LГЎszlГі Keviczky - Control Engineering: MATLAB Exercises

Frigyes Cski

(19211977)

This textbook is devoted to the memory of Frigyes Cski, who was the first professor of control in Hungary

This book is intended to aid students in their study of MATLAB/SIMULINK for use in solving control problems. Specifically, 16 labs for an introductory control course have been developed at the Department of Automation and Applied Informatics, Budapest University of Technology and Economics. This book is a collection of these labs. This exercise book is a supplement to the textbook Control Engineering, by Lszl Keviczky, Ruth Bars, Jen Hetthssy, and Csilla Bnysz [1], which is used in the control course. Each chapter of this exercise book is related to the corresponding chapter of the textbook.

The importance of accompanying textbooks by labs using CAD software was recognized decades ago at the department. At that time, a set of FORTRAN libraries supported the instruction both in control systems analysis and design. We still believe that learning control theory is best motivated by applications and simulations rather than by concepts alone. In fact, the use of MATLAB allows a lot of theoretical concepts to be easily implemented. If students can immediately show for themselves how certain concepts work in practice, they will go back to the theoretical considerations with greater confidence and an improved ability to move to the next field to study. Well, feedback is around us, anyway.

The problems discussed in this book are limited to linear, time-invariant control systems. Both continuous-time and discrete-time systems are considered, with deterministic inputs.

MATLAB/SIMULINK is useful only for those students, who master the tools offered. Though the application of MATLAB commands is simple and straightforward, a systematic introduction together with control-related examples is a must in our opinion. Time should be devoted to practicing fundamental MATLAB facilities, alternative command sequences, and visualization capabilities. An introductory lab is devoted to demonstrating the availability and power of MATLAB in this respect.

Frequency functions and transfer functions form essential tools in classical control theory. Interestingly enough, the frequency domain considerations gave a remarkable impetus to the postmodern control era, as well. Three labs have been devoted to discussing fundamental analysis of continuous-time systems including feedback and stability. As far as controller synthesis is concerned, three labs treat Y oula -parameterized control design, as well as PID compensation and series compensation for processes with dead time have also been elaborated. The case of controlling unstable processes is also involved. The theoretical discussion of state-space representations is supported by two labs offering a gentle introduction to the subject, as well as demonstrating the efficient algorithms and MATLAB commands available for state variable feedback.

These days, controllers are implemented as digital controllers. As most of the processes to be controlled are continuous time in nature, digital control needs additional tools to cover sampled data systems. Just to support the development of a proper view of discrete-time systems, an introductory lab has been added to this topic. Two labs are devoted to discrete controller design. One of them shows controller design using the Y oula parameterization, and the design of a S mith predictor as well as a deadbeat control as special cases of Y oula parameterization. The second lab discusses discrete-time PID controller design. State feedback control for discrete systems is also provided in a lab devoted to this topic.

Two labs deal with the polynomial design method for the compensation of unstable processes, both for the continuous and the discrete case.

In the last lab, the modelling and simulation of a heating process provide a case study.