PI Passivity-Based Control : Application to Physical Systems.

Rafael CISNEROS MONTOYA
Soutenance de thèse de doctorat le 13 Juillet 2016, 15h00 à CentraleSupelec (Gif-sur-Yvette) Salle des séminaires du L2S

One of the best known forms of feeding back a system is through a three-term control law called PID (Proportional-Integral-Derivative) controller. PID controllers are sufficient for many control problems, particularly when process dynamics are not highly nonlinear and the performance requirements are modest. Besides, because of its simple structure, the PID controller is the most adopted control scheme by industry and practitioners. Since, the PI(D) tuning methods are based on the linearization, commissioning a PI(D) to operate around a single operating point is relatively easy, however, the performance will be below par in wide operating regimes. To overcome this drawback the current practice is to re-tune the gains of the controllers based on a linear model of the plant evaluated at various operating points, a procedure known as gain-scheduling. There are several disadvantages of gain-scheduling including the need to switch (or interpolate) the controller gains and the non-trivial definition of the regions in the plants state space where the switching takes place --both problems are exacerbated if the dynamics of the plant is highly nonlinear. This makes impossible to guarantee the system stability.  In this context, the current thesis work is aimed at the designing of PI controllers, based on the passivity theory, such that the stability of the nonlinear model is guarantied in closed-loop. The approach here presented is constructive and motivated by the applicaton to physical systems.

Membres du jury :

M. Stanislav ARANOVSKIY, Maître de conférence, ITMO University, examinateur.
M. Robert GRIÑÓ, Professeur, Polytechnic University of Catalonia, rapporteur.
M. Hugues MOUNIER, Professeur, Laboratoire de Signaux et Systèmes, examinateur.
M. Romeo ORTEGA, Directeur de recherche au CNRS, directeur de thèse.
M. Jacquelien SCHERPEN, Professeur, University of Groningen, rapporteur.

Some results on the existence of equilibria and stability of dc linear networks with constant power loads.

Séminaire le 13 Juillet 2016, 11h30 à CentraleSupelec (Gif-sur-Yvette) Salle du conseil du L2S - B4.40
Prof. Robert Griñó, Polytechnic University of Catalonia


The presentation will show some results on the existence of equilibria in dc electrical networks that supply to constant power loads (CPLs). Specifically, a necessary condition for the existence, which is also sufficient for the case one and two CPLs, will be presented. Besides, a sufficient condition, based on the negative imaginary systems concept, that assures local stability for all the range of possible equilibria will be shown for the case of a dc linear network with a single ideal or finite-bandwidth CPL.

Singular perturbations for hyperbolic port-Hamiltonian and non-hyperbolic systems

Séminaire le 13 Juillet 2016, 10h30 à CentraleSupelec (Gif-sur-Yvette) Salle du conseil du L2S - B4.40
Prof. Jacquelien Scherpen, University of Groningen


In this talk we explore the methodology of model order reduction based on singular perturbations for a fexible-joint robot within the port-Hamiltonian framework. We show that a fexible-joint robot has a port-Hamiltonian representation which is also a singularly perturbed ordinary differential equation. Moreover, the associated reduced slow subsystem corresponds to a port-Hamiltonian model of a rigid-joint robot. To exploit the usefulness of the reduced models, we provide a numerical example where an existing controller for a rigid robot is implemented. In addition, we provide ideas on how to expand this to planar slow-fast systems at a non-hyperbolic point.  At these type of points, the classical theory of singular perturbations is not applicable and new techniques need to be introduced in order to design a controller that stabilizes such a point. We show for some class of nonlinear systems that using geometric  desingularization (also known as blow up), it is possible to design, in a simple way, controllers that stabilize non-hyperbolic equilibrium points of slow-fast systems. The results are exemplified on the Van der Pol oscillator.

Almost Lossless Variable-Length Source Coding on Countably Infinite Alphabets

Séminaire le 8 Juillet 2016, 14h00 à CentraleSupelec (Gif-sur-Yvette) Salle du conseil du L2S - B4.40
Prof. Jorge F. Silva, University of Chile, Santiago.


Motivated from the fact that universal source coding on countably infinite alphabets is not feasible, in this talk a notion of almost lossless source coding will be introduced.  This idea —analog to the  weak variable-length source coding proposed by Han 2000— aims at relaxing the lossless block-wise assumption to allow a distortion that vanishes asymptotically as the block-length goes to infinity.  In this almost lossless coding setting, new source coding results will be presented that on one hand show that Shannon entropy characterizes the minimum achievable rate (known statistics), while on the other,  that almost lossless universal source coding becomes feasible for the family of finite entropy stationary and memoryless sources with countably infinite alphabets.

