Systems research

Dynamical systems: reduction, optimization and control

The world is organized in complex systems. We aim to reduce their complexity and to optimize their control. For this purpose, we simplify models and design communication infrastructures.

 (c) David Ausserhofer
From automated control engineering: controlling two robots(Foto: David Ausserhofer).

Chaos and order

In observing termites, it is not immediately apparent that behind the supposed chaos hides a highly coordinated system of interactions and interconnections. Like termites that communicate via tap signals and scents, other types of autonomous agents – ranging from bacteria to swarm robots, smart grids or integrated manufacturing plants – also make use of complex and decentralized control and communication infrastructures and hence the ability to cope with changing circumstances “on the fly,” i.e., while carrying out a specific task.

In Project Network 3 (PN 3), our research focuses on developing methodologies and techniques that can handle complex spatial interconnections of dynamical systems encountered in the natural and engineering sciences. Modeling, analysis and controller design for such heterogeneously distributed interconnections together comprise our project network’s theme.

Dynamical feedback control

We emphasize the synthesis of structured dynamic feedback controllers for overall optimality and robustness of the controlled network. This includes designing controller components and their communication infrastructure – to the extent possible with self-adapting capabilities so they can cope with adverse circumstances or changing network settings. These aspects are also addressed during the implementation phase.

Tools for model reduction

Further, we develop tools for structured model reduction that play a key role in the efficient simulation of complex networks of systems and the subsequent optimal design of distributed controllers. The key issue here is reducing or redesigning models of subcomponents or of interconnection topologies. Both are supported by error quantification that takes into consideration global performance objectives or uncertainty.

Our principal objectives:

  • Develop powerful methods and tools for a structured reduction of models within coupled and distributed systems;
  • pursue systematic approaches to optimizing the distributed control of interacting heterogeneous systems.
+-

3-1
A Geometric Approach to Distributed Control of Heterogeneous Networked Systems
Project Coordinator: Prof. Dr.-Ing. Frank Allgöwer
Research Associate: Georg Seyboth
Institute for Systems Theory and Automatic Control

3-2
Distributed Control of Networked Systems by Convex Optimization
Project Coordinator: Prof. Dr. Carsten Scherer
Research Associate: Tobias Holicki
Institute for Mathematical Methods in the Engineering Sciences

3-3
Distributed Formation Control of Swarm Robots Using Ideas from Bacteria Populations
Project Coordinator: Prof. Dr.-Ing. Peter Eberhard
Research Associate: Dr.-Ing. Qirong Tong
Institute of Engineering and Computational Mechanics

3-4
Do separately reduced subsystems yield an optimally described global behaviour?
Project Coordinator: Prof. Dr.-Ing. Peter Eberhard
Mitarbeiter: Nicolai Wengert
Institute of Engineering and Computational Mechanics

3-5
Feedback Control of Parametric PDEs with Reduced Basis Surrogate Models
Projecrt Coordinator: Prof. Dr. Bernard Haasdonk
Research Associate: Andreas Schmidt
Institute for Applied Analysis und Numerical Simulation

3-6
Identification and Modelling of the Pose-Dependent Dynamic Behaviour of Large Machine Tool Structures in Consideration of Machining Processes
Project Coordinator: Prof. Dr.-Ing. Alexander Verl
Research Associate: Stefanie Apprich
Institute for Control Engineering of Machine Tools and Manufacturing Units

3-7
Improved Modelling of Cables for Kinematics and Dynamics of Leight-weight Robots
Project Coordinator: JP Dr.-Ing. Andreas Pott
Research Associate: Philipp Tempel
Institute for Control Engineering of Machine Tools and Manufacturing Units

3-8
Elastic Multibody Dynamics of Humans in Vehicle Safety
Project Coordinator: JP Dr.-Ing. Jörg Fehr
Research Associate: Christian Kleinbach
Institute of Engineering and Computational Mechanics

3-9 (completed)
Cooperative Behaviour of Autonomous Agents
Project Coordinator: Prof. Dr.-Ing. Frank Allgöwer
Research Associate: Matthias Müller
Institute for Systems Theory and Automatic Control

3-10
Error controlled nonlinear model reduction techniques for crash simulations
Project Coordinator: JP Dr.-Ing. Jörg Fehr
Research Associate: Dennis Grunert
Institute of Engineering and Computational Mechanics

3-11 (completed)
Model-based Identification and Active Suppression of Static and Dynamic Aberrations in High Performance Optics by Combination of Mechanical and Optical Simulation
Project Coordinator: Prof. Dr. Wolfgang Osten
Research Associate: Holger Gilbergs
Institute of Applied Optics

  • Engineering
  • Mathematics

Coordinators PN 3

Dieses Bild zeigt Allgöwer
Prof. Dr.-Ing.

Frank Allgöwer

Head of Graduate School, Coordinator Project Network 3, Coordinator Research Area C

Dieses Bild zeigt Haasdonk
Prof. Dr.

Bernard Haasdonk

Coordinator Project Network 3