Seminars Fall 2010

21.09.2010 – Prof. Dale Seborg – Department of Chemical Engineering University of California, Santa Barbara, U.S.A.

Type I diabetes mellitus (TIDM) is a metabolic disease characterized by the body’s inability to produce insulin for the regulation of blood glucose levels (glycemia). Consequently, exogenous insulin infusion is required to maintain blood glucose concentration within an acceptable range. But with the availability of small continuous glucose sensors and insulin infusion pumps that can be worn by an individual during normal daily activities, automated glucose control for T1DM has become feasible. However, this control problem is very challenging due to the variations in daily life that affect glucose concentration (e.g., meals, exercise, and stress) and potential instrumentation problems with the glucose sensor and insulin pump. This seminar will describe recent results from joint glucose control projects between UCSB and the Sansum Diabetes Research Institute in Santa Barbara. The research has involved the development and experimental evaluation of a variety of techniques for process identification, feedback-feedforward control and process monitoring.

24.09.2010 – Prof. O. Crisalle – Department of Chemical Engineering University of Florida, Gainesville, U.S.A.

Dynamic systems that exhibit a large variations in dead-time (also known as transport delay) effect are particularly challenging to control in a fashion that ensures both closed-loop stability and adequate performance. We review results characterizing the robustness of the Smith Predictor control scheme, a technique widely used in chemical engineering applications to address the challenges posed by time-delay systems. Practical experience and a vast literature record suggest that this control method may not be particularly robust. In fact, under certain conditions the control loop may become unstable when the estimated delay is affected by very small errors, a perplexing behavior. A complete framework is presented to assess in a quantitative fashion the robustness of Smith Predictor designs. Typical municipal wastewater chlorination systems are affected by large time-varying delays. Chlorine is added as a disinfectant in a fashion that ensures adequate retention time in the basin, while satisfying regulations on its concentration at the discharge point. This talk shows that typical control schemes are unable to effectively reject the adverse effect of inlet flow rate variations that lead to significant time-delay fluctuations. Instead of implementing a Smith Predictor scheme, which may lack robustness under these circumstances, an automatic control solution that involves engineering redesign of the plant is proposed and evaluated. A new residence-time controller is advocated as a solution that involves both feedback and feedforward architectures and that can be deployed using standard PID software.

15.10.2010 – Dr. A. Terrier – Laboratory of Biomechanical Orthopedics. EPFL

Musculoskeletal disorders are the most common causes of severe long-term pain, physical disability, and arthritis, which limit everyday activities about 25% of the European population. The classical total joint replacement is the gold standard to treat arthritis, but new solutions using tissue bioengineering are currently investigated. Modeling the neuro-musculoskeletal system can be useful to understand the occurrence of arthritis, to optimize classical treatments, and develop new ones. Based on a strong collaboration with orthopedic surgeons (CHUV-DAL), we have developed several musculoskeletal models of the main human joints, which we are now extending to add a controlled sensorimotor aspect, in collaboration with the Automatic Control Laboratory (EPFL-LA). This presentation will be an overview of the research is this field and our current work.

22.10.2010 – Dr. N. Othman – Laboratory of Control and Chemical Engineering, CNRS, Lyon, France.

Controlling polymerization processes is essential in order to ensure the process safety and product quality. These reactions are known to be exothermic, rapid and sensitive to impurities and they have nonlinear behavior with constraints on the inputs and the states.
In solution or emulsion polymerization, one of the main polymer properties to be controlled is the polymer molecular weight (MW). This property is related to the final product viscosity and mechanical and thermal properties.
In this presentation, a multivariable control strategy is used to control the polymer MW in emulsion polymerization processes while maximizing the reaction rate. First of all, an online estimate of the polymer MW is calculated by developing a cascade of two nonlinear observers. Note that real measurements of the polymer MW can only be done off line with important delay. Controlling the polymer MW is then done by using input-output linearizing control that takes into account the process nonlinearity. The strategy is validated in the case of styrene (St) emulsion polymerization for which the model is well known. The methodology is then validated for the monomer methyl methacrylate (MMA) where a simplified behavioral model is identified.

