Bratislava, Slovak Republic, July 8-11, 2015

Technical Program

Final Program

Full technical program of the conference is available at IFAC Papercept web.

Program Outline with Social Activities

Wednesday, July 8
14:00 17:00 Social Programme (Bratislava Old Town sightseeing - not included in the Registration Fee)
16:00 18:00 Registration
15:00 17:30 Tutorial (Melody Suite)
WeTTT Robust Control Design by the Performance Portrait (M. Huba, SK)
18:00     Walk from the hotel to the Primate´s Palace
18:30 19:30 Welcome Drink and Guided Tour of the Primate´s Palace
Thursday, July 9
7:30 8:30 Registration
8:30 9:00 ThOP1 Opening Ceremony (Radisson Blu Carlton Hotel Bratislava)
8:45 9:45 Plenary Session (Carlton Hall)
ThPL11 Stability of Time Delay Systems (V. Kharitonov, RU – Honorary Chair of ROCOND'15)
  Coffee Break
10:15 11:55 Technical Sessions
ThA1 H-Infinity and Robust Stabilization (Melody Suite)
ThA2 Computational Methods (Symphony Suite)
ThA3 Robust Adaptive Control (Rhapsody Suite)
12:00 13:00 Lunch
13:30 14:30 Plenary Session (Carlton Hall)
ThPL21 Control Theory and Its Impact on Society (V. Kučera, CZ - IFAC President's representative for ROCOND'15)
  Coffee Break
15:00 17:00 Technical Sessions
ThB1 LMI and Convex Optimization (Melody Suite)
ThB2 Robust Stability and Performance I (Symphony Suite)
13:30 17:00 Social Programme (Day Trips & Tours - not included in the Registration Fee)
19:30 24:00 Symposium Banquet (Radisson Blu Carlton Hotel Bratislava)
Performance of the Slovak State Traditional Dance Company SĽUK
Friday, July 10
7:45 8:30 Registration
8:30 9:30 Plenary Session (Carlton Hall)
FrPL11 LMI Hierarchies in Robust Control (D. Henrion and D. Peaucelle, FR)
  Coffee Break
10:00 12:00 Technical Sessions
FrA1 LMI and Convex Optimization II (Melody Suite)
FrA2 Robust Stability and Performance II (Symphony Suite)
12:00 13:00 Lunch
13:30 14:30 Plenary Session (Carlton Hall)
FrPL21 Convex Gain-Scheduling Design Algorithms (C. W. Scherer, DE)
  Coffee Break
15:00 17:00 Technical Sessions
FrB1 Robust Nonlinear Control (Melody Suite)
FrB2 Parametric Uncertainties (Symphony Suite)
13:30 17:00 Social Programme (Day Trips & Tours - not included in the Registration Fee)
19:30 20:30 Concert of the Hummel String Quartet
Saturday, July 11
8:30 9:30 Plenary Session (Carlton Hall)
SaPL11 Robust MPC Design, Future and Practical Applications (E. Camacho, ES)
  Coffee Break
10:00 12:00 Technical Sessions
SaA1 Robust Model Predictive Control (Melody Suite)
SaA2 Switched and Networked Control Systems (Symphony Suite)
8:30 12:00 Social Programme (Day Trips & Tours - not included in the Registration Fee)
12:00 13:00 Lunch
13:30 14:30 Plenary Session (Carlton Hall)
SaPL21 Sampled-Data Control Systems: Analysis, Design and Applications (J. C. Geromel, BR)
  Coffee Break
15:00 17:00 Technical Sessions
SaB1 Frequency Domain Methods (Melody Suite)
SaB2 Robust Control and Applications (Symphony Suite)
17:00   Closing Ceremony

Plenary Lectures

Carsten W. Scherer (Universität Stuttgart, Germany): Convex Gain-Scheduling Design Algorithms

Linear parameter varying (LPV) systems are described by linear differential equations whose describing matrices depend on time-varying parameters. The goal of the related synthesis problem is to design a controller of the very same structure such that the closed-loop system satisfies certain desired specifications on stability and performance for the entire set of admissible parameter trajectories. Since the parameters often admit the interpretation of describing the location of the system’s point of operation, LPV control methods are viewed as a viable alternative to classical gain-scheduling designs for controlling nonlinear systems.

In this talk we highlight the challenges in analyzing the stability and performance properties of LPV systems that result from the time-varying nature of the system description. We survey the key ideas how to develop systematic computational tools based on convex optimization that are suitable for controller synthesis.

In deviating from by now well-established approaches, we also emphasize the flexibility of the gain-scheduling design paradigm if using linear fractional system representations and dynamic integral quadratic constraints as tools for stability and performance guarantees. We sketch very recent developments in robust scheduled controller synthesis and present examples that illustrate the benefit of this design methodology both for synthesizing classical and distributed controllers.

