Events

Here you can find the list of seminars, meetings and events held by the Chair.

Guest Presentations


Date: 22.02.2024

Time: 10:00 CET (Central European Time)

Title of the Talk: Application of Deep Learning and Vision Transformer in Healthcare

Speaker: Amir A. Ghavifekr (Assistant Professor at University of Tabriz, Iran)

Profiles: LinkedIn, Google Scholar

Event Link: Zoom

Biography of the Speaker: Dr. Ghavifekr obtained his PhD along from the University of Tabriz in Iran. Since 2019, he is working as an assistant professor in University of Tabriz. His research interests include robotics and teleoperation systems among others.

Abschlusspräsentation des Projekts RADSPOT

Lieber Interessenten,

wir laden Sie herzlich zu unserer Präsentation im Projekt RADSPOT (Hochautomatisiertes und straßenschonendes Fahren auf Basis der Bodenradarsignale) ein.

Das Meeting wird als Web-Konferenz am 11. November 2021, von 9 – 13 Uhr stattfinden.


Kurze Projektbeschreibung

Entwicklungen zu automatisierten Fahrfunktionen werden derzeit weltweit mit hoher Intensität vorangetrieben. Insbesondere das hochautomatisierte Fahren rückt von einem Forschungsgegenstand immer mehr hin zur Produktentwicklung. Das vorliegende Forschungsvorhaben soll einen Beitrag hierzu durch die Digitalisierung der Verkehrsinfrastruktur und Vernetzung leisten. Im Mittelpunkt steht das innovative Konzept des „Digitalen Zwillings“ für Verkehrswege bzw. den Verkehr selbst. Dieses digitale „Wesen“ wächst kontinuierlich mit den Daten der Umfeldsensorik von fahrenden Fahrzeugen, insbesondere Bodenradarsignale. Durch den Einsatz moderner Verfahren des maschinellen Lernens und künstlicher Intelligenz ergeben sich hierbei denkbar viele nützliche Funktionen. Im aktuellen Forschungsprogramm wurden insbesondere Wartung von Verkehrswegen, Lokalisation und die intelligente, automatisierte straßenschonende Nutzung adressiert. Darüber hinaus werden die erforderliche digitale Infrastruktur und Werkzeuge entwickelt. Die Durchführung dieses komplexen Forschungsvorhabens erforderte umfangreiche Expertisen in den Bereichen Cyberphysik, Geophysik, Radartechnik und Telematik, die im Verbund abgedeckt werden.


Agenda

09:00  Eintreffen der Teilnehmer und Begrüßung

09:15  Vortrag der TUK, „Anomalieerkennung, Umwelterfassung, Kommunikationsinfrastruktur, Digitaler Zwilling und automatisiertes Fahren“

10:00  Vortrag GGD, „GPR: Technik, Prozessierung & Analyse“

10:30  kurze Pause

10:35  Vortrag 3D Mapping, „Referenzkartenerstellung und GPR-Integration in das 3D-Mapping-Messsystem“

10:55  Vortrag Dresden Elektronik, „Infrastruktur und Cloud-Dienste“

11:20  Vortrag Cmore/Luxoft, „Bindeglied für Car2X Kommunikation”

11:45 bis 12:00 kurze Pause

12:00 bis 13:00 ShowCase Präsentation und Diskussion

13:00  Verabschiedung


Zugangsdaten zum Zoom-Meeting

Den Einwahllink erhalten Sie nach vorheriger Anmeldung via Email. Wenden Sie sich hierfür bitte an Andrea.Scheck(at)mv.uni-kl.de

International Control e-Seminar

The International Control e-Seminar takes place monthly. For a possible participation, please contact Iqra Batool.


Date: 24.07.2023

Time: 15:00 CEST (Central European Summer Time)

Title of the Talk: Markov Decision Process for Emotional Behavior of Socially Assistive Robot

Speaker: Dr. Ali Nasir (Assistant ProfessorKing Fahd University of Petroleum & Minerals, Kingdom of Saudi Arabia)

Profiles: LinkedIn, Google Scholar

Event Link: Zoom

Abstract: This talk presents a Markov decision process-based model for a socially assistive robot. Problem addressed by the model is one to one correspondence of a robot with a human where robot has to convince the human about completing certain tasks. In this regard, emotions of the human and those of the robot are incorporated in the model. Furthermore, emotion transition probabilities and probabilities of robot being able to successfully convince the human are also incorporated. The resulting model however involves large state space. Computational complexity involved in calculation of optimal decision policy from the proposed model is discussed and a computational complexity reduction technique is presented that uses decomposition of the tasks to be performed into sub groups. An online learning framework is also proposed to account for un-modeled parameters in the problem. Behavior of decision-making optimal policy obtained from the proposed model has been demonstrated with the help of a simulation-based case study.

Biography of the Speaker: Dr. Ali Nasir obtained his PhD along with M.Sc. degrees in Electrical Engineering and Aerospace Engineering from the University of Michigan Ann Arbor in USA. He received his B.Sc. in Electrical Engineering from University of Engineering and Technology, Taxila. He is the recipient of Fulbright scholarship for MS leading to PhD. He is currently working on distributed control of multi-agent systems and human intent prediction. His research interests also include Approximate Dynamic programming, fault tolerant control, and nonlinear control and estimation.


