Download Applied Nonlinear Control Workshop: Presentations and Abstracts and more Study notes Design in PDF only on Docsity! Applied Nonlinear Control Workshop Saarbrücken Friday, May 28, 2010 Hosted by Prof. J. Rudolph Chair of Systems Theory and Control Engineering (Lehrstuhl für Systemtheorie und Regelungstechnik) Universität des Saarlandes Purpose The Applied Nonlinear Control Workshop aims to assem- ble experts and colleagues in the nonlinear control field to present and discuss current research activities. This in- formal event aims to further promote the national and in- ternational collaboration that exists among the attendees. The workshop is timed to benefit from Prof. Alan Lynch’s visit to Germany. Workshop Overview The event begins at 10:00 on Friday, May 28 with a wel- come address by Prof. J. Rudolph. He will also introduce the research activities of the Chair of Systems Theory and Control Engineering. Talks will be held throughout the day in Building A5 1 on the University of Saarland campus. Sessions are in room −1.22 (basement) before lunch and in room −1.03 (base- ment) after lunch. Coffee and lunch will be provided. Tentative Schedule Room −1.22 10:00 – 10:30 Welcome and Introduction (J. Rudolph) 10:30 – 11:00 T.M.P. Nguyen 11:00 – 11:15 Coffee 11:15 – 11:45 H. Fehr 11:45 – 12:15 S. K. Nguang 12:15 – 12:45 A. Lynch 12:45 – 14:00 Lunch and Discussions Room −1.03 14:00 – 14:30 M. Zeitz 14:30 – 15:00 J. Raisch 15:00 – 15:30 K. Listmann 15:30 – 16:00 Break/Discussions 16:00 – 16:30 A. Gensior 16:30 – 17:00 M. Barczyk 17:00 – 17:30 D. Kastelan Abstracts Magnetometer-plus-GPS-aided Inertial Navigation System Design for a Helicopter UAV M.Sc. Martin Barczyk (University of Alberta) We consider the Extended Kalman Filter (EKF)-based aided navigation system design used on the University of Alberta’s ANCL Helicopter UAV. We review the math- ematics of coordinate frame transformations, attitude parametrization and rigid-body kinematics. We cover the modeling and identification of the on-board inertial sensors corrupted by time-varying bias and white noise terms. We focus on the magnetometer and describe a novel calibration procedure applied to this sensor, experimentally evaluating its performance versus a conventional calibration and the uncalibrated case. We describe the topology of the aided navigation system and cover the details of the high-rate integration of the navigation dynamics, the low-rate cor- rection of the estimates using an EKF observer, and the initialization stage of the system. The INS performance in experiment is demonstrated for the cases of ground and flight testing, focusing on the issues of heading angle ob- servability and the importance of calibration. Flatness Based Switching Frequency Estimation of Sliding Mode Controllers for Single-Input Systems Dipl.-Ing. Hendrik Fehr (TU Dresden) A method is presented that allows an estimation of the switching frequency when sliding mode controllers are used for the control of single-input flat systems. The estimation of the switching frequency is done with respect to nominal trajectories for the flat output. That means, no simulation is required - even for non stationary trajectories - which is considered an outstanding feature of the proposed method. In order to illustrate the efficiency of the method a slid- ing mode controller for a boost converter is considered as an example. Finally the switching frequency estimation is compared to simulation and measurement results. Flatness Based Energy Optimal Control of a Dou- bly Fed Induction Generator System Dr.-Ing. Albrecht Gensior (TU Dresden) A system consisting of a doubly fed induction generator is considered where the stator windings are connected to the grid and the rotor windings are fed by a converter. The lat- ter shares the same DC-link with a second converter con- nected to the grid. Using this conguration it is possible to provide reactive power to the grid by both, the doubly fed machine and the grid connected converter. Using a flatness based approach this freedom is exploited to minimize the total power loss in the system. The findings are supported by an example. Inertial Sensor Model Identification Using Allan Variance M.Sc. David Kastelan (Universität des Saarlandes) The Allan variance may be used to identify and character- ize non-deterministic processes that affect the stability of a signal. This talk applies this analysis technique to drift test data from a Microstrain 3DM-GX1 inertial measure- ment unit (IMU). Allan variance expressions are outlined
