Local Bayesian optimization for controller tuning with crash constraints

Controller tuning is crucial for closed-loop performance but often involves manual adjustments. Although Bayesian optimization (BO) has been established as a data-efficient method for automated tuning, applying it to large and high-dimensional search spaces remains challenging. We extend a recently...

Full description

Saved in:
Bibliographic Details
Published inAutomatisierungstechnik : AT Vol. 72; no. 4; pp. 281 - 292
Main Authors von Rohr, Alexander, Stenger, David, Scheurenberg, Dominik, Trimpe, Sebastian
Format Journal Article
LanguageEnglish
Published De Gruyter 25.04.2024
Subjects
Bayesian optimization
Bayes’sche Optimierung
controller tuning
learning-based control
lernende Regler
Reglereinstellung
Online AccessGet full text

Cover

More Information
Summary:Controller tuning is crucial for closed-loop performance but often involves manual adjustments. Although Bayesian optimization (BO) has been established as a data-efficient method for automated tuning, applying it to large and high-dimensional search spaces remains challenging. We extend a recently proposed local variant of BO to include crash constraints, where the controller can only be successfully evaluated in an unknown feasible region. We demonstrate the efficiency of the proposed method through simulations and hardware experiments. Our findings showcase the potential of local BO to enhance controller performance and reduce the time and resources necessary for tuning.
ISSN:0178-2312
2196-677X
DOI:10.1515/auto-2023-0181