Date: 16.06.2023

Time: 15:00 CEST (Central European Summer Time)

Title of the Talk: Optimal Control of Switched Dynamical Systems under Dwell Time Constraints

Speaker: Dr. Usman Ali (Assistant Professor in School of Electrical Engineering and Computer Science at National University of Science and Technology, Islamabad, Pakistan)

Profiles: LinkedIn, Google Scholar

Event Link: Zoom

Abstract: This talk addresses the problem of optimal mode scheduling subject to dwell time constraints which is the minimum amount of time a mode has to spend in one mode before it can transition to another. The constraint is important since most physical systems cannot switch rapidly between different modes and its presence also eliminates the problem of chattering solutions by construction. Also we will look into investigation of the topology of the optimization space and show that it lacks structure to define local minima. A theoretical framework is then developed for defining the optimal solution as stationary points of optimality functions and an optimality function proposed for characterizing the necessary conditions of optimality. The challenges posed by dwell time constraints to algorithmic implementation are addressed by exploring the geometric properties of mode insertion gradient and a technique is developed for rapid update of the mode sequence. The algorithm’s convergence to optimal solution is proved and simulation results provided to demonstrate the algorithm’s application.

Biography of the Speaker: Dr. Usman Ali received the M.S. degree in electrical engineering and computer science, M.S. degree in Mathematics and Ph.D. degree in electrical engineering and computer science from Georgia Institute of Technology, GA, USA in 2012, 2014 and 2016 respectively. He is currently working as an assistant professor in School of Electrical Engineering and Computer Science at National University of Science and Technology, Islamabad, Pakistan. His research interests include control theory, robotics, cyber-physical systems and smart grids.

Dr. Ali was a Flubright Fellow (U.S Department of State) from 2010 to 2017 at Georgia Institute of Technology and a recipient of the Alumni of the Year Award from University of Engineering & Technology, Peshawar.


Date: 10.02.2023

Time: 15:00 CET (Central European Time)

Title of the Talk: Guaranteed data-driven representations of dynamical control systems

Speaker: Dr. Adnane SAOUD (Associate Professor at CentraleSupelec, University Paris-Saclay, France)

Event  Link: Zoom

Abstract: In the last few years, learning-based techniques have shown a great success to control complex cyber-physical systems (CPS), where learning-based tools are used to provide mathematical models of a system, based on which one can synthesize a controller. However, the use of learning-based techniques in the context of safety critical CPS is particularly problematic, since learning-based components are typically viewed as black box-type systems, lacking formal guarantees. In this talk, we first recall how symbolic control techniques can be used in the context of the control of CPS. Then, we focus on how to provide guarantees when learning is used at the model’s level, namely we show how to learn a symbolic model from data, while providing formal guarantees. This symbolic model will be then used to construct the controller for complex specifications, such as linear temporal logics, based on existing tools in the formal methods community.

Biography of the Speaker: Adnane SAOUD is an Associate Professor (Maitre de conférences) at CentraleSupelec, University Paris-Saclay, France, and an Affiliate Professor at the School of Computer Science at Mohammed VI Polytechnical University, Benguerir, Morocco. Between January and August 2021, he was a Postdoctoral Research Fellow in the Electrical and Computer Engineering Department at the University of California, Berkeley. Between January and December 2020, he was a Postdoctoral Research Fellow in the Electrical and Computer Engineering Department at the University of California, Santa Cruz. He received the Ph.D. degree in Control from CentraleSupelec, France, in 2019. In 2018, he was selected as one of the top three finalists for the Best student paper award at the European Control Conference, ECC. In 2020, he was selected as a finalist of the French national Best PhD award Discerned by the GdR MACS. In 2021, he was selected as one of the top three finalists of the EECI PhD award, recompensing the best PhD work in Europe in the field of Control. He obtained the M.Sc. degree in control from CentraleSupelec, France, in 2016, and Electrical Engineering degree from Ecole Mohammadia d’ingénieurs, (EMI), Morocco, in 2014. His current research interests include formal methods for Cyber-Physical systems, learning-based control, adaptive control and compositional analysis and synthesis of interconnected systems.


Date: 20.05.2022

Time: 15:00 CEST (Central European Summer Time)

Title of the Talk: On the Sufficiently Scattered Condition for Unique Simplex-Structured Matrix Factorizations

Speaker: Dr. Tim Marrinan (Postdoctoral research associate at Department of Electrical Engineering and Computer Science, Oregon State University)

Event  Link: Zoom

Abstract: A classical result of Freund and Orlin ("On the complexity of four polyhedral set containment problems'', Math. Prog., 1985) established that checking whether a ball is contained by a polytope described as a convex hull of a set of points is NP-complete. In the realm of simplex-structured matrix factorization (SSMF), which includes nonnegative matrix factorization (NMF) as a special case, checking the so-called sufficiently scattered condition (SSC) for uniqueness that is relevant in most practical applications is equivalent to this classical problem. The result of Freund and Orlin, therefore, implies that it is infeasible to check whether a factorization is unique for large-scale problems, even if the SSC  is  likely to be satisfied. On the other hand, stricter sufficient conditions exist that can be checked efficiently, but their assumptions are too restrictive for many settings. Since many downstream applications of SSMF require the decomposition to be essentially unique, we propose a novel sufficient condition that can be checked efficiently in many practical scenarios. In fact, the complexity of the method is a function of the desired rank of the factorization, rather than of the number of columns of the matrix to be factorized, and the complexity decreases in a stepped fashion as a function of the sparsity of the input data. Thus, the proposed conditions build a bridge between the "easy-to-check but infrequently applicable" conditions and the "frequently satisfied but intractable" conditions that currently exist in the literature.

