本书以控制系统的设计实验作为系统实例分析的重点，引领读者进行完整的控制系统设计：建模、识别、分析、设计、仿真和应用。本书内容涵盖连续与离散控制系统的各种分析和研究方法，列举大量的设计实例，且书中所涉及的问题均采用MATLAB求解。所附光盘内容包括书中问题的MATLAB求解、m文件，MATLAB指令的讨论与举例以及设计实例的测试数据。\r\n\r\n 本书可作为自动控制及相关专业师生选用的教学参考书。\r\n
\r\n
Chapter 1 Preliminaries \r\n\r\n Chapter 2 The Laplace Transform \r\n\r\n Chapter 3 The Transfer Function \r\n\r\n Chapter 4 Introducing Feedback \r\n\r\n Chapter 5 Root Locus Analysis \r\n\r\n Chapter 6 Quantifying Performance \r\n\r\n Chapter 7 Cascade Root Locus Design \r\n\r\n Chapter 8 Motor Speed Control:A Case Study \r\n\r\n Chapter 9 Frequency Response \r\n\r\n Chapter 10 Nyquist Criterion \r\n\r\n Chapter 11 Bode Design \r\n\r\n Chapter 12 Robust Control \r\n\r\n Chapter 13 Position Control:A Case Study \r\n\r\n Chapter 14 Discrete Systems \r\n\r\n Chapter 15 Digital Control \r\n\r\n Chapter 16 Aircraft Pitch Control:A Case Study \r\n\r\n Chapter 17 The Transportation Lag \r\n\r\n Chapter 18 The State Model \r\n\r\n Chapter 19 Observability and Controllabillty \r\n\r\n Chapter 20 The Controller/Observer \r\n\r\n Index \r\n\r\n \r\n
\r\n
When I first began teaching the senior elective in control systems at Georgia Tech twenty years ago, I had no interest in writing a textbook on the subject. There were plenty of good texts available, as there are today. Over lime, though, I began to typeset my own notes and hand them out to the students. Eventually, those notes became complete enough that the students seem to prefer them to the official text. Even at that point my attitude was simply that the notes offered the students an alternative point of view to that of the official text, thereby enhancing their understanding.
Around 1984 Abe Haddad joined the faculty and within a year or so had managed to pry some money out of the dean of engineering to create a control system laboratory. He asked me to build that lab and I agreed. It took a long time to get that laboratory to function the way I wanted, mainly because I had the students design all the equipment for the laboratory. I had no interest in off-the-shelf experiments or off-the-shelf equipment. The students designed power supplies, power amplifiers, and any number of test fixtures.
Meanwhile, I was trying to design a sequence of laboratory sessions that would take the students through the whole design cycle, beginning with the modeling and identification of the plant transfer function, proceeding through the analysis and design stages, and ending with the implementation of the control on the actual physical system. Eventually, I ended up with three basic devices or "systems'' and sixteen laboratory sessions. During those sixteen sessions the students who took the two quarter sequence in control went through the complete design cycle on each of the three systems. Those three basic systems, and a some of the content of those sixteen laboratory sessions appear in this book as three case studies that also take the reader through the entire design cycle.
So it was that when Catherine Fields showed up in my office a couple of years ago, asking if I wanted to write a book, I was somewhat less reluctant that I had been sixteen years previously. Catherine is a very good listener, and very patient, and she let me ramble on about control system education, laboratories, and cheetahs for a long time. In the end, her patience was rewarded because she convinced me to send her a manuscript.
When the initial reviews came back and were favorable, I began to overcome my natural pessimism and started thinking about how I was going to turn a set of class notes into a book in less than six months. Along the way I decided to add an introductory chapter on modem robust control, and after the second round of reviews I made some major revisions to answer the legitimate complaints of the reviewers. Like most writers I would have liked more time, but I kept remembering the character
Albert Camus' novel The Plague who spent twenty years trying to get the first sentence of his novel "just right."
I think the distinguishing feature of this book is that I have tried to lay out for the reader all the steps in what I call the design cycle: modeling, identification, analysis, design, simulation, and implementation. Chapters 8, 13, and 16 are case studies that take the reader through the entire design cycle for three different systems. My intent is to bring the reader as close to a real laboratory experience as I can within the confines of a textbook. In the end there is no substitute for getting in the laboratory and getting
your hands dirty, but I have tried to come as close to that as I could.