Biography: Jorge F. Silva is Associate Professor at the Electrical Engineering Department and director of the Information and Decision Systems (IDS) Group at the University of Chile, Santiago, Chile. He received the Master of Science (2005) and Ph.D. (2008) in Electrical Engineering from the University of Southern California (USC). He is IEEE member of the Signal Processing and Information Theory Societies and he is associate editor of the  IEEE Transactions on Signal Processing.  Dr. Silva is recipient of the Viterbi Doctoral Fellowship  2007–2008 and Simon Ramo Scholarship 2007–2008 at USC.   Dr. Silva general research interests include: detection and estimation, information theory and statistics, universal source coding, sparse and compressible models and  compressed sensing.

The Appointment Scheduling Problem: The Doctor, Her Patients and The Waiting Room

Séminaire le 8 Juillet 2016, 14h00 à CentraleSupelec (Gif-sur-Yvette) Salle du conseil du L2S - B4.40
Prof. Stijn De Vuyst, Ghent University (UGent), Belgium.


We consider the appointment scheduling problem in the case of one doctor who sequentially provides service to the patients in the waiting room, in particular with respect to the incurred waiting times for both the patients and the doctor. This problem is different from a classical single-service queueing system in at least two ways: (1) the arrivals happen at pre-determined instants instead of randomly and (2) equilibrium solutions are of no use here since we need to know the waiting time of each individual patient. Given the length of the session and the consultation time distribution of each of K scheduled patient, we obtain the moments of the patient's waiting time and of the doctor's idle times. We also discuss the complicating factors such as the impact of unpunctuality, i.e. what happens if patients do not arrive exactly as appointed as usually the case in practice. A mild degree of unpunctuality can be handled by our model, but problems arise as soon as patients can overtake each other. Finally, we use the our results to construct suitable heuristics for finding optimal optimal appointment schedules.

Biography: Stijn De Vuyst is currently assistant professor at the Faculty of Engineering and Architecture of Ghent University (UGent), Belgium, in the Department of Industrial Systems Engineering and Product Design. His expertise is in operations research, in particular stochastic modelling, simulation, queueing theory and scheduling with application to the design, planning and performance evaluation of production systems as well as telecommunication systems. He obtained a master degree in Electrical Engineering and a PhD degree in Engineering Sciences at Ghent University. Prior to 2012, he was a post-doctoral researcher affiliated with the department of Telecommunication and Information Processing and for 6 months with the Informatics department at Université Libre de Bruxelles. From 2012 to 2015 he presided the faculty's educational board for the Master program Industrial Engineering and Operations Research. He currently teaches various courses on stochastic simulation, quality engineering and industrial statistics.

Réseaux sans fil avec connaissance du canal imparfaite

Sheng YANG
Habilitation à Diriger des Recherches (HDR) le 30 Juin 2016, 11h30 à

La connaissance parfaite du canal dans un réseau sans fil est cruciale pour réduire de l’interférence entre les utilisateurs. Cependant, à cause des contraintes de délais et de ressource limité, cette connaissance est souvent bruitée ou retardée aux émetteurs/récepteurs. Etudier les limites fondamentales de la communications avec information adjacente imparfaite a donc une importance à la fois théorique et pratique.

 

La thèse d’HDR présente de nouveaux progrès dans cet axe. Notamment, l’influence des voies de retour et de la coopération sur les performances des réseaux de communication ainsi que l’impact de la manque de connaissance de phase du canal dans un canal MIMO ont été explorés en utilisant la théorie de l’information. Les résultats obtenus couvrent de différentes applications ,tel que les réseaux hétérogènes, la communication sécurisée, les réseaux de diffusion de contenu, et la communication par fibre optique.

