+ All Categories
Home > Documents > Systems Simulation: The Art and Science

Systems Simulation: The Art and Science

Date post: 24-Sep-2016
Category:
Upload: james-d
View: 212 times
Download: 0 times
Share this document with a friend
2
723 BOOK REVIEWS many useful results published originally in German, Russian, and other journals not easily available everywhere. The book is comprised of five chapters. The first chapter presents the connection between reliability design and tolerance analysis. The second chapter is introductory material to network analysis by matrix methods, including the mechanical-electrical-transport-thermal an- alogues. The third chapter presents the calculation of sensitivities. The sensitivities are determined by network analysis using artificial (auxiliary) excitations. Eigenvalue sensitivities are discussed as well. The fourth chapter is concerned with the most advanced methods. Interval analysis (see R. E. Moore's book of the same title) is applied for the solution of linear equations and extended to tolerance analysis of linear systems by tolerance contours. The fifth chapter gives a very short treatment of optimization tolerances. An appendix on matrix norms, references, and an index completes the small-sized book. In many places, discussions are rather brief, and some important subjects, for instance large change sensitivities and sensitivity measures, are omitted. In some places the numbering of equations and the use of symbols is not quite clear. The level of treatment assumes the reader's familiarity with math- ematics and network theory at the graduate level. The important applications and examples include electronics, mechanics, electro- mechanics, and others. In conclusion, engineers, mathematicians and economists would benefit by reading this book. Regelung in der elektrischen Antriebstechnik (Control Systems for Electrical Drives)-W. Leonhard (Stuttgart, West Germany: B. G. Teubner, 1974, 216 pp.). Reviewed by Heinrich J. Boenig, Los Alamos Scientific Laboratory, Los Alamos, NM 87545. This book is the contents of a course, developed by the author during the last ten years for seniors and first-year graduate students, at the University of Braunschweig, West Germany. Prerequisite for the understanding of the material is an introductory course in control systems and electrical machines. The book contains fourteen chapters. Chapters 1 to 4 review such topics of mechanics as the impulse theorem, the determination of the momentum of inertia, discussion and solution of the equation of motion, the torque-speed relationship in electrical machines and prime movers, and heating of electrical machines. Chapter 5 treats the separately excited dc machine. Differential equations and the block diagram with the transfer functions of the machine components are derived, and steady-state characteristics for variable terminal voltage and field current are developed. Similar derivations are made in Chapter 6 for the dc series motor. Block diagrams of practical control systems for a separately excited dc machine are given in Chapter 7. Chapter 8 gives a short introduction to the behavior and charac- teristics of silicon controlled rectifiers (SCR). The most commonly used SCR circuits for electrical drives are shown. The SCR controller for dc machines, both the line-commutated and forced-commutated type, are explained in Chapter 9. The next four chapters are a treatment of three-phase electrical drive systems. In Chapter 10 the dynamic equations and the block diagram of an induction machine are derived. Chapters 11, 12, and 13 cover control systems for asynchronous and synchronous machines, including control systems for an induction motor fed by a cyclo- converter or a pulse-width modulated inverter and slip energy recovery schemes. The reader is also introduced to the theory of decoupling the equa- tions of the three-phase machine by coordinate transformations, whereby both magnetizing current and torque can be controlled independently. For the induction and synchronous machine, multiloop block diagrams are developed, using the decoupling method. The final chapter contains examples, which have been tested in field applications, including time-optimal position control, rolling mill control systems for drives which provide methods to obtain steady tension of elastic media, etc. The strength of the book lies in its brief but thorough and precise description of dc and ac drives. The three major areas of electrical drives-machine, SCR controller, and control system-are described with thoroughness. The author has extensive academic and practical experience and writes with authority. The size of the book (216 pages) may be deceiving with respect to the scope of the material. Nevertheless, the reader finds much information and much practical experience in the book. The print and the figures are small but very legible. The book is a well-rounded introductory text to the topics of electrical drive systems. To the reviewer's knowledge the book is a unique treatment of modern dc and ac drives and has no counterpart in the English- language literature. The book is therefore recommended for students, teachers, and practicing engineers who have a reading knowledge of German. Systems Simulation: The Art and Science-Robert Shannon (Englewood Cliffs, NJ: Prentice-Hall, 1975, 387 pp.). Reviewed by James D. Johannes, Vanderbilt University, Nashville, TN. Systems Simulation: The Art and Science, as the title indicates, is a highly readable text, written with management and the student of simulation in mind. The text meets its goal of being intended for both those who design simulation models and those who will use the results of the studies based upon them. As the author points out, simulation modeling and experimentation remain very much intuitive processes-though he has presented a lucid description of the process of designing a model, running experiments with it, and analyzing the model's results. The reader is made aware of the critical issues and the benefits and possible pitfalls in a clear, yet concise manner. The major thrust of this text is on the art of modeling rather than on examining in detail already existing models. Since the text assumes both a varied background and an amount of knowledge in math- ematics, probability, and computer programming, it will appeal to a very wide range of readers. Extensive references at the end of each chapter refer the reader to the particular detailed literature on the primary sources of the subjects covered. This text, unlike others, covers the problems of managing the research effort and obtaining user acceptance of simulation models and their results. The text is suitable for a one-semester course at the first-year graduate, or even at the advanced undergraduate level in engineering, manage- ment science, computer science, or business administration programs. A basic background in calculus, probability theory, and high-level language computer programming is required. Because of its wealth of examples and explanatory material, it is an excellent text for self-study. Chapter 1, Fundamentals of Modeling, discusses general modeling philosophy and an overall view of simulation. Chapter 2, Systems Investigation, discusses some of the techniques that are useful in analyzing the system to be modeled and the design of the model. Chapter 3, Model Translation, is devoted to computer programming languages and how to choose the language to suit the problem. Several excellent decision aids are presented to the reader to help him select the appropriate language for the problem and to suit his preferences. Chapter 4, Design of Computer Simulation Experiments, presents an excellent treatment of the terminology and concepts of experimental design. Chapter 5, Tactical Planning, exposes some of the problems and pitfalls encountered in running and analyzing an experiment. Chapter 6, Validation and Analysis, treats the very important and significant aspects of model validation and the inferences drawn from the experiment from both a philosophical and practical viewpoint. Chapter 7, Management Aspects, confronts the problems in the user environment of managing the research activity and of gaining user acceptance. The text also contains three appendices: the first covering representative case studies, the second treating random variate genera- tion, and the third containing the more useful statistical tables. In summary, the text is an excellent introduction to simulation and modeling. Through the presentation style and the choice of analysis
Transcript

