The Theory-Practice Gap: Where Are We ?
Professor Edward J. Davison
University of Toronto

This talk examines the question : where are we, and where are we going in Control? The talk can be broken into three sections:

1) The Gap Goes Both Ways: An overview is made of a number of the most successful industrial control systems which exist today, and it will be pointed out that the techniques used to design such control systems are not in our standard tool box of “theory”. Here “physics”, which plays a key role, is missing! On the other hand, some examples from industry will be given to show that their controller design obtained can completely fail, and what is needed is a dose of “theory” from the toolbox. Thus the gap goes both ways. Industry often says that the tools of “theory” are inadequate – the response to this is that we need to construct “better theory”.

2) Bridging the Gap: Many problems have constraints associated with the output time response of the system, as well of course, control signal magnitude constraints; for example, it is often desired that the output response should not have any over-shoot and should ideally have zero interaction. A new cheap control performance index will be suggested which can deal with such output time response problem constraints. The notion of a “saturation index”(SI) is then discussed which gives a measure of the closed loop system’s operating range. In this case, a system with a saturation index of one implies that the controller makes optimal use of the control signal constraints, while a larger SI implies that the system’s operating range may be highly restrictive. The application of the new performance index in Model Predictive Control (MPC), which always results in a SI of one, is then discussed. MPC has the significant disadvantage that it can only be used on plants with slower dynamics … a new MPC algorithm which can be up to 40x faster than standard algorithms will then be illustrated.

3) New Areas of Research: At all times, we should be alert to new areas of research. A possible new direction “High Dimensional Modeling” will be suggested, which deals with the problem of modeling and simulation of very high dimensional state systems, where the system is nonlinear, very stiff, and dense. For example, in polymer engineering, systems of equations of order n = 25000 often arise, and on using a Cray computer, standard stiff integration software hopelessly fails. However, on using the physics of the process and system theory, we can convert the polymer model to a new representation, which allows it to be easily solved using your home computer.

Edward J. Davison received the A.R.C.T. degree in piano in 1958, the B.A.Sc. degree in Engineering-Physics and the M.A. degree in Applied Mathematics from the University of Toronto in 1960, 1961 respectively, and the Ph.D. degree and Sc.D. degree from Cambridge University in 1964 and 1977, respectively. He was appointed as University Professor of Electrical and Computer Engineering at the University of Toronto in 2001, and as University Professor Emeritus in 2004. He was inducted into the University of Toronto’s Engineering Alumni Hall of Distinction in 2003. Dr. Davison has received several awards including the National Research Council of Canada's E.W.R. Steacie Memorial Fellowship 1974-77, the Canada Council Killam Research Fellowship 1979-80, 1981-83, the Athlone Fellowship (Cambridge University) 1961-63, two IEEE Transactions on Automatic Control Outstanding Paper Awards, and a Current Contents Classic Paper Citation. He was elected a Fellow of the Canadian Academy of Engineering in 2005, a Fellow of the Institute of Electrical and Electronic Engineers in 1977, a Fellow of the Royal Society of Canada in 1977, an Honorary Professor of Beijing Institute of Aeronautics and Astronautics in 1986, and has been a designated Consulting Engineer of the Province of Ontario since 1979. In 1984, he received the IEEE Centennial Medal and was elected a Distinguished Member of the IEEE Control Systems Society, and in 1996 he received the Outstanding Member Service Award from IFAC. He has served on numerous positions in the IEEE Control Systems Society including President in 1983 and Consulting Editor of the IEEE Trans. on Automatic Control in 1985. He was Chairman of the IFAC Theory Committee in 1988-1990, Vice-Chairman of the IFAC Technical Board in 1990-1993, and a member of the IFAC Council in 1991-1996. He has served on numerous Editorial Boards of various journals. In 1993, he was awarded the triennial Quazza Medal from the International Federation of Automatic Control, in 1997, he received the IEEE Control System Society's Hendrick W. Bode Lecture Prize, and in 2003, he received the Canada Council Killam Prize in Engineering

The California Freeway Performance Measurement System (PeMS) database stores real-time data from 26,000 loop detectors. PeMS extracts useful information from these data and displays it in graphical or tabular form. The data provide an unparalleled opportunity to assess highway performance and discover ways to improve highway management. The talk illustrates this opportunity by suggesting approaches to reduce congestion. The approaches range from bottleneck identification and effective ramp metering to questioning the current operation of high-occupancy vehicle or carpool lanes.

