Logo TU Ilmenau

Analysis und Systemtheorie


Dr. Karl Worthmann


Telefon +49 3677 69-3624

E-Mail senden


Scientific Network

Gefördert von der Deutschen Forschungsgemeinschaft (DFG)

Regulation of Mobile Robots using Model Predictive Control
Set Point Stabilization, Path Following, and Distributed Control
(Grant WO 2056/1-1)

Regulation of Mobile Robots using Model Predictive Control
Beyond Set Point Stabilization
(Grant WO 2056/4-1)

Set Point Stabilisation, Path Following, and Distributed Control


Associated Members

Kick-off meeting

2015: July 12th - July 14th
at Technische Universität Ilmenau

Programme [PDF]

Accomodation: Hotel Garni am Kirchplatz

Do you want to participate? Please contact Jun.-Prof. Dr. Karl Worthmann

2nd Meeting

1st Symposium on Robotics and Model Predictive Control: Path Following
at BIBA, Universität Bremen
Dates: 13th - 16th of September 2015

Invited Speakers:

  • Boris Houska, Assistant Professor, PhD (School of Information Science and Technology ShanghaiTech University, China)


Organizers: Jürgen Pannek and Karl Worthmann

Accomodation: FIVE SEASONS designhotel Bremen
Do you want to participate? Please contact Jun.-Prof. Dr. Karl Worthmann


Program                                            Flyer

3rd Meeting

Workshop on Economic and Distributed Model Predictive Control
at École polytechnique fédérale de Lausanne, Switzerland
Dates: 21st - 22nd of March 2016


Organizers: Timm Fraulwasser, Colin Jones, and Karl Worthmann

Invited Speakers

Andrea Alessandretti, EPFL, Switzerland

Andrea Alessandretti: Continuous-Time Model Predictive Control for Economic Optimization: Theory, Design, and Applications to Motion Control of Underactuated Vehicles

This talk presents some recent results on the design of optimization-based control laws for the case where convergence to a desired set-point, minimization of an arbitrary performance index, or a combination of the two, is the desired objective. In recent years a growing attention has been dedicated to a new class of controllers that goes under the name of Economic-MPC, where. Here, the term economic is used to stress the fact that the performance index is a general index of interest that we wish to minimize, e.g., economic, which generally does not denote the distance to a desired set point. This setting makes full use of the potentialities of optimization-based control strategies. Although, it comes with some challenges. In fact, by choosing an arbitrary performance index, it is difficult to predict the evolution of the closed-loop system, which could potentially manifest undesirable behaviors. This talk presents analysis and certification of a variety of closed-loop behaviors stemming from the use Economic-MPC controllers. A set of tools for design of provably correct MPC controllers is provided for the case where the performance index is of the Tracking-MPC type, purely economic, or a combination of the two. The results focus the certification of both closed-loop economic performance and closed-loop state evolution. The proposed strategies are applied to a range of motion control problems for underactuated vehicles. An MPC controller for Trajectory-Tracking and Path-Following with convergence guarantees is first proposed and then extended, using the results presented on Economic-MPC, to address the control problems of distributed formation keeping, energy efficient trajectory-tracking, and target-following through highly observable trajectories.

Prof. Dr. Boris Houska, ShanghaiTech University, China

Prof. Dr. Houska: ALADIN---An Augmented Lagrangian Based Algorithm for Distributed Non-Convex Optimization and Control

This talk is about distributed derivative-based algorithms for solving optimization problems with a separable, potentially nonconvex objective function and coupled affine constraints. We propose an Augmented Lagrangian Based Alternating Direction Inexact Newton method (ALADIN), which combines ideas from the fields of sequential quadratic programming and augmented Lagrangian algorithms. In contrast to the alternating direction method of multipliers (ADMM), ALADIN has a superlinear convergence rate under suitable conditions and is applicable to nonconvex optimization problems. We illustrate the practical performance of ALADIN with applications from the field of distributed optimization and real-time model predictive control.