723BOOK REVIEWS

many useful results published originally in German, Russian, and otherjournals not easily available everywhere.The book is comprised of five chapters. The first chapter presents

the connection between reliability design and tolerance analysis. Thesecond chapter is introductory material to network analysis by matrixmethods, including the mechanical-electrical-transport-thermal an-alogues. The third chapter presents the calculation of sensitivities.The sensitivities are determined by network analysis using artificial(auxiliary) excitations. Eigenvalue sensitivities are discussed as well.The fourth chapter is concerned with the most advanced methods.Interval analysis (see R. E. Moore's book of the same title) is appliedfor the solution of linear equations and extended to tolerance analysisof linear systems by tolerance contours. The fifth chapter gives a veryshort treatment of optimization tolerances. An appendix on matrixnorms, references, and an index completes the small-sized book.

In many places, discussions are rather brief, and some importantsubjects, for instance large change sensitivities and sensitivity measures,are omitted. In some places the numbering of equations and the use ofsymbols is not quite clear.The level of treatment assumes the reader's familiarity with math-

ematics and network theory at the graduate level. The importantapplications and examples include electronics, mechanics, electro-mechanics, and others.

In conclusion, engineers, mathematicians and economists wouldbenefit by reading this book.

Regelung in der elektrischen Antriebstechnik (Control Systems forElectrical Drives)-W. Leonhard (Stuttgart, West Germany: B. G.Teubner, 1974, 216 pp.). Reviewed by Heinrich J. Boenig, Los AlamosScientific Laboratory, Los Alamos, NM 87545.

