Lecture Notes Electrical Engineering

Volume 18

Stanisaw Rosoniec

Fundamental Numerical Methods

for Electrical Engineering

123

Prof. Dr. Hab. Ing. Stanisaw Rosoniec

Institute of Radioelectronics
Warsaw University of Technology
Nowowiejska 15/19
00-665 Warsaw
Poland
[Link]@[Link]

ISBN: 978-3-540-79518-6

e-ISBN: 978-3-540-79519-3

Library of Congress Control Number: 2008927874

c 2008 Springer-Verlag Berlin Heidelberg

This work is subject to copyright. All rights are reserved, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting,
reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication
or parts thereof is permitted only under the provisions of the German Copyright Law of September 9,
1965, in its current version, and permission for use must always be obtained from Springer. Violations are
liable to prosecution under the German Copyright Law.
The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply,
even in the absence of a specific statement, that such names are exempt from the relevant protective laws
and regulations and therefore free for general use.
Cover design: eStudio Calamar S.L.
Printed on acid-free paper
9 8 7 6 5 4 3 2 1
[Link]

Contents

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
1 Methods for Numerical Solution of Linear Equations . . . . . . . . . . . . . . . .
1.1 Direct Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.1.1 The Gauss Elimination Method . . . . . . . . . . . . . . . . . . . . . . . . . .
1.1.2 The GaussJordan Elimination Method . . . . . . . . . . . . . . . . . . .
1.1.3 The LU Matrix Decomposition Method . . . . . . . . . . . . . . . . . . .
1.1.4 The Method of Inverse Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2 Indirect or Iterative Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2.1 The Direct Iteration Method . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2.2 Jacobi and GaussSeidel Methods . . . . . . . . . . . . . . . . . . . . . . .
1.3 Examples of Applications in Electrical Engineering . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1
5
5
9
11
14
17
17
18
23
27

2 Methods for Numerical Solving the Single Nonlinear Equations . . . . . .

2.1 Determination of the Complex Roots of Polynomial Equations
by Using the Lins and Bairstows Methods . . . . . . . . . . . . . . . . . . . . . .
2.1.1 Lins Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.2 Bairstows Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.3 Laguerre Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2 Iterative Methods Used for Solving Transcendental Equations . . . . . .
2.2.1 Bisection Method of Bolzano . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.2 The Secant Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.3 Method of Tangents (NewtonRaphson) . . . . . . . . . . . . . . . . . .
2.3 Optimization Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.4 Examples of Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

29
30
30
32
35
36
37
38
40
42
44
47

3 Methods for Numerical Solution of Nonlinear Equations . . . . . . . . . . . . .

3.1 The Method of Direct Iterations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2 The Iterative Parameter Perturbation Procedure . . . . . . . . . . . . . . . . . . .
3.3 The Newton Iterative Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