Keywords: multivariable control; input/output linearization; polymerization processes.

5.11.2010 – (new date) Prof. J. Schoukens – Faculty of Engineering, Department of Fundamental Electricity and Instrumentation, Vrije Universiteit Brussel, Belgium.

My dream is to develop robust identification methods that provide good models to the user without requesting advanced user interactions. High quality nonparametric frequency response function measurements together with nonparametric measurements of the disturbing noise power spectra turn out to be important steps to make this dream comes through. Nonparametric models can not only be used to initialize more advanced identification methods, the results can also be used directly in modern control design approaches.
In this presentation we study first the properties of the existing nonparametric methods for estimating the plant and noise transfer functions of a linear dynamic system. The analysis is based on the recent insight that leakage errors in the frequency domain have a smooth nature that is completely similar to the initial transients in the time domain. This not only allows us to understand better the existing classic methods, but also opens the road to new better performing algorithms. These will be presented in the second part of the presentation. The presentation includes the output error setup, the errors-in-variables setup, and measurements under feedback conditions. Eventually, some of the methods are illustrated on experimental data.

19.11.2010 – Prof. S. Skogestad – Department of Chemical Engineering, Norwegian University of Science and Technology (NTNU) Trondheim, Norway.

Self-optimizing control deals with the selection of measurements (y) or measurement combinations as controlled variables (CVs), c = Hy. The issue is to select H. This is an important decision which is usually not view as a decision at all, and certainly not treated systematically. The term “self-optimizing” refers to cases where one can keep constant setpoints for the CVs, without any need to reoptimize when disturbances (d) occur.
In the talk, some approaches for selecting self-optimizing variables are reviewed, including the very simple nullspace method which is to select H such that HF=0 where F = dyopt/dd is the optimal sensitivity. More generally, one should select H to minimize the norm of HF. Normally, F is obtained from the model, but it can also be used to find patterns in optimal data. It is interesting that the focus is on the small singular values of the data matrix F,and not on the large singular values as is normally the case with “chemiometric” methods.
It is argued that self-optimizing control is not an alternative to real-time optimization (RTO), NCO tracking or model predictive control (MPC), but is to be seen as complementary. In self-optimizing control we determine controlled variables (CV). Preferably, the CV set points are kept constant, but they may also be adjusted using RTO or NCO tracking. In any case, a good choice of CVs will reduce the frequency of setpoint changes by RTO or NCO tracking.
When selecting self-optimizing CVs, a set of disturbances has to be assumed, as unexpected disturbances are not rejected in SOC. On the other hand, RTO and NCO tracking adapt the inputs at given sample times without any assumptions on what disturbances occur. By using NCO tracking in the optimization layer and SOC in the control layer below, we demonstrate that the methods are complementary. This combination allows for fast optimal action for the expected disturbances (by SOC), while other disturbances are compensated by NCO tracking on a slower time scale.

26.11.2010 – Dr. J. Trygg – Department of Chemistry, Umea University, Umea, Sweden.

Experimental sciences, e.g. biology, chemistry and medicine have to a large extent become information sciences and in turn, bioinformatics and chemometrics are now prerequisites for experimental and applied research. In systems biology, the general trend is to perform increasingly comprehensive characterizations of organisms in order to study the associations between their molecular and cellular components in greater detail. Abundances of transcripts, proteins and metabolites are measured at a current state or over time and analyzed using advanced mathematical and computational techniques. One of the recent developments is the OPLS method and its extensions O2PLS, OPLS-DA that have been successfully applied for prediction of clinical endpoints and data integration of data sets from an array of profiling technologies. This allows for better class separation, simpler interpretation, and the opportunity to identify potential biomarkers as well as provide an understanding of experimental and biological variation. I will also describe to take into account the individual dynamics, for example, slow and fast responders. A novel approach is to use each individual as its own reference. This corresponds to modeling the dynamic behaviour over time based on each individual’s trajectory to identify interesting patterns or trends. Statistical modeling and data integration across platforms of each individual trajectory can then be used to summarize the global dynamic behaviour over all individuals. This makes it possible to evaluate and handle the different types of variations such as individual differences in kinetics, circadian rhythm, and fast and slow responders.

