Scribd Logo
Hochladen

Willkommen bei Scribd!

  • (Liviu Gr. Ixaru) Numerical Methods For Differentiation

    Hochgeladen von

    Aamer
    19 Ansichten358 Seiten
    Numerical methods

    Originaltitel

    [Liviu Gr. Ixaru] Numerical Methods for Differentiation

    Copyright

    © © All Rights Reserved

    Verfügbare Formate

    PDF oder online auf Scribd lesen

    Stufen Sie dieses Dokument als nützlich ein?

    Sind diese Inhalte unangemessen?

    Dieses Dokument melden
    Numerical methods

    Copyright:

    © All Rights Reserved
    Als PDF herunterladen oder online auf Scribd lesen
    Markieren Sie unangemessene Inhalte
    19 Ansichten358 Seiten

    (Liviu Gr. Ixaru) Numerical Methods For Differentiation

    Hochgeladen von

    Aamer
    Numerical methods

    Copyright:

    © All Rights Reserved
    Als PDF herunterladen oder online auf Scribd lesen
    Markieren Sie unangemessene Inhalte
    von 358
    Mathematics and Its Applications (East European Series) Managing Editor: M. HAZEWINKEL Centre for Mathematics and Computer Science, Amsterdam, the Netherlands Editorial Board: A. BIALYNICKI-BIRULDA, Institute of Mathematics PKIN, Warsaw, Poland J. KURZWEIL, Mathematics Institute, Academy of Sciences, Prague, Czecho slovakia L. LEINDLER, Bolyai Institute, Szeged, Hungary D. S. MITRINOVIC, Belgrade, Yugoslavia S. ROLEWICZ, Polish Academy of Sciences, Warsaw, Poland I. T. TODOROV, Academy of Sciences, Sofia, Bulgaria BL. H. SENDOV, Bulgarian Academy of Sciences, Sofia, Bulgaria Liviu Gr. Ixaru Institute of Physics and Nuclear Engineering Division of Fundamental Physics, Bucharest, Romania Numerical Methods for Differential Equations and Applications EDITURA ACADEMIEI A Bucuresti, Romania b D. Reidel Publishing Company Na A MEMBER OF THE KLUWER ACADEMIC PUBLISHERS GROUP Ly Dordrecht / Boston / Lancaster Library of Congress Cataloging in Publication Data CIP Txaru. Liviu Gr. Numerical methods for differential equations and applications. (Mathematics and its applications. East European series) Translation of: Metode numerice pentru ecuajii diferengiale cu aplicatii. Includes bibliographical references and index. 1. Differential equations— Numerical solutions. 2. Differential equations, Partial—Numerical solutions. I. Title. Il. Title: Numerical methods. IIT. Series: Mathematics and its ap- plications (D. Reidel Publishing Company). East European series. QA372.1913 1984 515.3'5 83-26948 ISBN 90—277—1597—1 (Reidel) Distributors for the U.S.A. and Canada Kluwer Academic Publishers 190 Old Derby Street, Hingham, MA 02043, U.S.A. Distributors for Albania, Bulgaria, Chinese People’s Republic, Cuba, Czechoslovakia, German Democratic Republic, Hungary, Korean People’s Republic, Mongolia, Poland, Romania, the U.S.S.R., Vietnam and Yugoslavia Editura Academiei, Bucharest, Romania Distributors for all remaining countries Kluwer Academic Publishers Group, P.O. Box 322, 3300 AH Dordrecht, Holland Original title: Metode numerice pentru ecuagii diferentiale cu aplicatii First edition published in 1979 by Editura Academici. First English edition published in 1984 by Editura Academiei and D, Reidel Publ. Co, All Rights Reserved © 1984 by Editura Academiei, Bucharest, Romania No part of the material protected by this copyright notice may be reproduced or utilized in any form or by any means, electronic or mechanical, including photo- copying, recording or by any information storage and retrieval system, without written permission from the copyright owner. Printed in Romania To the memory of my father Contents Editor’s Preface. © 2 2 1. ee ee ix WCULp og pol og doo oR Doo ob oOo odo ooo xi Preliminaries. © 2 2 1 ee ee ee Xi CHAPTER 1. Multistep Methods Strategies to Derive Algorithms ............ Elements of the Theory of Backward Differences 5 The General Algorithm of the Multistep Methods . The Open Methods Euler, Adams-Bashforth, Nystrém . The Closed Methods Adams-Moulton, Milne-Simpson, and Milne eee ae oe re 8 Error Analysis: Local Truncation Error and Consistency. . 10 Error Analysis: Error Propagation and Stability... . . 18 Practical Issues Concerning Multistep Algorithms... . . 36 Predictor-Corrector Procedure for Closed Miultistep Algorithms. 2... oe ee 40 1.10 User-oriented Investigation of the Accumulated Error for Multistep Methods... 2.2.0... ee ee ee 46 1.11 Multistep Methods for Second-Order Differential Equations of Special Form... 1. ee ee 63 1.12 The de Vogelaere Method... .. 2.2... 2... 79 1,13 Methods to Solve Linear Second-Order Differential Equations... 1... ee ee 103 RUC Ie oo og aoa oo oo oo Doo boo oad 112 abbr aane Cora CHAPTER 2. One-Step Methods 2.1 Generalities 2.6... ee ee 115 2.2. The Runge-Kutta Method; Various Versions ..... . 