49
49
51
52
v

vi

Contents

3.4 The Equivalent Optimization Strategies . . . . . . . . . . . . . . . . . . . . . . . . . 56

3.5 Examples of Applications in the Microwave Technique . . . . . . . . . . . . 58
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4 Methods for the Interpolation and Approximation of One
Variable Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
4.1 Fundamental Interpolation Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.1.1 The Piecewise Linear Interpolation . . . . . . . . . . . . . . . . . . . . . . 72
4.1.2 The Lagrange Interpolating Polynomial . . . . . . . . . . . . . . . . . . . 73
4.1.3 The Aitken Interpolation Method . . . . . . . . . . . . . . . . . . . . . . . . 76
4.1.4 The NewtonGregory Interpolating Polynomial . . . . . . . . . . . . 77
4.1.5 Interpolation by Cubic Spline Functions . . . . . . . . . . . . . . . . . . 82
4.1.6 Interpolation by a Linear Combination of Chebyshev
Polynomials of the First Kind . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.2 Fundamental Approximation Methods for One Variable Functions . . . 89
4.2.1 The Equal Ripple (Chebyshev) Approximation . . . . . . . . . . . . 89
4.2.2 The Maximally Flat (Butterworth) Approximation . . . . . . . . . . 94
4.2.3 Approximation (Curve Fitting) by the Method of Least Squares 97
4.2.4 Approximation of Periodical Functions by Fourier Series . . . . 102
4.3 Examples of the Application of Chebyshev Polynomials
in Synthesis of Radiation Patterns of the In-Phase Linear
Array Antenna . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
5 Methods for Numerical Integration of One and Two
Variable Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
5.1 Integration of Definite Integrals by Expanding the Integrand
Function in Finite Series of Analytically Integrable Functions . . . . . . 123
5.2 Fundamental Methods for Numerical Integration
of One Variable Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
5.2.1 Rectangular and Trapezoidal Methods of Integration . . . . . . . . 125
5.2.2 The Romberg Integration Rule . . . . . . . . . . . . . . . . . . . . . . . . . . 130
5.2.3 The Simpson Method of Integration . . . . . . . . . . . . . . . . . . . . . . 132
5.2.4 The NewtonCotes Method of Integration . . . . . . . . . . . . . . . . . 136
5.2.5 The Cubic Spline Function Quadrature . . . . . . . . . . . . . . . . . . . 138
5.2.6 The Gauss and Chebyshev Quadratures . . . . . . . . . . . . . . . . . . . 140
5.3 Methods for Numerical Integration of Two Variable Functions . . . . . . 147
5.3.1 The Method of Small (Elementary) Cells . . . . . . . . . . . . . . . . . 147
5.3.2 The Simpson Cubature Formula . . . . . . . . . . . . . . . . . . . . . . . . . 148
5.4 An Example of Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
6 Numerical Differentiation of One
and Two Variable Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
6.1 Approximating the Derivatives of One Variable Functions . . . . . . . . . . 157

Contents

vii

6.2

Calculating the Derivatives of One Variable Function

by Differentiation of the Corresponding Interpolating Polynomial . . . 163
6.2.1 Differentiation of the NewtonGregory Polynomial
and Cubic Spline Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
6.3 Formulas for Numerical Differentiation of Two
Variable Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
6.4 An Example of the Two-Dimensional Optimization Problem
and its Solution by Using the Gradient Minimization Technique . . . . 172
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
7 Methods for Numerical Integration of Ordinary
Differential Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
7.1 The Initial Value Problem and Related Solution Methods . . . . . . . . . . . 179
7.2 The One-Step Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
7.2.1 The Euler Method and its Modified Version . . . . . . . . . . . . . . . 180
7.2.2 The Heun Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
7.2.3 The RungeKutta Method (RK 4) . . . . . . . . . . . . . . . . . . . . . . . . 184
7.2.4 The RungeKuttaFehlberg Method (RKF 45) . . . . . . . . . . . . . 186
7.3 The Multi-step PredictorCorrector Methods . . . . . . . . . . . . . . . . . . . . . 189
7.3.1 The AdamsBashforthMoulthon Method . . . . . . . . . . . . . . . . 193
7.3.2 The MilneSimpson Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
7.3.3 The Hamming Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
7.4 Examples of Using the RK 4 Method for Integration
of Differential Equations Formulated for Some Electrical Rectifier
Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
7.4.1 The Unsymmetrical Voltage Doubler . . . . . . . . . . . . . . . . . . . . . 199
7.4.2 The Full-Wave Rectifier Integrated with the Three-Element
Low-Pass Filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
7.4.3 The Quadruple Symmetrical Voltage Multiplier . . . . . . . . . . . . 208
7.5 An Example of Solution of Riccati Equation Formulated
for a Nonhomogenous Transmission Line Segment . . . . . . . . . . . . . . . 215
7.6 An Example of Application of the Finite Difference Method
for Solving the Linear Boundary Value Problem . . . . . . . . . . . . . . . . . . 219
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
8 The Finite Difference Method Adopted for Solving Laplace
Boundary Value Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
8.1 The Interior and External Laplace Boundary Value Problems . . . . . . . 226
8.2 The Algorithm for Numerical Solving of Two-Dimensional Laplace
Boundary Problems by Using the Finite Difference Method . . . . . . . . 228
8.2.1 The Liebmann Computational Procedure . . . . . . . . . . . . . . . . . . 231
8.2.2 The Successive Over-Relaxation Method (SOR) . . . . . . . . . . . 238
8.3 Difference Formulas for Numerical Calculation
of a Normal Component of an Electric Field Vector
at Good Conducting Planes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242

