standard Finite Difference Method

Part IB Numerical Analysis

HIGH ORDER VARIABLE MESH EXPONENTIAL FINITE DIFFERENCE METHOD FOR THE NUMERICAL SOLUTION OF THE TWO POINTS BOUNDARY VALUE PROBLEMS

Numerical Methods. Scientists. Engineers

NUMERICAL SOLUTION OF ODE IVPs. Overview

PowerPoints organized by Dr. Michael R. Gustafson II, Duke University

arxiv: v1 [math.na] 31 Oct 2016

The Definition and Numerical Method of Final Value Problem and Arbitrary Value Problem Shixiong Wang 1*, Jianhua He 1, Chen Wang 2, Xitong Li 1

CHAPTER 5: Linear Multistep Methods

A NEW INTEGRATOR FOR SPECIAL THIRD ORDER DIFFERENTIAL EQUATIONS WITH APPLICATION TO THIN FILM FLOW PROBLEM

Ordinary Differential Equations

SOLUTION OF EQUATION AND EIGENVALUE PROBLEMS PART A( 2 MARKS)

Initial Value Problems

Transcription:

Global Journal of Pure and Applied Mathematics. ISSN 0973-1768 Volume 13, Number 6 (2017), pp. 2277 2290 Research India Publications http://www.ripublication.com/gjpam.htm Solution of First Order Initial Value Problem by Sixth Order Predictor Corrector Method R. M. Dhaigude P. G. Department of Mathematics, Government Vidarbha Institute of Science and Humanities, Amravati-444 604 (M.S) India. R. K. Devkate Department of Mathematics, Dr. Babasaheb Ambedkar Marathwada University, Aurangabad-431 004 (M.S) India. Abstract In this paper, we developed a new sixth order predictor-corrector method by using Newton s forward interpolation formula to obtain the solution of first order initial value problem. The basic properties of the derived corrector is investigated and found to be zero stable, consistent and convergent. The efficiency of the method is tested through numerical examples. Also, it found that this method gives better solution, comparative to other methods. AMS subject classification: 65L05, 65L06, 65L20. Keywords: Newton s forward interpolation formula, Initial value problem, Error analysis, Sixth order predictor-corrector method. 1. Introduction Initial value problems plays a very important role in the study of science, engineering and technology. Many problems in day today life forms as an initial value problem (IVP). This attracts many researchers have been involved in the field of theory, methods and applications of IVP s. It demands to study reliable and efficient technique to obtain the exact or approximate solution of IVP s. It is easy to solve linear IVP s and

2278 R. M. Dhaigude and R. K. Devkate for which analytical methods are available in literature too. The main difficulties lies between in nonlinear IVP s in such cases numerical methods will help to obtain the solution. The literature is existed on the solution of IVP s by using numerical techniques. James and Adesanya [1] discussed constant order predictor corrector algorithm for the solution of IVP. James et al. [2] developed starting order seven method accurately for the solution of first IVP s of first order ordinary differential equations. Adesanya et al. [3] obtained starting the five steps Stomer-Cowell method by Adams-Bashforth method for the solution of the first order ordinary differential equations. Sunday and Odekunle [5] employed new numerical integrator for the solution of IVP s in ordinary differential equations. Odekunle et al. [8] gave the direct integration of first order ordinary differential equations by using 4-Point block method. Furthermore, Abbas [10] employed fourth-order block method for the solution of first order IVP s. Though these method existed but accuracy, stability and convergence of the method matters. At the same time error analysis plays crucial role in numerical methods. In this paper, we made an attempt to obtain the solution of first order IVP by sixth order predictor corrector method (SOPCM). The plan of the paper is as follows. In section 2 we define some basic definitions and concepts. The section 3 is devoted to develop predictor and corrector by using Newton s forward interpolation formula. The consistency, zero stability and convergence of the sixth order corrector is proved in section 4. In last section test problems are solved. It shows that efficiency of the new method is good in comparison of the existing method. Also, conclusions is given at the end of the paper. 2. Preliminary In this section, we define some basic definitions and interpolation formulae [4, 6, 7, 9, 11] which are useful in this paper. Definition 2.1. equation with initial condition Initial value problem Consider the first order ordinary differential dy dx = f(x,y), (2.1) y(x 0 ) = y 0, (2.2) the equation (2.1) with (2.2) is called initial value problem of first order. The existence and uniqueness theorem for IVP (2.1) (2.2) is stated as follows. Theorem 2.2. Let f(x,y) be defined and continuous for all points (x, y) in the region D defined by {(x, y) : a x b, <y< }, where a and b finite, and let there exist a constant L such that for every x,y,y such that

