dc.contributor.author	Meindl, Michael
dc.contributor.author	Lehmann, Dustin
dc.contributor.author	Seel, Thomas
dc.date.accessioned	2022-08-01T13:53:03Z
dc.date.available	2022-08-01T13:53:03Z
dc.date.issued	2022-07-12
dc.date.updated	2022-07-26T10:22:03Z
dc.description.abstract	This work addresses the problem of reference tracking in autonomously learning robots with unknown, nonlinear dynamics. Existing solutions require model information or extensive parameter tuning, and have rarely been validated in real-world experiments. We propose a learning control scheme that learns to approximate the unknown dynamics by a Gaussian Process (GP), which is used to optimize and apply a feedforward control input on each trial. Unlike existing approaches, the proposed method neither requires knowledge of the system states and their dynamics nor knowledge of an effective feedback control structure. All algorithm parameters are chosen automatically, i.e. the learning method works plug and play. The proposed method is validated in extensive simulations and real-world experiments. In contrast to most existing work, we study learning dynamics for more than one motion task as well as the robustness of performance across a large range of learning parameters. The method’s plug and play applicability is demonstrated by experiments with a balancing robot, in which the proposed method rapidly learns to track the desired output. Due to its model-agnostic and plug and play properties, the proposed method is expected to have high potential for application to a large class of reference tracking problems in systems with unknown, nonlinear dynamics.	en
dc.description.sponsorship	DFG, 390523135, EXC 2002: Science of Intelligence (SCIoI)	en
dc.identifier.eissn	2296-9144
dc.identifier.uri	https://depositonce.tu-berlin.de/handle/11303/17286
dc.identifier.uri	http://dx.doi.org/10.14279/depositonce-16066
dc.language.iso	en	en
dc.rights.uri	https://creativecommons.org/licenses/by/4.0/	en
dc.subject.ddc	004 Datenverarbeitung; Informatik	de
dc.subject.other	autonomous systems	en
dc.subject.other	Gaussian processes	en
dc.subject.other	iterative learning control	en
dc.subject.other	nonlinear systems	en
dc.subject.other	reinforcement learning	en
dc.subject.other	robot learning	en
dc.title	Bridging Reinforcement Learning and Iterative Learning Control: Autonomous Motion Learning for Unknown, Nonlinear Dynamics	en
dc.type	Article	en
dc.type.version	publishedVersion	en
dcterms.bibliographicCitation.articlenumber	793512	en
dcterms.bibliographicCitation.doi	10.3389/frobt.2022.793512	en
dcterms.bibliographicCitation.journaltitle	Frontiers in Robotics and AI	en
dcterms.bibliographicCitation.originalpublishername	Frontiers	en
dcterms.bibliographicCitation.originalpublisherplace	Lausanne	en
dcterms.bibliographicCitation.volume	9	en
tub.accessrights.dnb	free	en
tub.affiliation	Fak. 4 Elektrotechnik und Informatik::Inst. Energie- und Automatisierungstechnik::FG Regelungssysteme	de
tub.affiliation.faculty	Fak. 4 Elektrotechnik und Informatik	de
tub.affiliation.group	FG Regelungssysteme	de
tub.affiliation.institute	Inst. Energie- und Automatisierungstechnik	de
tub.publisher.universityorinstitution	Technische Universität Berlin	en

Bridging Reinforcement Learning and Iterative Learning Control: Autonomous Motion Learning for Unknown, Nonlinear Dynamics

Files

Original bundle

License bundle

Collections