Nonlinear and Adaptive Control with Applications

302 Pages·2008·4.38 MB·English

by  Alessandro Astolfi PhD

Communications and Control Engineering
Series Editors
E.D. Sontag(cid:3) (cid:120)(cid:3) M. Thoma(cid:3) (cid:120)(cid:3) A. Isidori(cid:3) (cid:120)(cid:3) J.H. van Schuppen

Published titles include:
Stability and Stabilization of Infinite Dimensional Flow Control by Feedback
Systems with Applications Ole Morten Aamo and Miroslav Krsti(cid:252)
Zheng-Hua Luo, Bao-Zhu Guo and Omer Morgul
Learning and Generalization (Second edition)
Nonsmooth Mechanics (Second edition) Mathukumalli Vidyasagar
Bernard Brogliato
Constrained Control and Estimation
Nonlinear Control Systems II Graham C. Goodwin, María M. Seron and José A.
Alberto Isidori De Doná
L-Gain and Passivity Techniques in Nonlinear
2 Randomized Algorithms for Analysis and Control
Control
of Uncertain Systems
Arjan van der Schaft
Roberto Tempo, Giuseppe Calafiore and Fabrizio
Control of Linear Systems with Regulation and Dabbene
Input Constraints
Switched Linear Systems
Ali Saberi, Anton A. Stoorvogel and Peddapullaiah
Zhendong Sun and Shuzhi S. Ge
Sannuti
Subspace Methods for System Identification
Robust and H Control
(cid:102) Tohru Katayama
Ben M. Chen
Digital Control Systems
Computer Controlled Systems
Ioan D. Landau and Gianluca Zito
Efim N. Rosenwasser and Bernhard P. Lampe
Control of Complex and Uncertain Systems Multivariable Computer-controlled Systems
Stanislav V. Emelyanov and Sergey K. Korovin Efim N. Rosenwasser and Bernhard P. Lampe
Robust Control Design Using H Methods Dissipative Systems Analysis and Control
(cid:102)
Ian R. Petersen, Valery A. Ugrinovski and Andrey (2nd Edition)
V. Savkin Bernard Brogliato, Rogelio Lozano, Bernhard Maschke
and Olav Egeland
Model Reduction for Control System Design
Goro Obinata and Brian D.O. Anderson Algebraic Methods for Nonlinear Control Systems
Giuseppe Conte, Claude H. Moog and Anna M. Perdon
Control Theory for Linear Systems
Harry L. Trentelman, Anton Stoorvogel and Malo Polynomial and Rational Matrices
Hautus Tadeusz Kaczorek
Functional Adaptive Control Simulation-based Algorithms for Markov Decision
Simon G. Fabri and Visakan Kadirkamanathan Processes
Hyeong Soo Chang, Michael C. Fu, Jiaqiao Hu and
Positive 1D and 2D Systems
Steven I. Marcus
Tadeusz Kaczorek
Iterative Learning Control
Identification and Control Using Volterra Models
Hyo-Sung Ahn, Kevin L. Moore and YangQuan Chen
Francis J. Doyle III, Ronald K. Pearson and
Bobatunde A. Ogunnaike
Distributed Consensus in Multi-vehicle Cooperative
Non-linear Control for Underactuated Mechanical Control
Systems Wei Ren and Randal W. Beard
Isabelle Fantoni and Rogelio Lozano
Control of Singular Systems with Random Abrupt
Robust Control (Second edition) Changes
Jürgen Ackermann El-Kébir Boukas
Alessandro Astolfi • Dimitrios Karagiannis
Romeo Ortega

Nonlinear and
Adaptive Control
with Applications

123
Alessandro Astolfi, PhD Dimitrios Karagiannis, PhD
Department of Electrical and Electronic Department of Electrical and Electronic
Engineering Engineering
Imperial College London Imperial College London
London SW7 2AZ London SW7 2AZ
UK UK

and
Romeo Ortega, PhD
Dipartimento di Informatica, Sistemi Centre National de la Recherche Scientifique
e Produzione Laboratoire des Signaux et Systèmes
Università di Roma “Tor Vergata” Supélec
00133 Roma 91192 Gif-sur-Yvette
Italy France

