A Nonclassical Collocation Method For Solving Two-Point Boundary Value Problems Over Infinite Intervals

Similar documents
Quadrature Rules for Evaluation of Hyper Singular Integrals

New Expansion and Infinite Series

The Islamic University of Gaza Faculty of Engineering Civil Engineering Department. Numerical Analysis ECIV Chapter 11

Discrete Least-squares Approximations

Research Article Numerical Treatment of Singularly Perturbed Two-Point Boundary Value Problems by Using Differential Transformation Method

A Modified ADM for Solving Systems of Linear Fredholm Integral Equations of the Second Kind

A Bernstein polynomial approach for solution of nonlinear integral equations

Numerical quadrature based on interpolating functions: A MATLAB implementation

NUMERICAL INTEGRATION. The inverse process to differentiation in calculus is integration. Mathematically, integration is represented by.

Orthogonal Polynomials

On the Decomposition Method for System of Linear Fredholm Integral Equations of the Second Kind

Matrices, Moments and Quadrature, cont d

Application Chebyshev Polynomials for Determining the Eigenvalues of Sturm-Liouville Problem

Goals: Determine how to calculate the area described by a function. Define the definite integral. Explore the relationship between the definite

arxiv: v1 [math.na] 23 Apr 2018

Modification Adomian Decomposition Method for solving Seventh OrderIntegro-Differential Equations

Fredholm Integral Equations of the First Kind Solved by Using the Homotopy Perturbation Method

Monte Carlo method in solving numerical integration and differential equation

Solutions of Klein - Gordan equations, using Finite Fourier Sine Transform

Matrix Eigenvalues and Eigenvectors September 13, 2017

Numerical integration

Numerical Solutions for Quadratic Integro-Differential Equations of Fractional Orders

Orthogonal Polynomials and Least-Squares Approximations to Functions

Lecture 6: Singular Integrals, Open Quadrature rules, and Gauss Quadrature

Operations with Polynomials

Lecture 14: Quadrature

Quantum Physics II (8.05) Fall 2013 Assignment 2

Polynomial Approximations for the Natural Logarithm and Arctangent Functions. Math 230

Numerical Integration

Jordan Journal of Mathematics and Statistics (JJMS) 11(1), 2018, pp 1-12

u t = k 2 u x 2 (1) a n sin nπx sin 2 L e k(nπ/l) t f(x) = sin nπx f(x) sin nπx dx (6) 2 L f(x 0 ) sin nπx 0 2 L sin nπx 0 nπx

CLOSED EXPRESSIONS FOR COEFFICIENTS IN WEIGHTED NEWTON-COTES QUADRATURES

Solution to Fredholm Fuzzy Integral Equations with Degenerate Kernel

Arithmetic Mean Derivative Based Midpoint Rule

Lecture 19: Continuous Least Squares Approximation

3.4 Numerical integration

QUADRATURE is an old-fashioned word that refers to

63. Representation of functions as power series Consider a power series. ( 1) n x 2n for all 1 < x < 1

Advanced Computational Fluid Dynamics AA215A Lecture 3 Polynomial Interpolation: Numerical Differentiation and Integration.

Review of Calculus, cont d

Math 270A: Numerical Linear Algebra

The Algebra (al-jabr) of Matrices

Jim Lambers MAT 169 Fall Semester Lecture 4 Notes

Numerical Analysis. 10th ed. R L Burden, J D Faires, and A M Burden

Physics 116C Solution of inhomogeneous ordinary differential equations using Green s functions

Linear and Non-linear Feedback Control Strategies for a 4D Hyperchaotic System

An optimal 3-point quadrature formula of closed type and error bounds

Ordinary differential equations

Lecture 23: Interpolatory Quadrature

Introduction to Finite Element Method

Travelling Profile Solutions For Nonlinear Degenerate Parabolic Equation And Contour Enhancement In Image Processing

Construction of Gauss Quadrature Rules

A Direct Transformation of a Matrix Spectrum

Section 6.1 INTRO to LAPLACE TRANSFORMS

CBE 291b - Computation And Optimization For Engineers

Lecture 17. Integration: Gauss Quadrature. David Semeraro. University of Illinois at Urbana-Champaign. March 20, 2014