Prof. Dr. Melanie Zeilinger, ETH Zürich, Switzerland

Prof. Dr. Zeilinger: Distributed Predictive Control: Changing Topologies and Certified Computation

The control of a network of interacting dynamical systems is a central challenge for addressing a range of emerging application problems. Utilizing the connectivity and interactions in the network by exploiting advances in communication and computation technologies offers the potential for pushing these systems to higher performance while increasing efficiency of operation, which will reduce system over-design and associated costs. However, safety requirements and high system complexity represent key limiting factors for leveraging these new opportunities. This talk will present some of our recent work that brings high-performance control with hard guarantees on system safety to distributed systems, offering a scalable and modular approach that exploits interconnection effects and flexibly adjusts to network changes. A framework for plug and play distributed predictive control will be introduced and we will discuss essential theoretical and practical aspects for certifying distributed decision-making based on an optimization-in-the-loop paradigm. We will present an application example of these ideas to load shaping and voltage control in smart grids. Lastly, we will discuss some the computational aspects of the framework and present new results for certifying optimization with limited-precision computation or communication .


Dr.-Ing. Matthias Müller, University of Stuttgard, Germany

Dr.-Ing. Müller: Economic MPC: the role of dissipativity and application potentials in cooperative control

Economic model predictive control is a variant of MPC where in contrast to the classical control objective of stabilization, a more general performance criterion is considered which is possibly related to the economics of the considered system. Such a control objective arises in many applications such as, e.g., in the process industry, in wind turbine control, or builing climate control. In the first part of this talk, we show that dissipativity plays a crucial role in the context of economic MPC. In particular, we show that under mild technical assumptions, both cases where the optimal operating behavior is constant or periodic can equivalently be characterized by a suitable dissipativity condition. Furthermore, we discuss implications of these results for statements about the closed-loop system resulting from application of economic MPC schemes. In the second part of the talk, we present a distributed economic MPC algorithm which is suited for cooperative control problems in networks of self-interested systems. The proposed framework is such that the optimal cooperative behavior is synthesized online while the systems already take control actions, which makes it readily applicable in plug-and-play settings.

Mario Zanon, PhD, Chalmers University of Technology, Sweden

PhD Zanon: Tuning of Tracking MPC Based on Economic Criteria

The stability proof for economic Model Predictive Control (MPC) relies on strict dissipativity, which is in general difficult to establish. In contrast, tracking MPC has well-established and practically applicable stability guarantees, but can yield poor closed-loop performance in terms of the selected economic criterion. We propose a strategy to tune tracking MPC schemes so as to locally approximate the behaviour of economic MPC while guaranteeing stability of the closed-loop system.

4th Meeting

Workshop on Model Predictive Control: Real-time implementation and new applications
at Institute for Systems Theory and Automatic Control, University of Stuttgart, Germany

Dates: August 29 - 31, 2016

Program [pdf]

Organizers: Matthias Müller and Karl Worthmann

Invited Speakers

Prof. Dr. Boris Houska (ShanghaiTech University)

Prof. Dr. Houska: Self-reflective model predictive control

This talk is about a novel control scheme, named self-reflective model predictive control, which takes its own limitations in the presence of process noise and measurement errors into account. In contrast to existing output-feedback MPC and persistently exciting MPC controllers, self-reflective MPC controllers do not only propagate a matrix-valued state forward in time in order to predict the variance of future state-estimates, but they also propagate a matrix-valued adjoint state backward in time. This adjoint state is used by the controller to compute and minimize a second order approximation of its own expected loss of control performance in the presence of random process noise and inexact state estimates. A second part of the talk introduces a real-time algorithm, which can exploit the particular structure of the self-reflective MPC problems in order to speed-up the online computation time. It is shown that, in contrast to generic state-of-the-art optimal control problem solvers, the proposed algorithm can solve the self-reflective optimization problems with reasonable additional computational effort compared to standard MPC. The advantages of the proposed real-time scheme are illustrated by applying it to a benchmark predator-prey-feeding control problem.

Yuning Jiang (ShanghaiTech University)

Yuning Jiang: Distributed Optimization and Control with ALADIN

Structured nonlinear optimization problems arise in a variety of control applications ranging from nonlinear model predictive control via robust control for uncertain processes to distributed nonlinear control of hybrid systems. Recently, the Augmented Lagrangian based Alternating Direction Inexact Newton (ALADIN) method has been proposed to solve non-convex distributed optimization problems to local optimality. After reviewing the main idea of ALADIN, this talk focusses on three applications. The first one is about a real-time variant of ALADIN, which can be used to solve nonlinear model predictive control problems with long horizons. The performance of this real-time variant of ALADIN is illustrated by applying it to a continuously stirred tank reactor. The second application is about coordinating autonomous vehicles at traffic intersections. Finally, a third application of ALADIN in the field stochastic robust control is introduced, where an ensemble of uncertainty scenarios is optimized. We show how ALADIN can be used to robustly control an exothermic tubular plug flow reactor.