Several studies in plant biology and medicine will be presented and discussed.

(postponed in 2011) – Dr. D. Roberge – Department of Process Research, LONZA Ltd., Visp, Switzerland.

Microreactor Technology enables processes to be run in a continuous manner using a minimal quantity of reagent. Thus, it permits the rapid and scalable development of continuous processes in the fine chemical and pharmaceutical industries. Under such circumstances, advantages are associated to the continuous way of operation and to the micro structure as well such as the good thermal control. It is important to differentiate from where the advantage is coming from because it will significantly influence the scale up strategy for larger productions.
In the fine chemical industry, productions are made in multi-purpose plants of high flexibility. The integration of a continuous system is such an environment is feasible given appropriate modules are employed with (i) good chemical resistance (i.e. glass, Hastelloy, Teflon), with (ii) excellent material stability over a large range of temperatures, and with (iii) ease of connection avoiding dead volume. In addition the various modules need to take into account the physico-chemical properties of the reaction such as the reaction kinetics (==> residence time) and the reaction phases (solid – liquid – gas). This approach leads systematically to a toolbox concept.
A detailed analysis at Lonza [1] showed that ca. 50% of the reactions studied could fit into a microreactor based on their kinetic. However, by taking into account the reaction phases, this number shrinks to ca. 20% of potential candidates because a solid phase is present in more than 60% of the cases. In addition, the reactions were classified into 3 classes namely Type A (mixing controlled reactions), Type B (rapid but kinetically controlled reactions), and Type C reactions (batch reactions with thermal hazard).
The 3 types of reaction define of course 3 types of reactor modules required to operate in a flexible manner various pharmaceutical reactions. This talk will present the analysis of the industrial designs of reactors, review their use, and address in details the scale-up concept. The Lonza MicroReactors have already produced several tons of material. Thus, manufacturing examples will be presented based on Lonza experience showing the applicability of continuous processes and microreactors in an industrial environment.

References:
1. D.M. Roberge, L. Ducry, N. Bieler, P. Cretton, and B. Zimmermann, “Microreactor Technology: A Revolution for the Fine Chemical and Pharmaceutical Industries?”, Chem. Eng. Technol., 28 (3), pp.318-323, 2005.

17.12.2010 – Dr. S. Gros – MLS Intelligent Control Dynamics, Glasgow, Scotland.

The concept of massive wind farm was developed over 30 years ago in California as a response to the ongoing oil crisis. At the time, the understanding of wind turbines dynamics was sparse, and the first attempts were catastrophic, yielding a whole field of academic research. Through numerous studies, trial and errors, the intricate dynamics inherent to wind turbines were slowly discovered and understood. Nowadays the field is getting mature, yet the demand placed on the wind turbine industry grows fast and the machines size increases exponentially, resulting in new dynamic and control issues.
In this talk, Sébastien Gros, a R&D engineer from the MLS Intelligent Control Dynamics team will give a brief introduction to wind turbine dynamics and control, and present the future challenges and perspectives.

17.12.2010 (17h30 ~ 18h30)- Dr. S. Gros – un-peu-plus-loin.ch

En 2008, doctorat en poche, Sébastien Gros charge 30kg de matériel sur un vélo construit sur mesure et se lance sur la routes pour un pari démesuré: rejoindre le camp de base de l’Everest sur deux roues, en empruntant les routes les plus improbables et les cols les plus élevés. Aventure personnelle mais aussi humanitaire, le projet avait pour but de contribuer à un autre pari, celui de Nicole Niquille : faire survivre un hôpital à Lukla, au Népal, une petite ville reposant à 2700m, à cinq jours de marche de la route la plus proche.
Après 15 mois passés entre les plateaux anatoliens, les grands déserts d’Iran, les pistes enneigées des Pamirs, du Karakoram et du Kashmir, sur les cols stratosphériques du Ladakh et finalement dans les vallées profondes du Solokhumbu, Sébastien atteint son but.
Avec plus de 22’000km affichés au compteur, les sacoches remplies d’histoires rocambolesques et des centaines de photos, Sébastien vient partager avec nous son aventure…