Biography of the Speaker: Tim Marrinan is an applied mathematician focused on bridging computational mathematics and rigorous machine learning. This goal is achieved by advancing the understanding of factorization models, latent representation models, and constrained optimization methods to come up with theory-guaranteed data analytical/computational methods.  At the moment, Tim is an instructor and postdoctoral researcher at Oregon State University, where he works with Professor Xiao Fu. Prior to Oregon State, Tim got his PhD in mathematics from Colorado State University and did research appointments at both the University of Paderborn in Germany and the University of Mons in Belgium. Outside of research Tim is working to support the mental health of students in math and engineering and advocating for students from historically excluded groups. His hobbies include canoeing, basketball, ultimate frisbee, and taking walks with his Irish wolfhound, Reggie.


Date: 19.04.2022

Time: 16:00 CEST (Central European Summer Time)

Title of the Talk: Dynamic Equations on Time Scales

Speaker: Dr. Vipin Kumar (Max Planck Institute for Dynamics of Complex Technical Systems, Magdeburg, Germany)

Event  Link: Zoom

Abstract: In this seminar, we discuss the basics of time scales theory and the dynamic equations on time scales. More precisely, firstly we discuss some basic operators which are very important to study time scales calculus. We discuss the Hilger derivative or delta derivative followed by the delta integral. After that, we discuss the initial value problem on time scales and the variation of parameter formula. Furthermore, we also discuss the following research problem:
  Finite-Time Stability and Stabilization Results for Switched Impulsive Dynamical Systems on Time Scales:
Here, we study the finite-time stability (FTS) and finite time stabilization problems for a class of switched impulsive systems evolving on an arbitrary time domain. This problem is formulated using time scale theory where the time domain can be continuous, discrete, union of disjoint intervals with variable gaps and variable lengths or any combination of these. Using common Lyapunov-quadratic and Lyapunov-like functions, we establish sufficient conditions to ensure the FTS results. Further, to solve the stabilization problem, we design state feedback controllers. We have illustrated the effectiveness of the obtained analytical results though numerical examples.


Date: 28.01.2022

Time: 15:00 CET (Central European Time)

Title of the Talk: Multidimensional realization theory and polynomial system solving

Speaker: Dr. Philippe Dreesen  (KU Leuven, Belgium)

Event  Link: Zoom

Abstract: Multidimensional systems provide tools for estimation, simulation and control, which go beyond one-dimensional systems. In this talk, we will use multidimensional realization theory for solving a set of polynomial equations. We show that linear algebra suffices to solve the problem at hand. We view the polynomial equations as multidimensional difference equations that are associated with a Macaulay matrix, which is the multivariate generalization of the Sylvester matrix. The right null space of the Macaulay matrix can be viewed as a multidimensional observability matrix. The classical shift trick in Kung's method from realization theory reduces the task of finding all the solutions of the polynomial equations to solving an eigenvalue decomposition. We study multiple solutions and solutions at infinity.

It is a joint work with Kim Batselier and Bart De Moor.

Biography of the Speaker: Philippe Dreesen is a Mathematical Engineer, currently working as a research expert at KU Leuven, ESAT-STADIUS. He received an MSc degree in Electrical Engineering (Burgerlijk Werktuigkundig-Elektrotechnisch Ingenieur), and a PhD degree in Engineering Science from KU Leuven in 2007 and 2013. His doctoral work was a project on solving systems of polynomial equations with the use of linear algebra, funded through a fellowship of Flanders' Agency for Innovation by Science and Technology (IWT Vlaanderen). From October 2013 until July 2020 he held postdoctoral positions at Vrije Universiteit Brussel (VUB) on the development of tensor-based methods for nonlinear system identification and modelling, and from August 2020 until January 2021 at Antwerp University in CoSYS-lab. Since February 2021 he is with KU Leuven, ESAT-STADIUS, where he performs research on connections between polynomial algebra, linear algebra, tensor methods and multidimensional systems theory.


Date: 25.11.2021

Time: 15:00 CET (Central European Time)

Title of the Talk: The puzzle of initial data of systems of partial difference equations

Speaker:  Prof. Debasattam Pal (Indian Institute of Technology Bombay, India)

Event  Link: Zoom

Abstract: A cornerstone in control and systems theory is the fact that a system of ordinary difference/differential equation can be converted to a system of first order equations (the state space equations). The dimension of the minimal state space also formalises the quantity of initial data needed to get all solutions of the system of equations. Surprisingly, such construction is in general impossible for a system of partial difference/differential equations. Likewise the initial data needed to solve a general system of partial difference/differential equations remains an open issue. In this talk we address this issue and present our new findings that provide a direction that can potentially solve this problem for the special case of autonomous systems of linear partial difference equations.

Biography of the Speaker: Debasattam Pal received his Bachelor of Engineering degree from the Department of Electrical Engineering of Jadavpur University, Kolkata, in 2005. He received his M. Tech. and Ph.D. degrees from the Department of Electrical Engineering, Indian Institute of Technology  Bombay, in the years 2007 and 2012, respectively. He then worked as an Assistant Professor in Indian Institute of Technology Guwahati from 2012 to 2014. Since June, 2014, he has been working in the Electrical Engineering Department of Indian Institute of Technology Bombay where he is currently an Associate Professor. His research interests include multidimensional systems theory, algebraic analysis of systems, dissipative systems, optimal control and computational algebra.