New implementation of reproducing kernel Hilbert space method for solving a class of functional integral equations

A-Level Mathematics Transition Task (compulsory for all maths students and all further maths student)

5.7 Improper Integrals

Numerical Methods I Orthogonal Polynomials

SOLUTION OF QUADRATIC NONLINEAR PROBLEMS WITH MULTIPLE SCALES LINDSTEDT-POINCARE METHOD. Mehmet Pakdemirli and Gözde Sarı

Numerical Linear Algebra Assignment 008

Families of Solutions to Bernoulli ODEs

MAC-solutions of the nonexistent solutions of mathematical physics

Unit 1 Exponentials and Logarithms

1.2. Linear Variable Coefficient Equations. y + b "! = a y + b " Remark: The case b = 0 and a non-constant can be solved with the same idea as above.

37 Kragujevac J. Math. 23 (2001) A NOTE ON DENSITY OF THE ZEROS OF ff-orthogonal POLYNOMIALS Gradimir V. Milovanović a and Miodrag M. Spalević

M344 - ADVANCED ENGINEERING MATHEMATICS

Multiple Positive Solutions for the System of Higher Order Two-Point Boundary Value Problems on Time Scales

Application of Exact Discretization for Logistic Differential Equations to the Design of a Discrete-Time State-Observer

On Second Derivative-Free Zero Finding Methods

LYAPUNOV-TYPE INEQUALITIES FOR NONLINEAR SYSTEMS INVOLVING THE (p 1, p 2,..., p n )-LAPLACIAN

Math 131. Numerical Integration Larson Section 4.6

LINEAR ALGEBRA AND MATRICES. n ij. is called the main diagonal or principal diagonal of A. A column vector is a matrix that has only one column.

Chapter 6 Notes, Larson/Hostetler 3e

FUZZY HOMOTOPY CONTINUATION METHOD FOR SOLVING FUZZY NONLINEAR EQUATIONS

B.Sc. in Mathematics (Ordinary)

NUMERICAL INTEGRATION

Modern Methods for high-dimensional quadrature

SCHOOL OF ENGINEERING & BUILT ENVIRONMENT. Mathematics

Numerical Integration. 1 Introduction. 2 Midpoint Rule, Trapezoid Rule, Simpson Rule. AMSC/CMSC 460/466 T. von Petersdorff 1

Chapter 4 Contravariance, Covariance, and Spacetime Diagrams

The First Fundamental Theorem of Calculus. If f(x) is continuous on [a, b] and F (x) is any antiderivative. f(x) dx = F (b) F (a).

Solution of First kind Fredholm Integral Equation by Sinc Function

Oscillatory Quadratures Based on the Tau Method

An iterative method for solving nonlinear functional equations

CHAPTER 6b. NUMERICAL INTERPOLATION

AN INTEGRAL INEQUALITY FOR CONVEX FUNCTIONS AND APPLICATIONS IN NUMERICAL INTEGRATION

and that at t = 0 the object is at position 5. Find the position of the object at t = 2.

Chapter D05 Integral Equations

Math 520 Final Exam Topic Outline Sections 1 3 (Xiao/Dumas/Liaw) Spring 2008

Theoretical foundations of Gaussian quadrature

Sturm-Liouville Eigenvalue problem: Let p(x) > 0, q(x) 0, r(x) 0 in I = (a, b). Here we assume b > a. Let X C 2 1

1 The Lagrange interpolation formula

1 2-D Second Order Equations: Separation of Variables

Adomian Decomposition Method with Green s. Function for Solving Twelfth-Order Boundary. Value Problems

In Section 5.3 we considered initial value problems for the linear second order equation. y.a/ C ˇy 0.a/ D k 1 (13.1.4)

Lecture 12: Numerical Quadrature

4.5 THE FUNDAMENTAL THEOREM OF CALCULUS

Chapter 3 Polynomials

Transcription:

Austrlin Journl of Bsic nd Applied Sciences 59: 45-5 ISS 99-878 A onclssicl Colloction Method For Solving o-oint Boundry Vlue roblems Over Infinite Intervls M Mlei nd M vssoli Kni Deprtment of Mthemtics Mobreh Brnch Islmic Azd University Isfhn Irn Deprtment of Mthemtics Khorsgn Brnch Islmic Azd University Isfhn Irn Abstrct: A colloction method for solving to-point boundry vlue problems posed on n infinite intervl involving second order liner differentil eqution is proposed By reducing the infinite intervl to finite intervl tht is lrge nd pproimting the vrible using grnge interpoltion the resulting boundry vlue problem is reduced to n lgebric system by using nonclssicl pseudospectrl method he pplictions re demonstrted vi test emples Key ords: Colloction method o-point boundry vlue problem infinite intervl IRODUCIO During the lst fe yers much progress hs been mde in the numericl tretment of initil vlue problems nd boundry vlue problems over infinite intervls see for emple Ibiol ; Rvi Knth 3 nd references therein ypiclly these problems rise very frequently in fluid dynmics erodynmics quntum mechnics electronics nd other domins of science A fe notble emples re the Von Krmn sirling flos entini 98; Mroich 98 combined forced nd free convection over horizontl plte Schneider 979 nd eigenvlue problems for the Schrodinger eqution entini 98 In mny cses the domin of the governing equtions of these problems is infinite or semi-infinite so tht the specil tretment is required for these so clled infinite intervl problems he nlyticl solutions for these problems re not redily ttinble nd thus the problem is brought to the problem of finding efficient computtionl lgorithms for obtining numericl solution In the present pper e consider the liner to-point boundry vlue problem of the from y y y - Q y R ith y b y c or lim y c 3 here Q nd R re continuous functions nd Q We ssume tht -3 hve unique solution y to be determined Before computing the solution e plummet the infinite intervl to finite but lrge one so tht finite point represent infinity his is stndrd pproch of solving such problems tht re posed on infinite intervls nd nmed domin trunction he boundry condition t infinity is replced ith the sme conditions t finite vlue his provides very ccurte results provided tht is sufficiently lrge De Hoog nd Weiss De Hoog FR 979 proposed n nlyticl trnsformtion of the independent vrible tht reduces the originl problem to boundry vlue problem over finite intervl Usully tht produces singulrity of the second ind t the origin nd must be solved by suitble difference methods In this pper e introduce ne nonclssicl colloction method for solving -3 his method consists of reducing the solution of -3 to set of liner lgebric equtions by first epnding y in terms of grnge interpolting polynomils bsed on set of nonclssicl Guss-obtto G nodes hese nodes hich rise from nonclssicl orthogonl polynomils bsed on n rbitrry eight function over intervl [ ] re presented is then collocted t these G colloction points to evlute the unnon coefficients hich re the vlues of the function y t these colloction points his pper is orgnized s follos: the folloing section is devoted to the genertion of G colloction points nd function pproimtion In section 3 e eplin our method nd in section 4 e give our numericl findings by considering to emples to demonstrte the ccurcy nd pplicbility of the proposed technique Corresponding Author: M Mlei Deprtment of Mthemtics Mobreh Brnch Islmic Azd University Isfhn Irn E-mil: mm_mlei5@yhoocom 45

Aust J Bsic & Appl Sci 59: 45-5 46 onclssicl seudospectrl Method: gl oints nd Weights: In clssicl pseudospectrl methods Elngr 995; Elngr 998 the clssicl Guss-obtto colloction points re bsed on Chebyshev or egendre polynomils nd lie on the closed intervl [-] In the present or e consider the genertion of the G colloction points bsed on nonclssicl orthogonl polynomils ith respect to rbitrry eights in the intervl ] [ et be the number of colloction points nd t be the th-degree nonclssicl orthogonl polynomil ith respect to the eight t hich cn be obtined from the folloing three-term recurrence reltion Shizgl B 98 he recurrence coefficients in re given in Shizgl 996 by d d t d d d he G colloction points nd eights for re obtined by the method outlined by Golub Golub GH 973 he tridigonl Jcobi-obtto mtri of order is defined by 3 J here re the solution of the liner system heorem Golub 973 he Guss-obtto nodes re the eigenvlues of J nd the Guss-obtto eights re given by v here v is the normlized eigenvector of J corresponding to the eigenvlue v v e i nd v its firs component Function Approimtion: In order to interpolte function ] [ f t the point ] [ e use the grnge interpoltion of degree of the form f F f ] [ here re set of G colloction points interpoltion nodes in ] [ nd