126 vii Editor’s preface Growing specialization and diversification have brought a host of monographs and textbooks on increasingly specialized topics. However, the “tree” of knowledge of mathematics and related fields does not grow only by putting forth new branches. It also happens, quite often in fact, that branches which were thought to be completely disparate are suddenly seen to be related. Further, the kind and level of sophistication of mathematics applied in various sciences has changed drastically in recent years: measure theory is used (non-trivially) in regional and theoretical economics; algebraic geometry interacts with physics; the Minkowsky lemma, coding theory and the structure of water meet one another in packing and covering theory; quantum fields, crystal defects and mathematical programming profit from homotopy theory; Lie algebras are relevant to filtering; and prediction and electrical engineering can use Stein spaces. And in addition to this there are such new emerging subdisciplines as “completely integrable systems”, “chaos, synergetics and large-scale order’’, which are almost impossible to fit into the existing classification schemes. They draw upon widely different sections of mathematics. This program, Mathematics and Its Applications, is devoted to such (new) interrelations as exempli gratia: — a central concept which plays an important role in different mathematical and/or scientific specialized areas; — new applications of the results and ideas from one area of scientific endeavor into another; — influences which the results, problems and concepts of one field of enquiry have and have had on the development of another. The Mathematics and Its Applications programme tries to make available a careful selection of books which fit the philosophy outlined above. With such books, which are stimulating rather than definitive, intriguing rather than encyclopaedic, we hope to contribute something ix x EDITOR'S PREFACE towards better communication among the practitioners in diversified fields. Because of the wealth of scholarly research being undertaken in the Soviet Union, Eastern Europe, and Japan, it was decided to devote special attention to the work emanating from these particular regions. Thus it was decided to start three regional series under the umbrella of the main MIA programme. This book in the Eastern Europe series concerns numerical analysis and, more particularly, the numerical analysis needed to deal with the ordinary differential equations one meets in quantum physics, chemistry and engineering. The book was especially written with such potential active users of numerical methods in mind. The first Romanian edition (now sold out) was particularly well received, so it seemed a good idea to also make this work available to the English reading scientific community. The unreasonable effectiveness of mathematics in science... Enugene Wigner Well, if you knows of a better ‘ole, go to it. Bruce Bairnsfather What is now proved was once only imagined. William Blake Amsterdam, March 1983 As long as algebra and geometry proceeded along separate paths, their advance was slow and their applica~ tions limited. But when these sciences joined com- pany they drew from each other fresh vitality and thenceforward marched on at a rapid pace towards perfection. Joseph Louis Lagrange MicuteL HAZEWINKEL Preface This book is merely addressed to persons who, without being profes- sionals in applied mathematics, are often faced with the problem of numerically solving differential equations. Graduate students and persons who are active in the research in theoretical physics, chemistry or engineering are perhaps the most suited readers. In writing this book, special attention was paid to keep some balance between rigour and accessibility. Thus, use is made of only a background level of knowledge in mathematics which is typically achieved during training in the faculties of physics, chemistry and technical branches. At the same time, although the presentation is generally rigorous, occasions we deliberately sacrificed rigour in favour of ry. The presentation of various methods includes all aspects of interest: theory —- which is necessary for those who are rather deeply engaged in the field —, algorithms and programming advices — for hurried readers. The book also contains a lot of numerical examples treated up to the very end, sometimes just with programs included. The book is organized into five chapters. In each of the first three chapters a definite class of methods is discussed for the solution of the initial value problem for ordinary differential equations. It should be mentioned that, while the first two classes, that is of multistep and of one-step methods, are largely presented in the literature, we also consider a new class of methods (the so-called piecewise perturbation methods) which was developed only recently. As a matter of fact, it is now the first time, to our knowledge, that such methods are treated in a self- standing chapter of a book of numerical methods. The fourth chapter is mainly focussed on the boundary value problems for linear second-order equations. A special section is here devoted to the Schrédinger equation, the basic equation of quantum mechanics. xi xii PREFACE In the fifth chapter a numerical example is taken and solved com- pletely. Algorithms are given for six different methods, with the corres- ponding programs in FORTRAN-4. Special emphasis is here laid on some of the typical programming tricks that make computation faster. This version of the book is substantially modified and enlarged in comparison with the original Romanian edition in a way which is