Pravin Varaiya is Nortel Networks Distinguished Professor in the Department of Electrical Engineering and Computer Sciences at the University of California, Berkeley. From 1975 to 1992 he was also Professor of Economics at Berkeley. His research is concerned with transportation, communication networks, and hybrid systems. Prof. Varaiya has held a Guggenheim Fellowship and a Miller Research Professorship. He received an Honorary Doctorate from L’Institut National Polytechnique de Toulouse, and the Field Medal of the IEEE Control Systems Society. He is a Fellow of IEEE and a member of the National Academy of Engineering. He is on the editorial board of several journals, including Discrete Event Dynamical Systems and Transportation Research---C". He has co-authored three books and more than 250 technical papers. The second edition of High-Performance Communication Networks (with Jean Walrand) was published by Morgan-Kaufmann in 2000. “Structure and Interpretation of Signals and Systems” (with Edward Lee) was published in 2003 by Addison-Wesley.

We survey some recent research directions within the field of approximate dynamic programming (ADP), with a particular emphasis on rollout algorithms and model predictive control (MPC). We argue that while motivated by different concerns, these two methodologies are closely connected, and the mathematical essence of their desirable properties (cost improvement and stability, respectively) is couched on the central dynamic programming idea of policy iteration. In particular, among other things, we show that the most common MPC schemes can be viewed as rollout algorithms and are related to policy iteration methods. Furthermore, we embed rollout and MPC within a new unifying suboptimal control framework, based on a concept of restricted or constrained structure policies, which contains these schemes as special cases.

Dimitri Bertsekas obtained his doctorate at the Massachusetts Institute of Technology, where he is currently McAfee Professor of Engineering. His research includes optimization, control, large-scale computation, and data communication networks, and is closely tied to his teaching and book authoring activities. He has written numerous research papers in these fields, and he has authored or coauthored thirteen textbooks, including "Nonlinear Programming," "Dynamic Programming," "Network Optimization," "Neuro-Dynamic Programming," "Introduction to Probability," and "Convex Analysis and Optimization," all published within the last ten years.

Professor Bertsekas was awarded the INFORMS 1997 Prize for Research Excellence in the Interface Between Operations Research and Computer Science for his book "Neuro-Dynamic Programming" (co-authored with John Tsitsiklis), the 2000 Greek National Award for Operations Research, and the 2001 ACC John R. Ragazzini Education Award. In 2000, he was elected to the United States National Academy of Engineering.

Atomic force microscopes provide unprecedented access to surfaces at the nanometer level both for imaging and for local surface modifications. Precise positioning, accurate control of interaction forces, and speed are critical issues when operating these instruments. This lecture summarizes how modern model-based control strategies lead to higher permissible imaging speeds, improved control over the interaction forces, and better tracking of surface features compared with conventional proportional-integral-controlled atomic force microscopes. In particular, H∞- and l1-optimal methods are applied to control both lateral scanning motions and vertical positioning. Various experimental results verify the achieved performance.

Andreas Stemmer holds the Chair of Nanotechnology at the Swiss Federal Institute of Technology (ETH) in Zurich, Switzerland, where in 1995 he founded the Nanotechnology Group. Born 1962 in Basel, Switzerland, he studied at the University of Basel where he obtained his diploma in Physics in 1986. He continued his studies in molecular biology at the M.E. M�ller Institute of the Biocenter at the University of Basel and received certification from the Swiss Commission for Molecular Biology (SKMB). In 1990 he earned his doctorate in biophysics working on biological scanning tunneling microscopy. After conducting research as visiting scientist (1990-92) at the Medical Research Council Laboratory of Molecular Biology in Cambridge, UK, he was assistant scientist (1992-95) at the Marine Biological Laboratory in Woods Hole, MA, USA. He was elected to the Faculty of Mechanical and Process Engineering at ETH Zurich in 1995 and was promoted to full professor in 2004.

His research focuses on the development of novel tools and processes to create a link as direct as possible between the macro- and the nano-world. His Nanotechnology Group is particularly active in the fields of imaging tools and techniques to expand the limits of scanning probe and light microscopy, and in natural nanofabrication to pattern and assemble nanostructures outside of cleanrooms. Most recently he started a research vector in biological engineering aimed at harvesting electrical energy from living human cells and tissue.