ISBN 978-1-84800-065-0 e-ISBN 978-1-84800-066-7
DOI 10.1007/978-1-84800-066-7
Communications and Control Engineering Series ISSN 0178-5354
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Control Number: 2007941070
© 2008 Springer-Verlag London Limited
MATLAB® and Simulink® are registered trademarks of The MathWorks, Inc., 3 Apple Hill Drive,
Natick, MA 01760-2098, USA. http://www.mathworks.com
Apart from any fair dealing for the purposes of research or private study, or criticism or review, as
permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced,
stored or transmitted, in any form or by any means, with the prior permission in writing of the publishers,
or in the case of reprographic reproduction in accordance with the terms of licences issued by the
Copyright Licensing Agency. Enquiries concerning reproduction outside those terms should be sent to the
publishers.
The use of 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 laws and regulations and therefore free
for general use.
The publisher makes no representation, express or implied, with regard to the accuracy of the information
contained in this book and cannot accept any legal responsibility or liability for any errors or omissions
that may be made.
Cover design: LE-TEX Jelonek, Schmidt & Vöckler GbR, Leipzig, Germany
Printed on acid-free paper

9 8 7 6 5 4 3 2 1
springer.com
To Elisabetta and the kids (A.A.)
To Lara (D.K.)
To the memory of my father (R.O.)
Preface
Inthelastfew yearswehavewitnessedtheappearanceofaseriesofchalleng-
ing control engineering problems. Two common features of these new control
problems are that the interesting range of operationof the system is not nec-
essarily close to an equilibrium, hence nonlinear effects have to be explicitly
taken into account for a successful controller design, and that, even though
physicalmodellingallowstoaccuratelyidentifycertainwell-definednonlinear
effects, the controllerhasto cope witha highlevelofuncertainty,mainly due
to lack of knowledge on the system parameters and the inability to measure
the whole system state.
Thissituationjustifies the needforthe developmentoftoolsforcontroller
design for uncertain nonlinear systems, which is the main topic of this book.
Numerous theoretical control design methodologies for nonlinear sys-
tems have emerged over the last two decades. When viewed from a con-
ceptual standpoint, they can be broadly classified into analytically-oriented
and computationally-oriented. (The qualifiers analytical and computational
are used to distinguish between symbolical analysis and numerical computa-
tions.)
Theformerapproach,whichistheoneadoptedinthisbook,proceedsfrom
ananalyticalmodelofthe system,andthecontrollerdesignis theoutcomeof
asystematicprocessthatguaranteessomespecificbehaviour.Sincestabilityis
asine qua non condition,researchfollowingthisapproachusuallyrunsunder
the heading robust stabilisation, and it includes Lyapunov-based methods,
gain-assignmentmethods, and classical robust and adaptive tools.
Computationally-orientedtechniques,on the other hand, do not necessar-
ily require an analytical model, and they may be developed on the basis of a
numericalmodelofthesystemtobecontrolled—obtained,forinstance,bycol-
lecting large amounts of data to approximate its behaviour. Neural networks
based control, fuzzy control and intelligent control are the more conspicuous
representatives of this school. Recently, a second class of computationally-
orientedtechniques,thatreliesonanalyticalmodelsofthesystem,hasgained
some popularity. In an attempt to mimic the developments of linear systems
viii Preface
theory, piecewise linear (or linear parameter-varying) models are proposed
to capture nonlinear effects. Typically some optimal control objective is for-
mulated and the task of the controller design is to prove that, for the given
numerical values of the system model, the optimisation is feasible, e.g., it
can be translated into linear matrix inequalities, and a control signal can be
numerically computed. The optimal control approach suffers from two draw-
backs. First, the solutions are fragile with respect to plant uncertainty, e.g.,
lack of full state measurement and parametric uncertainty, which is the pre-
vailingconcerninmany,ifnotall,practicalapplications.Second,computation
of the optimal controllaw is feasible only for low-dimensionalsystems, which
puts a serious questionmark on the applicability ofthe method for nonlinear
systems. In addition there is not alwaysa clear reason,besides mathematical
convenience,to expressthe desiredbehaviourofadynamicalsysteminterms
of a scalar criterion to be optimised.
In summary, computationally-oriented approaches, while leveraging off a
swiftly growingcomputertechnology,providesolutionstosome specific prob-
lems,butdo notaimatexplainingwhy, how and when these solutionsindeed
work. At a more philosophical level, it is the authors’ opinion that controller
designshould not be reduced to the generationof numericalcode that imple-
mentsacontrollaw,withoutanyattempttotrytounderstandtheunderlying
mechanism that makes it work. This information is encoded in the nonlinear
system dynamics and revealed through a full-fledged nonlinear analysis.
We consider nonlinear control systems subject to various types of uncer-
tainty,includinglackofknowledgeontheparameters,partialmeasurementof
