Simple and reliable implementation of optimization-based controllers
Event details
| Date | 31.01.2014 |
| Hour | 10:00 |
| Speaker |
Prof. Michal Kvasnica, STU, Bratislava Bio: Michal Kvasnica was born in 1977 in Slovakia. He obtained his MSc degree in process control in 2000 from the Slovak University of Technology (STU) in Bratislava, Slovakia, and a PhD in electrical engineering in 2008 from ETH Zurich. Currently he is an associate professor at the Institute of Information Engineering, Automation, and Mathematics at STU in Bratislava. His research interests include model predictive control, modeling, control and analysis of hybrid systems, and development of software tools for control. He co-develops the Multi-Parametric Toolbox and HYSDEL packages. |
| Location | |
| Category | Conferences - Seminars |
Parametric optimization allows to implement model predictive control (MPC) strategies on a simple hardware, yet provide guarantees of worst-case execution time. This talk reviews the potential and main bottlenecks of MPC controllers synthesized by parametric optimization. We show how such controllers are constructed, analyzed and implemented. Specific attention will be devoted to simplification of such controllers to achieve low memory footprint. We will review several methods that allow to target hardware platforms with severe limitations on memory storage. Finally, we discuss rigorous certification schemes which provide guarantees that the closed-loop system operates correctly even when control commands are quantized.
Practical information
- Expert
- Free