Algorithms of Estimation for Nonlinear

Systems A Differential and Algebraic

Viewpoint 1st Edition Rafael
Martínez-Guerra
Visit to download the full and correct content document:
https://textbookfull.com/product/algorithms-of-estimation-for-nonlinear-systems-a-diffe
rential-and-algebraic-viewpoint-1st-edition-rafael-martinez-guerra/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Algebraic and Differential Methods for Nonlinear

Control Theory Elements of Commutative Algebra and
Algebraic Geometry Rafael Martínez-Guerra

https://textbookfull.com/product/algebraic-and-differential-
methods-for-nonlinear-control-theory-elements-of-commutative-
algebra-and-algebraic-geometry-rafael-martinez-guerra/

Algebraic and differential methods for nonlinear

control theory Martinez-Guerra R

https://textbookfull.com/product/algebraic-and-differential-
methods-for-nonlinear-control-theory-martinez-guerra-r/

Systems of Nonlinear Partial Differential Equations

Applications to Biology and Engineering Mathematics and
Its Applications A.W. Leung

https://textbookfull.com/product/systems-of-nonlinear-partial-
differential-equations-applications-to-biology-and-engineering-
mathematics-and-its-applications-a-w-leung/

Nonlinear Differential Equations in Micro/Nano

Mechanics: Application in Micro/Nano Structures and
Electromechanical Systems 1st Edition Ali Koochi

https://textbookfull.com/product/nonlinear-differential-
equations-in-micro-nano-mechanics-application-in-micro-nano-
structures-and-electromechanical-systems-1st-edition-ali-koochi/
Introduction to Data Science Data Analysis and
Prediction Algorithms with R 1st Edition By Rafael A.
Irizarry

https://textbookfull.com/product/introduction-to-data-science-
data-analysis-and-prediction-algorithms-with-r-1st-edition-by-
rafael-a-irizarry/

Numerical Solutions of Three Classes of Nonlinear

Parabolic Integro-Differential Equations 1st Edition
Jangveladze

https://textbookfull.com/product/numerical-solutions-of-three-
classes-of-nonlinear-parabolic-integro-differential-
equations-1st-edition-jangveladze/

Dynamic Estimation and Control of Power Systems 1st

Edition Abhinav K. Singh

https://textbookfull.com/product/dynamic-estimation-and-control-
of-power-systems-1st-edition-abhinav-k-singh/

Observer Design for Nonlinear Systems Pauline Bernard

https://textbookfull.com/product/observer-design-for-nonlinear-
systems-pauline-bernard/

Aspects of Integrability of Differential Systems and

Fields A Mathematical Primer for Physicists Costas J.
Papachristou

https://textbookfull.com/product/aspects-of-integrability-of-
differential-systems-and-fields-a-mathematical-primer-for-
physicists-costas-j-papachristou/
Understanding Complex Systems

Rafael Martínez-Guerra
Christopher Diego Cruz-Ancona

Algorithms
of Estimation
for Nonlinear
Systems
A Differential and Algebraic Viewpoint
Springer Complexity
Springer Complexity is an interdisciplinary program publishing the best research and
academic-level teaching on both fundamental and applied aspects of complex systems –
cutting across all traditional disciplines of the natural and life sciences, engineering,
economics, medicine, neuroscience, social and computer science.
Complex Systems are systems that comprise many interacting parts with the ability to
generate a new quality of macroscopic collective behavior the manifestations of which are
the spontaneous formation of distinctive temporal, spatial or functional structures. Models
of such systems can be successfully mapped onto quite diverse “real-life” situations like
the climate, the coherent emission of light from lasers, chemical reaction-diffusion systems,
biological cellular networks, the dynamics of stock markets and of the internet, earthquake
statistics and prediction, freeway traffic, the human brain, or the formation of opinions in
social systems, to name just some of the popular applications.
Although their scope and methodologies overlap somewhat, one can distinguish the
following main concepts and tools: self-organization, nonlinear dynamics, synergetics,
turbulence, dynamical systems, catastrophes, instabilities, stochastic processes, chaos, graphs
and networks, cellular automata, adaptive systems, genetic algorithms and computational
intelligence.
The three major book publication platforms of the Springer Complexity program are the
monograph series “Understanding Complex Systems” focusing on the various applications
of complexity, the “Springer Series in Synergetics”, which is devoted to the quantitative
theoretical and methodological foundations, and the “SpringerBriefs in Complexity” which
are concise and topical working reports, case-studies, surveys, essays and lecture notes of
relevance to the field. In addition to the books in these two core series, the program also
incorporates individual titles ranging from textbooks to major reference works.

Editorial and Programme Advisory Board

Henry Abarbanel, Institute for Nonlinear Science, University of California, San Diego, USA
Dan Braha, New England Complex Systems Institute and University of Massachusetts Dartmouth, USA
Péter Érdi, Center for Complex Systems Studies, Kalamazoo College, USA and Hungarian Academy
of Sciences, Budapest, Hungary
Karl Friston, Institute of Cognitive Neuroscience, University College London, London, UK
Hermann Haken, Center of Synergetics, University of Stuttgart, Stuttgart, Germany
Viktor Jirsa, Centre National de la Recherche Scientifique (CNRS), Université de la Méditerranée, Marseille,
France
Janusz Kacprzyk, System Research, Polish Academy of Sciences, Warsaw, Poland
Kunihiko Kaneko, Research Center for Complex Systems Biology, The University of Tokyo, Tokyo, Japan
Scott Kelso, Center for Complex Systems and Brain Sciences, Florida Atlantic University, Boca Raton, USA
Markus Kirkilionis, Mathematics Institute and Centre for Complex Systems, University of Warwick,
Coventry, UK
Jürgen Kurths, Nonlinear Dynamics Group, University of Potsdam, Potsdam, Germany
Ronaldo Menezes, Dept of Computer Science, Florida Institute of Technology, Melbourne, Florida, USA
Andrzej Nowak, Department of Psychology, Warsaw University, Poland
Hassan Qudrat-Ullah, School of Administrative Studies, York University, Toronto, ON, Canada
Peter Schuster, Theoretical Chemistry and Structural Biology, University of Vienna, Vienna, Austria
Frank Schweitzer, System Design, ETH Zurich, Zurich, Switzerland
Didier Sornette, Entrepreneurial Risk, ETH Zurich, Zurich, Switzerland
Stefan Thurner, Section for Science of Complex Systems, Medical University of Vienna, Vienna, Austria
Understanding Complex Systems
Founding Editor: S. Kelso

Future scientific and technological developments in many fields will necessarily

depend upon coming to grips with complex systems. Such systems are complex in
both their composition – typically many different kinds of components interacting
simultaneously and nonlinearly with each other and their environments on multiple
levels – and in the rich diversity of behavior of which they are capable.
The Springer Series in Understanding Complex Systems series (UCS) promotes
new strategies and paradigms for understanding and realizing applications of
complex systems research in a wide variety of fields and endeavors. UCS is
explicitly transdisciplinary. It has three main goals: First, to elaborate the concepts,
methods and tools of complex systems at all levels of description and in all scientific
fields, especially newly emerging areas within the life, social, behavioral, economic,
neuro- and cognitive sciences (and derivatives thereof); second, to encourage novel
applications of these ideas in various fields of engineering and computation such as
robotics, nano-technology and informatics; third, to provide a single forum within
which commonalities and differences in the workings of complex systems may be
discerned, hence leading to deeper insight and understanding.
UCS will publish monographs, lecture notes and selected edited contributions
aimed at communicating new findings to a large multidisciplinary audience.