Solution of First Order Initial Value Problem 2279 (x, y) and (x, y ) are both in D: f(x,y) f(x,y ) L y y. (2.3) Then, if η is any number, there exist a unique solution y(x) of the IVP (2.1)-(2.2), where y(x) is continuous and differentiable for all (x, y) in D. The inequality (2.3) is known as a Lipschitz condition where L called Lipschitz constant. Definition 2.3. Newton s forward interpolation formula n(n 1) f(x,y) = f 0 + n f 0 + 2 f 0 + 2! n(n 1)(n 2)(n 3) + 4 f 0 4! + n(n 1)(n 2) 3 f 0 3! n(n 1)(n 2)(n 3)(n 4) 5 f 0 +, 5! (2.4) this is the Newton s forward interpolation formula. 3. Methodology In this section, we are developed predictor and corrector formula by using Newton s forward interpolation formula. 3.1. Development of the Predictor Integrating IVP (2.1) (2.2) from x 0 to x 0 + 5h and using initial condition, we have y 5 = y 0 + x0 +5h x 0 f(x,y)dx. By using Newton s forward interpolation formula (2.4), we get x0 +5h [ ( n 2 n y 5 = y 0 + f 0 + n f 0 + x 0 2 ( n 4 6n 3 + 11n 2 ) 6n + 4 f 0 24 ( n 5 10n 4 + 35n 3 50n 2 + 24n + 120 ) 2 f 0 + ( n 3 3n 2 + 2n 6 ) ] 5 f 0 + dx, ) 3 f 0

2280 R. M. Dhaigude and R. K. Devkate put x = x 0 + nh dx = hdn, 5 [ ( n 2 n y 5 = y 0 + h f 0 + n f 0 + 0 2 ( n 4 6n 3 + 11n 2 ) 6n + 4 f 0 24 ( n 5 10n 4 + 35n 3 50n 2 + 24n + 120 ( = y 0 + h [nf 0 + n2 n 3 2 f 0 + 1 2 + 1 ( n 5 24 5 3n4 2 + 11n3 3 ( n 6 + 1 120 After simplifying, we have 6 2n5 + 35n4 4 3 n2 2 3n 2 ) 4 f 0 50n3 3 ) 2 f 0 + ( n 3 3n 2 + 2n 6 ) ] 5 f 0 + dn, ) 3 f 0 ) 2 f 0 + 1 ( n 4 ) 6 4 n3 + n 2 3 f 0 ] 5 + 12n ) 2 5 f 0 +. 0 [ y 5 = y 0 +h 5f 0 + 25 2 f 0 + 175 12 2 f 0 + 75 8 3 f 0 + 425 144 4 f 0 + 95 ] 288 5 f 0 +.... (3.5) Neglecting fifth and higher order differences and expressing f 0, 2 f 0, 3 f 0 and 4 f 0 in terms of the function, we get. [ y 5 = y 0 + h 5f 0 + 25 ( ) 175 f1 f 0 + 2 12 + 425 ( ) ] f4 4f 3 + 6f 2 4f 1 + f 0, 144 [( = y 0 + h 5 25 2 + 175 12 75 8 + 425 ) ( 25 f 0 + 144 2 175 6 + 225 8 425 36 ( 175 + 12 225 8 + 425 ) ( 75 f 2 + 24 8 425 ) f 3 + 425 ] 36 144 f 4, [ 95 = y 0 + h 144 f 0 25 72 f 1 + 25 6 f 2 175 72 f 3 + 425 ] 144 f 4, y 5 = y 0 + 5h [ ] 19f0 10f 1 + 120f 2 70f 3 + 85f 4, 144 which is called a predictor formula. ( ) 75( ) f2 2f 1 + f 0 + f3 3f 2 + 3f 1 f 0 8 ) f 1

