Learning Disturbances Online for Risk-Aware Control: Risk-Aware Flight with Less Than One Minute of Data
Abstract
Recent advances in safety-critical risk-aware control are predicated on apriori knowledge of the disturbances a system might face. This paper proposes a method to efficiently learn these disturbances online, in a risk-aware context. First, we introduce the concept of a Surface-at-Risk, a risk measure for stochastic processes that extends Value-at-Risk -- a commonly utilized risk measure in the risk-aware controls community. Second, we model the norm of the state discrepancy between the model and the true system evolution as a scalar-valued stochastic process and determine an upper bound to its Surface-at-Risk via Gaussian Process Regression. Third, we provide theoretical results on the accuracy of our fitted surface subject to mild assumptions that are verifiable with respect to the data sets collected during system operation. Finally, we experimentally verify our procedure by augmenting a drone's controller and highlight performance increases achieved via our risk-aware approach after collecting less than a minute of operating data.
Additional Information
Attribution 4.0 International (CC BY 4.0). The work of Prithvi Akella was supported by the Air Force Office of Scientific Research, grant FA9550-19-1-0302, and the National Science Foundation, grant 1932091. The work of Skylar Wei was supported in part by DARPA, through the Learning and Introspective Control program. We would also like to thank the Caltech Center for Autonomous Systems and Technologies for the use of the wind tunnel in our experiments.Attached Files
Accepted Version - 2212.06253.pdf
Files
Name | Size | Download all |
---|---|---|
md5:7097f1448570b18b48847361c9842586
|
5.3 MB | Preview Download |
Additional details
- Eprint ID
- 118471
- Resolver ID
- CaltechAUTHORS:20221219-234112304
- Air Force Office of Scientific Research (AFOSR)
- FA9550-19-1-0302
- NSF
- CNS-1932091
- Defense Advanced Research Projects Agency (DARPA)
- Created
-
2022-12-21Created from EPrint's datestamp field
- Updated
-
2023-06-02Created from EPrint's last_modified field
- Caltech groups
- Center for Autonomous Systems and Technologies (CAST)