More information about this series at http://www.springer.com/series/5394

Rafael Martínez-Guerra •
Christopher Diego Cruz-Ancona

Algorithms of Estimation
for Nonlinear Systems
A Differential and Algebraic Viewpoint

123
Rafael Martínez-Guerra Christopher Diego Cruz-Ancona
CINVESTAV-IPN CINVESTAV-IPN
Departamento de Control Automatico Departamento de Control Automatico
Mexico, Distrito Federal Mexico, Distrito Federal
Mexico Mexico

ISSN 1860-0832 ISSN 1860-0840 (electronic)

Understanding Complex Systems
ISBN 978-3-319-53039-0 ISBN 978-3-319-53040-6 (eBook)
DOI 10.1007/978-3-319-53040-6

Library of Congress Control Number: 2017937655

© Springer International Publishing AG 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, 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.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, express or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
To my wife and sons,
Marilen, Rafael, and Juan Carlos
To my parents and brothers,
Carlos, Virginia, Victor, Arturo, Carlos,
Javier, and Marisela
Remerciements:
Je tiens à remercier d’une manière très
spéciale au professeur Michel Fliess pour me
introduire dans le monde merveilleux de
l’algèbre différentielle avec des applications
dans la théorie du contrôle, ainsi que mon
ami professeur Sette Diop avec qui j’ai eu le
plaisir de collaborer dans certaines
publications je veux exprimer ma plus
profonde reconnaissance et ma gratitude.
Rafael Martínez-Guerra

To my family,
Evaristo, Carmen, Junior, and Robin,
and specially to
Alejandra, and Quico and Max, of course
Christopher Diego Cruz-Ancona
Preface

The purpose of this book is to acquaint the reader with current developments in
nonlinear systems theory and its applications. This book is directed particularly
to graduate students in the fields of control of physical processes and computing,
biophysics, and physical information science. A great deal of literature related
to the control problems of nonlinear systems has been developed over the past
decades. This book proposes several algorithms motivated by systems models that
can approximate a wide class of nonlinear control systems. They can be represented
either as a state space model or an input–output equation. There exists a class of
nonlinear systems called input–affine systems or bilinear systems containing the
simplest nonlinear systems, which can be used to represent a wide range of physical,
chemical, biological, and social systems as well as manufacturing processes that
cannot be effectively modeled under linearity assumptions. The concept of bilinear
systems was introduced in the 1960s. In automatic control theory, these are control
systems whose dynamics are jointly linear in the state and multiplicative control
variables. Their theory can be developed from the theories of time-variant linear
systems and Lie groups; they have been applied to many areas of service and
technology. The purpose of this book is to discuss the development of these ideas.
We emphasize the role of several disciplines that modified our outlook on nonlinear
systems theory, among them differential-algebraic theory, the modern control theory
of dynamical systems, observers for nonlinear systems, generalized canonical
forms, and parametric and state estimation. Affine systems can approximate a wide
class of nonlinear control systems. Input–affine systems can be represented as state
space models or systems in the form

xP .t/ D f .x/ C u .x/ g .x/

y .t/ D h .x/ ;

vii
viii Preface

where f .x/, g .x/, h.x/ are C1 functions, or in a simpler form,

!
X
xP .t/ D A C ui .t/ Bi x .t/ ; x .0/ D xo 2 M ;
i

y D Kx.t/;

with Ai and Bi matrices of appropriate sizes and M 2 Rn an algebraic manifold.

Martínez Guerra (1996) used Fliess generalized the nonlinear observable canonical
form and generalized nonlinear controllable canonical form to derive an observer-
based controller for nonlinear systems. Our contributions to this book have the
following intentions: to introduce a new differential-algebraic approach to observer
design that allows one to characterize a class of nonlinear systems transformable to
state affine systems or input–affine systems, modulo an output injection; establish a
generalized nonlinear observability canonical form with output injection (translated
GOCF) in the multivariable case (multi-input, multi-output); characterize explicitly
the type of coordinate transformation (using the differential primitive element) that
takes this class of nonlinear systems (NLS) into a translated generalized nonlinear
observability canonical form multi-input, multi-output, modulo an output injection
(single variable and multivariable cases); propose some type of observers, for
instance a high-gain Kalman-type observer, a Luenberger-type observer with time-
varying gains, a proportional reduced-order observer; design a control law based on
differential-algebraic tools that solves tracking and (local) stabilization problems by
introducing an observer to estimate a tracking error and its derivatives, as well as
an extension for the case with fault presence in actuators; introduce a simultaneous
parametric and state estimation approach; and finally, enhance these techniques in a
broader class of nonlinear systems with applications to chemical reactors. We give
some motivating examples to verify the effectiveness of the methodologies proposed
via numerical results. This book is intended to be self-contained and suitable for
self-study. A mathematical prerequisite is basic knowledge of dynamical systems,
stability, and modern algebra. Some appendices were added at the end of the book
to facilitate the readability of the text and concentrate on the main problems tackled
here. The text consists of ten chapters, whose sequencing is depicted in the following
diagram:
Preface ix

To grasp the essence of the differential-algebraic tools, it would be useful

to read the first four chapters of this book, which constitute an introduction
to the foundations and basic concepts of this field. Of course, more advanced
readers can simply scan the first chapters and focus on the remaining chapters,
which contain applications for nonlinear systems. More specifically, Chap. 1 is
concerned with problems of systems analysis of input–affine processes. Chapter 2
presents some concepts and elementary definitions from differential algebra that
are necessary for understanding subsequent chapters. In Chap. 3, the concept of
algebraic observability is established for a given nonlinear dynamics defined as a
finitely generated differential-algebraic extension, and the external behavior of a
universally observable system is discussed.
In Chap. 4, the concept of generalized observability canonical form (GOCF)
is introduced for the multivariable and single-variable cases. Then in Chap. 5,
the synthesis of some observers, for instance of Kalman and Luenberger type,
with time-varying gains is proposed to estimate the state for a class of nonlinear
systems. Chapter 6 contains the implementation of a observer–controller able to
solve tracking and (local) stabilization problems as well as an application to a
chemical reactor model. Chapter 7 deals with parametric and state estimation simul-
taneously with some different techniques, among them the immersion technique.
Also, we introduce the concept of parametric identification in terms of parametric
observability. Chapter 8 extends the observer synthesis problem to a more complex
class of nonlinear systems. Chapter 9 shows a separation principle for nonlinear
systems that is satisfied in the sense that overall systems of the plant and the
observer-based controller are uniformly asymptotically stable, and the effectiveness
of the methodology is illustrated with several examples: an academic example, a
continuously stirred tank biological reactor, a synchronous generator, and a single-
link flexible joint manipulator. Furthermore, a fault-tolerant control problem with
application to a three-tank system is given. And finally, Chap. 10 tackles some
observers that have not been much studied in the literature, such as bounded,
invariant, and proportional integral (PI) uncertain observers with some interesting
applications.