Date: 02.07.2021

Time: 15:00 CEST (Central European Summer Time)

Title of the Talk: Data informativity: a new perspective on data-driven analysis and control

Speaker:  Prof.Kanat Camlibel (Bernoulli Institute for Mathematics, Computer Science and Artificial Intelligence, University of Groningen, The Netherlands)

Event  Link:attendee.gotowebinar.com/register/7846454041399743759

Abstract: In this talk, we study data-driven analysis and control problems from the perspective of data informativity'. We collect data from an unknown dynamical system, assumed to be contained in a given model class. On the basis of these data, we want to assess system properties and to design controllers of the unknown system. Of course, this is only possible if the given data contain enough information. Our first contribution is to formalize what is meant byenough information' in the above context. In particular, we provide general definitions of the informativity of data for analysis and control. We then apply these definitions to a variety of analysis and design problems. For instance, in the case of input/state systems we characterize the information content required for the design of stabilizing and optimal controllers. In the same setting, we also derive new `data-driven Hautus tests' that enable controllability analysis directly using data. In the case of input/output data we study the data requirements for stabilization by dynamic measurement feedback. We also highlight several extensions including analysis and design from noisy measurements.
r-based control of 1D parabolic PDEs via modal decomposition will be presented.

Biography of the Speaker: Kanat Camlibel received the Ph.D. degree in mathematical theory of systems and control from Tilburg University, Tilburg, The Netherlands, in 2001. Currently, he is an Associate Professor at the Johann Bernoulli Institute for Mathematics and Computer Science, University of Groningen, The Netherlands, where he served as an Assistant Professor between 2007 and 2013. From 2001 to 2007, he held Post- Doctoral/Assistant Professor positions with the University of Groningen, Tilburg University, and Eindhoven Technical University, Eindhoven, The Netherlands. His research interests include differential variational inequalities, complementarity problems, optimization, piecewise affine dynamical systems, switched linear systems, constrained linear systems, multiagent systems, model reduction, and geometric theory of linear systems.


Date: 23.04.2021

Time: 15:00 CEST (Central European Summer Time)

Title of the Talk: Delayed and sampled-data control of PDEs

Speaker:  Prof. Emilia Fridman, Tel Aviv University, Israel

Event Software and related Link:attendee.gotowebinar.com/register/5229265621018200333

Abstract:  The talk will start with distributed control and estimation of semilinear 1D heat equations in the presence of input/output delays (which may include sampling in time). I will present sampling in space (or spatial decomposition) method, where the domain is divided into N subdomains with N sensors located in each subdomain, whereas the control is applied through distributed in space shape functions. Given upper bounds on the delays, sufficient conditions ensuring the stability and performance are established in terms of LMIs via appropriate Lyapunov functionals and Halanay's inequality. The results will be applied to a network-based deployment of multi-agents. Extensions to N-D heat, Kuramoto-Sivashiskii, damped wave, and other PDEs will be discussed.  For boundary control,  delay compensation via sub-predictors will be considered. Finally, recent results on finite-dimensional observer-based control of 1D parabolic PDEs via modal decomposition will be presented.

Biography of the Talk: Emilia Fridman received the M.Sc. degree from Kuibyshev State University, USSR, in 1981 and the Ph.D.degree from Voronezh State University, USSR, in 1986, all in mathematics. From 1986 to 1992, she was an Assistant and Associate Professor in the Department of Mathematics at Kuibyshev Institute of Railway Engineers, USSR. Since 1993 she has been at Tel Aviv University, where she's currently a professor of Electrical Engineering-Systems. She has held visiting positions at the Weierstrass Institute for Applied Analysis and Stochastics in Berlin (Germany), INRIA in Rocquencourt (France), Ecole Centrale de Lille (France), Valenciennes University (France), Leicester University (UK), Kent University (UK), CINVESTAV (Mexico), Zhejiang University (China), St. Petersburg IPM (Russia), Melbourne University (Australia), Supelec (France), KTH (Sweden). Her research interests include time-delay systems, networked control systems, distributed parameter systems, robust control, singular perturbations, and nonlinear control. She has published more than 180 articles in international scientific journals. She is the author of the two monographs including "Introduction to Time-Delay Systems: Analysis and Control" (Birkhauser, 2014). She serves/served as Associate Editor in Automatica, SIAM Journal on Control and Optimization, and IMA Journal of Mathematical Control and Information. In 2014, she was Nominated as a Highly Cited Researcher by Thomson ISI. Since 2018, she has been the incumbent for Chana and Heinrich Manderman Chair on System Control at Tel Aviv University. She is IEEE Fellow since 2019. She is currently a member of the IFAC Council.


Date: 26.03.2021

Time: 3:00 PM CET (Central European Time)

Title: Energy Management Strategy for Plug-in-Hybrid Electric Vehicles Based on Predictive PMP

Speaker: Roland Schmid (BMW Group, 80788 Munich, Germany)

Event Link: https://attendee.gotowebinar.com/register/5798984167585916176

Abstract: This talk presents an energy management strategy for parallel plug-in-hybrid electric vehicles based on Pontryagin's minimum principle (PMP) that is suitable for series-production vehicles. The PMP-strategy uses journey prediction data obtained from a series-production navigation unit. This introduces two important practical aspects that are addressed in detail in this work. First, the journey prediction provides clustered information such as estimated average speeds from the navigation map. This has a strong effect on the applicability of PMP due to so-called singularities that systematically appear whenever changes in the engine on/off command occur. Second, the prediction of the journey is necessarily uncertain. This implies that an open-loop PMP-solution is not sufficient and feedback has to be introduced. Two methods are discussed: one based on repeated optimization and the other based on an adaptive PMP approach. In this article, we furthermore present a novel solution method to deal with singular controls, which is easy enough to be relevant in automotive practice and is computationally efficient. The method achieves a performance very close to a benchmark solution, which is obtained from an offline application of PMP based on the information about the actual journey. An experimental study shows that there is considerable potential for energy consumption savings of 5.29% compared with a baseline control strategy.