the systemstatesand uncertaintyonthe systemorderandstructure.To deal
with all these situations we follow a common thread encrypted in the words
immersion and invariance (I&I).
In the I&I approach we propose to capture the desired behaviour of the
system to be controlled introducing a target dynamical system. The control
problem is then reduced to the design of a control law which guaranteesthat
the controlled system asymptotically behaves like the target system. More
precisely,theI&Imethodologyreliesonfindingamanifoldinstate-spacethat
canberenderedinvariantandattractive,withinternaldynamicsacopyofthe
desired closed-loop dynamics, and on designing a control law that robustly
steers the state of the system sufficiently close to this manifold.
I&I should be contrasted with the optimal control approach where the
objective is captured by a scalar performance index to optimise. In addition,
because of its two-step approach, it is conceptually different from existing
(robust) stabilisationmethodologiesthat rely on the use of controlLyapunov
functions. However, it resembles the procedure used in sliding-mode control,
where a given manifold—the sliding surface—is rendered attractive by a dis-
continuous controllaw. The key difference is that, while in sliding-mode con-
trol the manifold must be reached by the trajectories, in the proposed ap-
proachthemanifoldneednotbereached.(Thisfeatureisessentialinadaptive
control and in output feedback design.)
Preface ix
Thebookisorganisedasfollows.Afterabriefintroductionwherethemain
ideasofI&Iareillustratedbymeansofexamples(Chapter1),inChapters2
and 3 we introduce the I&I framework and show how it can be used to solve
stabilisation and adaptive control problems.
InChapter 4themethodisappliedtononlinearsystemswithparametric
uncertainties,where it is assumedthat the full state is availablefor feedback.
InChapter 5weshowthatI&Iprovidesanaturalframeworkforobserver
design for general nonlinear systems.
In Chapter 6 the problem of output feedback stabilisation is solved for
classes of nonlinear systems, which include systems with unstable zero dy-
namics and the well-known output feedback form. Furthermore, the method
isextendedtoallowunstructureduncertaintiestoenterthesystemequations.
In Chapter 7 the I&I approach is used to design a class of nonlinear
proportional-integral controllers,where the gains of the controller are nonlin-
earfunctionsthatarechosentoguaranteestabilityforsystemswithunknown
parameters and uncertain disturbances.
Chapters 8, 9 and 10 are devoted to applications from electrical, me-
chanical and electromechanical systems, including power converters, power
systems, electric machines and autonomous aircraft.
AppendixAprovidesthebasicdefinitionsandrecallsbrieflyresultsused
throughout the book. In particular, characterisations of Lyapunov stability,
input-to-state stability and a nonlinear versionof the small-gaintheorem are
given along with some useful lemmas.
Acknowledgements
Thisbookistheresultofextensiveresearchcollaborationsduringthelastfive
years. Some of the results of these collaborations have been reported in the
papers[20,19,14,15,157,162,21,99,105,16,92,186,95,106,100,103,101,
93, 94, 102, 96, 39, 17, 97, 40, 104, 98, 18]. We are gratefulto our co-authors,
Nikita Barabanov,DanieleCarnevale,GerardoEscobar,Micka¨elHilairet,Liu
Hsu, Zhong-Ping Jiang, Eduardo Mendes, Mariana Netto, Hugo Rodr´ıguez,
and Aleksandar Stankovi´c, for several stimulating discussions and for their
hospitality while visiting their institutions. We also thank the research staff
of the Laboratoire de G´enie E´lectrique de Paris, for rendering available their
experimental facilities.
Someofthetopicsofthisbookhavebeentaughtbytheauthorsinaseries
of one-week graduate control courses offered in Paris for the last four years.
These have been organisedby the EuropeanCommission’s Marie Curie Con-
trol Training Site (CTS) and by the European Embedded Control Institute
(EECI) in the framework of the European Network of Excellence HYCON.
We would like to thank Franc¸oise Lamnabhi-Lagarriguefor giving us the op-
purtunity to teach during these schools.
Workshops on the topics presented in this book were organised at the
IEEEConference onDecisionandControl,Las Vegas,USA, 2002,andatthe
x Preface
XII Latin-American Congress on Automatic Control, Bahia, Brazil, 2006. A
mini-tutorial was given at the European Control Conference, Kos, Greece,
2007. We have delivered lectures on selected topics of the book in the DISC
SummerSchool,Eindhoven,TheNetherlands,2007,andinnumerousresearch
seminars.
Finally, a large part of this work would not have been possible without
the financial support of several institutions. The first author would like to
thank the Engineeringand PhysicalSciences ResearchCouncil(EPSRC)and
the Leverhulme Trust. The second author’s work was supported first by the
European Commission’s Training and Mobility of Researchers (TMR) pro-
grammethroughtheNonlinearandAdaptiveControl(NACO2)network,then
byBAESystems andthe EPSRCviathe FLAVIIRIntegratedProgrammein
Aeronautical Engineering, and finally by EPSRC via the Control and Power
Portfolio Partnership. The third author would like to thank the European
Network of Excellence HYCON for supporting part of his work.
Rome, London, Paris Alessandro Astolfi
April 2007 Dimitrios Karagiannis
Romeo Ortega