Mexico, Distrito Federal, Mexico Rafael Martínez-Guerra

Mexico, Distrito Federal, Mexico Christopher Diego Cruz-Ancona
Contents

1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 1
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 5
2 Differential Algebra in Control Theory . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 7
2.1 Elements of Differential Algebra .. . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 8
2.1.1 Transcendental Fields and Algebraic Fields,
Subfields, Differential Field Extensions .. . . . . . . . . . . . . . . . . . 8
2.2 Differential Rings and Differential Fields . . . . . . .. . . . . . . . . . . . . . . . . . . . 9
2.2.1 Differential Rings . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 10
2.2.2 Notion of a Differential Field Extension:
Algebraic and Transcendental.. . . . . . . . .. . . . . . . . . . . . . . . . . . . . 12
2.2.3 Differential Primitive Element, Differential
Transcendence Basis, Differential
Transcendence Degree, and Tower
of Differential Field Extensions.. . . . . . .. . . . . . . . . . . . . . . . . . . . 12
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 13
3 Algebraic Observability for Nonlinear Systems . . . . .. . . . . . . . . . . . . . . . . . . . 15
3.1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 15
3.2 Algebraic Observability .. . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 16
3.2.1 Nonlinear Dynamics .. . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 16
3.2.2 Algebraic Observability Property
in Nonlinear Systems . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 17
3.2.3 Examples .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 18
3.3 On the External Behavior of a Universally
Observable System .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 19
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 24
4 Generalized Observability Canonical Form .. . . . . . . .. . . . . . . . . . . . . . . . . . . . 25
4.1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 25
4.2 Differential Primitive Element .. . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 27
4.3 Single-Variable Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 28

xi
xii Contents

4.4 Multivariable Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 29

References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 36
5 Observers for a Class of Nonlinear Systems . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 39
5.1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 39
5.2 Single-Variable Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 40
5.3 Multivariable Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 47
5.4 An Application to a Chemical Reactor Model ... . . . . . . . . . . . . . . . . . . . 49
5.4.1 Single-Variable Chemical Reactor . . . . .. . . . . . . . . . . . . . . . . . . . 49
5.4.2 Multivariable Chemical Reactor . . . . . . .. . . . . . . . . . . . . . . . . . . . 52
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 55
6 Observer-Based Local Stabilization and Asymptotic
Output Tracking .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 57
6.1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 57
6.2 Feedback Linearization of Nonlinear Dynamics.. . . . . . . . . . . . . . . . . . . 58
6.3 Linearization via a Dynamic State Feedback . . .. . . . . . . . . . . . . . . . . . . . 58
6.4 Separation Principle for Linear Systems . . . . . . . .. . . . . . . . . . . . . . . . . . . . 58
6.5 Output Tracking and (Local) Stabilization Problem .. . . . . . . . . . . . . . . 60
6.5.1 Application to a Chemical Reactor Model . . . . . . . . . . . . . . . . 70
6.5.2 Numerical Simulation and Results. . . . .. . . . . . . . . . . . . . . . . . . . 72
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 74
7 Parametric Identification of Time-Varying Nonlinear Systems . . . . . . . 77
7.1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 77
7.2 Reduced-Order Observer for Parameter Estimation . . . . . . . . . . . . . . . . 79
7.2.1 Motivating Example .. . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 81
7.3 Parametric and State Estimation .. . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 85
7.4 High-Gain Observer for Parametric and State Estimation.. . . . . . . . . 87
7.4.1 Application to Bioreactors . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 89
7.4.2 Numerical Results . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 89
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 91
8 More General Nonlinear Systems Case . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 93
8.1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 93
8.2 Synthesis of an Observer for a Nonlinear Dynamics . . . . . . . . . . . . . . . 95
8.3 Observer Synthesis .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 97
8.4 Application to Nonlinear Models .. . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 99
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 103
9 A Separation Principle for Nonlinear Systems . . . . . .. . . . . . . . . . . . . . . . . . . . 105
9.1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 105
9.2 Asymptotic Stabilization and Output Tracking .. . . . . . . . . . . . . . . . . . . . 107
9.2.1 Examples .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 116
9.3 Other Classes of Observers . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 127
9.3.1 Synchronous Generator . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 129
9.3.2 Single-Link Flexible Joint Manipulator . . . . . . . . . . . . . . . . . . . 134
Contents xiii

9.4 Fault-Tolerant Control . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 141

9.5 Fault-Tolerant Dynamic Controller with Numerical Results . . . . . . . 141
9.6 On the Fault Diagnosis Problem .. . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 146
9.7 Asymptotic Stability of the Closed-Loop System
in the Presence of Faults . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 147
9.7.1 Application to the Three-Tank System . . . . . . . . . . . . . . . . . . . . 150
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 155
10 Some Other Observers for Nonlinear Systems . . . . . .. . . . . . . . . . . . . . . . . . . . 159
10.1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 159
10.2 Bounded Observer.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 159
10.2.1 Motivational Example: Rössler Chaotic System . . . . . . . . . . 160
10.2.2 Convergence Analysis . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 162
10.3 PI Observers for Uncertainty Estimation in Continuous
Chemical Reactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 164
10.3.1 Problem Statement . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 165
10.3.2 Numerical Simulation . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 168
10.4 An Invariant Observer for Fault Diagnosis:
A Real-Time Application . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 170
10.4.1 The Left Invertibility Condition . . . . . . .. . . . . . . . . . . . . . . . . . . . 171
10.4.2 Invariant Observer . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 172
10.4.3 Construction of the Observers . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 174
10.4.4 Observer Design . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 178
10.4.5 Application to the Three-Tank System . . . . . . . . . . . . . . . . . . . . 179
10.4.6 Experimental Results . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 180
10.4.7 Identification .. . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 180
10.4.8 Fault Estimation Results . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 181
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 186

A Singularity Treatment .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 187

B Some Properties for Nonlinear Systems. . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 191

References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 193

Index . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 195
List of Figures

Fig. 2.1 Derivation in a commutative ring A . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 10

Fig. 5.1 Numerical simulation of a Luenberger-type
observer: convergence of estimates xO D .Ox1 xO 2 /T
to states x D .x1 x2 /T (left) and observation error
" D ."1 "2 /T convergence to zero (right) . . . . . .. . . . . . . . . . . . . . . . . . . . 50
Fig. 5.2 Numerical simulation of a reduced-order observer:
convergence of estimate xO 1 to x1 (left) and observation
error "1 convergence to zero (right) . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 52
Fig. 5.3 Numerical simulation of Kalman-type observer:
convergence of estimates xO D .Ox1 xO 2 /T to states x D .x1 x2 /T . . . 54
Fig. 5.4 Numerical simulation of Kalman-type observer:
convergence of estimates xO D .Ox3 xO 4 /T to states x D .x3 x4 /T . . . 54
Fig. 6.1 Product concentration (output) tracking .. . . . . . .. . . . . . . . . . . . . . . . . . . . 72
Fig. 6.2 Reactant concentration tracking . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 73
Fig. 6.3 Tracking error estimation, trajectories of eO 1 , e1 . . . . . . . . . . . . . . . . . . . . 73
Fig. 6.4 Tracking error estimation, trajectories of eO 2 , e2 . . . . . . . . . . . . . . . . . . . . 74
Fig. 6.5 Observer-based controller response u . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 74
Fig. 7.1 Convergence of estimate O to real time-varying
parameter  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 83
Fig. 7.2 Multiplicative state and parameter estimation,
uncertain parameter  vs. estimate O . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 83
Fig. 7.3 Auxiliary variable  .t/ . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 84
Fig. 7.4 Convergence of uncertain parameter error .t/ . . . . . . . . . . . . . . . . . . . . 84
Fig. 7.5 Convergence of the parameter estimate O to the real
parameter  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 87
Fig. 7.6 Convergence of parameter estimate O to real parameter  . . . . . . . . 90

xv
xvi List of Figures

Fig. 7.7 Convergence of parameter estimate O to real parameter  . . . . . . . . 91

