Safe Learning in Control – Leveraging Online Data for Performance with System Theoretic Guarantees
Event details
| Date | 27.03.2015 |
| Hour | 10:00 › 11:00 |
| Speaker |
Dr. Melanie Zeilinger, Berkeley, California, USA Bio : Melanie Zeilinger is a Postdoctoral Researcher in a joint program with the Empirical Inference Department at the Max Planck Institute for Intelligent Systems in Tuebingen, Germany and the Department of Electrical Engineering and Computer Sciences at the University of California, Berkeley. From 2011-2012 she was a postdoctoral fellow at the École Polytechnique Fédérale de Lausanne (EPFL) in Switzerland. She received the Ph.D. degree in Electrical Engineering from ETH Zurich in Switzerland in 2011, and the diploma in Engineering Cybernetics from the University of Stuttgart in Germany in 2006. She conducted her diploma thesis research at the University of California at Santa Barbara in 2005-2006. She received the ETH medal for her dissertation in 2012 and was awarded a Marie Curie Fellowship for Career Development by the European Commission in 2011. Her research interests revolve around distributed control and optimization, as well as safe learning-based control, with applications to energy distribution and management systems and human-in-the-loop control. |
| Location | |
| Category | Conferences - Seminars |
Abstract: Demanding performance requirements combined with increasing complexity, uncertainty and human interaction in many emerging application problems, e.g. in robotic, transportation, or power systems, are pushing traditional control methods to their limits. A new opportunity to address these challenges is offered by sensor technologies with the ability to collect large amounts of data online. While machine learning provides powerful techniques to analyze and utilize such large-scale data, safety concerns when integrating them in a closed-loop, automated decision-making process represent a key limitation for leveraging their potential.
In this talk, we will discuss some of our recent work towards an automatic controller synthesis that utilizes online data to enhance system performance, while ensuring satisfaction of safety conditions at all times. We show how a predictive controller can be systematically tailored to the particular system at hand by improving predictions, quantifying uncertainties and/or tailoring the objective function online based on data, providing a high performance controller with reduced development times. Then, a safety wrapper is introduced that exploits reachability analysis to ensure satisfaction of constraints for any online control scheme. The key novelty is the learning capability of the wrapper itself, utilizing data to find the largest region of safe operation where a performance-maximizing controller can be employed. Finally, experimental results are shown for a quad-rotor safely learning to fly.
In this talk, we will discuss some of our recent work towards an automatic controller synthesis that utilizes online data to enhance system performance, while ensuring satisfaction of safety conditions at all times. We show how a predictive controller can be systematically tailored to the particular system at hand by improving predictions, quantifying uncertainties and/or tailoring the objective function online based on data, providing a high performance controller with reduced development times. Then, a safety wrapper is introduced that exploits reachability analysis to ensure satisfaction of constraints for any online control scheme. The key novelty is the learning capability of the wrapper itself, utilizing data to find the largest region of safe operation where a performance-maximizing controller can be employed. Finally, experimental results are shown for a quad-rotor safely learning to fly.
Practical information
- Informed public
- Free
Organizer
- IGM-GE
Contact
- Géraldine Palaj