viii

Contents

8.4

Examples of Computation of the Characteristic Impedance

and Attenuation Coefficient for Some TEM Transmission Lines . . . . . 245
8.4.1 The Shielded Triplate Stripline . . . . . . . . . . . . . . . . . . . . . . . . . . 246
8.4.2 The Square Coaxial Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
8.4.3 The Triplate Stripline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
8.4.4 The Shielded Inverted Microstrip Line . . . . . . . . . . . . . . . . . . . . 253
8.4.5 The Shielded Slab Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
8.4.6 Shielded Edge Coupled Triplate Striplines . . . . . . . . . . . . . . . . 263
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
A Equation of a Plane in Three-Dimensional Space . . . . . . . . . . . . . . . . . . . . 269
B The Inverse of the Given Nonsingular Square Matrix . . . . . . . . . . . . . . . . 271
C The Fast Elimination Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
D The Doolittle Formulas Making Possible Presentation of a
Nonsingular Square Matrix in the form of the Product of Two
Triangular Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
E Difference Formula for Calculation of the Electric Potential at
Points Lying on the Border Between two Looseless Dielectric Media
Without Electrical Charges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
F Complete Elliptic Integrals of the First Kind . . . . . . . . . . . . . . . . . . . . . . . 279
Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281

About the Author

Stanisaw Rosoniec received his [Link]. degree in electronic engineering from the
Warsaw University of Technology, Warsaw, in 1972. After graduation he joined
the Department of Electronics, (Institute of Radioelectronics), Warsaw University
of Technology where in 1976 he was granted with distinction his doctors degree
(Ph.D). The thesis has been devoted to nonlinear phenomena occurring in microwave oscillators with avalanche and Gunn diodes. In 1991, he received Doctorate
in Science degree in electronic engineering from the Warsaw University of Technology for a habilitation thesis on new methods of designing linear microwave circuits.
Finally, he received in 2001 the degree of professor of technical science. In 1996,
he was appointed as associate professor in the Warsaw University of Technology,
where he lectured on Fundamentals of radar and radionavigation techniques,
UHF and microwave antennas, Numerical methods and Methods for analysis
ix

About the Author

and synthesis of microwave circuits. His main research interest is computer-aided

design of different microwave circuits, and especially planar multi-element array
antennas. He is the author of more than 80 scientific papers, 30 technical reports
and 6 books, viz. Algorithms for design of selected linear microwave circuits (in
Polish), Wk, Warsaw 1987, Mathematical methods for designing electronic circuits with distributed parameters (in Polish), WNT, Warsaw 1988, Algorithms for
computer-aided design of linear microwave circuits, Artech House, Inc. Boston
London 1990, Linear microwave circuits methods for analysis and synthesis
(in Polish), WK, Warsaw 1999 and Fundamentals of the antenna technique ( in
Polish), Publishing House of the Warsaw University of Technology, Warsaw 2006.
The last of them is the present book Fundamental Numerical Methods for Electrical Engineering. Since 1992, Prof. Rosloniec has been tightly cooperating with
the Telecommunications Research Institute (PIT) in Warsaw. The main subject of his
professional activity in PIT is designing the planar, in-phase array antennas intended
for operation in long-range three-dimensional (3D) surveillance radar stations. A
few of two-dimensional (planar) array antennas designed by him operate in radars
of type TRD-12, RST-12M, CAR 1100 and TRS-15. These modern radar stations
have been fabricated by PIT for the Polish Army and foreign contractors.