Fig. 8.1 Ex. 1: convergence of the estimated state xO 1 to the real
state x1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 101
Fig. 8.2 Ex. 1: convergence of the estimated state xO 2 to the real
state x2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 101
Fig. 8.3 Ex. 2: convergence of the estimated states xO 1 and xO 2 to
the real states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 102
Fig. 8.4 Ex. 2: convergence of the estimated states xO 1 and xO 2 to
the real states with output noise . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 103
Fig. 9.1 The state 1 converges to the desired reference
trajectory yR .t/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 119
Fig. 9.2 The derivative reference trajectory dyR =dt and the
actual controlled dynamical response 2 . . . . . . .. . . . . . . . . . . . . . . . . . . . 119
Fig. 9.3 The derivative reference trajectory d2 yR .t/=dt2 and
the actual controlled dynamical response 3 . . .. . . . . . . . . . . . . . . . . . . . 120
Fig. 9.4 The dynamic controller uR .t/ . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 120
Fig. 9.5 The desired trajectory yR .t/ D X2 and the actual
dynamically controlled responses x2 and xO 2 . . . .. . . . . . . . . . . . . . . . . . . . 122
Fig. 9.6 The trajectories of states x1 and xO 1 . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 123
Fig. 9.7 The dynamic controller ueO .t/ . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 123
Fig. 9.8 CSTR’s controlled composition . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 125
Fig. 9.9 Tracking error dynamics . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 126
Fig. 9.10 Control input: flow of pure B . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 126
Fig. 9.11 The power angle and the reference signal . . . . . .. . . . . . . . . . . . . . . . . . . . 133
Fig. 9.12 The rotor angular speed . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 133
Fig. 9.13 The stator voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 134
Fig. 9.14 The control law applied to the generator . . . . . . .. . . . . . . . . . . . . . . . . . . . 134
Fig. 9.15 The desired trajectory yR .t/ D x1 . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 139
Fig. 9.16 The desired trajectory y0R .t/ D x1 . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 140
Fig. 9.17 The dynamic controller ue .t/ . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 140
Fig. 9.18 Schematic diagram of TTS. . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 150
Fig. 9.19 Output 1 tracking of the TTS . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 154
Fig. 9.20 Output 2 tracking of the TTS . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 154
Fig. 9.21 Fault estimation 1 of the TTS . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 155
Fig. 9.22 Fault estimation 2 of the TTS . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 155
Fig. 10.1 Convergence of estimates: (a) x1 vs xO 1 , (b) x2 vs xO 2
and (c) x3 vs. xO 3 .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 164
Fig. 10.2 Relative errors, eN 2 , eN 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 164
Fig. 10.3 Filtering process of the reactor temperature
measurements considering sustained disturbances
(m D 0:01) .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 168
Fig. 10.4 Online estimation of the heat of reaction, considering
sustained disturbances (m D 0:01) . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 169
List of Figures xvii

Fig. 10.5 Online estimation of the heat of reaction (m D 0:001) . . . . . . . . . . . . 170

Fig. 10.6 On-line estimation of the heat of reaction (m D 0:1) .. . . . . . . . . . . . . 171
Fig. 10.7 Parameter identification .. . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 181
Fig. 10.8 Reduced-order observer: (a) Level estimation, (b)
Fault reconstruction f1 and (c) Fault reconstruction f2 .. . . . . . . . . . . . 182
Fig. 10.9 Invariant observer: (a) Level estimation, (b) Fault
reconstruction f1 and (c) Fault reconstruction f2 . . . . . . . . . . . . . . . . . . . 183
Fig. 10.10 LTV observer: (a) Level estimation, (b) Fault
reconstruction f1 and (c) Fault reconstruction f2 . . . . . . . . . . . . . . . . . . . 184
Fig. 10.11 Performance evaluation of observers for the
estimation error of fault f1 . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 185
Fig. 10.12 Performance evaluation of observers for the
estimation error of fault f2 . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 185
Fig. A.1 Effect of singularity treatment on the estimate x1e .. . . . . . . . . . . . . . . . 188
Fig. A.2 Effect of singularity treatment on the estimate x2e .. . . . . . . . . . . . . . . . 189
Symbols and Acronyms

_ Disjunction, or conditional OR
, By definition
 Approximately equal
WD Defined as
 Equivalent to
¤ Different from
< .>/ Less (greater) than
 ./ Less (greater) than or equal to
8 For all
2 Belongs to
 Subset of
[ Union of sets
fg Set
! Tends to
) Implies
”
P Equivalent to, if and only if
Summation
R The set of real numbers
RC Positive real numbers
C The set of complex numbers
C Complex numbers with negative real part
Rmn The set of all m  n matrices with elements from R
Rmn
C The set of all m  n matrices with elements from RC
Œaij  m  n matrix with entries aij , 1  i  m, 1  j  n
Amn Matrix with m rows and n columns
det A The determinant of a square matrix A 2 Rnn
jAj The determinant of a matrix A
Tr A The trace of a matrix A
AT The transpose of a matrix obtained by interchanging the rows
and columns of A

xix
xx Symbols and Acronyms

.; /T The transpose of a vector

.; /0 The transpose of a vector
col.; / Column vector
rank A The minimal number of linearly independent rows or columns
of A 2 Rmn
A1 Inverse of A
max Maximum
min Minimum
sup Supremum, the least upper bound
inf Infimum, the greatest lower bound
Br The ball fx 2 Rn j k x k rg
f W S1 ! S2 A function f mapping a set S1 into a set S2
max .P/(min .P/) The maximum (minimum) eigenvalue of a symmetric matrix P
P>0 A positive definite matrix P
yP First derivative of y with respect to time
yR Second derivative of y with respect to time
«
y Third derivative of y with respect to time
@
f Partial derivative of function f
@x
DC y Upper Dini derivative of y
y.i/ ith derivative of y with respect to time
lim Limit
h ! 0C h approaches zero from above
jaj Absolute value of a scalar a
kxk The Euclidean norm of a vector
kxk2P The quadratic form xT Px with weight matrix P
h; i Dot product
diff tr dı Differential transcendence degree
1 Infinity
 Designation of the end of a proof
lim sup Limit superior
Khui Differential field generated by the field K, the input u.t/, and
the time derivatives of u
Khu; yi Differential field generated by the field K, the input u.t/, the
output y.t/, and the time derivatives of u and y
xO Estimated value of x
AOC Algebraic observability condition
SISO Single input single output
MIMO Multiple input multiple output
@ .x/
Jacobian matrix of  .x/
@x
K=k Differential field extension .k  K/
C1 Smooth functions
Chapter 1
Introduction

