TU Darmstadt / ULB / TUbiblio

Reinforcement learning of optimal active particle navigation

Nasiri, Mahdi ; Liebchen, Benno (2022)
Reinforcement learning of optimal active particle navigation.
In: New Journal of Physics, 24 (7)
doi: 10.1088/1367-2630/ac8013
Artikel, Bibliographie

Dies ist die neueste Version dieses Eintrags.

Kurzbeschreibung (Abstract)

The development of self-propelled particles at the micro- and the nanoscale has sparked a huge potential for future applications in active matter physics, microsurgery, and targeted drug delivery. However, while the latter applications provoke the quest on how to optimally navigate towards a target, such as e.g. a cancer cell, there is still no simple way known to determine the optimal route in sufficiently complex environments. Here we develop a machine learning-based approach that allows us, for the first time, to determine the asymptotically optimal path of a self-propelled agent which can freely steer in complex environments. Our method hinges on policy gradient-based deep reinforcement learning techniques and, crucially, does not require any reward shaping or heuristics. The presented method provides a powerful alternative to current analytical methods to calculate optimal trajectories and opens a route towards a universal path planner for future intelligent active particles.

Typ des Eintrags: Artikel
Erschienen: 2022
Autor(en): Nasiri, Mahdi ; Liebchen, Benno
Art des Eintrags: Bibliographie
Titel: Reinforcement learning of optimal active particle navigation
Sprache: Englisch
Publikationsjahr: 2022
Ort: Darmstadt
Verlag: IOP Publishing
Titel der Zeitschrift, Zeitung oder Schriftenreihe: New Journal of Physics
Jahrgang/Volume einer Zeitschrift: 24
(Heft-)Nummer: 7
Kollation: 7 Seiten
DOI: 10.1088/1367-2630/ac8013
Zugehörige Links:
Kurzbeschreibung (Abstract):

The development of self-propelled particles at the micro- and the nanoscale has sparked a huge potential for future applications in active matter physics, microsurgery, and targeted drug delivery. However, while the latter applications provoke the quest on how to optimally navigate towards a target, such as e.g. a cancer cell, there is still no simple way known to determine the optimal route in sufficiently complex environments. Here we develop a machine learning-based approach that allows us, for the first time, to determine the asymptotically optimal path of a self-propelled agent which can freely steer in complex environments. Our method hinges on policy gradient-based deep reinforcement learning techniques and, crucially, does not require any reward shaping or heuristics. The presented method provides a powerful alternative to current analytical methods to calculate optimal trajectories and opens a route towards a universal path planner for future intelligent active particles.

Freie Schlagworte: active matter physics, colloids, soft matter physics, microswimmers, optimal navigation, reinforcement learning, optimization
Sachgruppe der Dewey Dezimalklassifikatin (DDC): 500 Naturwissenschaften und Mathematik > 530 Physik
Fachbereich(e)/-gebiet(e): 05 Fachbereich Physik
05 Fachbereich Physik > Institut für Physik Kondensierter Materie (IPKM)
Hinterlegungsdatum: 02 Aug 2024 12:42
Letzte Änderung: 02 Aug 2024 12:42
PPN:
Export:
Suche nach Titel in: TUfind oder in Google

Verfügbare Versionen dieses Eintrags

Frage zum Eintrag Frage zum Eintrag

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