TU Darmstadt / ULB / TUbiblio

Optimal Control and Inverse Optimal Control by Distribution Matching

Arenz, Oleg ; Abdulsamad, Hany ; Neumann, Gerhard (2022)
Optimal Control and Inverse Optimal Control by Distribution Matching.
2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). Daejeon, Korea (09.-14.10.2016)
doi: 10.26083/tuprints-00022929
Konferenzveröffentlichung, Zweitveröffentlichung, Postprint

Kurzbeschreibung (Abstract)

Optimal control is a powerful approach to achieve optimal behavior. However, it typically requires a manual specification of a cost function which often contains several objectives, such as reaching goal positions at different time steps or energy efficiency. Manually trading-off these objectives is often difficult and requires a high engineering effort. In this paper, we present a new approach to specify optimal behavior. We directly specify the desired behavior by a distribution over future states or features of the states. For example, the experimenter could choose to reach certain mean positions with given accuracy/variance at specified time steps. Our approach also unifies optimal control and inverse optimal control in one framework. Given a desired state distribution, we estimate a cost function such that the optimal controller matches the desired distribution. If the desired distribution is estimated from expert demonstrations, our approach performs inverse optimal control. We evaluate our approach on several optimal and inverse optimal control tasks on non-linear systems using incremental linearizations similar to differential dynamic programming approaches.

Typ des Eintrags: Konferenzveröffentlichung
Erschienen: 2022
Autor(en): Arenz, Oleg ; Abdulsamad, Hany ; Neumann, Gerhard
Art des Eintrags: Zweitveröffentlichung
Titel: Optimal Control and Inverse Optimal Control by Distribution Matching
Sprache: Englisch
Publikationsjahr: 2022
Ort: Darmstadt
Verlag: IEEE
Buchtitel: 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
Kollation: 14 ungezählte Seiten
Veranstaltungstitel: 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
Veranstaltungsort: Daejeon, Korea
Veranstaltungsdatum: 09.-14.10.2016
DOI: 10.26083/tuprints-00022929
URL / URN: https://tuprints.ulb.tu-darmstadt.de/22929
Zugehörige Links:
Herkunft: Zweitveröffentlichungsservice
Kurzbeschreibung (Abstract):

Optimal control is a powerful approach to achieve optimal behavior. However, it typically requires a manual specification of a cost function which often contains several objectives, such as reaching goal positions at different time steps or energy efficiency. Manually trading-off these objectives is often difficult and requires a high engineering effort. In this paper, we present a new approach to specify optimal behavior. We directly specify the desired behavior by a distribution over future states or features of the states. For example, the experimenter could choose to reach certain mean positions with given accuracy/variance at specified time steps. Our approach also unifies optimal control and inverse optimal control in one framework. Given a desired state distribution, we estimate a cost function such that the optimal controller matches the desired distribution. If the desired distribution is estimated from expert demonstrations, our approach performs inverse optimal control. We evaluate our approach on several optimal and inverse optimal control tasks on non-linear systems using incremental linearizations similar to differential dynamic programming approaches.

Freie Schlagworte: Optimal control, Entropy, Heuristic algorithms, Trajectory, Cost function, Learning (artificial intelligence)
Status: Postprint
URN: urn:nbn:de:tuda-tuprints-229290
Sachgruppe der Dewey Dezimalklassifikatin (DDC): 000 Allgemeines, Informatik, Informationswissenschaft > 004 Informatik
Fachbereich(e)/-gebiet(e): 20 Fachbereich Informatik
20 Fachbereich Informatik > Intelligente Autonome Systeme
Hinterlegungsdatum: 25 Nov 2022 12:51
Letzte Änderung: 08 Aug 2023 12:13
PPN: 503350850
Zugehörige Links:
Export:
Suche nach Titel in: TUfind oder in Google
Frage zum Eintrag Frage zum Eintrag

Optionen (nur für Redakteure)
Redaktionelle Details anzeigen Redaktionelle Details anzeigen