This book is the contents of a course, developed by the authorduring the last ten years for seniors and first-year graduate students,at the University of Braunschweig, West Germany. Prerequisite forthe understanding of the material is an introductory course in controlsystems and electrical machines.The book contains fourteen chapters. Chapters 1 to 4 review such

topics of mechanics as the impulse theorem, the determination of themomentum of inertia, discussion and solution of the equation ofmotion, the torque-speed relationship in electrical machines and primemovers, and heating of electrical machines. Chapter 5 treats theseparately excited dc machine. Differential equations and the blockdiagram with the transfer functions of the machine components arederived, and steady-state characteristics for variable terminal voltageand field current are developed. Similar derivations are made inChapter 6 for the dc series motor. Block diagrams of practical controlsystems for a separately excited dc machine are given in Chapter 7.

Chapter 8 gives a short introduction to the behavior and charac-teristics of silicon controlled rectifiers (SCR). The most commonlyused SCR circuits for electrical drives are shown. The SCR controllerfor dc machines, both the line-commutated and forced-commutatedtype, are explained in Chapter 9.The next four chapters are a treatment of three-phase electrical

drive systems. In Chapter 10 the dynamic equations and the blockdiagram of an induction machine are derived. Chapters 11, 12, and13 cover control systems for asynchronous and synchronous machines,including control systems for an induction motor fed by a cyclo-converter or a pulse-width modulated inverter and slip energy recoveryschemes.The reader is also introduced to the theory of decoupling the equa-

tions of the three-phase machine by coordinate transformations,whereby both magnetizing current and torque can be controlledindependently. For the induction and synchronous machine, multiloopblock diagrams are developed, using the decoupling method. The finalchapter contains examples, which have been tested in field applications,including time-optimal position control, rolling mill control systemsfor drives which provide methods to obtain steady tension of elasticmedia, etc.

The strength of the book lies in its brief but thorough and precisedescription of dc and ac drives. The three major areas of electricaldrives-machine, SCR controller, and control system-are describedwith thoroughness. The author has extensive academic and practicalexperience and writes with authority. The size of the book (216 pages)may be deceiving with respect to the scope of the material. Nevertheless,the reader finds much information and much practical experience inthe book. The print and the figures are small but very legible. Thebook is a well-rounded introductory text to the topics of electricaldrive systems.To the reviewer's knowledge the book is a unique treatment of

modern dc and ac drives and has no counterpart in the English-language literature. The book is therefore recommended for students,teachers, and practicing engineers who have a reading knowledge ofGerman.

Systems Simulation: The Art and Science-Robert Shannon (EnglewoodCliffs, NJ: Prentice-Hall, 1975, 387 pp.). Reviewed by James D.Johannes, Vanderbilt University, Nashville, TN.

Systems Simulation: The Art and Science, as the title indicates, is ahighly readable text, written with management and the student ofsimulation in mind. The text meets its goal of being intended for boththose who design simulation models and those who will use theresults of the studies based upon them. As the author points out,simulation modeling and experimentation remain very much intuitiveprocesses-though he has presented a lucid description of the processof designing a model, running experiments with it, and analyzing themodel's results. The reader is made aware of the critical issues and thebenefits and possible pitfalls in a clear, yet concise manner.The major thrust of this text is on the art of modeling rather than on

examining in detail already existing models. Since the text assumesboth a varied background and an amount of knowledge in math-ematics, probability, and computer programming, it will appeal to avery wide range of readers. Extensive references at the end of eachchapter refer the reader to the particular detailed literature on theprimary sources of the subjects covered. This text, unlike others,covers the problems of managing the research effort and obtaininguser acceptance of simulation models and their results.The text is suitable for a one-semester course at the first-year graduate,

or even at the advanced undergraduate level in engineering, manage-ment science, computer science, or business administration programs.A basic background in calculus, probability theory, and high-levellanguage computer programming is required. Because of its wealthof examples and explanatory material, it is an excellent text forself-study.