expected to better serve its purpose. In writing this book I benefited from help and encouragement from many colleagues and collaborators. Special thanks are addressed to Dr. Gh. Adam and Dr. M. Rizea for fruitful discussions, especially in the early stages of my work, and to Dr. I. Brandus and Dr. M. I. Cristu for permanent support and encouragement. Finally I wish to thank Silvia Mihiilescu for beneficial stylistic adjustments and Mariana Pripas for carefully typing the manuscript. Bucharest, March 1983 Liviu Ixaru Preliminaries The first two chapters of this book are merely devoted to the numerical methods of solving the initial value problem associated to first-order ordinary differential equations. GENERALITIES, CONVENTIONS AND NOTATION Let f(x, y) be a real-valued function of two variables defined for the real x ¢(a, 6] and all real y. Suppose that y is a real valued function defined on (a, 6], The equation ¥ =flx y) M is called an ordinary differential equation of the first order. Any real valued function y(x) which is differentiable and satisfies (1) for a< x< bis said to be a solution of this differential equation. In general, such a solution, if it exists, will not be unique because an arbitrary constant is introduced on integrating (1). To make the solution unique we must impose an extra condition on the solution. This condition usually takes the form of requiring (x) to have a specified value at a particular value c of the argument x, that is WO=5, (2) where cefa, 6] is the particular value of x and s is the preset value of y(c). Equations (1) and (2) are collectively called an initial value (or Cauchy) problem. Equation (2) is called the initial condition and s is the initial value of y. xiii xiv PRELIMINARIES For practical reasons we will further assume that the point ¢ is just the left-hand end of the domain of the equation, i.e., c= a. The initial value will be denoted by yo so that the initial condition reads H(A) = You Q) This assumption is not restrictive at all for our purpose which is to derive algorithms to numerically solve the initial value problem (1)—@Q). In fact, once we have such algorithms, we can equally apply them to solve the (theoretically more general) problem (1)—(2). To illustrate this, consider the problem a 1+) with a =0, b= 10, ¢=5 and y(5) =s. To solve it numerically by an algorithm for problem (1)—(3) we divide the domain [0, 10] into two subdomains /, = 0, b = 5] and J, = [a, = 5, b = 10]. The numerical integration of it on /, is straight forward because the problem is manifestly of the very form (1)—(3). To solve the problem on I, we must first transform the equation by introducing a new variable y X=5—.x and a new function Y(X)= (x). Now, since ¥(X) = = —y'(x) the problem to be solved becomes yo-1_ 1+¥? for X€ [0,5], with the initial condition ¥(0) = y(5) =. ‘The latter is also of the form (1)—(3) and thus it can be integrated numerically by means of the same algorithm. Another mathematical problem of interest is the one of the systems of first-order differential equations, formulated as follows. A system of N ordinary differential equations of the first order is a system of equations of the form vy =fi(x,yyy,....y%), 1=1,2,...,M, x€ [a, 5], 4) in which f1,f%,...,f are known real functions of their N+ 1 real arguments x, y,..., The initial value (or Cauchy) problem for system (4) consists of finding a set of functions y*(x), y*(x),...,y(x) defined and differentiable on [a, b], which assume some preset initial conditions, vi: yo) =s', §=1,2,...,.N, io) PRELIMINARIES xv for given c ¢[a, 6] and real numbers s4, s*,..., s%, and satisfy system (4) identically in x. Problem (4)—(5) can be more compactly written in vector form, y(x) = f(, y), (6) yo) =, ™ where y,f and s are real N dimensional vectors, y LPP IN) st y=] |, fay =] POY I) sa]? |. (8) yy) "es PPV) sv Once again we will focus on the particular case when c =a, that is y =f, y), x€ [a, 5], (9) ¥(@) = Yo- (10) If eqs (1)—(3) are compared with eqs (9)—(10) it clearly appears that the initial value problems for single equations and for systems of equations are mathematically identical except for the fact that f, y, and s are scalars in the first case, but N-dimensional vectors in the second case. The same holds also true for the mathematical approaches to the two problems in the sense that the mathematical methods of investigation for the second problem, resulting theorems, numerical algorithms, etc. are simply generalizations in vector language of the ones in the scalar case. In connection with each of the problems (1)—(3) and (9)—(10) mathe- maticians usually say that the solution, if it exists, is unique. The question is when does the solution actually exist, or in other words, under which conditions to be satisfied by function f (or the vector of functions f) does the solution exist? The question is answered. by a well-known theorem, the so-called existence theorem. For the scalar case this theorem says that the solution of problem (1)—(3) exists and is unique provided the real function (x, y) satisfies the following two conditions: (AS) f(x, y) is defined and continuous in the strip a¥ fon =tfx,y for —Y¥

    Das könnte Ihnen auch gefallen