Introduction

Stormy development of electronic computation techniques (computer systems and

software), observed during the last decades, has made possible automation of data
processing in many important human activity areas, such as science, technology,
economics and labor organization. In a broadly understood technology area, this
development led to separation of specialized forms of using computers for the design
and manufacturing processes, that is:
computer-aided design (CAD)
computer-aided manufacture (CAM)
In order to show the role of computer in the first of the two applications mentioned above, let us consider basic stages of the design process for a standard piece
of electronic system, or equipment:
formulation of requirements concerning user properties (characteristics, parameters) of the designed equipment,
elaboration of the initial, possibly general electric structure,
determination of mathematical model of the system on the basis of the adopted
electric structure,
determination of basic responses (frequency- or time-domain) of the system, on
the base of previously established mathematical model,
repeated modification of the adopted diagram (changing its structure or element
values) in case, when it does not satisfy the adopted requirements,
preparation of design and technological documentation,
manufacturing of model (prototype) series, according to the prepared documentation,
testing the prototype under the aspect of its electric properties, mechanical durability and sensitivity to environment conditions,
modification of prototype documentation, if necessary, and handing over the
documentation to series production.
The most important stages of the process under discussion are illustrated in
Fig. I.1.
xi

xii

Introduction

Fig. I.1
Recognition
of need

Create
a general
design

Prepare
a mathematical
model
Improve
the design

Evaluate the
frequency or
time responses

No

?
Test

Yes
Prepare a
technical
documentation

Transfer the
design to
manufacture

According to the diagram presented above, the design process begins with the
formulation of user requirements, which should be satisfied by the designed system
in presence of the given construction and technological limitations. Next, among
various possible solutions (electrical structures represented by corresponding structures), the ones, which best satisfy the requirements adopted at the start are chosen.
During this stage, experience (knowledge and intuition) of the designer has decisive
influence on the design process. For general solution chosen in this manner (values
of system elements can be changed), mathematical model, in the form of transfer
function, insertion losses function or state equations, is next determined. On the

Introduction

xiii

base of the adopted mathematical model, frequency- or time-domain responses of

the designed system are then calculated. These characteristics are analyzed during
the next design stage. In case when the system fully satisfies the requirements taken
at the start, it is accepted and its electric structure elaborated in this manner can be
considered as the base for preparation of the construction and technological documentation. In the opposite case, the whole design cycle is repeated for changed
values of elements of the adopted electrical structure. When modification of the
designed system is performed with participation of the designer (manual control),
the process organized in this way is called interactive design. It is also possible to
modify automatically the parameters of the designed system, according to appropriate improvement criterions (goal function), which should take usually minimal
or maximal values. Design process is then called optimization. During the stage of
constructing mathematical model of the designed system, as well as during the stage
of analysis, there is a constant need for repeated performing of basic mathematical
procedures, such as:

solving systems of linear algebraic equations,

solving systems of nonlinear algebraic equations,
approximation or interpolation of one or many variable functions,
integration of one or many variable functions,
integration of ordinary differential equations,
integration of partial differential equations,
solving optimization problems, the minimax problem included.

The second process mentioned above, namely the CAM, can be considered in
a similar way. The author is convinced that efficient use of computer in both processes considered, requires extensive knowledge of mathematical methods for solving the problems mentioned above, known commonly under the name of numerical
methods. This is, among other things the reason, why numerical methods became
one of the basic courses, held in technical universities and other various kinds of
schools with technical profile Considerable cognitive virtues and specific beauty of
this modern area of mathematics is the fact, which should also be emphasized here.
This book was worked out as education aid for the course Numerical Methods in
Radio Electronics lead by the author on the Faculty of Electronics and Information
Technology of Warsaw University of Technology. During its elaboration, considerable emphasis was placed on the transparency and completeness of discussed issues,
and presented contents constitute sufficient base for writing calculation programs in
arbitrary programming language, as for example in Turbo Pascal. Each time, when it
was justified for editorial reasons, vector notation of the equation systems and vector operations were deliberately abandoned, the fact that facilitates undoubtedly the
understanding of methods and numerical algorithms explained in this book. Numerous examples of engineering problems taken from electronics and high-frequency
technology area serve for the same purpose.