Date: 19.03.2021

Time: 3:00 PM CET (Central European Time)

Title: Relative Stability and Input-to-state Stability in the Spatial Sup-norm for Parabolic PDEs

Speaker: Dr. Jun Zheng (Department of Electrical Engineering, Ecole Polytechnique de Montreal, Montreal, Quebec, Canada)

Event Link:attendee.gotowebinar.com/register/8283630857220159504

Abstract: I will introduce the notion of relative $\mathcal{K}$-equi-stability  (RKES) to characterize the uniformly continuous dependence of solutions on external disturbances for nonlinear parabolic PDE systems. Based on the RKES and the De Giorgi iteration, I will show a result on the input-to-state stability (ISS) in the spatial sup-norm for a class of nonlinear parabolic PDEs with either Dirichlet or Robin boundary disturbances. Two examples, which are concerned with a super-linear parabolic PDE with Robin boundary condition and a $1$-D parabolic PDE with an unstable term, respectively, will be provided to illustrate the obtained ISS results. Besides, I would like to show how to conduct stability analysis for cascaded parabolic PDEs, which are coupled over the domain or on the boundary of the domain by using the proposed notion of RKES and the De Giorgi iteration.  This is joint work with Prof. Guchuan. Zhu from the Department of Electrical Engineering, Polytechnique Montréal, and Prof. Sergey Dashkovskiy from the Institute of Mathematics, University of Würzburg.


Date: 19.02.2021

Time: 3:00 PM CET (Central European Time)

Title: Interpolation-based Design of Stabilizing Controllers for Retarded and Neutral Delay Systems

Speaker: Hitay Özbay, Bilkent University, Ankara, Turkey

Event Link:https://attendee.gotowebinar.com/register/8497667287637907723

Abstract: A class of retarded and neutral time-delay systems is considered: the plant may have finitely many or infinitely many unstable poles. Stabilizing controllers are obtained from two different types of interpolation-based designs. Robustness to uncertainties in the interpolation is discussed. A parameterization of all stabilizing integral action controllers is obtained. Examples are given to illustrate this simple design procedure.

Biography: Hitay Özbay is a Professor of Electrical and Electronics Engineering at Bilkent University, Ankara Turkey. He received the B.Sc., M.Eng. and Ph.D. degrees from Middle East Technical University (1985), McGill University (1987), and University of Minnesota (1989), respectively. His prior academic affiliations include the University of Rhode Island (1989-1990), and The Ohio State University (1991—2006). Professor Özbay served as Associate Editor for many journals, including IEEE Transactions on Automatic Control (1997-1999), SIAM Journal on Control and Optimization (2011-2014), and Automatica (2001-2007, 2012 - 2019). He was a Vice-Chair of the IFAC Technical Committee on Networked Control Systems (2005-2011); and currently is a Vice-Chair of the IFAC Technical Committee on Linear Control Systems (2017-2023). He is was a member of the Board of Governors of the IEEE Control Systems Society (2017-2019), and also was a General Assembly member of the European Control Association (EUCA), (2013-2019). Dr. Özbay is a Fellow of IEEE.


Date: 15.01.2021

Time: 3:00 PM CET

Title: Optimality of Stationary and Deterministic Policies for Zero-Delay Coding of Rd-Valued Linear Markov Sources with Quadratic Cost

Speaker:  Meysam Ghomi (Department of Mathematics and Statistics, Queen's University, Kingston, Ontario, Canada)

Abstract: We study the optimal causal zero-delay coding of a stable linear Markov process, under a quadratic cost with a fixed communication rate constraint. The goal is to minimize the infinite horizon average distortion problem.
a) We establish the existence and structure of optimal quantization policies provided that the realized quantizers are those with convex code cells. In particular, an optimal quantization policy is shown to be stationary and deterministic.
b) We show that a finite time horizon performance of an optimal (Markov and deterministic) T-horizon coding policy approaches the infinite time horizon average-cost optimal cost at a rate of O(1/T). This result establishes explicit bounds on the performance of finite-memory or sliding-block encoders.


Date: 13.11.2020

Time: 3:00 PM CET

Title: Modeling and Control of COVID-19 Epidemic through Testing Policies

Speaker:  M. Umar b. Niazi (Gipsa-Lab (CNRS) and Inria Grenoble Rhone-Alpes)

Abstract: Testing to detect the infected cases in a population is one of the most important mechanisms to control an epidemic. It enables the government to isolate the detected infected cases that limit the disease transmission to the susceptible population. However, despite the significance of testing in the control of an epidemic, the recent literature on testing policies lacks the control-theoretic perspective. In this paper, an epidemic model that incorporates the testing rate as a control input is presented. The proposed model differentiates the undetected infected from the detected cases, where the detected cases are assumed to be removed from the disease spreading process in the population. First, the model is estimated and validated for the COVID-19 data of France. Then, two testing policies, best-effort strategy for testing (BEST) and constant optimal strategy for testing (COST), are proposed. The BEST policy is a suppression strategy that provides a lower bound on the testing rate such that the epidemic switches from a spreading to a non-spreading phase. The COST policy is a mitigation strategy that provides an optimal value of testing rate that minimizes the peak value of the infected population when the total stockpile of tests is limited. Both testing policies are evaluated by predicting the number of active intensive care unit (ICU) cases and the cumulative number of deaths due to COVID-19.

About the speaker:  Muhammad Umar B. Niazi is currently pursuing the Ph.D. degree in Automatic Control from Grenoble INP, Universite Grenoble-Alpes, France, and is affiliated with Gipsa-Lab (CNRS) and Inria Grenoble Rhone-Alpes. His research interests include largescale network systems, aggregated state observers, model reduction, and differential game theory. Mr. Niazi was a finalist for the Best Student Paper Award in European Control Conference 2019.


