Safe risk-averse bayesian optimization for controller tuning

König, Christopher; Ozols, Miks; Makarova, Anastasia; Balta, Efe C.; Krause, Andreas; Rupenyan-Vasileva, Alisa

doi:10.1109/LRA.2023.3325991

Please use this identifier to cite or link to this item: https://doi.org/10.21256/zhaw-29008

Publication type:	Article in scientific journal
Type of review:	Peer review (publication)
Title:	Safe risk-averse bayesian optimization for controller tuning
Authors:	König, Christopher Ozols, Miks Makarova, Anastasia Balta, Efe C. Krause, Andreas Rupenyan-Vasileva, Alisa
et. al:	No
DOI:	10.1109/LRA.2023.3325991 10.21256/zhaw-29008
Published in:	IEEE Robotics and Automation Letters
Issue Date:	13-Oct-2023
Publisher / Ed. Institution:	IEEE
ISSN:	2377-3766 2377-3774
Language:	English
Subjects:	Optimization; Bayesian method; Artificial Intelligence; Probabilistic model; Risk-averse Bayesian optimization; Heteroscedastic noise
Subject (DDC):	006: Special computer methods
Abstract:	Controller tuning and parameter optimization are crucial in system design to improve both the controller and underlying system performance. Bayesian optimization has been established as an efficient model-free method for controller tuning and adaptation. Standard methods, however, are not enough for high-precision systems to be robust with respect to unknown input-dependent noise and stable under safety constraints. In this work, we present a novel data-driven approach, RAGoOSe, for safe controller tuning in the presence of heteroscedastic noise, combining safe learning with risk-averse Bayesian optimization. We demonstrate the method for synthetic benchmark and compare its performance to established BO-based tuning methods. We further evaluate RaGoose performance on a real precision-motion system utilized in semiconductor industry applications and compare it to the built-in auto-tuning routine.
URI:	https://digitalcollection.zhaw.ch/handle/11475/29008
Fulltext version:	Accepted version
License (according to publishing contract):	Licence according to publishing contract
Departement:	School of Engineering
Organisational Unit:	Centre for Artificial Intelligence (CAI)
Appears in collections:	Publikationen School of Engineering

Files in This Item:

File	Description	Size	Format
2023_Koenig-etal_Safe-risk-averse-bayesian-optimization-for-controller-tuning_IEEE_AAM.pdf	Accepted Version	5.07 MB	Adobe PDF	View/Open

Show full item record

König, C., Ozols, M., Makarova, A., Balta, E. C., Krause, A., & Rupenyan-Vasileva, A. (2023). Safe risk-averse bayesian optimization for controller tuning. IEEE Robotics and Automation Letters. https://doi.org/10.1109/LRA.2023.3325991

König, C. et al. (2023) ‘Safe risk-averse bayesian optimization for controller tuning’, IEEE Robotics and Automation Letters [Preprint]. Available at: https://doi.org/10.1109/LRA.2023.3325991.

C. König, M. Ozols, A. Makarova, E. C. Balta, A. Krause, and A. Rupenyan-Vasileva, “Safe risk-averse bayesian optimization for controller tuning,” IEEE Robotics and Automation Letters, Oct. 2023, doi: 10.1109/LRA.2023.3325991.

KÖNIG, Christopher, Miks OZOLS, Anastasia MAKAROVA, Efe C. BALTA, Andreas KRAUSE und Alisa RUPENYAN-VASILEVA, 2023. Safe risk-averse bayesian optimization for controller tuning. IEEE Robotics and Automation Letters. 13 Oktober 2023. DOI 10.1109/LRA.2023.3325991

König, Christopher, Miks Ozols, Anastasia Makarova, Efe C. Balta, Andreas Krause, and Alisa Rupenyan-Vasileva. 2023. “Safe Risk-Averse Bayesian Optimization for Controller Tuning.” IEEE Robotics and Automation Letters, October. https://doi.org/10.1109/LRA.2023.3325991.

König, Christopher, et al. “Safe Risk-Averse Bayesian Optimization for Controller Tuning.” IEEE Robotics and Automation Letters, Oct. 2023, https://doi.org/10.1109/LRA.2023.3325991.