Nonlinear Systems Vol 1 Mathematical

Theory and Computational Methods

Victoriano Carmona
Visit to download the full and correct content document:
https://textbookfull.com/product/nonlinear-systems-vol-1-mathematical-theory-and-co
mputational-methods-victoriano-carmona/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Computational Statistics and Mathematical Modeling

Methods in Intelligent Systems Proceedings of 3rd
Computational Methods in Systems and Software 2019 Vol
2 Radek Silhavy
https://textbookfull.com/product/computational-statistics-and-
mathematical-modeling-methods-in-intelligent-systems-proceedings-
of-3rd-computational-methods-in-systems-and-
software-2019-vol-2-radek-silhavy/

Software Engineering Perspectives in Intelligent

Systems: Proceedings of 4th Computational Methods in
Systems and Software 2020, Vol.1 Radek Silhavy

https://textbookfull.com/product/software-engineering-
perspectives-in-intelligent-systems-proceedings-of-4th-
computational-methods-in-systems-and-software-2020-vol-1-radek-
silhavy/

Nonlinear Analysis: Problems, Applications and

Computational Methods Zakia Hammouch

https://textbookfull.com/product/nonlinear-analysis-problems-
applications-and-computational-methods-zakia-hammouch/

Software Engineering Perspectives in Intelligent

Systems: Proceedings of 4th Computational Methods in
Systems and Software 2020, Vol.2 Radek Silhavy

https://textbookfull.com/product/software-engineering-
perspectives-in-intelligent-systems-proceedings-of-4th-
computational-methods-in-systems-and-software-2020-vol-2-radek-
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/

Nonlinear Systems and Their Remarkable Mathematical

Structures Volume 2 1st Edition Norbert Euler (Editor)

https://textbookfull.com/product/nonlinear-systems-and-their-
remarkable-mathematical-structures-volume-2-1st-edition-norbert-
euler-editor/

Computational and Statistical Methods in Intelligent

Systems Radek Silhavy

https://textbookfull.com/product/computational-and-statistical-
methods-in-intelligent-systems-radek-silhavy/

Cooperative Control of Nonlinear Networked Systems

Infinite time and Finite time Design Methods Yongduan
Song

https://textbookfull.com/product/cooperative-control-of-
nonlinear-networked-systems-infinite-time-and-finite-time-design-
methods-yongduan-song/

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/
Understanding Complex Systems

Victoriano Carmona
Jesús Cuevas-Maraver
Fernando Fernández-Sánchez
Elisabeth García-Medina Editors

Nonlinear
Systems,
Vol. 1
Mathematical Theory and
Computational Methods
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 “Springer Briefs 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, Florida Institute of Technology, Computer Science Department, 150 W. University Blvd,
Melbourne, FL 32901, USA
Andrzej Nowak, Department of Psychology, Warsaw University, Poland
Hassan Qudrat-Ullah, School of Administrative Studies, York University, Toronto, ON, Canada
Linda Reichl, Center for Complex Quantum Systems, University of Texas, Austin, USA
Peter Schuster, Theoretical Chemistry and Structural Biology, University of Vienna, Vienna, Austria
Frank Schweitzer, System Design, ETH Zürich, Zürich, Switzerland
Didier Sornette, Entrepreneurial Risk, ETH Zürich, Zürich, 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

Victoriano Carmona Jesús Cuevas-Maraver
•

Fernando Fernández-Sánchez
Elisabeth García-Medina
Editors

Nonlinear Systems, Vol. 1

Mathematical Theory and Computational
Methods

123
Editors
Victoriano Carmona Fernando Fernández-Sánchez
Departamento de Matemática Aplicada II Departamento de Matemática Aplicada II
Universidad de Sevilla Universidad de Sevilla
Sevilla, Spain Sevilla, Spain

Jesús Cuevas-Maraver Elisabeth García-Medina

Departamento de Física Aplicada I Departamento de Matemática Aplicada II
Universidad de Sevilla Universidad de Sevilla
Sevilla, Spain Sevilla, Spain

ISSN 1860-0832 ISSN 1860-0840 (electronic)

Understanding Complex Systems
ISBN 978-3-319-66765-2 ISBN 978-3-319-66766-9 (eBook)
https://doi.org/10.1007/978-3-319-66766-9
Library of Congress Control Number: 2018946603

© Springer International Publishing AG, part of Springer Nature 2018

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 the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
To Prof. Antonio Castellanos Mata, founder
and head of the Group of
Electrohydrodynamics and Cohesive
Granular Media of the University of Seville.
Antonio directed research projects for more
than 30 years, and this made it possible for
him to organize two laboratories at the
University. Dedicating a lot of efforts to pure
science, Antonio was also interested in
practical problems and collaborated with
industry (Xerox Corporation, Novartis, Dow
Corning, IFPRI).
Antonio belonged to a generation that
played an important role in the revival of
physics in Spain. In 2013, he was awarded
the Prize FAMA for the research career by
the University of Seville. In his last years,
Antonio worked on thermodynamics in
relativity and on triboelectricity in fine
powders As a researcher, Antonio combined a
strong theoretical mind, experimental
intuition, profound understanding of physics
of phenomena, and passionate love for
science.
Preface

This book commemorates the conference Nolineal2016: International Conference

on Nonlinear Mathematics and Physics, that took place in Sevilla, Spain, from 7 to
10 July 2016. There were delegates from many different countries in Europe and
also three other continents.
At the end of the conference, the decision was taken of writing a book to provide
the readers with a landscape of the many different fields in which nonlinear science
is being developed with great success. Contributions would not be proceedings but
present an introduction to the different subjects, provide context, present the state of
art and certainly the own research of the authors in the field.
Although the mathematics and physics of nonlinear systems are closely inter-
twined, it has been considered convenient to divide the matter in two volumes:
• Nonlinear Systems, Vol. 1. Mathematical Theory and Computational Methods in
Nonlinear Systems, edited by Victoriano Carmona, Jesús Cuevas-Maraver,
Fernando Fernández-Sánchez and Elisabeth García-Medina
• Nonlinear Systems, Vol 2. Nonlinear Phenomena in Biology, Optics and
Condensed Matter, edited by Juan F. R. Archilla, Faustino Palmero, M. Carmen
Lemos, Bernardo Sánchez-Rey and Jesús Casado-Pascual
The present book is the first volume, it deals with the theory of nonlinear
systems, especially from mathematical and computational approaches, and it is
divided into four large areas of knowledge, namely bifurcation analysis, wave
equations, differential or difference equations and computational methods. From
Lorenz system and its bifurcations to applications of cellular automata to laser
dynamics, from fast–slow systems to dark matter, from integrability and normal
forms to gravitational waves and solitons, this book oscillates between theoretical
analysis of nonlinear phenomena and the most current topics in science today.

vii
viii Preface

The outline of the book is as follows:

1. Bifurcation Analysis
• “A Review on Some Bifurcations in the Lorenz System” by A. Algaba,
M. C. Domínguez-Moreno, M. Merino and A. J. Rodríguez-Luis
• “Normal Form for a Class of Three-Dimensional Systems with Free-
Divergence Principal Part” by A. Algaba, N. Fuentes, E. Gamero and
C. García
• “Piecewise-Linear (PWL) Canard Dynamics: Simplifying Singular
Perturbation Theory in the Canard Regime Using Piecewise-linear
Systems” by M. Desroches, S. Fernández-García, M. Krupa, R. Prohens
and A. E. Teruel
2. Wave Equations
• “Solitary Waves in the Nonlinear Dirac Equation” by J. Cuevas-Maraver,
N. Boussaїd, A. Comech, R. Lan, P. G. Kevrekidis and A. Saxena
• “On Nonlinear Schrödinger Equation as a Model for Dark Matter:
Comments on Galactic Collisions, Supermassive Black Holes and
Analogue Laboratory Implementations” by A. Paredes and H. Michinel
• “Adiabatic Invariants of Second Order Korteweg-de Vries Type Equation”
by P. Rozmej and A. Karczewska
• “Nonlinear Gravitational Waves and Solitons” by F. R. Villatoro
3. Differential and Difference Equations
• “Local Integrability for Some Degenerate Nilpotent Vector Fields” by
A. Algaba, I. Checa and C. García
• “A Logistic Non-linear Difference Equation with Two Delays” by
F. Balibrea
• “Diffusive Limits of the Master Equation in Inhomogeneous Media” by
L. Salasnich, A. Bonato and F. Sattin
4. Computational Methods
• “Anticipating Abrupt Changes in Complex Networks: Significant Falls in the
Price of a Stock Index” by A. Cordoba, C. Castillejo, J. J. García-Machado
and A. M. Lara
• “On the Numerical Approximation to Generalized Ostrovsky Equations: I”
by Ángel Durán
• “On the Numerical Approximation to Generalized Ostrovsky Equations: II”
by Ángel Durán
• “Simulating Laser Dynamics with Cellular Automata” by
F. Jiménez-Morales, J. L. Guisado and J. M. Guerra
Preface ix

Chapters will provide an opportunity for the readers to understand subjects

which are normally dispersed in different journals for specialists. We expect them to
feel the fascination of nonlinear physics and its broad applicability, stimulating their
curiosity and perhaps extending their own research to unexpected territory.
The editors of this book do not want to miss the opportunity to acknowledge the
University of Seville and the Instituto de Matemáticas de la Universidad de Sevilla
Antonio Castro Brzezicki for the financial and administrative support.

Sevilla, Spain Victoriano Carmona

July 2017 Jesús Cuevas-Maraver
Fernando Fernández-Sánchez
Elisabeth García-Medina
In memoriam of Prof. Antonio Castellanos Mata (07.03.1947–27.01.2016), Full Professor of
Electromagnetism at the University of Seville, and Director of the group of Electrohydrodynamics
and Cohesive Granular Media
Contents

Part I Bifurcation Analysis

A Review on Some Bifurcations in the Lorenz System . . . . . . . . . . . . . . 3
Antonio Algaba, M. Cinta Domínguez-Moreno, Manuel Merino
and Alejandro J. Rodríguez-Luis
Normal Form for a Class of Three-Dimensional Systems
with Free-Divergence Principal Part . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Antonio Algaba, Natalia Fuentes, Estanislao Gamero and Cristóbal García
Piecewise-Linear (PWL) Canard Dynamics . . . . . . . . . . . . . . . . . . . . . . 67
Mathieu Desroches, Soledad Fernández-García, Martin Krupa,
Rafel Prohens and Antonio E. Teruel

Part II Wave Equations

Solitary Waves in the Nonlinear Dirac Equation . . . . . . . . . . . . . . . . . . 89
Jesús Cuevas-Maraver, Nabile Boussaïd, Andrew Comech, Ruomeng Lan,
Panayotis G. Kevrekidis and Avadh Saxena
On Nonlinear Schrödinger Equation as a Model for
Dark Matter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Angel Paredes and Humberto Michinel
Adiabatic Invariants of Second Order Korteweg-de Vries
Type Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Piotr Rozmej and Anna Karczewska
Nonlinear Gravitational Waves and Solitons . . . . . . . . . . . . . . . . . . . . . 207
Francisco R. Villatoro

xi
xii Contents

Part III Differential and Difference Equations

Local Integrability for Some Degenerate Nilpotent Vector Fields . . . . . . 243
Antonio Algaba, Isabel Checa and Cristóbal García
A Logistic Non-linear Difference Equation with Two Delays . . . . . . . . . 269
Francisco Balibrea
Diffusive Limits of the Master Equation in Inhomogeneous Media . . . . 295
Luca Salasnich, Andrea Bonato and Fabio Sattin

Part IV Computational Methods

Anticipating Abrupt Changes in Complex Networks: Significant Falls
in the Price of a Stock Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
Antonio Cordoba, Christian Castillejo, Juan J. García-Machado
and Ana M. Lara
On the Numerical Approximation to Generalized Ostrovsky
Equations: I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
Ángel Durán
On the Numerical Approximation to Generalized Ostrovsky
Equations: II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
Ángel Durán
Simulating Laser Dynamics with Cellular Automata . . . . . . . . . . . . . . . 405
Francisco Jiménez-Morales, José Luis Guisado and José Manuel Guerra
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423
Contributors

Antonio Algaba Departamento de Ciencias Integradas, Centro de Investigación de

Física Teórica y Matemática FIMAT, Universidad de Huelva, Huelva, Spain
Francisco Balibrea Group of Dynamical Systems, Facultad de Matemáticas,
Universidad de Murcia, Murcia, Spain
Andrea Bonato Dipartimento di Fisica e Astronomia Galileo Galilei, Universitá di
Padova, Padova, Italy
Nabile Boussaïd Université Bourgogne Franche-Comté, Besançon CEDEX,
France
Christian Castillejo Instituto IBT, Parque Empresarial Nuevo Torneo, Sevilla,
Spain
Isabel Checa Departamento de Matemáticas, Centro de Investigación de Física
Teórica y Matemática FIMAT, Universidad de Huelva, Huelva, Spain
Andrew Comech Department of Mathematics, Texas A&M University, College
Station, TX, USA; IITP, Moscow, Russia
Antonio Cordoba Departamento de Física de la Materia Condensada, Universidad
de Sevilla, Sevilla, Spain
Jesús Cuevas-Maraver Departamento de Física Aplicada I, Escuela Politécnica
Superior, Grupo de Física No Lineal, Universidad de Sevilla, Sevilla, Spain;
Instituto de Matemáticas de la Universidad de Sevilla (IMUS), Sevilla, Spain
Mathieu Desroches MathNeuro Team, Inria Sophia Antipolis Research Centre,
Sophia Antipolis Cedex, France
M. Cinta Domínguez-Moreno Departamento de Ciencias Integradas, Centro de
Investigación de Física Teórica y Matemática FIMAT, Universidad de Huelva,
Huelva, Spain

xiii
xiv Contributors

Ángel Durán Applied Mathematics Department, University of Valladolid,

Valladolid, Spain
Soledad Fernández-García Departamento EDAN, Facultad de Matemáticas,
University of Sevilla, Sevilla, Spain
Natalia Fuentes Department of Integrated Sciences, Investigation Center of
Theoretical Physics and Mathematic FIMAT, Huelva University, Huelva, Spain
Estanislao Gamero Department of Applied Mathematic II, E.T.S.I. Sevilla
University, Sevilla, Spain
Juan J. García-Machado Departamento de Economía Financiera, Contabilidad y
Dirección de Operaciones, Universidad de Huelva, Huelva, Spain
José Manuel Guerra Departamento de Óptica, Facultad de CC. Físicas,
Universidad Complutense de Madrid, Madrid, Spain
Cristóbal García Departamento de Matemáticas, Centro de Investigación de
Física Teórica y Matemática FIMAT, Universidad de Huelva, Huelva, Spain
José Luis Guisado Departamento de Arquitectura y Tecnología de Computadores,
Universidad de Sevilla, Sevilla, Spain
Francisco Jiménez-Morales Departamento de Física de la Materia Condensada,
Universidad de Sevilla, Sevilla, Spain
Anna Karczewska Faculty of Mathematics, Computer Science and Econometrics,
University of Zielona Góra, Zielona Góra, Poland
Panayotis G. Kevrekidis Department of Mathematics and Statistics, University of
Massachusetts, Amherst, MA, USA
Martin Krupa MathNeuro Team, Inria Sophia Antipolis Research Centre, Sophia
Antipolis Cedex, France; Université Côte d’Azur (UCA), Nice, France; Laboratoire
J. A. Dieudonné, Université de Nice Sophia Antipolis, Nice Cedex 02, France
Ruomeng Lan Department of Mathematics, Texas A&M University, College
Station, TX, USA
Ana M. Lara Instituto IBT, Parque Empresarial Nuevo Torneo, Sevilla, Spain
Manuel Merino Departamento de Ciencias Integradas, Centro de Investigación de
Física Teórica y Matemática FIMAT, Universidad de Huelva, Huelva, Spain
Humberto Michinel Área de óptica, Escola de Enxeñaría Aeroespacial, Ourense,
Spain
Angel Paredes Área de óptica, Departamento de Física Aplicada, Ourense, Spain
Rafel Prohens Departament de Matemàtiques i Informàtica, Universitat de les
Illes Balears, Palma de Mallorca, Spain
Contributors xv

Alejandro J. Rodríguez-Luis Departamento de Matemática Aplicada II, E.S.