Composition du Jury

Prof. Philippe CIBLAT, Telecom ParisTech (Examinateur)

Prof. Mérouane DEBBAH, CentraleSupélec (Examinateur)

Dr. Pierre DUHAMEL, L2S, CNRS  (Examinateur)

Dr. Walid HACHEM, LTCI, CNRS (Rapporteur)

Prof. Syed Ali JAFAR, University of California, Irvine (Rapporteur)

Prof. Gerhard KRAMER, TU Munich   (Rapporteur)

Prof. Hikmet SARI, CentraleSupélec (Invité)

Topological Interference Management

Séminaire le 28 Juin 2016, 14h00 à CentraleSupelec (Gif-sur-Yvette) Salle du conseil du L2S - B4.40
Dr Syed Ali Jafar


Abstract:

Studies of the degrees of Freedom (DoF) of wireless communication networks often focus on clever ways to exploit an abundance of channel knowledge which is rarely available in practice while ignoring topological aspects that are the basis of most robust interference management schemes. Topological interference management refers to a complementary perspective where the focus is on exploiting network topology under limited channel knowledge. Progress in this direction includes the discovery that optimal interference avoidance is essentially the index coding problem, that interference alignment plays a central role in this problem even though no precise knowledge of channel realizations is available, a new set of conditions for the approximate optimality of treating interference as noise, novel outer bounds based on aligned image sets, and connections to network coding problems such as distributed storage repair, multiple unicasts and private information retrieval. This talk will summarize the advances in the broad area of topological interference management and highlight some of the key open problems.

Biography:

Syed Ali Jafar received his B. Tech. from IIT Delhi, India, in 1997, M.S. from Caltech, USA, in 1999, and Ph.D. from Stanford, USA, in 2003, all in Electrical Engineering. His industry experience includes positions at Lucent Bell Labs, Qualcomm Inc. and Hughes Software Systems. He is a Professor in the Department of Electrical Engineering and Computer Science at the University of California Irvine, Irvine, CA USA. His research interests include multiuser information theory, wireless communications and network coding.

Dr. Jafar received the Blavatnik National Laureate in Physical Sciences and Engineering in 2015, the UCI Academic Senate Distinguished Mid-Career Faculty Award for Research in 2015, the School of Engineering Mid-Career Excellence in Research Award in 2015, the School of Engineering Maseeh Outstanding Research Award in 2010, the ONR Young Investigator Award in 2008, and the NSF CAREER award in 2006. His co-authored papers received the IEEE Information Theory Society Best Paper Award in 2009, IEEE Communications Society Best Tutorial Paper Award in 2013, IEEE Communications Society Heinrich Hertz Award in 2015, IEEE Signal Processing Society Young Author Best Paper Award (to student co-authors) in 2015, an IEEE GLOBECOM Best Paper Award in 2012 and an IEEE GLOBECOM Best Paper Award in 2014. Dr. Jafar received the UC Irvine EECS Professor of the Year award five times, in 2006, 2009, 2011, 2012, and 2014, from the Engineering Students Council and the Teaching Excellence Award in 2012 from the School of Engineering. He was a University of Canterbury Erskine Fellow in 2010 and an IEEE Communications Society Distinguished Lecturer for 2013-2014. Dr. Jafar was recognized as a Thomson Reuters Highly Cited Researcher and included by Sciencewatch among The World's Most Influential Scientific Minds in 2014 and again in 2015. He served as Associate Editor for IEEE Transactions on Communications 2004-2009, for IEEE Communications Letters 2008-2009 and for IEEE Transactions on Information Theory 2009-2012. Dr. Jafar was elevated to IEEE Fellow, Class of 2014, for contributions to analyzing the capacity of wireless communication networks.

Quelques problèmes d'automatique rencontrés dans l'automobile et les neurosciences

William PASILLAS LEPINE
Habilitation à Diriger des Recherches (HDR) le 15 Juin 2016, 14h30 à

Des résultats sur l'analyse des cycles limite de l'ABS et sur l'observation de l'XBS seront présentés.

Jury:

Maria Domenica DI BENEDETTO, Professeur à l'Université de l'Aquila, Rapporteur
Gildas BESANÇON, Professeur à l'INP de Grenoble, Rapporteur
Isabelle QUEINNEC, Directeur de recherche au CNRS, Rapporteur
Françoise LAMNABHI-LAGARRIGUE, Directeur de recherche au CNRS, Examinateur
Manfredi MAGGIORE, Professeur à l'Université de Toronto, Examinateur
Wim MICHIELS, Professeur à KU Leuven, Examinateur

A Delay-Based Sustained Oscillator : Oregonator Based Model

Séminaire le 14 Juin 2016, 15h00 à CentraleSupelec (Gif-sur-Yvette) Salle du conseil du L2S - B4.40
Hakki Ulas UNAL (L2S, Anadolu University)


Many metabolic and  physiological processes occur in some periodic fashion. The phenomena has been known for a long time,  however, the underlying mechanism of such oscillatory behaviour has not been  fully understood.  Belousov-Zhabotinskii reaction, which exhibits oscillatory behaviour that are analogous to ones observed in certain biological systems, is often utilized to better understand the  oscillatory mechanism in  these systems. The reaction is very complicated, however, its oscillatory behaviour is described by a simple model, called Oregonator.  By the mass-action kinetics, the model can be described  by three  variables, which correspond to concentration of some chemical reactants.  In this talk,  delay-based Oregonator model obtained by the use of delay-mass-action kinetics will  be discussed. Some qualitative analysis on the  model will also be presented.