Date: 23.10.2020

Time: 3:00 PM CEST

Title: Modeling by Infinite Petri Nets

Speaker:  Dmitry A. Zaitsev (Odessa State Environmental University, Ukraine)

Abstract: Infinite Petri nets were introduced for modeling modern networks, clusters, computing grids, and clouds that also concerns automated manufacture, cellular automata, and biological systems. A finite specification of infinite nets has been offered in the form of a parametric multiset rewriting system that takes into consideration spatial structure on a plane and in multidimensional lattices. A composition and analysis technique has been developed for the investigation of infinite Petri nets. A case study of a square grid structure composition and analysis is presented. Parametric description of Petri nets, parametric representation of infinite systems for the calculation of place/transition invariants, and solving them in parametric form allowed the invariance proof for infinite Petri net models. Some additional analysis techniques based on graphs of transmissions and blockings are presented. Complex deadlocks has been revealed and classifies as: a loop of blockings; a chain of blockings ended on an already blocked vertex; because of isolation by neighboring blocked devices. Further generalization on multidimensional structures such as hypercube and hypertorus implemented. Torus structures play a key role in communication subsystems of super computers, clusters, and networks on chip. Generators of Petri net models developed and put on GitHub for public use. As a result of complex deadlocks disclosure, a possibility of network blocking via ill-intentioned traffic has been revealed. Prospects for further development of infinite Petri net theory are outlined.

About the speaker:  Dmitry A. Zaitsev received the Eng. degree in Applied Mathematics from Donetsk Polytechnic Institute, Donetsk, Ukraine, in 1986, the Ph.D. degree in Automated Control from the Kiev Institute of Cybernetics, Kiev, Ukraine, in 1991, and the Dr.Sc. degree in Telecommunications from the Odessa National Academy of Telecommunications, Odessa, Ukraine, in 2006. He is a Professor of Information Technology at Odessa State Environmental University, Ukraine since 2019. He developed the analysis of infinite Petri nets with regular structure, the decomposition of Petri nets in clans, generalized neighborhood for cellular automata, and the method of synthesis of fuzzy logic function
given by tables. He developed Opera-Topaz software for manufacture operative planning and control; a new stack of networking protocols E6 and its implementation within Linux kernel; Petri net analysis software http://daze.ho.ua.


Date: 16.10.2020

Time: 3:00 PM CEST

Title: Delay Compensation in Control Systems

Speaker: Professor Michael Malisoff (Department of Mathematics, Lousiana State University, Baton Rouge, LA)

Abstract: Many control systems are subject to time delays. For instance, input delays may be caused by time-consuming information gathering, or gestation delays in biological processes. One approach to solving control problems under input delays involves solving the problems with the delays set to zero, and then computing upper bounds on the input delays that the systems can tolerate while still realizing the control objective. This is often done by finding Lyapunov-Krasovskii functions. For longer delays, the reduction model approach is often used but can lead to implementation challenges because of the presence of distributed terms in the control. A third approach to delay compensation involves sequential predictors, which can compensate for arbitrarily long input delays using stacks of differential equations instead of distributed terms. This talk will review recent developments in this area, and is based in part on the speaker’s joint work with Miroslav Krstic, Frederic Mazenc, Fumin Zhang, and several students.

About the speaker: Michael Malisoff earned his Ph.D. in Mathematics at Rutgers University in New Brunswick, NJ in 2000. He received the First Place Student Best Paper Award at the 1999 IEEE Conference on Decision and Control. In 2001, he joined the professorial faculty in the Department of Mathematics at Louisiana State University in Baton Rouge, LA, where he holds the Roy P. Daniels Professorship # 3 in the College of Science. His research is on systems and control, with an emphasis on engineering applications. He is an associate editor for Asian Journal of Control, European Journal of Control, Discrete and Continuous Dynamical Systems Series B, and SIAM Journal on Control and Optimization.


Date: 11.09.2020

Time: 3:00 PM CEST

Title: Modelling and Approximation in Complex Networked Systems

Speaker: Dr. Xiaodong Cheng (Department of Electrical Engineering, Eindhoven University of Technology, The Netherlands)

Abstract: Complex networked systems are becoming ever more prevalent in our society. The term ‘complex’ refers to large-scale topological features of the interactions as well as high-dimensional dynamics from different physical domains. The overwhelming complexity of these systems poses significant challenges in the systems and control domain of how to effectively and efficiently analyze and control these systems. In this talk, I will briefly introduce my research on scalable modeling and complexity reduction in dynamic networked systems. The first part considers the identifiability problem in complex networks, which aims to allocate external excitation signals such that all the dynamics in a network can be identified from data. The second part is focused on model-order reduction of structured networked systems that is to reduce the dimension of a complex network system while retaining all the salient structures and dynamics in the network avoiding unnecessary redundancy.

 


Date: 14.08.2020

Time: 3:00 PM CEST

Title: Shared Control Between Pilots and Autopilots: Illustration of a Cyber-Physical Human System

Speaker: Emre Eraslan (Systems Laboratory, Department of Mechanical Engineering, Bilkent University, Turkey)

Abstract: The control when two distinct decision-makers, a human operator and an advanced automation working together, face severe uncertainties and anomalies is a challenging problem. We focus on shared control architectures (SCAs) that allow an advantageous combination of their abilities and provide a desired resilient performance. One of the major challenges with two decision-makers in the loop is bumpy transfer when control responsibility switches between them. We propose the use of a common metric that enables a smooth, bumpless transition when severe anomalies occur. This common metric is termed Capacity for Maneuver (CfM), that is rooted in human behavior and can be identified in control systems as the actuator’s proximity to its limits of saturation. Two different SCAs are presented, both of which use CfM, and describe how human-experts and automation can participate in a shared control action and recover gracefully from anomalous situations. Both of the SCAs are validated using human-in-the-loop experiments. The experimental results show that in the context of flight control, these SCAs result in a bumpless, resilient performance.