Jose C. Geromel (University of Campinas - UNICAMP, Brazil): Sampled-Data Control Systems: Analysis, Design and Applications

Slides from presentation

Sampled-data control systems analysis and design is revisited. Stability and H2 and Hinf optimal control design conditions are expressed through linear matrix inequalities (LMI) opening the possibility to handle nonconstant sampled-data rate and Markov jump linear systems. Our intention is twofold, first to develop some basic theoretical results and second to put in evidence problems that can be treated in this context as, for instance, networked control systems operating under bandwidth constraints and packet dropout. Usefulness and limitations towards practical applications are discussed and illustrated.

Eduardo Camacho (Escuela Superior de Ingenieros, Sevilla, Spain): Robust MPC design, Future and Practical Applications

Slides from presentation

Model Predictive Control (MPC) has developed considerably in the last decades both in industry and in academia. This success is due to the fact that Model Predictive Control is perhaps the most general way of posing the control problem in the time domain. Although the technique originated in industry, the academic research community has contributed, during the last two decades, important results, especially in the stability domain, where stability and robustness conditions for MPC have been well established. It is widely accepted that the control of linear processes with linear constraints (i.e. linear MPC) is relatively mature research field. One of the main advantages of MPC is that model uncertainties can be taken explicity into account and this allows for the design of robust MPC. The receding control strategy used in MPC can be extended to the case of system identification by parameter bounding and furthermore to determine if a model is consistent with the obtained data in a receding horizon manner and this implictly allows for fault detection. The talk addresses these issues, it shows how concepts arising from the fault detection and fault tolerant design methods can be incorporated in an MPC framework and the advantages that can be gained by using MPC in this context.

Didier Henrion and Dimitri Peaucelle (LAAS-CNRS, Toulouse, France): LMI Hierarchies in Robust Control

Slides from presentation: Peaucelle, Henrion

Linear matrix inequalities (LMIs) were developed during the 1990s as a modeling and optimization framework especially relevant in robust control. First we survey their historical developments related to Lyapunov methods, the S-procedure and its generalization to quadratic separation. Then, we describe more recent extensions to LMI hierarchies aimed at gradually trading off between conservatism and computational cost. This will allow us to emphasize the link with LMI hierarchies in polynomial optimization since their inception in the early 2000s, building on results of real algebraic geometry (representation of non-negative polynomials as sums of squares) and functional analysis (representation of non-negative measures with their moments). Finally, we report on recent developments in polynomial optimal control and semi-algebraic approximations of regions of attraction and controlled invariant sets.

Vladimír Kučera (Czech Technical University in Prague, Czech Republic): Control Theory and its Impact on Society

Slides from presentation

The best way to understand the impact of control theory on society is to examine the evolution of automatic control throughout the centuries in the context of social needs and developments. The earliest examples of control were empirical feedback devices; ingenious but science lacking. They responded to practical needs in the antiquity and during the middle ages. The industrial revolution brought steam and the problem of speed control. The main tool was the centrifugal governor, which draw attention of scientists and gave rise to systematic studies of systems and control.

At that stage, control theory emerged and was based on mathematical models; control devices were replaced by universal controllers that made no longer an integral part of the inventions. During the second half of the last century, the computer technology had a tremendous impact on control theory and its application. Today, as a result of this evolution, it is possible to implement advanced control methodologies.

Control theory is a key facilitator of the modern technological developments, with great impact on product performance, cost, reliability, and energy consumption. Manufacturing, transportation, medicine, communications, economy, and other fields of human activity strongly benefit from control theory. The truly exciting developments in any field occur where there is a confluence of application drivers and disciplinary development of the subject. Automatic control is no exception. Control theory emerged from practical needs, and gradually replaced the control art to become the key facilitator of modern technological developments.

Vladimir Kharitonov (Saint-Petersburg State University, Russia): Stability of Time Delay Systems

Slides from presentation

Lyapunov-Krasovskii functional approach is a powerful tool for stability and robust stability analysis of time-delay systems. The effectiveness of the approach depends on the availability of functionals suitable for the analysis of a given time-delay system. It is assumed usually that the functionals have a certain structure with free matrix parameters. Then, after computing the time derivative of the functionals along the solutions of the time-delay system, the free matrix parameters are tried to choose to guarantee the negativity of the time derivative. This is how a great deal of LMI type stability and robust stability results reported in the literature have been obtained.

In this talk we present a different approach to the computation of functionals. The approach is based on the classical idea of Lyapunov: Choose first an appropriate time derivative, and then compute a functional, such that it’s time derivative along the solutions of the system coincides with the chosen one. It is shown that the new functionals are defined by special matrix valued functions, that are natural extension of classical Lyapunov matrices to the case of time-delays. An effective numerical scheme for the computation of these matrix valued functions is provided, as well. Application of the new class of the functionals to the robust stability analysis of time-delay systems is discussed in detail.