Mass-Action Kinetic Models

Séminaire le 14 Juin 2016, 14h00 à CentraleSupelec (Gif-sur-Yvette) Salle du conseil du L2S - B4.40
VijaySekhar Chellaboina, (Mahindra Ecole Centrale)


In this talk, we present a general construction of the mass-action kinetic equations in a state-space form. Next, we discuss the nonnegativity of solutions to the kinetic equations and the inverse problem of constructing a reaction network having specified essentially non- negative dynamics. The problem of reducibility of the kinetic equations is next considered as well as the stability of the equilibria of the kinetic equations. Specifically, Lyapunov methods are employed to show boundedness and convergence of solutions. Finally, the zero deficiency result for mass-action kinetics in standard matrix terminology is presented.

Développement et Application des bornes issus de la théorie de l'information à certaines types de problèmes de coordination

Achal AGRAWAL
Soutenance de thèse de doctorat le 14 Juin 2016, 14h00 à CentraleSupelec (Gif-sur-Yvette) Amphi F3-06

Avec la montée de la connectivité entre les appareils (internet des objets), nouvelles possibilités de coordination entre les différentes entités ont ouvert. En même temps, des résultats récents, issus de la théorie de l'information, ont fourni des limites pour la performance que tout système de coordination pourrait atteindre sous certaines structures d'information. Dans cette thèse, nous développons ces résultats théoriques dans le but de les rendre plus facilement applicable aux problèmes pratiques. À cet égard, la contribution de cette thèse est double: 1) En outre développer les résultats théoriques pour fournir un aperçu de la structure des solutions au problème d'optimisation posés dans les travaux anterieurs , ainsi que la généralisation des résultats. 2 ) Développer des algorithmes qui exploitent le cadre théorique fourni par les

travaux antérieurs pour concevoir des mécanismes de coordination pratiques, décentralisées et robustes. La généralité de l'approche se prête à diverses applications, dont les éléments suivants ont été traités: optimisation de puissance dans les réseaux sans fil, planification de la consommation d'énergie dans les applications de réseau intelligent, ainsi que Witsenhausen contre-exemple, un problème important issu de la théorie du contrôle. Diverses possibilités sont encore à venir pour exploiter le cadre et les outils développés ici.En effet, ils pourraient être utiles même dans des domaines qui ne sont pas abordés dans cette thèse, mais qui nécessitent une coordination entre les agents avec des informations différentes à la disposition de chacun.

 

Jury:

M, Assaad, Mohamad Professeur Adjoint, CentraleSupelec Examinateur
M, Chahed, Tijani Professeur, Telecom SudParis Examinateur
M, Gesbert, David Professeur, EURECOM Rapporteur
M, Lasaulce, Samson Directeur de Recherche, CNRS Directeur de thèse
M, Millerioux, Gilles Professeur, Université de Lorraine Examinateur
M, Yezekael, Hayel Maître de Conférences, Université d'Avignon Rapporteur

Séminaire d'Automatique du Plateau de Saclay : Robust perfomance by a stable controller for infinite-dimensional plants

Séminaire le 8 Juin 2016, 11h00 à CentraleSupelec (Gif-sur-Yvette) Salle du conseil du L2S - B4.40
Hakki Ulas Unal (Anadolu University)


In a feedback system, besides the stabilization, the controllers are often designed to meet some performance specifications defined by H∞ norm minimization of corresponding sensitivity functions. From the practical point of view, if it is possible, it is desired the controller to be designed is stable. In this work, stable controller design to minimize the H∞ norm of the corresponding sensitivity function in a feedback system with a single-input single-output biproper infinite-dimensional real plant is considered. The plant may have infinitely many poles and simple zeros in the right-half-plane, however, its zeros are assumed to satisfy some growth condition. Interpolation-based approach will be used to design such a controller and a numerical example will be presented.

