Loading...
Please wait, while we are loading the content...
Similar Documents
A Safe Learning Framework based on Hamilton-Jacobi-Isaacs Reachability
| Content Provider | Semantic Scholar |
|---|---|
| Author | Fernandez-Fisac, Jaime Tomlin, Claire J. |
| Copyright Year | 2015 |
| Abstract | Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission. would like to also thank my advisor, Claire J. Tomlin, for her support, guidance, and patience. The autonomy she has given me in developing and exploring my own research ideas has contributed significantly to my growth. Lastly, I want to thank my parents, Anayo and Josephine Akametalu. Moving from Nigeria to the U.S. in 1994 certainly had its fair share of growing pains, but from school to sports my parents always afforded me the best opportunities. Second Reader Date 1 Abstract Reinforcement learning for robotic applications faces the challenge of constraint satisfaction, which currently impedes its application to safety critical systems. Recent approaches successfully introduce safety based on reachability analysis, determining a safe region of the state space where the system can operate. However, overly constraining the freedom of the system can negatively affect performance, while attempting to learn less conservative safety constraints might fail to preserve safety if the learned constraints are inaccurate. We propose a novel method that uses a principled approach to learn the system's unknown dynamics based on a Gaussian process model and iteratively approximates the maximal safe set. A modified control strategy based on real-time model validation preserves safety under weaker conditions than current approaches. Our framework further incorporates safety into the reinforcement learning performance metric, allowing a better integration of safety and learning. We demonstrate our algorithm on simulations of a cart-pole system and on an experimental quadrotor application and show how our proposed scheme succeeds in preserving safety where current approaches fail to avoid an unsafe condition. 2 Acknowledgements Though my journey in graduate school is not yet complete, I have grown a lot in my time here at Berkeley. In my two and a half years I have become more confident in proposing, conducting, and presenting research. I would be remiss in not taking time to acknowledge those that have contributed to my progress. There are a number of people that I owe my gratitude, but for my first thesis I will keep this … |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://digitalassets.lib.berkeley.edu/techreports/ucb/text/EECS-2015-2.pdf |
| Alternate Webpage(s) | http://www.eecs.berkeley.edu/Pubs/TechRpts/2015/EECS-2015-2.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