Date: 17.07.2020

Time: 3:00 PM CEST

Title: Adaptive control allocation for constrained systems

Speaker: Shahabaldin Tohidi (Systems Laboratory, Department of Mechanical Engineering, Bilkent University, Turkey)

Abstract: Control allocation is the method of distributing control signals among redundant actuators. In the presence of actuator uncertainty, most existing control allocation strategies require estimation methods and/or excessive assumptions which might be hard to satisfy in practice. The proposed adaptive control allocation method does not require uncertainty estimation and/or persistency of excitation assumption. Actuator constraints are also respected by employing the projection algorithm. A human-in-the-loop stability analysis, in the presence of the proposed control allocation method, implies proper stability properties. In addition, a modified version of this method can be applied for pilot induced oscillation (PIO) mitigation in uncertain over-actuated systems.

 


Date: 26.06.2020

Time: 2:00 PM CEST

Title: Finite-time Estimation for Time Varying Systems with Delay

Speaker: Dr. Saeed Ahmed

Abstract: Constructing asymptotic observers for engineering systems is an important topic that is motivated by the difficulty of measuring state variables of systems. However, less attention has been paid to constructing finite-time observers, whose objective is to find values for states of the system after a predetermined finite time. Such a problem is important in engineering processes with deadlines and solving it can make it possible to design output feedback stabilizing controls. Moreover, measurement delays are present in many practical applications, such as chemical processes, aerodynamics, and communication networks, and they are time-varying. Therefore, the problem of finite-time estimation of systems with a time-varying delay is strongly motivated. The main aim of this talk is to present the design of finite-time observers for time-varying systems in the presence of a time-varying delay.


Date: 12.06.2020

Time: 2:00 PM CEST

Title: Self-triggered H-infinite control for Markov jump system

Speaker: Haiying Wan (Polytechnical University of Milano)

Abstract: Given the rapid development related to networked cyber-physical systems, new challenges arise in the context of systems with limited communication/computation bandwidth and/or limited energy resources in general. As one of the few aperiodic control strategies, self-triggered control can truly balance the usage of computational/communication resources on the one hand and control performance on the other. This talk will report recent work on self-triggered H-infinity control of Markov jump systems (MJSs). A resource-aware self-triggering scheme is proposed for MJSs to fulfill both disturbance rejection and computational resources saving goals. Besides, considering the finite frequency domain characteristics of the disturbance, the issue of a multi-frequency controller design based on the self-triggered strategy for MJSs is addressed.


Date: 15.05.2020

Time: 2:00 PM CEST

Title: Optimal Scheduling and Model Predictive Control for Trajectory Planning of Cooperative Robot Manipulators

Speaker: Dipl.-Ing. Argtim Tika

Abstract: The use of cooperating robots on existing assembly or packing lines relies on balancing the robots' workload and ensuring collision-free trajectory planning and execution. In this talk, we present a hierarchical control architecture with two-layer optimization-based control policies involving task scheduling in the top layer and path planning in the bottom one. For task allocation, a distance minimization algorithm is used, leading to an integer optimization problem with linear constraints and a bilinear cost function. For the trajectory planning, MPC-based methods are considered, which can be realized as a centralized architecture common to both robots, or as a distributed architecture where each robot is considered separately, as a local system.

Postdocs Presentations

Date: 26.05.2023

Title of the Talk: : Reachability analysis for optimal control using Koopman operator theory

Speaker: Dr. Duvan Andres Tellez Castro (26.05.2023)

Profiles: LinkedIn, Google Scholar

Biography of the Speaker: Dr. Castro is a visiting researcher in Umesh Vaidya's group at Clemson university. He received his PhD degree in Electrical Engineering from Universidad Nacional de Colombia in 2022, under the supervision of Eduardo Mojica-Nava and Jorge Sofrony. In 2015, he received a master’s degree in Electronic Engineering from Universidad de los Andes, Colombia, where he worked with Nicanor Quijano.


Date: 08.08.2022

Title: A Study on Parameter Dependent Stabilization of Networked Control Systems with Variable Sampling

Speaker: Dr. Seungyong Han

Profiles: Google Scholar

Biography: Seungyong Han received the B.S. degree in physics from Yeungnam University, Gyeongsan, South Korea, in 2016, and the M.S. degree in electronics engineering, in 2019, from Kyungpook National University, Daegu, South Korea, where he is currently working toward the Ph.D. degree in embedded systems and control engineering with the School of Electronic and Electrical Engineering, Kyungpook National University.,His research interests include cyber-physical systems, networked control systems, nonlinear systems, and control of manipulators and mobile robots.


Date: 19.04.2022

Title: Dynamic Equations on Time Scales

Speaker: Dr. Vipin Kumar

Profiles: Google Scholar

Abstract: In this seminar, we discuss the basics of time scales theory and the dynamic equations on time scales. More precisely, firstly we discuss some basic operators which are very important to study time scales calculus. We discuss the Hilger derivative or delta derivative followed by the delta integral. After that, we discuss the initial value problem on time scales and the variation of parameter formula. Furthermore, we also discuss the following research problem: Finite-Time Stability and Stabilization Results for Switched Impulsive Dynamical Systems on Time Scales

Here, we study the finite-time stability (FTS) and finite time stabilization problems for a class of switched impulsive systems evolving on an arbitrary time domain. This problem is formulated using time scale theory where the time domain can be continuous, discrete, union of disjoint intervals with variable gaps and variable lengths or any combination of these. Using common Lyapunov-quadratic and Lyapunov-like functions, we establish sufficient conditions to ensure the FTS results. Further, to solve the stabilization problem, we design state feedback controllers. We have illustrated the effectiveness of the obtained analytical results though numerical examples.