Séminaire d'Automatique du Plateau de Saclay :On Control Lyapunov-Krasovskii Functionals and Stabilization in the Sample-and-Hold Sense of Nonlinear Time-Delay Systems

Séminaire le 8 Juin 2016, 10h00 à CentraleSupelec (Gif-sur-Yvette) Salle du conseil du L2S - B4.40
Pierdomenico Pepe (Università degli Studi dell'Aquila)


This talk deals with the stabilization in the sample-and-hold sense of nonlinear systems described by retarded functional differential equations. The notion of stabilization in the sample-and-hold sense has been introduced in 1997 by Clarke, Ledyaev, Sontag and Subbotin, for nonlinear delay-free systems. Roughly speaking, a state feedback (continuous or not) is said to be a stabilizer in the sample-and-hold sense if, for any given large ball and small ball of the origin, there exists a suitable small sampling period such that the feedback control law obtained by sampling and holding the above state feedback, with the given sampling period, keeps uniformly bounded all the trajectories starting in any point of the large ball and, moreover, drives all such trajectories into the small ball, uniformly in a maximum finite time, keeping them in, thereafter. In this talk suitable control Lyapunov-Krasovski functionals will be introduced and suitable induced state feedbacks (continuous or not), and it will be shown that these state feedbacks are stabilizers in the sample-and- hold sense, for fully nonlinear time-delay systems. Moreover, in the case of time-delay systems, implementation by means of digital devices often requires some further approximation due to non availability in the buffer of the value of the system variables at some past times, as it can be frequently required by the proposed state feedback. In order to cope with this problem, well known approximation schemes based on first order splines are used. It is shown, for fully nonlinear retarded systems, that, by sampling at suitable high frequency the system (finite dimensional) variable, stabilization in the sample-and-hold sense is still guaranteed, when the holden input is obtained as a feedback of the (first order) spline approximation of the (infinite dimensional) system state, whose entries are available at sampling times, and the state feedback is Lipschitz on any bounded subset of the Banach state space

Séminaire d'Automatique du Plateau de Saclay : Optimal control and Lyapunov functions applied to the satellite attitude control

Séminaire le 24 Mai 2016, 11h00 à CentraleSupelec (Gif-sur-Yvette) Salle du conseil du L2S - B4.40
Nadjim Horri (Coventry University)


The use of Lyapunov functions is generally limited to proving the stability of a system with a given control law. In this presentation, Lyapunov functions are used to formulate optimal control problems as pointwise nonlinear programmes. These optimisation problems are equivalent to inverse optimal control problems. This approach is applied to satellite attitude control. The optimal attitude control problems under consideration will be the minimisation of the norm of the control torque subject to constraints on the convergence rate of a Lyapunov function. This approach improves the tradeoff between rapidity and energy consumption compared to a benchmark controller, which is taken to be a PD type controller without loss of generality. The phase space trajectories show that the solutions to some fundamental open loop optimization problems are particular cases of optimal control problem formulations based on the convergence rates of Lyapunov functions. This is the case of the minimum time single axis attitude control problem, which is a special case of the problem of maximizing the convergence rate of a Lyapunov function under maximum torque limitations. It is also the case of the problem of minimising toque for fixed manoeuvre time. The solution to this problem is a particular case of the problem of minimizing the norm of the control torque under a Lyapunov convergence rate constraint.

S³: Condition monitoring using vibration signals

Séminaire le 24 Mai 2016, 10h30 à CentraleSupelec (Gif-sur-Yvette) Salle des séminaires du L2S
Asoke K. Nandi


Condition monitoring of machines is an essential part of smooth, efficient, safe, and productive operation of machines. In this presentation, focus will be on rotating machines and in the use of vibration signals. Classification of vibration signals to different states of machines has been achieved through the developments and applications of signal processing and machine learning. This presentation will cover research efforts and some case studies carried out over many years.