Consider a dynamical system to be controlled. Commonly, it is necessary to

reconstruct the state of the system in order to apply a designed control strategy.
When we are heading in this direction, there are some difficulties we have to
take into account: the use of sensors and gauges to measure state variables could
be difficult and impractical, or even impossible in the case of state variables
without physical meaning. That is why the estimation of these variables becomes an
important issue, and from observers theory we have found a solution to this problem.
An observer is an auxiliary dynamical system that allows one to estimate an
unknown state of another system with information about the input and output of
that system. In other words, the observer synthesis is explained in the following
way.
Given a system ˙ described by differential equations with input u.t/ and a
function describing its output y.t/, take another system O whose inputs and outputs
are identical to (those of) the original system ˙. This new system O is called an
observer for system ˙ that allows one to estimate an unknown state of system ˙.
The idea of observability, in the case of linear systems, is a sufficient condition
linked to the reconstruction of a state of the system. There are two well-known
cases related to observer synthesis:
• Luenberger observer [21],
• Kalman observer [16].
In 1963, Luenberger studied observers theory with aim to reconstruct a system state
for linear dynamical systems with constant coefficients. The use of observers and
their relationship to the fundamentals of systems theory make observers theory an
important area of research interest. Since Kalman introduced the methods of state
variables, observers theory has appeared in causal multivariable systems.
It is well known that a system is observable if and only if from observations of its
output and input it is possible to reconstruct the initial state of the system. (Later, we
will introduce a mathematical formulation of this concept.) There has been a great

© Springer International Publishing AG 2017 1

R. Martínez-Guerra, C.D. Cruz-Ancona, Algorithms of Estimation
for Nonlinear Systems, Understanding Complex Systems,
DOI 10.1007/978-3-319-53040-6_1
2 1 Introduction

deal of interest shown with regard to the observation problem, and examples of the
literature devoted to this topic from different perspectives for nonlinear systems and
linear systems include [1, 2, 5, 17]. Here, our perspective is different, since we adopt
a differential-algebraic approach to overcome this problem.
Given a input–affine system class, the observer synthesis is not obvious. In this
case, according to the input, we can distinguish two types of systems: those that are
observable for any input (universally observable system) and those with singular
inputs (namely, inputs that make the system unobservable, i.e., “bad inputs”). In
designing an observer, bad inputs are clearly inconvenient, because they add a
singularity problem. In the case of linear systems, the absence of singularities
explains the design and use of asymptotic observers (with arbitrary convergence
estimation speed), such as Kalman and Luenberger observers.
The first results obtained for observable systems for any input lead us to the case
in which the class of input–affine systems may be transformable to an observable
time-varying linear system with an output injection.
The observability property for nonlinear systems is associated with inputs u.t/.
Hence, the concept of “singular” inputs is implicit. That is to say, for a nonlinear
observable system, certain inputs u.t/ make unobservable all pairs of points in the
state space and complicate the synthesis of an observer. Then it is of interest to know
and describe these inputs (topological properties) with the aim of avoiding them.
Bilinear systems, besides being the simplest nonlinear systems, have great
practical importance, since many systems are modeled in this form (ecological
systems, nuclear reactors, petrochemistry, etc.). In [28], a model order reduction
algorithm for a given SISO bilinear system based on optimality conditions is
proposed. It is shown in [4, 12] that under a nonlinear transformation (e.g., via
immersion [20, 23]; output injection, etc.), it is possible to describe a nonlinear
system by an input–affine system (it may have greater dimension than the initial
system in the case of an immersion) and to reconstruct the nonlinear system state
from an observer for the input–affine system. Consequently, the study of input–
affine systems is fundamental. Now, a logical approach is to design observers for
the simplest case (input–affine systems ˙1 ) before tackling the problem of a more
general class of nonlinear systems.
Consider the class of input–affine systems that can be modeled by means of the
following relationships:

˙1 W xP D .A0 C ˙im Ai ui /x C Bu;

y D Cx C Du;

with x 2 Rn , u 2 Rm , y 2 Rp , and matrices Ai , 0  i  m, B, C, and D of appropriate

dimensions.
1 Introduction 3

Such systems can be transformed into a class of systems represented by ˙2 [22]:

˙2 W zP D Au z C '.u; y/;
y D Cz;

where system ˙2 is observable independently of the value of the input u if and only
if z is observable with respect to fu; yg , i.e., z is algebraic over K hu; yi (z satisfies a
differential-algebraic equation on K hu; yi). Note that the system ˙2 can be viewed
as a time-varying linear system with output injection; the time-variant character
is due to matrix Au having its entries in K hui(elements of matrix Au depend on
input). Notice that the synthesis of an observer is simpler for the system ˙2 than
for the system ˙1 . So if we design an observer for the transformed system, we can
obtain estimates for system ˙1 . Therefore, a question that arises is this: what kind
of observers are valid for ˙2 ? For instance, we can think of the following observers.
Considering that ˙2 is observable, we can find Lu such that the matrix expression
(Au Lu C) has its spectrum in the complex left half- plane (.Au Lu C/ C ).
Then we get a differential-algebraic polynomial in j , u, and their time derivatives:

Lu D ˇNj . j ; u; uP ; : : :/; 1  j  n; ˇNj D constant; (1.1)

where . j ; u; uP ; : : :/ is a differential-algebraic polynomial in Khui and j D

fLui ; 1  i  j  ng.
Choosing Lu as above, it is possible to design an exponential observer given by

zPO D Au zO C '.u; y/ Lu .y yO /: (1.2)

However, a Kalman-type observer can be proposed. This can be described as

zPO D Au zO C '.u; y/ S1 CT ŒCOz y (1.3)

SP D S ATu S SAu C CT C; .Lyapunov equation/

where  is a strictly positive constant and S satisfies a time-varying Lyapunov

equation. In both cases, the estimate zO converges exponentially to the state z.t/ as t
tends to infinity .t ! 1/.
When it is necessary to estimate only part of the state, a more intuitive synthesis
of an observer is given when we do not need to know beforehand the system
itself and the reconstruction of the state should be given only using the system’s
output. Then a proportional reduced-order observer (so-called free model type) can
be synthesized, which is described as follows:

zPOi D ki .zi zOi /

4 1 Introduction

with a constant gain ki > 0 for 2  i  n. Since zi cannot be obtained from the
model, an auxiliary system