Chapter 1, Fundamentals of Modeling, discusses general modelingphilosophy and an overall view of simulation. Chapter 2, SystemsInvestigation, discusses some of the techniques that are useful inanalyzing the system to be modeled and the design of the model.Chapter 3, Model Translation, is devoted to computer programminglanguages and how to choose the language to suit the problem. Severalexcellent decision aids are presented to the reader to help him selectthe appropriate language for the problem and to suit his preferences.Chapter 4, Design of Computer Simulation Experiments, presentsan excellent treatment of the terminology and concepts of experimentaldesign. Chapter 5, Tactical Planning, exposes some of the problemsand pitfalls encountered in running and analyzing an experiment.Chapter 6, Validation and Analysis, treats the very important andsignificant aspects of model validation and the inferences drawn fromthe experiment from both a philosophical and practical viewpoint.Chapter 7, Management Aspects, confronts the problems in the userenvironment of managing the research activity and of gaining useracceptance. The text also contains three appendices: the first coveringrepresentative case studies, the second treating random variate genera-tion, and the third containing the more useful statistical tables.

In summary, the text is an excellent introduction to simulation andmodeling. Through the presentation style and the choice of analysis

IEEE TRANSACTIONS ON SYSTEMS, MAN, AND CYBERNETICS, OCTOBER 1976

tools, the reader should develop an intuitive feeling for the art andscience of simulation and its many applications. The book is well-designed to serve as a text for a first course in simulation and modeling.

Digital Signal Analysis-S. D. Stearns (Rochelle Park, NJ: HaydenBook Company, 1975, 280 pp.). Reviewed by N. Ahmed, Departmentof Electrical Engineering, Kansas State University, Manhattan, KS66506.

This is an excellent book which introduces the engineer, mathe-matician, and computing scientist to the fundamentals of signalanalysis, with an emphasis on digital signals. It has evolved from aone-year course in digital signal analysis taught at the Universityof New Mexico, beginning in 1969, and at the Sandia Laboratoriesin Albuquerque, NM, starting in 1970.The book consists of fourteen chapters, the first three of which are

devoted to a review of linear signal analysis, least squares, Fourierseries, continuous transforms, transfer functions, etc. Topics relatedto sampling, analog-to-digital conversion, the fast Fourier transform,and spectral computations with digital signals are discussed in Chapters4-7. Throughout these chapters, the author does a fine job of em-

phasizing the relationships between continuous and discrete waveformsand spectra.

Topics pertaining to recursive and nonrecursive digital systems,digital simulation of continuous systems, and the design of digitalfilters are considered in Chapters 8-12. Some discussion pertaining to

errors due to finite word lengths is also included. A topic whichdeserves special mention is the design of cascaded digital Butterworthand Chebyshev filters via the bilinear transformation; the relatedtreatment is one of the best this reviewer has seen. A set of Fortranprograms which enable one to readily design cascaded Butterworthfilters is included in an appendix. Finally, a concise but well-writtendiscussion related to random digital signals and their spectral estima-tion is presented in Chapters 13 and 14.

In summary, this is a carefully prepared book which consists ofnumerous illustrative examples and thought-provoking exercises. Itis essentially self-contained and is of great value to any serious studentof digital signal processing.

System Dynamics: A Unified Approach-D. Karnopp and R. Rosenberg(New York: Wiley-Interscience, 1975, 402 pp.). Reviewed by Alan S.

Perelson, Los Alamos Scientific Laboratory, University of California,Los Alamos, NM 87545.

It is now generally recognized that most lumped-parameter systemshave a common mathematical structure. Viewing a system as a collec-tion of interacting parts leads to system descriptions in which there are

sets of equations describing the properties of each component, i.e.,constitutive equations and sets of equations which describe the mannerin which the system components are connected together-the system'stopology. The separability of the topological and constitutive aspectsof a system was probably first noted by Kirchhoff when he formulatedhis now famous current and voltage laws, KCL and KVL. The modern

graph theoretic approach to electrical network theory is based uponthis separability.