Bio: Professor Asoke K. Nandi received the degree of Ph.D. in Physics from the University of Cambridge, Cambridge (UK). He held academic positions in several universities, including Oxford (UK), Imperial College London (UK), Strathclyde (UK), and Liverpool (UK) as well as Finland Distinguished Professorship in Jyvaskyla (Finland). In 2013 he moved to Brunel University London (UK), to become the Chair and Head of Electronic and Computer Engineering. Professor Nandi is a Distinguished Visiting Professor at Tongji University (China) and an Adjunct Professor at University of Calgary (Canada).
In 1983 Professor Nandi contributed to the discovery of the three fundamental particles known as W+, W− and Z0 (by the UA1 team at CERN), providing the evidence for the unification of the electromagnetic and weak forces, which was recognized by the Nobel Committee for Physics in 1984. His current research interests lie in the areas of signal processing and machine learning, with applications to communications, gene expression data, functional magnetic resonance data, and biomedical data. He has made many fundamental theoretical and algorithmic contributions to many aspects of signal processing and machine learning. He has much expertise in “Big Data”, dealing with heterogeneous data, and extracting information from multiple datasets obtained in different laboratories and different times. He has authored over 500 technical publications, including 200 journal papers as well as four books, entitled Automatic Modulation Classification: Principles, Algorithms and Applications (Wiley, 2015), Integrative Cluster Analysis in Bioinformatics (Wiley, 2015), Automatic Modulation Recognition of Communications Signals (Springer, 1996), and Blind Estimation Using Higher-Order Statistics (Springer, 1999),. Recently he published in Blood, BMC Bioinformatics, IEEE TWC, NeuroImage, PLOS ONE, Royal Society Interface, and Signal Processing. The h-index of his publications is 63 (Google Scholar).

Professor Nandi is a Fellow of the Royal Academy of Engineering and also a Fellow of seven other institutions including the IEEE and the IET. Among the many awards he received are the Institute of Electrical and Electronics Engineers (USA) Heinrich Hertz Award in 2012, the Glory of Bengal Award for his outstanding achievements in scientific research in 2010, the Water Arbitration Prize of the Institution of Mechanical Engineers (UK) in 1999, and the Mountbatten Premium, Division Award of the Electronics and Communications Division, of the Institution of Electrical Engineers (UK) in 1998.

Séminaire d'Automatique du Plateau de Saclay : Inverse optimal control: the sub-Riemannian case

Séminaire le 24 Mai 2016, 10h00 à CentraleSupelec (Gif-sur-Yvette) Salle du conseil du L2S - B4.40
Frédéric Jean (ENSTA)


An inverse control problem is formulated as follows: given a set of trajectories and a control system, find a cost such that these paths are optimal. The first question to ask is the uniqueness of the solution of such a problem. For general classes of costs the problem appears to be very difficult, even with a trivial dynamics. We are therefore interested in this issue for the class of costs which are quadratic in the control, when the dynamics depend linearly in the control (Riemannian and sub-Riemannian case). In this case we can reduce the problem to the question of the existence of geodesically equivalent metrics and the existing results will be described, from the theorem of Levi-Civita (1890) to those we obtained recently with Sofya Maslovskaya and Igor Zelenko.

S³: Time Frequency Array Signal Processing: Multi-Dimensional processing for non-stationary signals

Séminaire le 20 Mai 2016, 10h30 à CentraleSupelec (Gif-sur-Yvette) Salle du conseil du L2S - B4.40
Adel Belouchrani


Conventional time-frequency analysis methods are being extended to data arrays, and there is a potential for a great synergistic development of new advanced tools by exploiting the joint properties of time-frequency methods and array signal processing methods. Conventional array signal processing assumes stationary signals and mainly employs the covariance matrix of the data array. This assumption is motivated by the crucial need in practice for estimating sample statistics by resorting to temporal averaging under the additional hypothesis of ergodic signals. When the frequency content of the measured signals is time varying (i.e., nonstationary signals), this class of approaches can still be applied. However, the achievable performances in this case are reduced with respect to those that would be achieved in a stationary environment. Instead of considering the nonstationarity as a shortcoming, Time Frequency Array Processing   takes advantage of the nonstationarity by considering it as a source of information in the design of efficient algorithms in such environments. This talk deals with this  relationship between time-frequency methods and array signal processing methods. Recent results on the performance analysis of the Time Frequency MUSIC algorithm will be also presented.

Bio: Adel Belouchrani was born in Algiers, Algeria, on May 5, 1967. He received the State Engineering degree in 1991 from Ecole Nationale Polytechnique (ENP), Algiers, Algeria, the M.S. degree in signal processing from the Institut National Polytechnique de Grenoble (INPG), France, in 1992, and the Ph.D. degree in signal and image processing from Télécom Paris (ENST), France, in 1995. He was a Visiting Scholar at the Electrical Engineering and Computer Sciences Department, University of California, Berkeley, from 1995 to 1996. He was with the Department of Electrical and Computer Engineering, Villanova University, Villanova, PA, as a Research Associate from 1996 to 1997. From 1998 to 2005, he has been with the Electrical Engineering Department of ENP as Associate Professor. He is currently and since 2006 Full Professor at ENP. His research interests are in statistical signal processing, (blind) array signal processing, time-frequency analysis and time-frequency array signal processing with applications in biomedical and telecommunications. Professor Adel Belouchrani is an IEEE Senior Member and has published over 180 technical publications including 48  journal papers, 4 book chapters and 4 patents that have been cited over 5400 times according Google Scholar  and  over 2000 time according to ISI Web Of Science. He has supervised over 19 PhD students. Professor Adel Belouchrani is currently Associated Editor of the IEEE Transactions on Signal Processing and Editorial board member of the Digital signal processing Journal (Ed. Elsevier).  He has been recently nominated  as a founding member of the Algerian Academy of Science  and Technology.