Solution of First Order Initial Value Problem 2281 3.2. Development of the Corrector In the equation (3.5) neglect sixth and higher order differences and expressing f 0, 2 f 0, 3 f 0, 4 f 0 and 5 f 0 in terms of the function, we get. [ y 5 = y 0 + h 5f 0 + 25 ( ) 175( ) 75( ) f1 f 0 + f2 2f 1 + f 0 + f3 3f 2 + 3f 1 f 0 2 12 8 + 425 ( ) 95 ( ) ] f4 4f 3 + 6f 2 4f 1 + f 0 + f5 5f 4 + 10f 3 10f 2 + 5f 1 f 0, 144 288 [( = y 0 + h 5 25 2 + 175 12 75 8 + 425 144 95 ) ( 25 f 0 + 288 2 175 6 + 225 8 425 36 + 475 ) f 1 288 ( 175 + 12 225 8 + 425 24 950 ) ( 75 f 2 + 288 8 425 36 + 950 ) ( 425 f 3 + 288 144 475 ) f 4 + 95 ] 288 288 f 5, [ 95 = y 0 + h 288 f 0 + 125 96 f 1 + 125 144 f 2 + 125 144 f 3 + 125 96 f 4 + 95 ] 288 f 5, y 5 = y 0 + 5h [ ] 19f0 + 75f 1 + 50f 2 + 50f 3 + 75f 4 + 19f 5, 288 which is called a corrector formula. Hence, the following theorem is obvious. Theorem 3.1. If the first order initial value problem with initial condition dy dx = f(x,y), y(x 0 ) = y 0, and f(x,y) is continuous and satisfies a Lipschits condition in the variable y, and f 0 = f(x 0,y 0 ), f 1 = f(x 0 + h, y 1 ), f 2 = f(x 0 + 2h, y 2 ), f 3 = f(x 0 + 3h, y 3 ) and f 4 = f(x 0 + 4h, y 4 ) then predictor y n+1 = y n 4 + 5h [ ] 19fn 4 10f n 3 + 120f n 2 70f n 1 + 85f n, (3.6) 144 and corrector y n+1 = y n 4 + 5h [ ] 19fn 4 + 75f n 3 + 50f n 2 + 50f n 1 + 75f n + 19f n+1, (3.7) 288 for n = 4, 5, 6, 7, 8,... Remark 3.2. Sixth order predictor-corrector method is not a self-starting method. Four additional starting values y 1,y 2,y 3 and y 4 must be given. They are usually computed using the Picard s method or Taylor⣠s series method or Runge Kutta method of fourth order.

2282 R. M. Dhaigude and R. K. Devkate 4. Analysis of the Basic Properties of the Corrector In this section, we will discuss the basic properties of the corrector. 4.1. Order and Error Constant of the Corrector Consider the linear operator L{y(x n ); h} associated with the linear multistep method (3.7) be defined as L{y(x n ); h} =y(x n+1 ) ay(x n 4 ) h 5 b i f n i+1. (4.8) Expanding (4.8) using Taylor s series expansion and comparing the coefficients of h gives where c 0 = 1 a L{y(x n ; h} =c 0 y(x n ) + c 1 hy (x n ) + c 2 h 2 y (x n ) + + c p h p y p (x n ) + c p+1 h p+1 y p+1 (x n ) +, (4.9) c q = 1 q! [1 a( 4)q ] 1 (q 1)! i=0 5 b i (1 i) q 1,q = 1, 2, 3,...,p. i=0 Definition 4.1. The linear operator L and the associated continuous linear multistep method (3.7) is said to be of order p if, in (4.9) c 0 = c 1 = c 2 = = c p = 0 and c p+1 = 0. c p+1 /σ (1) is called the error constat and implies that the local truncation is given by In this method: T n+1 = c p+1 h p+1 y p+1 (x) + O(h p+2 ). (4.10) c 0 = c 1 = c 2 = c 3 = c 4 = c 5 = c 6 = 0 and c 7 = 0 Therefore, the new predictor corrector method is of order six with error constant 247, where σ(1) = 5. 60480 4.2. Zero Stability of the Corrector Definition 4.2. A linear multistep method (3.7) is said to be zero stable, if all the roots of the first characteristic polynomial ρ(ξ) satisfies ξ 1. In this method all the roots of the first characteristic polynomial ρ(ξ) = ξ 5 1 = 0 satisfies ξ 1. Hence SOPC method is zero stable.

Solution of First Order Initial Value Problem 2283 4.3. Consistency of the Corrector Definition 4.3. A linear multistep method (3.7) is said to be consistent, if it has order p 1 and if ρ(1) = 0, ρ (1) = σ(1). Where ρ(ξ) is the first characteristic polynomial and σ(ξ) is the second characteristic polynomial for this method defines as follows respectively. ρ(ξ) = ξ 5 1, σ(ξ) = 95 288 ξ 5 + 125 96 ξ 4 + 125 144 ξ 3 + 125 144 ξ 2 + 125 96 ξ + 95 288. Therefore ρ(1) = 0, ρ (1) = σ(1). Hence SOPC method is consistent. 4.4. Convergence The necessary and sufficient condition for a linear multistep method to be convergent is that it must be consistent and zero stable. Hence new sixth order predictor corrector method is convergent. 5. Numerical Test Problems The following notations are used in the following tables. RKFO Runge-Kutta Method of Fourth Order. MPCM Milne s Predictor-Corrector method. SOPCM Sixth Order Predictor-Corrector method. NNI A new Numerical Integrator (2012). Problem 5.1. Consider a first order initial value problem dy dx The exact solution of problem (5.1) is = x + y, y(0) = 1, 0 x 1, h = 0.1. y(x) = (1 + x) + 2e x. In this Problem result in table 1 shows that solution by RKFO, MPCM and SOPCM and the comparison with the existing method is shown in table 2. Problem 5.2. Consider a first order initial value problem dy dx = xy, y(0) = 1, 0 x 1, h = 0.1.

