Path Following – On Tailored Normal Forms and NMPC Design

Timm Faulwasser

Institut für Angewandte Informatik, Karlsruher Institut für Technologie; Laboratoire d’Automatique, École Polytechnique Fédérale de Lausanne

Wednesday, May 11, 2016, 11:00

Room 02-012, Georges-Köhler Allee 102, Freiburg 79110, Germany

Abstract
Recently, path-following formulations have been suggested as alternatives to the classic control problems of set-point stabilization and trajectory tracking. While set-point stabilization refers to the task of stabilizing a fixed point in the state space, trajectory tracking describes the design of controllers that ensure tracking of a time-dependent reference signals. However, not all control tasks arising in applications fit well into the framework of stabilization and tracking. For instance, consider tasks such as steering an autonomous vehicle along a given reference track, precise machine tooling or control of autonomous underwater vehicles. All these applications have in common that a system should be steered along a pre-specified geometric curve in an (position) output space, whereby the speed to move along the curve is not fixed a priori.
In this talk, we give an overview of recent progress on optimization-based solutions to constrained path-following problems. We present a geometric problem analysis, which enables different optimization-based solution approaches. Furthermore, we show that this analysis simplifies the design of predictive controllers and the computation of optimal feed-forward controls. We present experimental results from different applications: predictive path-following control of a robot arm and feed-forward path following for an autonomous helicopter. The talk concludes with an outlook on open questions.

References
Faulwasser, T. & Findeisen, R. Nonlinear model predictive control for constrained output path following; IEEE Trans. Automat. Contr., 2016, 61, 1026-1039
Faulwasser, T.; Hagenmeyer, V. & Findeisen, R. Constrained Reachability and Trajectory Generation for Flat Systems; Automatica, 2014, 50, 1151-1159
Faulwasser, T.; Weber, T.; Zometa, J. & Findeisen, R. Implementation of Constrained Nonlinear Model Predictive Path-Following Control for an Industrial Robot; arXiv:1506.09084, 2015
Dauer, J.; Faulwasser, T. & Lorenz, S. Run-to-Run Disturbance Rejection for Feedforward Path Following of an Adaptively Controlled Unmanned Helicopter; Control Applications (CCA), IEEE Conference on, 2015, 1779-1785
Dauer, J.; Faulwasser, T.; Lorenz, S. & Findeisen, R. Optimization-based Feedforward Path Following for Model Reference Adaptive Control of an Unmanned Helicopter; Proc. of AIAA Guidance, Navigation and Control Conference, 2013

Short CV
Timm Faulwasser has received a diploma (master equivalent) in Engineering Cybernetics from the University Stuttgart in 2006. In 2007 he joined the group of Rolf Findeisen at the Institute of Automation Engineering at the Otto-von-Guericke University Magdeburg. From 2008-2012 he was a member of the International Max Planck Research School for Analysis, Design and Optimization in Chemical and Biochemical Process Engineering Magdeburg. He obtained his PhD from the Faculty of Electrical Engineering and Information Engineering, Otto-von-
Guericke University Magdeburg in 2012. Since 2013 he is working as a Postdoc at the Laboratoire d’Automatique, Ecole Polytechnique Fédérale de Lausanne, Switzerland in the group of Dominique Bonvin. Since April 2015 he is also leading the Advanced Control Group at the Institut für Angewandte Informatik at Karlsruher Institut für Technologie.
His main research interests are optimization-based and predictive control of nonlinear systems, path-following and trajectory-tracking problems, nonlinear control and real-time optimization with applications in energy systems, aeronautics, mechatronics/robotics, physics and process engineering.