Identification for control: from the early achievements to the revival of experiment design

This lecture presents the author's views on the development of identification for control. The lecture reviews the emergence of this subject as a specific topic over the last 15 years, at the boundary between system identification and robust control. It shows how the early focus on identification of control-oriented nominal models has progressively shifted towards the design of control-oriented uncertainty sets. This recent trend has given rise to an important revival of interest in experiment design issues in system identification. Some recent results on experiment design are presented.

Michel Gevers was born in Antwerp, Belgium, in 1945. He obtained an Electrical Engineering degree from the Universit� Catholique de Louvain, Belgium, in 1968, and a Ph.D. degree from Stanford University, California, in 1972. He is a Fellow of the IEEE, a Distinguished Member of the IEEE Control Systems Society, and he holds a Honorary Degree (Doctor Honoris Causa) from the University of Brussels. He has been President of the European Union Control Association (EUCA) from 1997 to 1999, and Vice President of the IEEE Control Systems Society in 2000 and 2001. In 2004, he initiated a petition within the IEEE that successfully brought an end to the embargo applied by the IEEE towards its members in Cuba, Iran, Libya and Sudan. Michel Gevers is Professor at the Universit� Catholique de Louvain in Louvain-la-Neuve, Belgium. He is the coordinator of the Belgian Interuniversity Pole on Dynamical Systems and Control, funded by the Federal Ministry of Science. He has spent long term visits at the University of Newcastle, Autralia, the Technical University of Vienna, and the Australian National University, where he has also held a 3-year position as Senior Research Fellow. His present research interests are in system identification and its interconnection with robust control. Michel Gevers has been Associate Editor of Automatica and of the IEEE Transactions on Automatic Control. He is presently Associate Editor at Large of the European Journal of Control, and Associate Editor of Mathematics of Control, Signals, and Systems. He has published over 200 papers and conference papers, and two books: "Adaptive Optimal Control - The Training Man's GPC", by R.R. Bitmead, M. Gevers and V. Wertz (Prentice Hall, 1990), and "Parametrizations in Control, Estimation and Filtering Problems: Accuracy Aspects", by M. Gevers and G. Li (Springer-Verlag, 1993).

This lecture is devoted to new challenging control problems arising in the automotive industry as a consequence of the customer-driven performance specifications adopted by car builders which have dramatically increased the number of new proposed automated features where feedback interacts with the driver. The notion of "Fun-to-Drive by Feedback" relates, here, to the ability to design a control scheme resulting in good ride comfort behavior as well as acceptable safe operation. The paper shows how control techniques can be used to solve some of these problems, and discusses how these subjective notions can be formalized thanks to concepts such as passivity and model matching control. The paper presents a series of examples concerning systems that provide assisted automated devices (i.e. electrical power steering and assisted clutch synchronization), as well as systems with fully automated features (i.e. steer-by-wire system, stop-and-go), in which these aspects are assessed. The material of the lecture was prepared in collaboration with the following individuals: Hubert Bechart, Xavier Claeys, Pietro Dolcini, and John-Jairo Martinez. Support from Renault, CNRS, and the ARCOS-program is gratefully acknowledged.

Carlos Canudas-de-Wit was born in Villahermosa, Tabasco, Mexico in 1958. He received his B.Sc. degree in electronics and communications from the Technological Institute of Monterrey, Mexico in 1980. In 1984 he received his M.Sc. in the Department of Automatic Control,Grenoble, France. He was visitor researcher in 1985 at Lund Institute of Technology, Sweden. In 1987 he received his Ph.D. in automatic control from the Polytechnic of Grenoble (Department of Automatic Control), France. Since then he has been working at the same department as Director of research at the National Center for Scientific Research (CNRS), where He teaches and conducts research in the area of nonlinear control of mechanical systems. His research topics includes: vehicle control, adaptive control, identification, control of robots, systems with friction, AC and CD drives, and networked controlled systems. His research publications includes more than 120 conference papers, and more than 47 published papers in international journals. He has been pass associate editor for the IEEE-Transaction on Automatic Control 1992-1997, and for AUTOMATICA 1999-2002

44^th IEEE Conference on Decision and Control

and

European Control Conference 2005

December 12-15, 2005, Sevilla, Spain

Plenary Sessions

44th IEEE Conference on Decision and Control

and

European Control Conference 2005

December 12-15, 2005, Sevilla, Spain

Plenary Sessions

44^th IEEE Conference on Decision and Control