Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Gottschalk, Simon, Lanza, Lukas, Worthmann, Karl, Lux-Gottschalk, Kerstin
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2402.08306
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866909271333011456
author Gottschalk, Simon
Lanza, Lukas
Worthmann, Karl
Lux-Gottschalk, Kerstin
author_facet Gottschalk, Simon
Lanza, Lukas
Worthmann, Karl
Lux-Gottschalk, Kerstin
contents We propose a two-component data-driven controller to safely perform docking maneuvers for satellites. Reinforcement Learning is used to deduce an optimal control policy based on measurement data. To safeguard the learning phase, an additional feedback law is implemented in the control unit, which guarantees the evolution of the system within predefined performance bounds. We define safe and safety-critical areas to train the feedback controller based on actual measurements. To avoid chattering, a dwell-time activation scheme is implemented. We provide numerical evidence for the performance of the proposed controller for a satellite docking maneuver with collision avoidance.
format Preprint
id arxiv_https___arxiv_org_abs_2402_08306
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Reinforcement Learning for Docking Maneuvers with Prescribed Performance
Gottschalk, Simon
Lanza, Lukas
Worthmann, Karl
Lux-Gottschalk, Kerstin
Optimization and Control
We propose a two-component data-driven controller to safely perform docking maneuvers for satellites. Reinforcement Learning is used to deduce an optimal control policy based on measurement data. To safeguard the learning phase, an additional feedback law is implemented in the control unit, which guarantees the evolution of the system within predefined performance bounds. We define safe and safety-critical areas to train the feedback controller based on actual measurements. To avoid chattering, a dwell-time activation scheme is implemented. We provide numerical evidence for the performance of the proposed controller for a satellite docking maneuver with collision avoidance.
title Reinforcement Learning for Docking Maneuvers with Prescribed Performance
topic Optimization and Control
url https://arxiv.org/abs/2402.08306