Prof. Dr. Jürgen Pannek (Bremer Institut für Produktion und Logistik)

Prof. Dr. Pannek: Model Predictive Control as Enabler for (Inter-)Industrial Applications

As the potential of embedded control systems appears to be reached for many applications, information and communication technology (ICT) offers new possibilities to extend the potential of single applications in networks. Within the scope of Industrie 4.0 and Industrial Internet of Things, embedded systems have been opened up to cyber-physical systems (CPS), which allow for both governed and coordinated concepts of control. Until now, most of these ideas were applied in scenarios within one company or between competitors leading to cooperative and non cooperative control schemes. Within this presentation, we show how Model Predictive Control (MPC) can be used as a holistic approach on operational, tactical and strategic level. We illustrate these by examples from distributed material flow systems in in-house logistics, reconfigurable machine tools in job shop systems and human robot collaboration in manufacturing on operational level, the amazon problem in last mile transport as well as perishable goods in transocean transport on tactical level, and decision support in transport and manufacturing on strategic level. To conclude and extend the ideas and possibilities offered by MPC, ICT and CPS, we last show an application where the economical players are even from different industrial sectors. In particular, we show how mobility and energy networks can be combined to allow for synergies for both industrial sectors.

Do you want to participate? Please contact Dr.-Ing. Matthias Müller

Project description

In the context of autonomous mobile robots, the application of nonlinear model predictive control (MPC) has  been proposed for different control tasks such as set-point stabilization, trajectory tracking, and path following. For these application areas the question of sufficient stability conditions can be answered with the help of additional terminal constraints and / or costs. Since these additional (artificial) ingredients are restrictive as soon as obstacles or several mobile robots are taken into account, avoiding these stability enforcing constraints is of practical interest.

Within the scientific network our goal is to develop rigorous stability guarantees that require neither terminal regions nor terminal constraints. Herein, we follow a three stage program:
First, we focus on set point stabilization for continuous time formulations as well as for the respective sampled-data implementations. Secondly, based on the obtained results, we consider path following tasks in order to enable the mobile robots to circumvent static obstacles. And last, we investigate a setting with several mobile robots. Flanking this theoretical part, we plan to implement and test the proposed MPC schemes as well as measure their performance utilizing both simulations and prototypes.

The proposed project shall establish a collaboration platform for early stage researchers working on different aspects of MPC, which will allow for identifying and fostering future joint research projects.

Beyond Set Point Stabilization


Associated Members

5th Meeting

Workshop on Model Predictive Control

September 4 - 6, 2017, Chalmers University of Technology, Gothenburg, Sweden

Program [pdf]

Organizers: Boris Houska, Karl Worthmann and Mario Zanon



Meeting with Industry

  • Dr.-Ing. K.D. Listmann (ABB Corporate Research, Ladenburg)
  • Dr.-Ing. U. Münz (Siemens Corporate Technology, Princeton, NJ, USA)
  • Dr.-Ing. M. Reble (BASF Advanced Process Control, Ludwigshafen)
  • Dr.-Ing. M. Werling (BMW Research and Technology, München)

Do you want to participate? Please contact Dr.-Ing. Timm Faulwasser

Project description

The project serves as a stage for scientific interaction between up-and-coming scientists, who are devoted to current trends in model predictive control (MPC):

  • Distributed MPC of interconnected systems is concerned with automatic control of cyber physical systems. The different approaches are mainly distinguished with respect to the intensity of collaboration (level of cooperation).
  • Self-tuning MPC exhibits additional flexibility and, thus, enlarges the range of potential applications. This allows for fault-tolerant control and the adaptation of the algorithm during its runtime to realize the control task.

The study and analysis of interconnected, spatially distributed or differently coupled systems opens up new perspectives, which are, e.g., essential in multi-agent systems (mobile robots), in logistics from the viewpoint of Industrie 4.0 or within the control of renewable energy systems. Within this paradigm shift <em>economic MPC</em> plays a major role. Here, the stage costs are directly taken from the respective application. Hence, set points or periodic orbits are the outcome of the optimization and not set a priori as optimization objective.
Thereby, a key objective is to build up a network to tackle joint, innovative research projects, which are both scientifical sound and applicable to industrially relevant control problems.