Modélisation et observabilité des machines électriques en vue de la commande sans capteur mécanique

Mohamad KOTEICH
Soutenance de thèse de doctorat le 18 Mai 2016, 14h00 à CentraleSupelec (Gif-sur-Yvette) Amphi F3-06

Les stratégies de commande avancée des machines électriques nécessitent une connaissance précise de la position et/ou de la vitesse du rotor. Ces grandeurs mécaniques sont traditionnellement mesurées par des capteurs, ce qui augmente le coût et diminue la fiabilité et la robustesse du système. D’où l’importance de la commande sans capteurs mécaniques, dite commande sensorless : elle consiste à remplacer les capteurs physiques par un observateur d’état, qui estime les grandeurs mécaniques en se basant sur un modèle de la machine et à partir de ses entrées (tensions) et ses sorties (courants mesurés). Toutefois, avant d’entamer la synthèse d’un observateur, il faut examiner l’observabilité du système, c.à.d. vérifier si les grandeurs à estimer peuvent être reconstruites à partir des mesures et des entrées du système.

Ce travail de thèse concerne la modélisation et l’étude d’observabilité des machines électriques en vue de la commande sensorless. Dans un premier temps, les modèles des machines électriques sont établis, et il est montré qu’une modélisation unifiée des machines à courant alternatif est possible. Ensuite, en se basant sur ces modèles, l’observabilité des machines électriques est étudiée en vue de la commande sensorless. La théorie d’observabilité instantanée locale est appliquée, ce qui permet de formuler des conditions analytiques, faciles à interpréter et à vérifier en temps-réel, et qui font le lien avec les phénomènes physiques dans les machines.

La validité des conditions d’observabilité est confirmée par des simulations et sur des données expérimentales, en appliquant un observateur de type Kalman étendu.

Cette thèse contribue à une nouvelle vision des machines à courant alternatif commandées sans capteurs mécaniques, et à une compréhension plus profonde de leurs propriétés. Elle contribue ainsi à la synthèse de stratégies d’observation plus performantes dans les régimes de fonctionnement critiques (à vitesse nulle et/ou à fréquence d’entrée nulle).

Les nouveaux concepts proposés dans cette thèse, tels que le flux équivalent et le vecteur d’observabilité, avec les résultats obtenus, ouvrent de nouveaux horizons dans un domaine qui paraît devenir assez mature.

 

M. Jean-Pierre BARBOT          Professeur des Universités, ECS-Lab, ENSEA                                          Rapporteur    

M. Guy CLERC                         Professeur des Universités, AMPERE, Univ. Claude Bernard Lyon I        Rapporteur    

M. Maurice FADEL                   Professeur, ENSEEIHT, Laboratoire LAPLACE                                         Examinateur  

M. Philippe MARTIN                Enseignant-Chercheur, CAS, Ecole des Mines de Paris                          Examinateur

M. Eric MONMASSON            Professeur des Universités, SATIE, Université de Cergy-Pontoise         Examinateur

M. Gilles DUC                         Professeur, L2S, CentraleSupélec                                                           Directeur de thèse

M. Guillaume SANDOU          Professeur, L2S, CentraleSupélec                                                            Co-encadrant 

M. Abdelmalek MALOUM      Docteur, Renault                                                                                       Co-encadrant

Cooperative Relaying Protocols and Distributed Coding Schemes for Wireless Multiterminal Networks

Abdulaziz MOHAMAD
Soutenance de thèse de doctorat le 10 Mai 2016, 14h30 à CentraleSupelec (Gif-sur-Yvette) Amphi F3-06

Avec la croissance rapide des appareils et des applications mobiles, les besoins en débit et en connectivité dans les réseaux sans fil augmentent rapidement. Il est prouvé que les communications coopératives peuvent augmenter significativement l’efficacité spectrale et la fiabilité des transmissions entre les nœuds extrémaux. Le concept de coopération dans un réseau sans fil compte parmi les sujets de recherche les plus actifs en télécommunications, le but étant d'identifier les stratégies de coopération qui maximiseraient les gains en efficacité spectrale et en puissance d'émission. Pour coopérer, les nœuds du réseau partagent leurs ressources (énergie, bande de fréquence, etc.) pour améliorer mutuellement leurs transmissions et leurs réceptions. Dans les réseaux sans fil avec relais, les relais sont des nœuds dédiés à améliorer la qualité de la communication entre les nœuds sources et destination.