Ingenieros, Universidad de Sevilla, Sevilla, Spain
Piotr Rozmej Faculty of Physics and Astronomy, Institute of Physics, University
of Zielona Góra, Zielona Góra, Poland
Luca Salasnich Dipartimento di Fisica e Astronomia Galileo Galilei, Universitá di
Padova, Padova, Italy
Fabio Sattin Consorzio RFX (CNR, ENEA, INFN, Universitá di Padova,
Acciaierie Venete SPA), Padova, Italy
Avadh Saxena Los Alamos National Laboratory, Center for Nonlinear Studies
and Theoretical Division, Los Alamos, NM, USA
Antonio E. Teruel Departament de Matemàtiques i Informàtica, Universitat de les
Illes Balears, Palma de Mallorca, Spain
Francisco R. Villatoro Department Lenguajes y Ciencias de la Computación,
Escuela de Ingenierías Industriales, Ampliación del Campus de Teatinos,
Universidad de Málaga, Málaga, Spain
 Part I
Bifurcation Analysis
A Review on Some Bifurcations
in the Lorenz System

Antonio Algaba, M. Cinta Domínguez-Moreno, Manuel Merino

and Alejandro J. Rodríguez-Luis

Abstract In this chapter, we review some bifurcations exhibited by the classical

Lorenz system, where the parameters can have any real value. Analytical results on
the pitchfork, Hopf and Takens–Bogdanov bifurcations of the origin, as well as the
Hopf bifurcation of the nontrivial equilibria, are summarized. These results serve as
a guide for the numerical study that reveals other important organizing centers of
the dynamics: Takens–Bogdanov bifurcations of periodic orbits, torus bifurcations
and the resonances associated, homoclinic and heteroclinic connections with several
degeneracies, etc. We also point out that the analysis of the Hopf-pitchfork and the
triple-zero bifurcations of the origin cannot be performed with the usual tools and
propose a way to carry out this study avoiding the structural singularities exhibited
by the Lorenz system.

Keywords Lorenz · Bifurcation · Pitchfork · Hopf · Takens–Bogdanov · Torus

Resonances · Hopf-pitchfork · Triple-zero

1 Introduction

The famous Lorenz system was derived from a simplified model of convection in
the atmosphere: a two-dimensional fluid cell is warmed from below and cooled
from above and the resulting convective motion is modeled by a partial differential

A. Algaba · M. C. Domínguez-Moreno (B) · M. Merino

Departamento de Ciencias Integradas, Centro de Investigación de Física Teórica y Matemática
FIMAT, Universidad de Huelva, 21071 Huelva, Spain
e-mail: mcinta.dominguez@dmat.uhu.es
A. Algaba
e-mail: algaba@uhu.es
M. Merino
e-mail: merino@uhu.es
A. J. Rodríguez-Luis
Departamento de Matemática Aplicada II, E.S. Ingenieros, Universidad de Sevilla,
Camino de los Descubrimientos s/n, 41092 Sevilla, Spain
e-mail: ajrluis@us.es

© Springer International Publishing AG, part of Springer Nature 2018 3

V. Carmona et al. (eds.), Nonlinear Systems, Vol. 1,
Understanding Complex Systems, https://doi.org/10.1007/978-3-319-66766-9_1
4 A. Algaba et al.

equation. This system is obtained after a Galerkin approximation, that is, the variables
are expanded into an infinite number of modes and all except three of them are put
to zero [86, 97]: ⎧
⎨ ẋ = σ (y − x),
ẏ = ρx − y − x z, (1)
⎩
ż = −bz + x y.

Thus, the variable x is proportional to the intensity of convective motion, y is propor-

tional to the temperature difference between ascending and descending currents and
z is proportional to the distortion from linearity of the vertical temperature profile.
The dimensionless parameters have a physical interpretation: σ is a Prandtl number
(the ratio of kinematic viscosity and thermal diffusivity), ρ is a Rayleigh number
(proportional to the temperature difference across the fluid layer and the gravita-
tional acceleration acting on the fluid) and b is a ratio of the height and width of the
fluid layer. Consequently, the three parameters considered by Lorenz were positive.
The Lorenz equations also arise in simplified models in a variety of fields, for
instance, lasers [73], dynamos [77], thermosyphons [71], chemical reactions [93],
electric circuits [49] and brushless DC motors [75]. They have even been proposed
very recently in the thermodynamic modelling of leukaemia malignancy [1].
The complex dynamics exhibited by the Lorenz system has fascinated to a large
number of scientists in such a way that, in the last 50 years, hundreds of studies
have examined this emblematic dynamical system. To illustrate this fact without
pretending to be too exhaustive, we cite several of them, indicating briefly the topic
considered: Lorenz chaotic attractor [39–41, 100, 103], T-point heteroclinic cycle
and some degeneracies [16, 48, 69], global invariant manifolds and chaos visualiza-
tion [54–56, 88], dimension of the global attractor [83], its relation with Kolmogorov
systems [90], numerical modeling of its dynamics [91] and its preeminence over other
proposed Lorenz-like systems [11, 13]. However, among so many papers, there are
some that present incorrect results. For example, in Ref. [104, Appendices A and B]
the authors claim to have proved the presence of a Shilnikov heteroclinic orbit in the
Lorenz system, via the undetermined coefficient method. Unfortunately, this method
(used for the first time in Refs. [105, 106]) is wrong as it is demonstrated in Refs.
[7, 9, 10, 12, 15] (see also references therein).
On the other hand, even if this fact is less known, the Lorenz system also appears,
when σ < 0, in the study of traveling-wave solutions in the Maxwell-Bloch equations
[57] and in the analysis of a thermosolutal convection model [78]. This is one of
the reasons why the Lorenz system is also considered for negative values of the
parameters. Moreover, from a dynamical point of view, it is also stimulating to
analyze the behavior of this iconic system when the parameters can take any real
value. In this context, several aspects on the dynamics of the Lorenz system have
been investigated, for instance, in the following works: invariant algebraic surfaces
[43, 84, 85, 98], degenerate heteroclinic cycles [79, 87], Hopf bifurcation [2, 14,
101], Takens–Bogdanov and global bifurcations [3], resonances of periodic orbits
[24] and superluminal periodic orbits [36].
A Review on Some Bifurcations in the Lorenz System 5