In recent years a number of undergraduate texts have attempted to

present a broad unified view of system dynamics by discussing electrical,mechanical, hydraulic, and thermal components and graph theoreticprocedures for describing their interconnection. The classification of

variables as "through" and "across" (i.e., obeying a generalized KCLand KVL, respectively) is basic to this unity of description.The Karnopp-Rosenberg book is notable because it is the first

text to solely rely upon bond graphs as the means of representingsystem topology. Bond graphs provide a convenient description for

systems with multiport components and energy transductions. Con-

sequently, electromechanical, electrothermal, or thermodynamicsystems can be depicted and analyzed with unified notations and

techniques. Biological and chemical processes have also been treated

via bond graphs. Much of the book deals with bond graphs andtechniques of bond-graph modeling. As such it is probably the bestavailable primer on bond graphs and is highly recommended to anyoneinterested in bond-graph modeling.

Linear Control System Analysis and Design J. D'Azzo and C. Houpis(New York: McGraw-Hill, 1975, 636 pp.). Reviewed by RichardL. Brunson, Air Force Institute of Technology, Wright Patterson AirForce Base, OH.

This text is unique in the control system field. Professors D'Azzoand Houpis address the conventional control theory topics normallypresented in an undergraduate text and in addition relate them tomodern control theory. This correlation between conventional andmodern control theory makes this text particularly invaluable for use

in an intermediate course bridging the gap between the conventionaland modern control disciplines. Also, the text is excellent for use in anintroductory control course since the modern control discussionprovides the student with additional insight into control systemformulation and solution techniques.The 15 chapters of this text can be classified into the following

three categories: 1) introductory course material (Chapters 1 through 9);2) control system compensation (Chapters 10 through 12); and 3) ad-vanced topics (Chapters 13 through 15). Chapters 1 through 9 presentcontrol system representation and analysis methods which includeroot locus, frequency response, and state variable formulationmethods. Chapters 10 and 11 introduce the design of cascade andfeedback compensation through the use of root locus and frequencyresponse methods. Chapters 13 through 15 cover quadratic forms,Liapunov stability, the Riccati equation, performance indexes, andoptimal linear system design.An especially significant feature evident throughout the text is the

liberal use of example problems. Each example problem is very wellconceived and illustrates the intended concept quite well. Logical andorderly equation development is another key feature of the text.

Only the obvious steps are deleted during the course of an equationdevelopment. These two features suggest the book as an excellentself-study text for the practicing engineer.

In summary, this text is especially well suited for either an intro-ductory or intermediate level undergraduate course, as well as a self-study text for the practicing engineer. The text displays a superb jobof editing since the original printing had only very minor errors whichwere corrected in the errata sheet to the text.

Inners and Stability of Dynamic Systems-E. 1. Jury (New York:

Wiley, 1974, 308 pp.). Reviewed by Lawrence Stark anld V. V. Krishnan,University of Califbrnia, Berkeley, CA.

In this exciting contribution, Professor E. 1. Jury presents his unifiedtreatment of both continuous and discrete systems. The history of

dynamical systems has been, to a considerable extent, the developmentof various criteria and tests of root clustering and root distributions.The inners approach now combines consideration of stability,aperiodicity, observability, realizability, controllability, optimality,integral-square-measures, and others, in a beautiful, elegant, and

parsimonious form. Living as we do in the explosive growth phase ofscience and technology, where even reading titles of new articles is

beyond a man's capability, it is reassuring to know that scholars likeProfessor Jury are working to create simple order and unity where

diversity and dissimilarity prevailed.The classical theory of differential equations was of paramount

importance in Jury's background; contrariwise, during his academiccareer he has been immersed in difference equations sampled-datatheory, the Z-transform, and the modified Z-transform. Thus it was

not unnatural that the unified inners approach should have sprunginto Jury's head as he was "unoccupied" one April afternoon in 1970on his daily three-mile stroll around beautiful Lake Merritt near theBerkeley campus. Among the pleasures this development has given

724


Recommended