Dans la première partie de la thèse, nous nous concentrons sur un réseau sans fil avec relais spécifique où l'ensemble de sources (mobiles) veulent communiquer leurs messages à une destination commune (station de base) avec l'aide d'un ensemble de relais (contexte cellulaire, sens montant). Nous étudions, sur les plans théorique et pratique, un schéma coopératif dans lequel les relais, après une durée d'écoute fixée a priori, essayent de décoder les messages des sources et commencent à transmettre des signaux utiles pour ceux qui sont décodés correctement. Ces signaux utiles sont le résultat d'un codage canal-réseau conjoint.

Une des limitations du système coopératif précédent est précisément que le temps d'écoute des relais est figé et ne peut pas être adapté à la qualité fluctuante (aléatoire) des liens instantanés sources-relais. Pour pallier cette difficulté, nous proposons et analysons, dans une seconde partie de la thèse, un schéma de coopération plus avancé où le temps d'écoute de chaque relais peut être dynamique. Dans ces conditions, un relais bénéficiant d'une meilleure qualité de réception des sources peut commencer à coopérer plus tôt que d'autres relais ayant une qualité de réception moindre.

Enfin, dans la troisième et dernière partie de la thèse, nous considérons la présence d'une information de retour limitée (limited feedback) entre la destination et les sources et les relais, et tentons de caractériser l'efficacité spectrale d'un tel système.

Jury:

Antoine O. BERTHET  Directeur de thèse  CS – L2S CNRS UMR 8506 - Gif sur Yvette
Alister BURR  Examinateur  University of York – UK
Jean-Pierre CANCES  Rapporteur  XLIM CNRS UMR 7252 - Limoges
Pierre DUHAMEL  Examinateur  L2S CNRS UMR 8506 – Gif sur Yvette
Florian KALTENBERGER  Rapporteur  Rapporteur, Eurecom - Biot
Soon Xin NG (Michael)  Examinateur  University of Southampton - UK
Raphaël VISOZ  Co-directeur de thèse  Orange Labs – Issy les Moulineaux

Control of Fuel Cell Hybrid Power Systems

Séminaire le 9 Mai 2016, 15h00 à CentraleSupelec (Gif-sur-Yvette) Salle du conseil du L2S - B4.40
Prof. Jian Chen, College of Control Science and Engineering, Zhejiang University, Hangzhou, China


with the environment and energy crisis worsening, fuel cells have shown significant potential as an alternative energy conversion device with high efficiency and low environmental impact. In particular, Fuel cells are promising for new energy automotive applications. Fuel cell hybrid power systems usually include fuel cells and lithium-Ion batteries. In this presentation, I will talk about control of fuel cell systems, SOC Estimation of Lithium-Ion Batteries, and adaptive control strategies of fuel cell/battery hybrid power systems.

Biography: Jian Chen (M’06-SM’10) received the B.E. and M.E. degrees from Zhejiang University, Hangzhou, China, in 1998 and 2001, respectively, and the Ph.D. degree from Clemson University, South Carolina, USA, in 2005. From 2006 to 2008, he was as a research fellow working on fuel cell modeling and ­control at the University of Michigan, Ann Arbor, MI, USA. He joined IdaTech LLC, Oregen, USA, working on fuel cell back power systems and Proterra Inc., South Carolina, USA, working on the National Fuel Cell Bus Program, in 2008 and 2012, respectively. In 2013, he joined the Department of Control Science and Engineering, Zhejiang University, Hangzhou, China. Dr. Chen is currently a professor with the College of Control Science and Engineering, Zhejiang University, Hangzhou, China. He has been supported by the Chinese Recruitment Program of Global Youth Experts since 2012 and received the Major Program of National Natural Science Foundation of China on Modeling and Control of Fuel Cell vehicles in 2014. His research interests include fuel cell modeling and control, battery management, visual servo techniques, and nonlinear control.

Pages