Pi D ki i ki2 zNi1 ;
zOi D i C ki zNi1 ;

is a stable observer that can be regarded as a reduced-order observer for the ith state
of system ˙2 , where zN D .y; zO2 ; : : : ; zOn1 / is the output and the estimates of the
output derivatives. In this case, the estimation error zi zOi converges asymptotically
to a closed ball centered at the origin (compact set).
Some authors have worked with the development of state estimators for restricted
classes of nonlinear systems, either from a geometric viewpoint (theory mainly
attributed to Isidori, Hermann, Krener, Van der Schaft, Nijmeijer, Gauthier, Bornard,
Hammouri, Zeitz, Williamson, Levine and Marino, Khalil) or from a differential-
algebraic viewpoint (introduced by Fliess in 1986). Among other approaches, we
can mention the work in [26], where a method to synthesize an observer for an
uncertain Takagi–Sugeno fuzzy system is proposed; the case of observers for time
delay input–affine systems is studied in [9]; in [15], a predictor input–affine observer
is developed for a class of nonlinear systems subject to delayed outputs; [24]
considers the problem of optimal input signal estimation for input–affine systems
under output measurements and parametric estimation.
The foundations of the algebraic approach are based on differential algebra [7],
from which it has been attempted to capture the algebraic nature of the observability
concept. Preliminary work in this direction has been published [6, 13, 25].
In [18] and [19], the authors characterized input–affine systems and in general
nonlinear systems. These are carried out by means of a diffeomorphism into some
nonlinear canonical form of the type given in ˙2 .
For input–affine systems, asymptotic observers are found in some works such as
Hara–Furuta [14] and Funahashi [8], where some sufficient conditions are given in
order to guarantee asymptotic convergence of the estimation error.
In [27], Williamson shows that input–affine systems without singular inputs have
an observability canonical form, and from that structure he gives an appropriate
observer. A generalization to the former result is given by J.P. Gauthier and G.
Bornard [10] for observable nonlinear systems with respect to any input.
Using the nonlinear canonical form, it is shown in [11] that those systems have
a high-gain observer. An extension for observable nonlinear systems regarding any
input are found in [3].
In general, input–affine systems, modulo an output injection, have singular
inputs. Despite this fact, there are observers that work for some classes of inputs.
These observers are a modification of the Kalman filter in the deterministic case.
References 5

References

1. Balatif, et al.: Optimal control problem for a class of bilinear systems via block pulse functions.
IMA J. Math. Control Inform. (2016), dnw005
2. Besançon, G.: Nonlinear Observers and Applications. Lecture Notes in Control and Informa-
tion Sciences. Springer, New York (2007)
3. Bornard, G., Hammouri, H.: A high gain for a class of nonlinear systems under locally regular
inputs. In: IEEE Conf. on Dec. and Control CDC Brighton G.B. (1991)
4. Bornard, G., Couenne, N., Celle, F.: Regularly persistent observers for bilinear systems.
In: Proceedings of the 29 International Conference on nonlinear Systems. New Trends in
Nonlinear System Theory, Vol 122. Springer, New York (1988)
5. Fan, L., Wang, W.: A GPI based sliding mode control method for boost DC-DC converter.
IEEE Int Conference on Industrial Technology, pp. 1826–1831, March, 2016
6. Fliess, M.: Quelques Remarques sur les observateurs nonlinéaires. Proc. Colloque Gretsi Trait.
Signal Images, Nice pp. 169–172, 1987
7. Fliess, M.: Automatique et Corps différentiel. Forum Math. 1, 227–238 (1989)
8. Funahashi, Y.: Stable state estimator for bilinear systems. Int. J. Control 29(2), 181–188 (1979)
9. Gamiochipi, J., Ghanes, M., Aggome, W., De Leon, J., Barbot, J-P.: Observer design for a class
of nonlinear systems under a persistent excitation. In: 10th IFAC Symposium on Nonlinear
Control Systems, NOLCOS, 2016
10. Gauthier, J.P., Bornard, G.: Observbility for any u(t) of a class of nonlinear systems. IEEE
Trans. Automatic Control AC-26(4), 922–926 (1981)
11. Gauthier, J.P., Hammouri, H., Othman, S.: A simple observer for nonlinear systems applica-
tions to bioreactors. IEEE Trans. Automatic Control 36(6), 875–880 (1992)
12. Gauthier, J.P., Kazakos, D.: Observabilite et observateurs de systemes non lineaires. RAIRO
APII Analyse des Systemes 21, 201–212 (1987)
13. Glad, S.T.: Differential algebraic modelling of nonlinear systems. In: Kaashoek, M.A., Van
Schuppen, J.H., Ran, A.C.M. (eds.), Realization and Modelling in System Theory MTNS-89,
pp. 97–105. Birkhäuser, Boston (1989)
14. Hara, S., Furuta, K.: Minimal order state observers for bilinear systems. Int. J. Control. 24(5),
705–718 (1976)
15. Ibrir, S., Hunte, K., Bettayeb, M.: State observation of a class of bilinear system with large
sensor delay. IFAC Procc. 47(3), 5023–5028 (2014)
16. Kalman, R.E., Bucy, R.: New results in linear filtering and prediction theory. J. Basic Eng. 82
D, 35–40 (1960)
17. Korovin, S.K., Fomichev, V.V.: State observers for linear systems with uncertainty. De Gruyter
Expositions in Mathematics 51, p. 242 (2009)
18. Krener, A.J., Isidori, A.: Linearization by output injection and nonlinear observers. Syst.
Control Lett. 3, 47–52 (1983)
19. Krener, A.J., Respondek, W.: Nonlinear observer with linear error dynamics. SIAM J. Control
Optim. 23, 197–216 (1985)
20. Levine, J., Marino, R.: Nonlinear systems immersion observers and finite dimension filters.
Syst. Control Lett. 7, 133–142 (1986)
21. Luenberger, D.G.: Observers for multivariable systems. IEEE Trans. Automatic Control 11,
190–197 (1966)
22. Martínez-Guerra, R.: Observer synthesis for a class of bilinear systems. In: 32nd Annual
Allerton Conference on Communication Control and Computing, pp. 523–524, 1994
23. Martínez-Guerra, R.: Inmersión de un sistema no lineal y construcción de un observador no
lineal para un sistema TPA. In: Congreso Latinoamericano de Control Automático, Habana,
Cuba, pp. 251–255, 1992
24. Pardalos, P.M., Knopov, P.S., Urysev, S.P., Yatsenko, V.S.; Optimal estimation of signal
parameters using bilinear observation. In: Rubinov, A., Glover, B. (eds.), Optimization and
Related topics. Kluwer Academic, Boston (2001)
6 1 Introduction

25. Pommaret, J.F.: Géométrie différentielle algébrique et théorie du contrôle. C.R. Acad. Sc. Paris,
t. 302(Serie I, No.15), 547–550 (1986)
26. Vu, V.P., Wang, W.J.: Observer synthesis for uncertain Takagi–Sugeno fuzzy systems with
multiple output matrices. IET Control Theory Appl. 10(2), 151–161 (2016)
27. Williamson, D.: Observation of bilinear systems with application to biological control.
Automatica 13, 243–254 (1977)
28. Yang, P., Xu, K.L., Jiang, Y.L.: H2 model order reduction for bilinear systems based on the
cross gramian. IMA J. Math. Control Inform. 32, 1–16 (2016)
Chapter 2
Differential Algebra in Control Theory

This chapter presents some elementary definitions and ideas from differential
algebra on which this book is based.
In automatic control, the use of techniques from differential algebra are increas-
ingly important. This is the main reason why it is not only necessary to remember
the useful elements of this language. In addition, such techniques will serve as a
descriptive language to prove various lemmas and corollaries related to this theory.
Some contributions to theory can be found in [3, 6, 11].
Roughly speaking, differential algebra is a theory of structures that generalizes
the structures that have been used in traditional mathematics and its applications
for a long time. Such algebraic structures include rings, fields, field extensions, and
vector spaces (see, for example, [2]).
Fundamental differential-algebraic structures are essentially differential rings,
differential fields, differential field extensions, etc. These have corresponding
underlying structures that are usual algebraic structures. Their construction is given
generally by adding operations to the underlying structures, that is, additional laws
such as derivation, subject to consistency axioms yielding richer structures that are
also more complex than the original structures.
The key concepts in this chapter are differential polynomials and their handling
as well as elementary structures from differential algebra such as differential rings,
differential fields, and differential field extensions.