2284 R. M. Dhaigude and R. K. Devkate Table 1: Numerical results for problem 5.1 x Exact solution RKFO MPCM SOPCM 0.1 1.110341836 1.110341667 1.110341667 1.110341667 0.2 1.242805516 1.242805142 1.242805142 1.242805142 0.3 1.399717616 1.399716994 1.399716995 1.399716995 0.4 1.583649396 1.583648480 1.583649220 1.583648482 0.5 1.797442542 1.797441277 1.797442197 1.797442312 0.6 2.044237600 2.044235924 2.044237745 2.044237208 0.7 2.327505414 2.327503253 2.327505486 2.327504783 0.8 2.651081856 2.651079126 2.651082484 2.651080939 0.9 3.019206222 3.019202827 3.019206897 3.019205009 1 3.436563656 3.436559488 3.436564986 3.436563067 Figure 1: Comparison of solutions of problem 5.1 The exact solution of problem (5.2) is y(x) = e x2 2 In this Problem result in table 3 shows that solution by MPCM and SOPCM and the comparison with the existing method is shown in table 4. Problem 5.3. (Logistic Model) The logistic model finds applications in various fields, among which are; neural networks, Statistics, Medicine, Physics and so on. In Neural

Solution of First Order Initial Value Problem 2285 Table 2: Absolute error in problem 5.1 x RKFO MPCM SOPCM 0.1 0.000000169 0.000000169 0.000000169 0.2 0.000000374 0.000000374 0.000000374 0.3 0.000000622 0.000000621 0.000000621 0.4 0.000000916 0.000000176 0.000000914 0.5 0.000001265 0.000000345 0.000000230 0.6 0.000001676 0.000000145 0.000000392 0.7 0.000002161 0.000000072 0.000000631 0.8 0.000002730 0.000000628 0.000000917 0.9 0.000003395 0.000000675 0.000001213 1 0.000004168 0.000001330 0.000000589 Table 3: Numerical results for problem 5.2 x Exact solution MPCM SOPCM 0.1 1.005012521 1.005012521 1.005012521 0.2 1.020201340 1.020201340 1.020201340 0.3 1.046027860 1.046027860 1.046027860 0.4 1.083287068 1.083287632 1.083287066 0.5 1.133148453 1.133149301 1.133148521 0.6 1.197217363 1.197219137 1.197217472 0.7 1.277621313 1.277623878 1.277621476 0.8 1.377127764 1.377131933 1.377128001 0.9 1.499302500 1.499308427 1.499302840 1 1.648721271 1.648730134 1.648721829 networks for example, the logistic model is used to introduce nonlinearity in the model and/or to clamp signals to within a specific range. In statistics, they are used to model how the probability p of an event may be affected by one or more explanatory variables. In medicine, they are being used to model the growth of tumors. In chemistry, the concentration of reactants and products in autocatalytic reactions follows the logistic function. If the special initial conditions are applied to the logistic model then it forms logistic nonlinear first order IVP dy dx = y(1 y), y(0) = 0.5, 0 x 1, h = 0.1.

2286 R. M. Dhaigude and R. K. Devkate Table 4: Absolute error in problem 5.2 x MPCM SOPCM 0.1 0.000000000 0.000000000 0.2 0.000000000 0.000000000 0.3 0.000000000 0.000000000 0.4 0.000000564 0.000000002 0.5 0.000000848 0.000000068 0.6 0.000001774 0.000000109 0.7 0.000002565 0.000000163 0.8 0.000004169 0.000000237 0.9 0.000005927 0.000000340 1 0.000008863 0.000000558 Figure 2: Comparison of solutions of problem 5.2 The exact solution of problem (5.3) is y(x) = 1 1 + e x This problem was solved by NNI [5]. The numerical result it shown in table 5 and the comparison with the existing method is shown in table 6.