Aust J Bsic & Appl Sci 59: 45-5 Differentiting the series of to times nd evluting the result t the collo-ction points gives m m d m F [ ] f D f m 3 m d here the coefficients D re entries of n squre differentition mtri m 3 Solution of o-oint Boundry Vlue roblems: In order to solve -3 ith nonclssicl colloction method e first choose n intervl [ is ny positive integer et y y y here y [ y y y ] nd [ ] Differentiting 3 nd using 3 e obtin 3 y y y D 33 y y y D 33 here D [ ] nd D [ ] By substituting 3 33 in e get y y D y D Q y R 34 hich cn be reritten s y y [ D D Q ] R 35 We no collocte 35 t G colloction points s y y [ D D Q ] R 36 he sets of G colloction points re defined on the intervls [ ] In ddition using e get e here e is n vector hose th component is one nd other components re zero Furthermore using 3 e pproimte the boundry conditions in 3 s follos: y y b 37 y y c 38 here is chosen so tht the computed solution pproimtes the ctul solution Using 36 38 e obtin system of liner equtions Solving this system nd substituting the obtined vlues of vector y in 3 the pproimte solutions cn be obtined 4 Computtionl Results: In this section e hve implnted the present method on to emples he numericl results re compred ith ect results nd other results in the literture here re mny orthogonl eights functions tht cn be used Mlei ; Chen In this pper e use only to eights he cses re summrized in ble ble : Different Cse e 47

Aust J Bsic & Appl Sci 59: 45-5 Emple : Consider the boundry vlue problem y y y y e 4 ith y 4 y 43 his problem hs erlier been considered in Kdlboo 984 nd Rvi Knth 3 nd its ect solution is 3 ye e e he boundry vlue problem given by 4-43 hs been solved using nonclssicl colloction method for the cses given in ble he numericl results re presented in bles -3 nd compred ith the method in Rvi Knth 3 nd the ect solutions ote tht in Rvi Knth 3 defines domin trunction [ ] Define E m ye y : here y nd y e re pproimted nd ect solution respectively In tble 4 the mimum bsolute errors E for different vlues of nd re given ble 4 shos tht in this method by chnging the eight function the obtined results cn be improved ble : Computtionl results of y for emple = 8 Ect Method in Rvi Knth resent resent 3 8 h Cse 64 5 Cse 9 759-776 - 733-759 - 3 3773-37748 - 37634-3773 - 4 7674-3 76737-3 74839-3 7676-3 5 3839-4 388-4 544-4 385-4 6 3-5 35-5 9983-5 3653-5 7 4838-6 47943-6 647676-6 483-6 75-6 5358-6 -6 84398-6 499 ble 3: Computtionl results of y for emple = Ect Method in Rvi Knth resent resent 3 h Cse 64 Cse 6 759-776 - 759-759 - 3 3773-37748 - 3773-3773 - 4 7674-3 76737-3 76757-3 7674-3 5 3839-4 3883-4 3555-4 3839-4 6 3-5 368-5 3968-5 37-5 7 4838-6 489-6 4477-6 483-6 8 56486-7 563886-7 6963-7 56966-7 9 76547-8 758-8 563386-8 7347-8 95-8 847-8 88984-8 333-8 Emple : A second emple y y y 44 ith y 45 y 46 48