The first two steps in the design cycle, modeling and identification, are to my mind the most crucial. Once the plant has been determined with sufficient accuracy, the design, simulation, and implementation normally follow without too much trouble.
In this book modeling means finding the structure of the transfer function, by which I mean deciding how many poles and zeros the transfer function has. Once that decision is made, the more difficult problem of parameter identification, deciding on the exact locations of the poles and zeros, has to be addressed.
Most books do a good job with the first step, but I don't know of many, at least at the undergraduate level, that put much emphasis on parameter identification, which, to me, is the most crucial step in the whole design cycle. Beginning in Chapter 3, and again in Chapters 8, 9, 13, and 16, I provide ,the student with a variety of techniques for parameter identification, and some experience in implementing these techniques.
Parameter identification is not easy. In describing it to students in the laboratory I often paraphrase the great running back Jim Brown, who was once asked by a sportswriter, who had probably never carried a football in his life, to describe the techniques and the finesse he had used to gain all those yards. Brown snorted and replied something like, "it's not technique, it's not finesse, I just go out there and scuffle and claw for every yard I can get." That to me is the perfect description of parameter identification, and in Chapters 8, 13, and 161 have endeavored to show the reader how to scuffle and claw his way to a viable model of a physical system. Once the parameters of the transfer function are identified, the control tums out to be very straightforward, almost anticlimactic.
Aside from Chapters 8, 13, and 16, the book contains the same material found in most undergraduate control books, although in a slightly different arrangement. My goal was to get the student designing control systems as quickly as possible. Thus, after disc. ussing transfer functions and feedback in Chapters 3 and 4, I introduce root locus analysis in Chapter 5 and then follow it with a chapter on specifications. Most books go about it the other way around, which is fine, but my goal was to get the root locus technique in front of the student and then pose the question: now that we have a picture of all the possible sets of closed-loop poles, which set do we want to choose? Having answered that question in Chapter 6, the student begins designing systems in Chapter 7, using the root locus method. Thus, 164 pages into the book, the student is designing.
In Chapter 8, I build on Chapter 7 by taking the student through the whole design cycle. First two models, i.e., transfer functions, both based on the analysis of a dc motor from Chapter 3, are chosen. Then the parameters of these two transfer functions are identified using two identification techniques, one that was introduced in Chapter 3, and a second developed in Chapter 8. Then a control is designed, implemented, and
compared with simulation results for both systems. In Chapter 8, I try to give the student something close to an actual laboratory experience. Of course it isn't a real laboratory experience, but I hope the student can at least see the basic steps that have to be followed to implement the control.
Chapters 9 through 13 might be called the heart of the book. Chapter 9 begins with frequency response analysis and ends with the design of compensators that provide disturbance rejection for disturbances at the output. Along the way I spend some time discussing some of the other positive benefits of feedback, a discussion originally begun in Chapter 4. Chapter 10 covers the Nyquist criterion, in preparation for Bode
design in Chapter 11 and robust control in Chapter 12. Chapter 13 is again a case study, taking the student through the entire design cycle for a simple robotic system, namely, a positioning table.
Chapters 10, 11, and 12 cover a good deal of the history of control, starting with Nyquist's contributions in Chapter 10; continuing with the contributions of Bode, Nichols, Chestnut, Axleby, Truxal, and many others in Chapter 11; and finishing with the contributions of Doyle, Francis, Tannenbaum, and others in Chapter 12. I think these chapters fit together pretty well, and the student should come away with a sense of the importance of the ideas first introduced in the 1930s and 1940s. Chapter 12 is really only an introduction to robust control, as I try to point out at the beginning of the chapter. But I think it does give the student an idea of how robust control builds on previous work, and hence some sense of the evolution of control design.
Chapter 14 is a review of discrete systems and is followed by Chapter 15 on the design of sampled data systems. Chapter 16 is the third, and final, case study, and is based on an apparatus that captures the essential pitch dynamics of an airplane. This time the modeling and parameter identification are done using frequency domain methods, and the control design and implementation are for a sampled data
system.
Chapter 17 is devoted to systems that have a transportation lag. Both continuous and sampled data systems are discussed. The pure delay, or transportation lag, and the the zero-order hold have similar effects on the stability of a system. Both introduce negative phase, which tends to destabilize the closed-loop system, and that was the motivation for having a separate chapter devoted to this subject. It is possible, however, to use only the portion of the chapter devoted to continuous systems, or that on sampled
data systems.