Date: 04.02.2022

Title: Nonlinear system sparse identification

Speaker: Dr. Lazaros Moysis

Profiles: LinkedIn, Google Scholar

Abstract: In this talk, we will delve into the fascinating realm of chaos, exploring its dynamics and implications. We will then shift our focus to system identification techniques, specifically discussing the application of linear regression methods. Drawing inspiration from the fundamental paper by Brunton et al. (2016), we will explore the sparse identification of nonlinear dynamical systems from data. Finally, we will conclude by highlighting the key considerations for future research and development in this field.

Biography: Lazaros Mozsis is a post-doctoral researcher at the Laboratory of Nonlinear Systems - Circuits & Complexity, situated in the Physics Department of Aristotle University of Thessaloniki. With a strong background in control-related subjects, Lazaros has held associate teaching positions at various universities, including AUTH, UOWM, and IHU, working on Control courses under yearly contracts sponsored by NSRF/ΕΣΠΑ ΕΔΒΜ. His research focuses on a diverse range of areas, such as chaotic systems, descriptor systems, chaos control, and synchronization. Lazaros applies his expertise in these fields to explore applications in encryption, secure communications, pseudo-random number generation, and area exploration.


Date: 17.09.2021

Title: Stability verification, scheduling, and synthesis of cyber-physical systems

Speaker: Dr. Mohammad Al-Khatib

Abstract: We present a study conducted on cyber-physical systems on three main aspects: stability verification, scheduling, and parameter synthesis. Embedded control systems (ECS) acting under timing contracts are the considered class of cyber-physical systems in the talk. ECS refers to integrations of a computing device with the physical system. As for timing contracts they are time constraints on the instants where some events happen such as sampling, actuation, and computation. These contracts are used to model issues that arise in modern embedded control systems: uncertain sampling to actuation delays, uncertain sampling periods, and interaction of several physical systems with shared computational resources (CPUs). Now given an ECS and a timing contract we reformulate the system into an impulsive one and verifies stability of the system, under all possible bounded uncertainties given by the contract, using safe convex approximation techniques and new generalized results for the problem on a class of systems modeled in the framework of difference inclusions. Second given a set of controllers implemented on a common computational platform (CPUs), each of which is subject to a timing contract, and best and worst case execution times on each CPU, we synthesize a dynamic scheduling policy, which guarantees that each timing contract is satisfied and that each of the shared CPUs are allocated to at most one embedded controller at any time. The approach is based on a timed game formulation that allows us to write the scheduling problem as a timed safety game. Then using the tool UPPAAL-TIGA, a solution to the safety game provides a suitable scheduling policy. In addition, we provide a novel necessary and sufficient condition for schedulability of the control tasks based on a simplified timed game automaton. Last, we solve a parameter synthesis problem which consists of synthesizing an under-approximation of the set of timing contracts that guarantee at the same time the schedulability and stability of the embedded controllers. The synthesis is based on a re-parameterization of the timing contract to make it monotonic, and then on a repeatedly sampling of the parameter space until reaching a predefined precision of approximation.

Biography: Dr. Mohammad Al-Khatib is an accomplished researcher with a diverse educational background. He holds a Ph.D. degree from Université Grenoble Alpes, specializing in the verification and synthesis of cyber-physical systems. Dr. Al-Khatib's expertise expanded during his postdoctoral research, where he focused on compositional synthesis and verification of hybrid systems at the Technical University of Munich in Germany. He further enhanced his knowledge in control synthesis, specifically in interconnected food production units, during his second postdoctoral research position at the Technical University of Chemnitz, also in Germany. Dr. Al-Khatib's impressive academic background showcases his dedication to advancing the fields of system control, information technology, and applied mathematics.


Date: 01.03.2021

Title: Data-driven controllability à la Jan Willems

Speaker: Dr. Vikas Kumar Mishra

Profiles: LinkedIn, Google Scholar

Abstract: In this talk, first, the behavioral controllability introduced by Prof. Jan C. Willems is recalled. Then, a data-driven criterion is developed to verify it for LTI systems. Finally, the identifiability of uncontrollable LTI systems is discussed and the results are illustrated with some examples.


Date: 27.11.2020

Title: Spatio-temporal prediction of patterns in Images

Speaker: Dr. Sandesh Anthani Hiremath

Profiles: LinkedIn

Abstract: In this presentation we shall cover different techniques of predicting spatial and temporal evolution of patterns as observed in camera images. Relevant to this we shall consider two different domains of applications namely automotive and cancer biology. In the former we consider the problem of dense depth estimation of a scene captured by a fisheye camera and for the latter we shall consider the problem of pseudopalisade pattern formation in Glioblastoma. For the problem of dense depth estimation we shall present a self-supervised DNN based algorithm that is able make fast predictions to meet real-time requirements. For the latter we shall present a classical gradient descent optimization of a nonlinear PDE that is able to predict the self-organizing behavior of cancer cells leading to the formation of pseudopalisades.


 

Regelungstechnisches Seminar

das KL-Seminar der Regelungstechnik, wobei interne und externe Vortragende in regelmäßigen monatlichen Abständen eigene aktuelle Forschungsthemen zur Methodik und Anwendungen der Regelungs- und Systemtheorie vorstellen.

Interessierte Forschungsassistenten, Doktoranden, Dozenten und Professoren sind herzlich zur Teilnahme eingeladen. Das Seminar findet am ersten Dienstag des Monats statt.