Aust J Bsic & Appl Sci 59: 45-5 ble 4: Mimume bsolute error Cse E for emple 8 cse nd 5-3 5-5 56-8 8 cse nd 8-5 45-5 86-8 cse nd 5-3 5 64-5 -7 6-9 cse nd - 5 65-4 5-6 6 3-9 his emple hs erlier been considered in Robertson 97 nd lter in Rvi Knth 3 For this problem e consider intervls of the form [ We pplied the method presented in this pper nd solved 44-46 nd then evluted the different vlues of y hich lso ere evluted in Rvi Knth 3 by using forth order finite difference method he computtionl results re presented in bles 5-6 nd comprison is mde ith the method outlined in Rvi Knth 3 ble 5: Computtionl results of y for emple =6 Method in [] resent resent 5 h 3 Cse 5 Cse 5 987 983 987 3 85347-8588 - 8535-5 3987-39546 - 3987-7 439-364 - 437-9 9537-967 - 9545-7965 - 7975-7938 - 3 54554-559 - 55633-5 45-937 - 9358 - ble 6: Computtionl results of y for emple 6 Method in [] 5 h 3 resent Cse E resent Cse 3 3 85347-85334 - 85345-5 3987-39566 - 39875-7 443-464 - 443-9 957-964 - 957-7954 - 79456-795 - 3 568469-568998 - 568464-5 4693-47353 - 7 33359-33369 - 33369-4377 - 485-5 844953-3 3544-344 - 49

Aust J Bsic & Appl Sci 59: 45-5 Conclusion: A nonclssicl colloction method hs been used for the pproimte solution of to-point boundry vlue problems over infinite intervls he orthogonl eight function Wnd the intervl of definition of orthogonl polynomils cn be chosen rbitrrily hich me this method computtionlly very ttrctive becuse vriety rnge of Guss-obtto colloction points cn be utilized Emples sho the efficiency nd ccurcy of this method REFERECES Chen H BD Shizgl A spectrl solution of the Sturm-iouville eqution: comprison of clssicl nd nonclssicl bsis sets J Comput Appl Mth 36: 7-35 De Hoog FR R Weiss 979 he numericl solution of boundry vlue problems ith n essentil singulrity SIAM J umer Anl 6: 637-669 Elngr G MA Kzemi 998 seudospectrl Chebyshev optiml control of constrined nonliner dynmicl systems Comput Opt Appl : 95-7 Elngr G MA Kzemi M Rzzghi 995 he pseudospectrl egendre method for discretizing optiml control problems IEEE rns Automt Cont 4: 793-796 Golub GH 973 Some modified mtri eigenvlue problems SIAM Rev 5: 38-334 Ibiol EA RB Ogunrinde One ne numericl scheme for the solution of Initil vlue problems Ivps in ordinry differentil equtions Aust J Bsic Appl Sci 4: 577-58 Kdlboo MK KS Rmn 984 Discrete invrint imbedding for the numericl solution of boundry vlue problems over infinity intervls Appl Mth Comput 5: 35-34 entini M HB Keller 98 Boundry vlue problems over semi-infinite intervls nd their numericl solution SIAM J umer Anl 7: 577-64 entini M Keller HB 98 he Von Krmn sirling flos SIAM J Appl Mth 38: 5-64 Mlei M M Mshli-Firouzi A numericl solution of problems in clculus of vrition using direct method nd nonclssicl prmeteriztion J Comput Appl Mth 34: 364-373 Mroich A 98 Asymptotic nlysis of Von Krmn flos SIAM J Appl Mth 4: 549-557 Rvi Knth ASV Y Reddy 3 A numericl method for solving to-point boundry vlue problems over infinite intervls Appl Mth Comput 44: 483-494 Robertson 97 he liner to point boundry vlue problems on n infinite intervl Mth Comput 5: 475-483 Schneider 979 A similrity solution for combined forced nd free convection horizontl plte Int J Het Mss rnsf : 4-46 Shizgl B 98 A Gussin qudrture procedure for use in the solution of the Boltzmn eqution nd relted problems J Comput hys 4: 39-38 Shizgl B H Chen 996 he qudrture discretiztion method QDM in the solution of the Schrodinger eqution ith nonclssicl bsis functions J Chem hys 4: 437-45 5