Chapters 18, 19, and 20, cover continuous and discrete state models, and the controller/observer formulation of feedback control for the single input single output case. This is pretty standard fare. In Chapter 18 the continuous state model is derived from a transfer function. Second-order state models are then studied in depth using phase portraits to provide some intuition for stability. The extension to systems of higher dimension is quite easy. Chapter 19 begins with the discretization of the state model. This is followed by an informal discussion of pole placement. Controllability, observability, and the associated canonical forms are then introduced. Chapter 20 covers the controller/observer for both the regulator and tracking problems. Current as well as prediction estimators are discussed. At the end of Chapter 201 tie the state model and transfer functions methods together by showing that the controller/observer
and plant can be turned back into a closed-loop transfer function.
It is not realistic to try to cover the entire book in one semester. There are, however, several "one-semester books" within the book. Which one you choose depends on the background of the students. Some of the different courses that could be taught from the book are now discussed.
INTRODUCTION TO DESIGN
For students with no background in control, the first thirteen chapters represent a thorough treatment of continuous systems, beginning with an introduction to the Laplace transform, ending with modem robust control, and offering the student two case studies that cover the entire design cycle. This is not likely to be an option for electrical engineers (EEs), but it might be a good one for mechanical or civil engineers.
SENIOR ELECTIVE FOR EEs--OPTION A
Most electrical engineering students get a reasonably thorough treatment of transformtheory and linear systems at the junior level. This means that a senior-level course
could begin in Chapter 4, or even or Chapter 5. In that case it should be possible tocover the material up through Chapter 16 in one semester. By omitting some of the design methods in Chapter 15, Chapter 17 could also be included.
SENIOR ELECTIVE FOR EEs--OPTION B
If the students have seen the Z transform, then Chapter 14 will be mostly a review and won't take up much time. Then by being selective in Chapters 15, 17, and 18 it should be possible to cover Chapters 5-20. Some places where the instructor can be selective are:
1. In Chapter 15 a variety of design techniques are presented. A couple of techniques can be omitted without much loss.
2. Chapter 17 is a very thorough study of transportation lags for both continuous and discrete systems. Some time could be saved by simply omitting transportation lags for discrete systems.
3. The material in Chapter 18 on the method of isoclines and generalized eigenvectors and the Jordan form does not impact the presentations in Chapters 19 and 20 and could be omitted.
4. Chapter 20 covers both prediction and current observers. Just covering the prediction observer provides the basics of controller/observer design. One could also omit the development on the tracking problem and concentrate onregulators.
SENIOR ELECTIVE FOR EEs--OPTION C
It is also possible to teach a traditional senior level course by leaving out the case studies in Chapters 8, 13, and 16. In that case I think Chapters 5, 6, 7, 9, 10, 11, 12, 14, 15, 17, 18, 19, and 20 could be covered in one semester by omitting some topics in Chapters 15, 17, 18 and 20, as discussed under option B.
DIGITAL CONTROL
Chapters 14-20 provide a thorough coverage of digital control and are suitable for a one quarter course on that subject. Chapter 18 covers state models for continuous systems, and can be included or omitted as the instructor sees fit. I would probably include it, but it is really a matter of taste.
I consider the book MATLAB-friendly but not MATLAB-crazy. I think the MATLAB control toolbox is of great use, once the students have done some hand calculations and know what they are doing. For me, the best thing about MATLAB is not the "canned" routines, but the fact that I can very quickly write a fifteen- or twenty-line m-file that lets me see the effect of varying a parameter, such as a pole or zero, over a range of values. This is something that would be hard to do with hand calculations, and is, I think, the most valuable feature of MATLAB. Many of the problems are designed to get the student to write this kind of program. The CD-ROM that comes with the book discusses the MATLAB commands relevant to each chapter and provides some examples of how to use these commands.
The CD-ROM also includes some actual test data for the three systems introduced in Chapters 8, 13, and 16. In the solution manual, I provide the results of the actual implementation of the control designs for the problems in these chapters.
In summary, I think the book offers some unique features, most notably attention tothe parameter identification problem at a level that an undergraduate can understand, plus three complete trips through the design cycle. Not being much of a salesman I will stop short of saying your life will change if you adopt this book, and merely add that the students at Georgia Tech, who are about the same as the students at Nebraska,
Penn State, Michigan State, or Kansas, or for that matter MIT, seem to like the notes on which this book is based.
无封面
第二书店&China-pub战略联盟提供专业服务
第二书店联系方式 010-64348411 webmaster@dearbook.com