© Springer International Publishing AG 2017 7

R. Martínez-Guerra, C.D. Cruz-Ancona, Algorithms of Estimation
for Nonlinear Systems, Understanding Complex Systems,
DOI 10.1007/978-3-319-53040-6_2
8 2 Differential Algebra in Control Theory

2.1 Elements of Differential Algebra

2.1.1 Transcendental Fields and Algebraic Fields, Subfields,

Differential Field Extensions

Definition 2.1 Let K be a field. A subfield K 0 of K is a subring of K provided with

the internal composition laws (addition and multiplication) induced from those of
K. This gives a structure to the field, and K is called a field extension of K 0 .
Definition 2.2 Let L be a field extension of the field K and let S be a subset of L.
The subfield generated by K [ S is represented by K .S/, and it is called the subfield
generated by the adjunction of elements of S to K.
If S is reduced to a single element ˛ 2 L (and ˛ … K), then K.˛/ is called a
simple extension.

2.1.1.1 Algebraic Dependence

Definition 2.3 Let L be a field extension of the field K and let .xi /i2I be a family of
elements of L with I a set of indices. The family .xi /i2I is said to be algebraically
dependent over K if there exist a finite subset fi1 ; : : : ; in g of I and a nonzero
polynomial P of AŒxi1 ; : : : ; xin , the ring of polynomials, such that P .xi1 ; : : : ; xin / D
0. Such a family is said to be algebraically independent over K if every nonzero
polynomial in AŒ.xi /i2I  is zero for .xi /i2I (this is the exact analogue of vectorial
linear independence in vector spaces).

2.1.1.2 Transcendental and Algebraic Elements

Definition 2.4 An element ˛ of a field extension L of a field K is said to be

algebraic over K if ˛ is algebraically dependent over K, i.e., if there exists a nonzero
polynomial P 2 KŒx such that P.˛/ D 0. Otherwise, ˛ is said to be transcendental
over K.
We illustrate the above definition with the following examples:
p p
Example 2.1 Consider the field K D Q, and let 2 2 L D R. Then 2 isp algebraic
over Q, since there exists a polynomial, namely P.x/ D x2 2, such that P. 2/ D 0.
Example 2.2 Every element of L D Q is algebraic over K D Z.
Example 2.3 Consider i 2 C. Then i is algebraic over R, since i satisfies the
polynomial P.x/ D x2 C 1, namely, P.i/ D 0.
Example 2.4 Let L D R. Then ; e 2 R are transcendental over K D Q.
2.2 Differential Rings and Differential Fields 9

2.1.1.3 Transcendental Extensions

Definition 2.5 An extension L of a field K it is called algebraic if all its elements are
algebraic over K, and it is called transcendental if there exists at least one element
that is transcendental over K.

2.1.1.4 Transcendence Basis

Given a field extension L of a field K, it is possible to prove the existence of a subset

x of L that is algebraically independent over K such that L is algebraic over K.x/.
Definition 2.6 Let L be a field extension of K and assume there exists a subset x of
L algebraically independent over K such that L is algebraic over K.x/. This subset
is called a transcendence basis of the field extension L=K (this is the exact analogue
of a basis in vector spaces). The cardinality of a maximal such subset is called the
transcendence degree of the extension L=K, and it is denoted by tr dı L=K.
To illustrate the above definition, consider the following examples:
Example 2.5 The extension L=K is algebraic if and only if tr dı L=K D 0.
Example 2.6 Let K D R, and consider the field of rational functions in a single
indeterminate s (one variable), denoted by R.s/:

a0 C a1 s C    C an sn
f .s/ D ; (2.1)
b0 C b1 s C  C bm sm

where ai , bi 2 R. Hence, if s is transcendental over K D R, then tr dı R.s/=R D 1.

Definition 2.7 Let L be an extension of the field K. One says that the extension
L=K is finite if there exists a finite subset x of L such that L D K.x/.

2.2 Differential Rings and Differential Fields

Differential algebra was introduced by the American mathematician Joseph F. Ritt

[10, 11]. This subject contributed tools to the theory of differential equations like
those proposed in commutative algebra for algebraic equations [1].
Definition 2.8 Let A be a commutative ring with unit element. The derivative
(differentiation) @ of A is a closed mapping @ W A ! A such that all pairs .a; b/,
a, b 2 A, satisfying

C W AA!A
.a; b/ 7! a C b
10 2 Differential Algebra in Control Theory

Fig. 2.1 Derivation in a

commutative ring A

and

 W AA!A
.a; b/ 7! a  b

also satisfy

@ .a C b/ D @a C @b

and

@ .a  b/ D .@a/ b C a .@b/ :

This definition is illustrated in Fig. 2.1. Note that a  b; a C b 2 A, since A is a ring.

2.2.1 Differential Rings

Definition 2.9 A differential ring A is a commutative ring with a unit element that is
provided with a finite set of differentiation operators over A such that 8@1 ; @2 2 ,
8a 2 A,

@1 @2 a D @2 @1 a:

Definition 2.10 A differential ring is called ordinary or partial if the finite set
respectively contains a single or more than one differentiation operators. These are
useful in the study of ordinary algebraic differential equations and partial algebraic
differential equations, respectively.
Definition 2.11 A differential subring of a differential ring A is a subring of A such
that with the differential operator in , A is a differential ring.
Definition 2.12 A differential subring is a differential ring that is also a differential
field.
2.2 Differential Rings and Differential Fields 11

Definition 2.13 A constant a of a differential ring is an element such that 8@ 2 ,

@ .a/ D 0.
We have the following result in the form of a corollary.
Corollary 2.1 The set of constants of a differential ring (differential field) is a
differential subring (differential subfield).
Proof Let A be a differential ring and let C the set of constants, with 0, 1 2 C. For
all a 2 A,

a C 0 D a:

Applying the differential operator @ 2 yields

@ .a C 0/ D @a C @ .0/ D @ .a/ :

Hence

@ .0/ D 0:

Next, applying the differential operator @ 2 to

a D a  1;

we obtain

@a D @ .a  1/ D 1  @a C a@1 D @a C a@1:

Hence

@.1/ D 0:

Let c1 and c2 2 C. Then

@ .c1 C c2 / D @ .c1 / C @ .c2 / D 0;

@ .c1 c2 / D @ .c1 / c2 C @ .c2 / c1 D 0:

Let c 2 C, c ¤ 0. Then 1=c is a constant, i.e.,

 
1 @c
8@ 2 @ D D 0:
c c2


12 2 Differential Algebra in Control Theory

2.2.2 Notion of a Differential Field Extension: Algebraic

and Transcendental

2.2.2.1 Differential Field Extension

Definition 2.14 If L and K are two differential fields such that K is a differential
subfield of L (K L), then L is called a differential field extension of K.
Definition 2.15 The intersection of a set of differential subfields of L is also a
differential subfield of L.
Definition 2.16 An element a 2 L is called differentially algebraic over K if
there exists an algebraic differential equation with coefficients in K, i.e., there
exists a polynomial with coefficients in K, P x; xP ; : : : ; x.˛/ , with a finite number
 
of derivatives of x such that P a; aP ; : : : ; a.˛/ D 0.
To illustrate the above definition, consider the following examples.
Example 2.7 K D Q, a D et 2 L, satisfies the differential equation xP x D 0.
Example 2.8 K D R. Let L be the field R.x/, and xP D e , xR D xP e . The element x
x x

satisfies the differential equation with coefficients in R given by xR xP 2 D 0.