The aim of this survey is to comment some results on the Lorenz system obtained
by means of the Local Bifurcation Theory (see, for instance, [72, 82, 102]), in the way
we briefly summarize below. When an autonomous system is analyzed an usual target
is the knowledge of its dynamics in certain zones of the parameter space. In practice,
the parameter space is divided in regions, bounded by bifurcation loci, and the goal
is to determine the qualitative behaviour in each of such regions. This can be done in
several steps. First, the detection of the equilibrium points and the analysis of the lin-
earization around such equilibria, allow to show the presence of linear degeneracies
(nonhyperbolicities) for some values of the control parameters. Second, the computa-
tion of approximations of the center manifold (and also of the reduced system on the
center manifold) enables to reduce the dimension of the problem, transforming the
reduced system into the corresponding normal form by means of changes of variables
(sometimes a reparametrization of the time is also needed). Symbolic computation
algorithms greatly facilitate this task. Third, the analysis of the unfolding of the nor-
mal form in the nondegenerate cases, provides local information on the bifurcation
sets. Furthermore, possible nonlinear degeneracies giving rise to a higher codimen-
sion bifurcation problem can be detected at this step. Finally, from the information
achieved in the study of local bifurcations, good starting points for the application
of adequate numerical techniques can be obtained. This will provide a global picture
of the dynamics of the system in the parameter space (see, for instance, [64]).
To illustrate how the method described above allows to obtain a deep knowledge of
the dynamical system under consideration, we mention now two three-dimensional
systems with a very rich dynamics. On the one hand, for a modified van der Pol-
Duffing electronic oscillator, interesting information can be found in the following
references on some local and global bifurcations: Hopf and Takens–Bogdanov [17],
Hopf-pitchfork [18, 19, 21], triple-zero [67], periodic orbits bifurcations [4, 32,
59], homoclinic connections and some degeneracies [45, 65], T-points and some
degeneracies [5, 8, 58, 60, 62]. Secondly, in the case of the widely studied Chua’s
equation, the following references clarify how to deal with the corresponding local
and global bifurcations: Hopf [27], Takens–Bogdanov [20], Hopf-pitchfork [28],
triple-zero [29], homoclinic connections and some degeneracies [33–35], T-points
and some degeneracies [6, 25, 26, 61].
This work is organized as follows. In Sect. 2 we enumerate the linear degeneracies
that the equilibrium at the origin of the Lorenz system can exhibit. The analysis of the
pitchfork bifurcation is considered in Sect. 3. In Sect. 4 we present results on Hopf
bifurcations. Section 5 is devoted to Takens–Bogdanov bifurcations of equilibria as
well as of periodic orbits. Section 6 is dedicated to the study of resonances, whose
presence is motivated by the existence of torus bifurcations of periodic orbits. Finally,
some conclusions are reported in Sect. 7.
6 A. Algaba et al.

2 Linear Degeneracies

Along this work we consider Lorenz system (1) where σ , ρ and b are real parameters.
We exclude two degenerate situations: the system is linear if σ = 0 and non-isolated
equilibria on the z-axis exist for b = 0.
The Lorenz system (1) is invariant to the change (x, y, z) → (−x, −y, z). The
origin E 0 = (0, 0, 0) is always one equilibrium
√ √ for b(ρ − 1) > 0, two
point and,
symmetric nontrivial equilibria, E ± = (± b(ρ − 1), ± b(ρ − 1), ρ − 1), exist.
The linearization matrix of system (1) at the origin is
⎛ ⎞
−σ σ 0
⎝ ρ −1 0 ⎠ , (2)
0 0 −b

whose characteristic polynomial is

p = λ3 + p 1 λ 2 + p 2 λ + p 3 ,

where
p1 = b + 1 + σ, p2 = σ (1 + b − ρ) + b, p3 = −bσ (ρ − 1).

It is easy to check that the following nonhyperbolic situations may arise:

• A simple zero eigenvalue, and the other two with nonvanishing real part. This case
comes up if p1 = 0, p2 = 0, p3 = 0, that, in terms of the parameters, occurs into
the set ρ = 1, σ = 0, −1, b = 0.
The corresponding codimension-one pitchfork bifurcation gives rise to the appear-
ance of the two nontrivial equilibria E ± in the region b(ρ − 1) > 0.
• A pair of imaginary eigenvalues and the third one nonzero. This degeneration
occurs when p1 p2 = p3 , p2 > 0, p3 = 0, i.e., for σ = −1, ρ > 1, b = 0.
The analysis of the corresponding Hopf bifurcation of the origin was carried out
in Ref. [2], where the Hopf bifurcation exhibited by the nontrivial equilibria E ±
is also studied.
• A double-zero eigenvalue and the third one nonzero. This case appears when
p2 = p3 = 0, p1 = 0, or, in terms of the σ , ρ, b parameters, in the set σ = −1,
ρ = 1, b = 0.
The associated Takens–Bogdanov bifurcation, both in the homoclinic and in the
heteroclinic case, has been analyzed in Ref. [3].
• A pair of imaginary eigenvalues and the third one zero. This situation corresponds
to p1 = p3 = 0, p2 > 0, that is, it occurs when σ = −1, b = 0, ρ > 1. A Hopf-
pitchfork bifurcation is exhibited but, as non-isolated equilibria appear when b =
0, this bifurcation cannot be analyzed by the standard procedures.
• A triple-zero eigenvalue. This situation arises if p1 = p2 = p3 = 0, which, in the
parameter space, corresponds to the point σ = −1, ρ = 1, b = 0.
This bifurcation cannot be studied by the standard methods because, for b = 0,
the origin is a non-isolated equilibrium.
A Review on Some Bifurcations in the Lorenz System 7

3 Pitchfork Bifurcation

For ρ = 1, the linearization matrix (2) has the eigenvalues 0, −(σ + 1), −b. There-
fore, as a consequence of its symmetry, the Lorenz system (1) exhibits a pitchfork
bifurcation. To study this bifurcation, we examine the Lorenz system at the critical
values of the parameters and use the linear change of variables given by
⎛ ⎞ ⎛ ⎞⎛ ⎞
x 1 σ 0 u
⎝ y ⎠ = ⎝ 1 −1 0 ⎠ ⎝ v ⎠ (3)
z 0 0 1 w

in order to obtain the linearization matrix in canonical form.

Thus, the Lorenz system is transformed into
⎛ ⎞ ⎛ ⎞⎛ ⎞ ⎛ σ ⎞
u̇ 0 0 0 u − σ +1 w(u + σ v)
⎝ v̇ ⎠ = ⎝ 0 −(σ + 1) 0 ⎠ ⎝ v ⎠ + ⎝ 1 w(u + σ v) ⎠ . (4)
σ +1
ẇ 0 0 −b w (u − v)(u + σ v)

Assuming that σ + 1 = 0, b = 0, the second-order approximation to the center

manifold is given by

1 2
v = 0 + O(u3 ), w = u + O(u3 ),
b
and the third-order reduced system on the center manifold is
σ
u̇ = − u3 .
b(σ + 1)

Consequently, the bifurcation is supercritical when the coefficient of u3 is negative

and subcritical if it is positive.
The above results can be summarized in the following theorem.

Theorem 1 The locus in the (σ, ρ, b)-parameter space where the origin of the
Lorenz system undergoes a nondegenerate pitchfork bifurcation is defined by

ρ = 1, σ = 0, −1, b = 0.

This bifurcation is supercritical if:

(i) σ ∈ (−1, 0), b < 0;
(ii) σ ∈ (−∞, −1) ∪ (0, ∞), b > 0.
It is subcritical when:
(i) σ ∈ (−∞, −1) ∪ (0, ∞), b < 0;
(ii) σ ∈ (−1, 0), b > 0.
8 A. Algaba et al.