Solution of First Order Initial Value Problem 2287 Table 5: Numerical results for problem 5.3 x Exact solution NNI SOPCM 0.1 0.52497919 0.52497977 0.52497919 0.2 0.54983398 0.54983515 0.54983401 0.3 0.57444252 0.57444417 0.57444253 0.4 0.59868768 0.59868979 0.59868768 0.5 0.62245932 0.62246192 0.62245933 0.6 0.64565632 0.64565927 0.64565631 0.7 0.66818777 0.66819102 0.66818778 0.8 0.68997446 0.68997794 0.68997449 0.9 0.71094948 0.71095312 0.71094952 1 0.73105860 0.73106223 0.73105857 Table 6: Absolute error in problem 5.3 x NNI SOPCM 0.1 0.00000058 0.00000000 0.2 0.00000117 0.00000003 0.3 0.00000165 0.00000001 0.4 0.00000211 0.00000000 0.5 0.00000260 0.00000001 0.6 0.00000295 0.00000001 0.7 0.00000325 0.00000001 0.8 0.00000348 0.00000003 0.9 0.00000364 0.00000004 1 0.00000363 0.00000003 Problem 5.4. Consider a nonlinear first order initial value problem dy dx = xy2, y(0) = 2, 0 x 0.5, h = 0.05. The exact solution of problem (5.4) is y(x) = 2 x 2 1 6. Conclusion In this paper, we have developed a new sixth order predictor-corrector method for the solution of first order initial value problem. It has been shown that the method is zero

2288 R. M. Dhaigude and R. K. Devkate Figure 3: Comparison of solutions of problem 5.3 Table 7: Comparison of numerical results for problem 5.4 x Exact solution Computed Solution Absolute error Relative error % error 0.05 2.005012532 2.005012534 0.000000002 0.000000001 0.0000001 0.10 2.020202020 2.020202030 0.000000010 0.000000005 0.0000005 0.15 2.046035806 2.046035830 0.000000024 0.000000012 0.0000012 0.20 2.083333334 2.083333380 0.000000046 0.000000022 0.0000022 0.25 2.133333334 2.133333581 0.000000247 0.000000116 0.0000116 0.30 2.197802198 2.197802650 0.000000452 0.000000206 0.0000206 0.35 2.279202280 2.279203086 0.000000806 0.000000354 0.0000354 0.40 2.380952380 2.380953839 0.000001459 0.000000613 0.0000613 0.45 2.507836990 2.507839701 0.000002711 0.000001081 0.0001081 0.50 2.666666666 2.666672078 0.000005412 0.000002030 0.0002030 stable and consistent hence convergent. The numerical problems have shown that sixth order predictor-corrector method gives better and consistent approximate solution than existing methods.

Solution of First Order Initial Value Problem 2289 Figure 4: Comparison of solutions of problem 5.4 References [1] A. A. James and A. O. Adesanya (2014). A Note on the construction of constant order predictor corrector algorithm for the solution of dy/dx=f(x,y), J. British Math. and Comp. Sci., 4(6), 886 895. [2] A. A. James, A. O. Adesanya and M. K. Fasasi (2013). Starting order seven method accurately for the solution of first initial value problems of first order ordinary differential equations, Prog. in Appl. Math., 6(1), 30 39. [3] A. O. Adesanya, M. R. Odekunle, M. A. Alkali and A. B. Abubakar (2013). Starting the five steps Stomer-Cowell method by Adams-Bashforth method for the solution of the first order ordinary differential equations, J. African Math. and Comp. Sci. Res., 6(5), 89 93. [4] J. D. Lambert (1973). Computational Methods in Ordinary Differential Equations, John Willey and Sons, New York. [5] J. Sunday and M.R. Odekunle (2012). A New Numerical integrator for the solution of initial value problems in ordinary differential equations, J. The Pacific Sci. and Tech., 13(1), 221 227. [6] K. S. Rao (2004). Numerical Methods for Scientist and Engineers, Second Edition, Prentice-Hall of India, New Delhi.

2290 R. M. Dhaigude and R. K. Devkate [7] M. K. Jain, S. R. K. Iyengar and R. K. Jain (2008). Numerical Methods for Scientific and Engineering Computation, Fifth Edition, New age International Private Ltd Publishers, New Delhi. [8] M. R. Odekunle, A. O. Adesanya and J. Sunday (2012). 4-Point block method for direct integration of first-order ordinary differential equations, J. Int. Engg. Research and Appl., 2(5), 1182 1187. [9] P. B. Patil and U. P. Verma (2009). Numerical Computational Methods, Revised Edition, Narosa Publishing House Private Ltd, New Delhi. [10] S. H. Abbas (2006). Fourth-order block methods for the numerical solution of first order initial value problems, Sarajevo J. of Math., 2(15), 247 258. [11] S. S. Sastry (1997). Introductry Methods of Numerical Analysis, Second Edition, Prentice-Hall of India Private Ltd, New Delhi.