Decision making under uncertainty can be framed as a partially observable Markov decision process (POMDP). Finding exact solutions of POMDPs is generally computationally intractable, but the solution can be approximated by sampling-based approaches. These sampling-based POMDP solvers rely on multi-armed bandit (MAB) heuristics, which assume the outcomes of different actions to be uncorrelated. In some applications, like motion planning in continuous spaces, similar actions yield similar outcomes. In this paper, we utilize variants of MAB heuristics that make Lipschitz continuity assumptions on the outcomes of actions to improve the efficiency of sampling-based planning approaches. We demonstrate the effectiveness of this approach in the context of motion planning for automated driving.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Efficient Sampling in POMDPs with Lipschitz Bandits for Motion Planning in Continuous Spaces


    Beteiligte:
    Tas, Omer Sahin (Autor:in) / Hauser, Felix (Autor:in) / Lauer, Martin (Autor:in)


    Erscheinungsdatum :

    11.07.2021


    Format / Umfang :

    1684411 byte




    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Adaptive Discretization Using Voronoi Trees for Continuous-Action POMDPs

    Hoerger, Marcus / Kurniawati, Hanna / Kroese, Dirk et al. | Springer Verlag | 2022


    Adaptive Discretization Using Voronoi Trees for Continuous-Action POMDPs

    Hoerger, Marcus / Kurniawati, Hanna / Kroese, Dirk et al. | TIBKAT | 2023


    Handling Prediction Model Errors in Planning for Automated Driving Using POMDPs

    Bey, Henrik / Sackmann, Moritz / Lange, Alexander et al. | IEEE | 2021


    Real-Time Decision Making for Large POMDPs

    Paquet, S. / Tobin, L. / Chaib-draa, B. et al. | British Library Conference Proceedings | 2005