4 Hopf Bifurcations

In this section we precis the principal results obtained in Ref. [2], devoted to the
analysis of Hopf bifurcations and their degeneracies in the Lorenz system (to do
that, the computation of some Lyapunov coefficients of the corresponding normal
form is needed [70, 72, 82, 102]). First, we consider the Hopf bifurcation of the
origin and later, the Hopf bifurcation exhibited by the nontrivial equilibria E ± .
As was mentioned in Sect. 2, the origin E 0 undergoes a Hopf bifurcation if σ =
−1, ρ > 1, b = 0. The corresponding normal form to third order, obtained with
the recursive algorithm developed in Ref. [66] is (see the details in Ref. [2, Sect. 2])

ṙ = a1 r 3 + · · · ,
(5)
θ̇ = 1 + b1r 2 + · · · ,

where the first Lyapunov coefficient is given by

−b − 2
a1 = √ .
8 ρ − 1 4(ρ − 1) + b2

A degeneracy occurs if a1 = 0, i.e., if b = −2.

In this case, as the center manifold is an algebraic invariant surface, the Hopf
bifurcation has infinite codimension: the Lorenz equations have a center at the origin
(see Fig. 1). In fact, it can be easily proved that x 2 + 2z = 0 is the only polynomial
center manifold for the Hopf bifurcation of the origin in the Lorenz system. To prove
this fact it is enough to consider the six invariant algebraic surfaces the Lorenz system
has [85, 98] and apply the conditions for the Hopf bifurcation (b = −2, σ = −1,
ρ > 1). Thus, it is easily obtained that the only solution is x 2 + 2z = 0.

−1 −1
z z
−2

−3
−2 4

2
5 0
2
0 y −2 1
0
y −5 1 2 −4 −1
−2 −1 0
−2 x
x

Fig. 1 Two different perspectives of the phase space of the Lorenz system (1) for b = −2, σ = −1,
ρ = 2 where the origin undergoes a degenerate Hopf bifurcation of infinite codimension. Some
periodic orbits on the center manifold are drawn. Reproduced with permission from [2]. Copyright
(2015) by Springer
A Review on Some Bifurcations in the Lorenz System 9

The following statement sum up all the results on the Hopf bifurcation of the
origin.
Theorem 2 ([2, Theorem 1]) The locus in the (σ, ρ, b)-parameter space where the
origin of the Lorenz system undergoes a Hopf bifurcation is defined by

σ = −1, ρ > 1, b = 0.

This bifurcation is supercritical when b > −2 and subcritical if b < −2. A degen-
erate Hopf bifurcation of infinite codimension occurs if b = −2.
The rest of this section is devoted to the Hopf bifurcation of the nontrivial equilibria
(all the details can be found in Ref. [2, Sect. 3]). The standard techniques used in
the study of a Hopf bifurcation allow to determine, in a first step, the locus where it
occurs and to compute, in a second step, the Lyapunov coefficients that lead to the
detection of all the degeneracies this bifurcation can have. The results obtained are
summarized below.
Proposition 1 ([2, Proposition 2]) The nontrivial equilibria of the Lorenz system
experiment a Hopf bifurcation in the surface parameterized in explicit form by

−σ 2 − (3 − ρ)σ − ρ
Shnt = σ, ρ, b = : (σ, ρ) ∈ Ω ,
σ +ρ

with Ω = Ω1 ∪ Ω2 ∪ Ω3 ∪ Ω4 ∪ Ω5 , where

σ 2 + 3σ
Ω1 = (σ, ρ) ∈ R : σ < −1, ρ < ,
σ −1
Ω2 = {(σ, ρ) ∈ R : σ = −1, ρ < 1} ,
Ω3 = {(σ, ρ) ∈ R : −1 < σ < 0, ρ < −σ } ,

σ 2 + 3σ
Ω4 = (σ, ρ) ∈ R : 0 < σ < 1, ρ < ,
σ −1

σ 2 + 3σ
Ω5 = (σ, ρ) ∈ R : σ > 1, ρ > .
σ −1

The region Ω is drawn in Fig. 2.

The study of the first Lyapunov coefficient of the third-order normal form for the
reduced system [see Eq. (5)], given by
√
(σ + 1) −Δ (σ + ρ) N 1(σ, ρ)
a1 = , (6)
D1(σ, ρ)
10 A. Algaba et al.

Fig. 2 Region Ω which 15

corresponds to the projection ρ=σ(σ+3)/(σ−1)
Ω
onto the (σ, ρ)-plane of the σ+ρ=0 5
surface Shnt where the Hopf 10 Ω−Ω5
bifurcation of the nontrivial Ω
5
equilibria occurs. To get that
surface 2recall that 5 σ=0
b = −σ −(3−ρ)σ −ρ
. ρ=1
σ +ρ
Reproduced with permission ρ σ=1
from [2]. Copyright (2015) 0
by Springer

−5
Ω − Ω5

−10
−10 −5 0 5
σ

where

Δ = σ 2 + (3 − ρ)σ + ρ,
N 1(σ, ρ) = 6σ 4 + 15σ 3 + 9σ 3 ρ + 35σ 2 ρ + σ 2 ρ 2
+21σρ 2 + 2σρ + σρ 3 + 5ρ 3 + ρ 4 ,
 
D1(σ, ρ) = 8(ρ − 1) σ 2 (ρ − 1)2 − (σ + ρ)2 Δ
 
× 4σ 2 (ρ − 1)2 − (σ + ρ)2 Δ ,

and of the second Lyapunov coefficient a2 of the fifth-order normal form for the
reduced system (whose expression appears in Ref. [2, Appendix A]) allow to find all
the degeneracies this Hopf bifurcation may experiment. This information is summa-
rized below.
Theorem 3 The nontrivial equilibria of the Lorenz system undergo a degenerate
Hopf bifurcation in the following cases:
1. The first Lyapunov coefficient a1 vanishes for all the values (σ, ρ) ∈ Ω where
the polynomial N 1(σ, ρ) is zero. A codimension-two bifurcation occurs in this
case when the second Lyapunov coefficient a2 is nonzero.
2. On the two points (σ, ρ, b) ∈ Shnt given by

P1 ≈ (−0.646547, −6.605871, −1.709567)

and
P2 ≈ (−0.0100012, −0.0396965, −1.408456)
A Review on Some Bifurcations in the Lorenz System 11

15 0.1
ρ=σ(σ+3)/(σ−1) Ω ρ=σ(σ+3)/(σ−1)
5
σ+ρ=0 (A) σ+ρ=0
Ω−Ω5 Ω−Ω5
10
Ω 0.05
5 σ=0
σ=0 a =01
5 a1=0
ρ ρ=1 ρ 0
σ=1 (F)
0
(D)
−0.05
−5 (C) (D)
(B)
Ω − Ω5 (E)
−10 −0.1
−10 −5 0 5 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
σ σ

Fig. 3 (Left) Projection onto the (σ, ρ)-plane of the locus where the first Lyapunov coefficient a1
is zero. When this curve is inside the region Ω, it corresponds to a degenerate Hopf bifurcation.
(Right) Zoom in a neighborhood of the origin. Reproduced with permission from [2]. Copyright
(2015) by Springer

a codimension-three Hopf bifurcation occurs because the first and the second
Lyapunov coefficient vanish simultaneously and the third one a3 is nonzero.
These are the unique codimension-three Hopf bifurcation points.
3. On the half-line given by σ = −1, b = −2, ρ < 1 a Hopf bifurcation of codi-
mension infinite occurs because the reduced system on the center manifold is
Hamiltonian (centers on center manifolds).
We would like to do several remarks on the above result. First, the region Ω is
split in six zones (see Fig. 3). A subcritical Hopf bifurcation occurs in the zones (A),
(C), (E) and (F), while on the contrary it is supercritical in the zones (B) and (D).
Remark that it is well-known that the Hopf bifurcation in the region where the three
parameters are positive (our region (A) that corresponds to Ω5 ) is always subcritical
[89, 95, 99].
Second, to guarantee the existence of the two codimension-three points P1 and P2
the Poincaré-Miranda theorem was used [81]. A detailed analysis of the roots of a
polynomial of degree 104 (it appears in the computation of the resultant of N 1(σ, ρ)
and the numerator of a2 ) is also needed.
Finally, as it occurs for the origin, the Hopf bifurcation of the nontrivial equilibria
has infinite codimension because the center manifold is an algebraic invariant surface,
namely x 2 + 2z = 0 (see Fig. 4). Moreover, the Hopf bifurcation of the nontrivial
equilibria only has this polynomial center manifold in the Lorenz system.
In the following we provide the results of some numerical continuations, obtained
with AUTO [52], in order to illustrate the dynamical consequences of the presence
of a Hopf bifurcation of codimension-three (that occurs at P1 and at P2 ). From this
degeneracy, a curve of cusp bifurcations of periodic orbits appears [70]. Thus, two
curves of saddle-node bifurcation contact tangentially at the cusp point, giving rise to
a semicubic parabola. Three periodic orbits exists in the system for proximate param-
12 A. Algaba et al.