Definition 2.17 An extension L=K is called differentially algebraic if all element
of L are differentially algebraic over K (hyperalgebraic).
Definition 2.18 An element a 2 L is said to be differentially transcendental if it is
not differentially algebraic.
Definition 2.19 An extension L=K it is said to be differentially transcendental
if there exists at least one element of L differentially transcendental over K
(hypertranscendental).
Definition 2.20 A set fi ji 2 Ig of elements in L is called differentially
n K-
.i /
algebraically dependent if the set of derivatives of every order i ji 2 I; i D
0; 1; 2; : : :g is K-algebraically dependent (see Definition 2.4).
Definition 2.21 A set that is not differentially K-algebraically dependent is called
differentially K-algebraically independent.

2.2.3 Differential Primitive Element, Differential

Transcendence Basis, Differential Transcendence
Degree, and Tower of Differential Field Extensions

Definition 2.22 Every maximal differentially K-algebraically independent family

with respect to inclusion is called a differential transcendence basis of L=K. If we
Another random document with
no related content on Scribd:
 Prophylaxis,

531-533

Symptoms,

514-521

Cramp and spasm, characters and muscles affected,

515-517

Electrical reactions in,

521

General,

520

Lock-spasm in,

517

Pain and modification of sensation in,

519
 Paresis and paralysis, characters and seat,

518

Tremor in,

518

Vaso-motor and trophic disturbances,

520

Treatment,

533

543

Arsenic, use,

537

538
 Atropia, hypodermically,

537

Electricity, use,

534-536

Gymnastics and massage,

536

537

Local applications in,

538

Medication, value of,

537

Mechanical apparatus in,

540-542
 Rest in,

533

534

Strychnia, use,

537

538

Tenotomy and nerve-stretching in,

538

540

Z.

Zinc salts, use, in chorea,

455

in hysteria,

278

in hystero-epilepsy,

313

Valerianate, use of, in migraine,

1232

GENERAL INDEX.
 A.

Abdominal Glands, Diseases of, ii. 1182

Abortion, iv. 467

Acne, iv. 641

rosacea, iv. 647

Addison's disease, iii. 939

Alalia, iv. 569

Albinismus, iv. 647

Albuminuria, iv. 34

Alcoholism, v. 573

Alimentation, rectal, ii. 928

Alopecia, iv. 678

areata, iv. 680

Amblyopia, toxic, iv. 803

Amenorrhœa, iv. 183

Anæmia, iii. 887

progressive pernicious, iii. 898

Anæsthesia of peripheral origin, v. 1198

Aneurism, abdominal, iii. 821

thoracic, iii. 801

Angina pectoris, iii. 755

Angioma, iv. 688

Anidrosis, iv. 584

Ankyloglossia, ii. 349

Anthrax, i. 926; iv. 606

Anus, cutaneous eruptions and parasitic conditions of, ii. 892

syphilis of, ii. 900

Anus and Rectum, Diseases of, ii. 877

fissure of, ii. 888

prolapse of, ii. 881

Aorta, aneurism of, iii. 801, 821

atheroma of, iii. 800

perforation and occlusion of, iii. 824

rupture of, iii. 823

stenosis of, iii. 825

Aortic insufficiency, iii. 659

obstruction or stenosis, iii. 654

Apoplexy, cerebral, v. 917

 pulmonary, iii. 293

Artery, coronary, diseases of, iii. 828

hepatic, aneurism of, iii. 840

inferior mesenteric, diseases of, iii. 839

pulmonary, diseases of, iii. 833

superior mesenteric, diseases of, iii. 836

Arthritis, gonorrhœal, ii. 102

rheumatoid, ii. 78

Ascaris lumbricoides, ii. 952

Ascites, ii. 1153

Asthma, Bronchial, iii. 184

Hay, iii. 210

Atelectasis, iii. 251

Athetosis, v. 457

Atrophy, v. 1266

facial, progressive unilateral, v. 694

Auditory canal, external, diseases of, v. 811

Auricle, external ear, diseases of, iv. 810

B.

Baldness, iv. 678

Battey's operation, iv. 290

Barber's itch, iv. 723

Basedow's disease, iii. 761

Bed-bugs, iv. 733

Bell's palsy, v. 1202

Beriberi, i. 1038

Bile-ducts, catarrh of, ii. 1051

Biliary concretions, ii. 1058

Biliary passages, affections of, ii. 1051

occlusion of, ii. 1082

Biliousness, ii. 965

Bladder, atony and paralysis of, iv. 133

atrophy and hypertrophy of, in women, iv. 348

diseases of, iv. 123

functional diseases of, in women, iv. 349

hemorrhage of, iv. 134, 340

hyperæmia of, in women, iv. 339

inflammation of, acute and chronic, iv. 123, 126, 128, 341

neuroses of, iv. 132

 new growths of, iv. 136

paralysis of, in women, iv. 350

Blood and Blood-glandular System, Diseases of, iii. 882

Boils, iv. 604

Bowel, compression and contraction of, ii. 857

stricture of, ii. 854

Bowels, hemorrhage from, ii. 830

Brain, abscess of, v. 792

atrophy of, v. 993

hypertrophy of, v. 996

inflammation of, v. 790

softening of, v. 989

syphilis of, v. 1003, 1017

and its Envelopes, Tumors of, v. 1028

Spinal Cord, Concussion of, v. 907

Diseases of Membranes of, v. 703

Hyperæmia and Anæmia of, v. 763, 774

Bright's disease, iv. 72

Bromidrosis, iv. 584

Bronchi, Diseases of, iii. 164

Bronchial Asthma, iii. 184

tubes, dilatation of, iii. 227

Bronchitis, acute and chronic, capillary, mechanical, pseudo-

membranous, and rheumatic, iii. 169, 170, 171, 172, 173, 174

C.
Caisson Disease, the, iii. 855

Calculi, hepatic, ii. 1058

renal, iv. 42

Callositas, iv. 662

Cancrum oris, ii. 338

Canities, iv. 678

Cannabis indica, habitual addiction to, v. 669

Carbuncles, iv. 606

Cardiac murmurs and their relation to valvular heart disease, iii. 651

Cardiac Thrombosis, iii. 718

Cardialgia, ii. 459

Cardicentesis, iii. 798

Caruncle, urethral, iv. 403

Catalepsy, v. 314