1 5

0
0
−1 z
−2 −5

−3
−10
z −4 8
−5 6
4
−6
2
−7
y 0
−8 −2
−4
10
−6
0 5
0 −8 2 3 4 5
−10 −5 −2 −1 0 1
y −5 −4 −3
x x

Fig. 4 Two different perspectives of the phase space of the Lorenz system (1) for b = −2, σ =
−1, ρ = −2, where the nontrivial equilibria undergo a degenerate Hopf bifurcation of infinite
codimension. Some periodic orbits on the center manifold appear. Reproduced with permission
from [2]. Copyright (2015) by Springer

-0,52 -0,65

sn2

σ σ sn 2 h
h
Dh1
h sn 1
-0,53
cu 1 h Dh1
-0,66
-6,6 -6,4 -6,2 -6,6 -6,4 -6,2
ρ ρ

Fig. 5 Two partial bifurcation sets in the (ρ, σ )-plane in a neighborhood of the point P1 : (Left)
for b = −1.6; (Right) for b = −1.72. Reproduced with permission from [2]. Copyright (2015) by
Springer

eter values. These orbits disappear in pairs by means of saddle-node bifurcations.

The existence of hysteresis phenomenon is associated with the cusp bifurcation.
As similar results are obtained in the vicinity of P1 and P2 , we only consider here
the point P1 . Thus, we show for b = −1.6 and b = −1.72 (values at both sides of
b1 ≈ −1.709567) the corresponding partial bifurcation sets in the (ρ, σ )-plane.
A degeneracy in the Hopf curve of the nontrivial equilibria h occurs at the point
Dh1 when b = −1.6 > b1 (see Fig. 5 (Left)). Thus, if we move along h from left
to right, the Hopf bifurcation changes from subcritical to supercritical. A curve of
saddle-node of periodic orbits sn1 arises from Dh1 and remains below h. It collapses
with the saddle-node curve sn2 into the cusp cu1 .
To be placed on the other part of the point P1 , a value b < b1 has to be taken, for
instance b = −1.72 (see Fig. 5 (Right)). In this situation, as a consequence of the
A Review on Some Bifurcations in the Lorenz System 13

degenerate point Dh1 on the curve h, the Hopf bifurcation varies from subcritical
(on the left) to supercritical (on the right). Now the saddle-node curve sn2 emerges
from Dh1 and remains above h.

5 Takens–Bogdanov Bifurcations

In this section we summarize the results obtained in Ref. [3], devoted to the analysis of
Takens–Bogdanov bifurcations in the Lorenz system. In the first part we mention the
analytical results in the case of the Takens–Bogdanov bifurcation of the equilibrium
at the origin. Secondly, we precis some numerical results on the existence of Takens–
Bogdanov bifurcations exhibited by periodic orbits.
As was stated in Sect. 2, the origin E 0 exhibits a Takens–Bogdanov bifurcation
when
σ = −1, ρ = 1, b = 0. (7)

The corresponding normal form to third order for the reduced system on the center
manifold, obtained with the recursive algorithm developed in Ref. [66], is (see the
details in Ref. [3, Sect. 2])
u̇ = v,
(8)
v̇ = a3 u3 + b3 u2 v,

with
1 −2 − b
a3 = , b3 = .
b b2
Whereas the coefficient a3 cannot vanish, a degenerate Takens–Bogdanov bifur-
cation occurs when b3 = 0, i.e. when b = −2.
As it is well known (see, for instance, [47, 72]), when b3 = 0, the nondegenerate
Takens–Bogdanov bifurcation is of heteroclinic type if a3 > 0 and of homoclinic
type for a3 < 0. Therefore, a nondegenerate Takens–Bogdanov of heteroclinic type
exists in the Lorenz system if b > 0 and of homoclinic type for b < 0 (b = −2).
In symmetric systems, the Takens–Bogdanov point TB appears when a curve of
pitchfork bifurcations of the origin Pi collapses with a curve of Hopf bifurcations of
the same equilibrium H. In the heteroclinic case (see Fig. 7 (Left)), a curve of hetero-
clinic connections of the nontrivial equilibria He emerges from TB. In the homoclinic
case (see Fig. 10 (Left)), three curves emerge from TB: h (of Hopf bifurcations of
the nontrivial equilibria), Ho (of homoclinic connections to the origin) and SN (of
saddle-node bifurcations of symmetric periodic orbits).
When the coefficient b3 of the normal form (8) vanishes (if b = −2) a nonlinear
degeneracy appears. Specifically, as the center manifold is an algebraic invariant
surface, the Takens–Bogdanov bifurcation has infinite codimension: the origin is a
center in the Lorenz system when b = −2, σ = −1 and ρ = 1. This fact is illustrated
in Fig. 6.
14 A. Algaba et al.

0
z −1
−0.5
−2
z −1 4

−1.5 2

−2 y 0
5
2 −2
1
0 0
y −1 −4 1 2
−5 −2 x −2 −1 0
x

Fig. 6 Two different perspectives of the phase space of the Lorenz system (1) for b = −2, σ = −1,
ρ = 1, where the origin undergoes a degenerate Takens–Bogdanov bifurcation of infinite codimen-
sion. Some periodic orbits on the center manifold x 2 + 2z = 0 appear. Reproduced with permission
from [3]. Copyright (2016) by Elsevier

18
Pi
H 16
-1 TB
DHe1
DHe 2 14

z
σ 12

Pi He
DHe3 10
-2

0 1 2 3 4 5 -6 -4 -2 0 2 4 6
ρ x

Fig. 7 (Left) For b = 1, partial bifurcation set on the (ρ, σ )-plane in a neighborhood of the Takens–
Bogdanov point TB (heteroclinic case). Three degeneracies DHe1 , DHe2 and DHe3 are present
on the curve of heteroclinic connections He. (Right) For b = 1 and ρ = 15, projection onto the
(x, z)-plane of the Shilnikov heteroclinic loop He existing for σ ≈ −3.874338. Reproduced with
permission from [3]. Copyright (2016) by Elsevier

All the information on the Takens–Bogdanov bifurcation is condensed in the

following result.
Theorem 4 The locus in the (σ, ρ, b)-parameter space where the origin of the
Lorenz system undergoes a Takens–Bogdanov bifurcation is defined by

σ = −1, ρ = 1, b = 0.

This bifurcation is nondegenerate if b = −2: of heteroclinic type when b > 0 and

of homoclinic type when b ∈ (−∞, −2) ∪ (−2, 0). A degenerate Takens–Bogdanov
bifurcation of infinite codimension occurs if b = −2.
We summarize now the results mentioned in Sects. 4 and 5 on bifurcations of
infinite codimension. All three cases appear on the straight line b = −2, σ = −1:
A Review on Some Bifurcations in the Lorenz System 15

a Hopf of the origin if ρ > 1, a Hopf of the nontrivial equilibria when ρ < 1 and a
Takens–Bogdanov of the origin for ρ = 1.
On the other hand, it is interesting to comment that, in these three situations,
it is possible to find analytical expressions for the period of the orbits existing in
the center manifold x 2 + 2z = 0 (see Figs. 1, 4 and 6). Thus, by taking limit in the
corresponding expressions, for finite values of the parameter ρ, the existence of
superluminal periodic orbits (periodic orbits with unbounded amplitude and whose
period tends to zero) is demonstrated. All the details can be found in Ref. [36]. In
this work, it is also numerically shown that superluminal periodic orbits also exists
in other situations of physical interest when the parameter ρ tends to infinity.
In the rest of this section, we highlight the most important numerical results on the
Takens–Bogdanov bifurcation of the origin presented in Ref. [3, Sect. 3], which have
been obtained with AUTO [52]. Specifically, we present two partial bifurcation sets
in the (ρ, σ )-parameter plane. The first one, for b = 1, illustrates the heteroclinic
case (b = 1) whereas the second one, for b = −1.6, corresponds to the homoclinic
case. Note that, as for b = 0 a triple-zero degeneracy occurs, we obtain information
on both sides of such rich bifurcation.
In Fig. 7 (Left), for b = 1, a partial bifurcation set is drawn in a neighborhood
of the Takens–Bogdanov point on the (ρ, σ )-plane. According to the well-known
results in the heteroclinic case [47, 72], the Takens–Bogdanov point TB is placed on
the curve where the origin exhibits a pitchfork bifurcation Pi. From that point a curve
of Hopf bifurcation of the origin H emerges. As it is a supercritical Hopf bifurcation, a
stable symmetric periodic orbit arises at H. This periodic orbit disappears in the curve
He, where a heteroclinic orbit to the nontrivial equilibria occurs. In Fig. 7 (Right) a
heteroclinic loop is drawn for ρ = 15.
Three degeneracies are numerically detected on the curve He. For their descrip-
tion, the eigenvalues of the nontrivial equilibria are denoted by α ± βi, λ, and the
saddle-quantity δ = |α/λ| is considered. The first degeneracy He1 appears when the
nontrivial equilibria change from real saddle to saddle-focus. As δ > 1, this global
connection remains tame [44] and, in this way, a symmetric stable periodic orbit is
born from the curve He. A second degeneracy, He2 , is present when δ = 1, namely
the eigenvalues are resonant. At this point the heteroclinic orbit changes from tame
to chaotic Shilnikov [82, 102].
The third degeneracy He3 occurs because δ = 1/2 (null divergence). Since the
expression of the divergence in the Lorenz system (1) is divF = −(b + σ + 1) and
we have fixed b = 1, then divF = 0 along the straight line σ = −2. Observe that
divF has no dependence on the spatial variables but only on the system parameters.
This fact has important consequences on the bifurcations of the periodic orbits as we
briefly explain in the following. If γ (t) denotes a periodic orbit in the autonomous
system ẋ = F(x), then the variational equation is defined by the linear system

ẏ = D F(γ (t))y = A(t)y.

Another random document with
no related content on Scribd:
glycerin, gr. 1/100, has also been of benefit. For the paræsthesiæ and
anæsthesia of the early stage the faradic brush has been applied,
and for the motor weakness the faradic current, with the effect of
relieving but not curing these symptoms. During the attacks of
neuralgia morphine may be employed.

Inflammation.—The most important trophic disturbances are those

which consist of an actual disintegration of tissue. These are very
numerous. Herpes zoster, certain forms of eczema and pemphigus,
lepra anæsthetica, scleroderma, acute ulceration of the fingers,
perforating ulcer of the foot, acute bed-sores, keratitis with
anæsthesia of the cornea, sympathetic ophthalmia, and various
forms of joint disease, are examples of such trophic diseases.88 For
the discussion of the skin and eye diseases mentioned the reader is
referred to special articles on those subjects. In regard to lepra
anæsthetica, it may be mentioned that in this disease lesions have
been found both in the sensory nerves and in the posterior cornua of
the spinal cord.89
88 Leloir, “Tropho-neuroses,” Nouveau Dictionnaire de Médecine et Chirurgie
practiques, 1885.

89 Virchow, “Nerven Lepra,” Geschwülste, ii. 521; Tschirijew, Travaux du Laboratoire

de Vulpian, 1879.

I have reported90 a case of acute ulceration of the ends of the fingers

and toes occurring in an anæmic girl aged twelve, and persisting for
more than a year, associated with cyanosis of the hands and feet,
and with a tendency to rapid ulceration of any part of the body which
happened to be exposed to pressure. The hair of the eyebrows,
eyelids, and to a less extent that of the scalp, fell out during the
disease. There was no local asphyxia and no appearance of
gangrene; hence the case was not one of Raynaud's disease. Its
symmetrical distribution, the lack of any constitutional organic
disease or of any local cause for the production of the ulceration,
and the anæmic and nervous condition present, all pointed to a
disturbance of the central nervous system. It did not yield to
treatment, either local or general or electrical. Mills has seen a
similar case.91
90 Journal Nerv. and Ment. Disease, Feb., 1886, clinical cases.

91 Mills, Amer. Journ. Med. Sci., Oct., 1878.

Perforating ulcer of the foot begins on the sole, beneath any of the
metatarso-phalangeal articulations, preferably the first or the fifth, or
under the heel, as a small pustule under the epidermis. This
ruptures, and the ulcer which results begins to extend in a direction
vertical to the surface, involving the deeper tissues or even opening
into the joint and destroying the bone. It appears rather like a sinus
than an ulcer, and is remarkable from the fact that it is not painful
and is insensitive to touch, although it may prevent the patient from
walking on account of extreme tenderness. The epidermis around
the sinus is thickened and insensitive, and there may be anæsthesia
of the entire sole of the foot, and even of the leg, although this is due
to the neuritis present, which is also the cause of the ulcer, rather
than to the ulceration. The circulation is sluggish in the affected
extremity; it becomes cyanotic on exposure to cold, and seems
peculiarly liable to become œdematous. The œdema may go on to
suppuration, and involve the articulations, and ankylosis of the
smaller joints may follow. This, too, is to be traced to the neuritis.
The skin of the foot becomes pigmented, and may be dry or covered
with offensive sweat. As the patient cannot walk while the ulcer
remains, the condition demands treatment. Rest, moist warm
applications, antiseptic lotions, scraping out the sinus, and other
surgical means appropriate to the treatment of ulcers and sinuses do
not often prove of benefit, and in obstinate cases recourse has been
had to amputation of the foot. Electrical treatment has been tried in
vain.

Such perforating ulcers may occur symmetrically on both feet, and

may be numerous. They also occur rarely on the hands. They have
been observed not infrequently in locomotor ataxia, occurring either
early or late in the disease,92 and have appeared in patients suffering
from dementia paralytica. The only constant lesion found is a
degeneration of the peripheral nerve supplying the affected part.
92 For bibliography see Ross, Diseases of Nervous System, 2d ed., i. p. 256, and
Blum, Berliner klin. Wochen., 1875, Nos. 13-15.

Acute bed-sores (decubitus) develop in many cases of spinal and

cerebral disease, but not in all. It is argued that anæsthesia of a part
or simple pressure upon a part or irritation of an anæsthetic part will
not explain their occurrence, and that they must be traced to a
destruction of trophic centres in the cord or brain. No amount of
attention to the position and cleanliness of the patient is thought to
be sufficient to prevent them in some cases, and their progress is
often so rapid as to remove them from the category of simple
ulceration. Their most frequent seat is upon the buttocks, over the
sacrum, on the heels, and over the scapulæ; but it is believed that in
the conditions in which they appear in these places pressure on any
part may cause one. They begin within a few hours after the
occurrence of the lesion as irregular mottled patches of redness, in
which there soon appear small or large vesicles filled with dark
serum. These rupture, leaving little areas of ulcerated surface, which
soon unite to form a large ulcer with softened, infiltrated, and bloody
base and ragged edges. The tissues around and within the ulcer
have a tendency to slough and to become gangrenous, and the
process goes on rapidly until a considerable area of the surface is
completely destroyed, together with the deeper structures, fat,
muscles, and fasciæ, down to the bone, which may become
necrosed. Acute cystitis is so frequently associated with acute bed-
sore that it is traced to the same nervous cause, though here, again,
another explanation is possible. The constitutional disturbance
produced by these two affections may be increased by a true
septicæmia, to which the patient succumbs; or metastatic abscesses
may be formed, and pyæmia cause death; or, lastly, the ulcer on the
back may lay bare the vertebræ and erode their ligaments, thus
opening a way for the extension of suppuration to the meninges of
the spinal cord. The duration may be from a few days to two weeks,
and the prognosis is very grave.
Chronic bed-sores develop in a similar manner, although all the
processes described occur more slowly and are much less severe.
They do not show the tendency to extend so widely or so deeply, nor
to become gangrenous. They remain stationary after attaining a
certain size, and the tissues seem to lack all tendency to
reproduction and repair. When the central lesion which they follow
begins to pass away they heal; sometimes local treatment is
successful; but in many cases they persist for years, always
threatening the life of the patient from the possibility of their sudden
aggravation or from the occurrence of the complications already
mentioned in connection with the acute process.

Since neither form of decubitus occurs after lesion of the anterior

cornua of the spinal cord, nor in sclerosis of the posterior or lateral
columns, nor in those cases of locomotor ataxia in which extensive
invasion of the posterior cornua is present, while they are particularly
frequent as a complication of general myelitis, hemorrhage in the
cord, and syringo-myelitis, it is argued that the trophic centres whose
destruction is the cause of these bed-sores lie in the central gray
matter of the cord, near to the central canal. In unilateral lesion of the
cord the bed-sore occurs on the anæsthetic side of the body—i.e. on
the side opposite to the seat of the lesion—a fact which may indicate
a decussation of the trophic nerve-fibres in the cord,93 but which has
also been urged in proof of the theory that the sore is always due to
irritation of an insensitive surface.
93 Recklinghausen believes that these cases of disturbance of nutrition can be
explained by injury of anæsthetic parts in which vaso-motor disturbances of central
origin exist (Handbuch d. Path. d. Ernahrung, pp. 236, 237).

The treatment of bed-sores belongs to the domain of surgery, the

same methods being pursued as in the case of any large ulcer. The
application of electricity to the sore has been tried, but the results are
not sufficiently encouraging to warrant its recommendation.

Joint affections of neurotic origin are discussed in the articles upon

Locomotor Ataxia, Hemiplegia, and Hysteria.
It is evident from this review that the nervous centres have some
influence upon the activity of the cells of which the body is made up,
and that they control the processes of growth, function, repair, and
reproduction. The facts are too numerous, too varied, and too
positive to admit of any other explanation. Trophic disturbances must
therefore be considered as a set of symptoms referable to various
lesions of the nervous system. It is evident from the preceding
discussion that they may be produced by disease of the peripheral
nerves; by disease of the ganglionic cells, which nourish those
nerves; by disease of the spinal cord, especially in the region of the
central gray matter; and, finally, by disease of the brain. Whether in
the last condition the effect is a direct one, or is produced
secondarily by an irritation of the spinal centres, cannot yet be
determined. There are no trophic centres as yet localized in the
cerebrum, but the pathological facts already mentioned warrant the
conclusion that such centres will not long elude search. Certain facts
observed in cases of infantile hemiplegia point to the motor area of
the cortex as the seat of trophic centres for the motor mechanisms;
since it is found that when the motor cortex is destroyed in early life
the bones and muscles which it controls fail to develop properly.
Trophic centres for sensory mechanisms are not yet discovered.
Certain investigations of Luciani recently published94 point to the
cerebellum as the part of the brain which governs the general
nutrition of the body, but these need confirmation.
94 Alienist and Neurologist, July, 1885.

CONCLUSION.—While an attempt has been made here to consider

vaso-motor and trophic neuroses separately, it must be admitted that
in very many conditions the two are coincident. This follows
inevitably from what has been stated regarding the localization of the
vaso-motor and trophic centres in the spinal cord, and regarding the
course of the vaso-motor and trophic nerves from the spinal centres
to the periphery. These two classes of centres and nerves lie side by
side in the central and peripheral organs, and it is less surprising that
they should be jointly affected than that one should ever be involved
alone. In any case of lesion of the peripheral nerves or of the central
nervous system they may be expected. In all cases they are to be
regarded as symptoms of such lesions rather than as distinct
diseases.

INDEX TO VOLUME V.

A.

Abdomen, state of, in lead colic,

684

in tubercular meningitis,

726

727
Abscess of the brain,

791

792

of the spinal cord,

808

Achromatopsia, hysterical,

247

Aconite and aconitia, use of, in neuralgia,

1224

1227
 hypodermically in exacerbations of cerebral syphilis,

1015

Acoustic sensory disturbances in nervous diseases,

40

41

Acrodynia,

1254

Acute affections produced by heat,

387

alcoholism,

586
 cerebral anæmia,

776

encephalitis,

791

myelitis, spinal,

810

poisoning by alcohol in lethal doses,

596

simple meningitis,

716

spinal meningitis,

749

pachymeningitis,

747
Æsthesodic system of encephalon, localization of lesions in,

81

of spinal cord, localization of lesions in,

69

Affective mental disease,

153

Age, influence on causation of acute spinal meningitis,

750

of catalepsy,

315

of cerebral anæmia,
777

of chorea,

441

of chronic lead-poisoning,

680

686

688

of disseminated sclerosis,

883

of embolism and thrombosis of cerebral veins,

951

of epilepsy,

470
 of family form of tabes dorsalis,

871

of general paralysis of the insane,

77

of hæmatoma of the dura mater,

707

of hysteria,

216

of hystero-epilepsy,

293

of insanity,

116

117
 of intracranial hemorrhage and apoplexy,

927

of labio-glosso-laryngeal paralysis,

1173

of migraine,

406

1230

of myxœdema,

1271

of neuralgia,

1217

of paralysis agitans,

433
 of progressive unilateral facial atrophy,

694

of syphilitic affections of nerve-centres,

1000

of tabes dorsalis,

856

of the opium habit,

650

of thrombosis of cerebral veins and sinuses,

983

of tubercular meningitis,

725

of tumors of the brain,

1029

spinal cord,

1090

of writers' cramp,

512

relation of, to limitation of the myelitis in infantile paralysis,

1147

Agraphia in hemiplegia,

957

Akinesis in nervous diseases, definition of,

42-44
Alalia (see

Speech, Disorders of

).

Alexia in nervous diseases,

31

Albuminuria in chronic lead-poisoning,

682

689

Alcohol, influence on causation of acute meningitis,

716

of cerebral hyperæmia,

765
 of intracranial hemorrhage and apoplexy,

929

933

of tabes dorsalis,

854

of writers' cramp,

512

use of, in alcoholism,

642

644

in cerebral anæmia,

789
 in heat-exhaustion,

388

in insomnia,

380

381

in the chloral habit,

677

in the opium habit,

673

675

in thermic fever,

397
Alcoholic abuse as a cause of epilepsy,

472

insanity,

175

202

630-633

LCOHOLISM

573

Classification,

573

Definition,
573

Diagnosis,

637

of acute alcoholism,

637

of chronic alcoholism,

638

of dipsomania,

639

of hereditary alcoholism,

639

Etiology,

575