This paper presents a search-based global motion planning method, called the two-phase A*, with an adaptive heuristic weight. This method is suitable for planning a global path in real time for a car-like vehicle in both indoor and outdoor environments. In each planning cycle, the method first estimates a proper heuristic weight based on the hardness of the planning query. Then, it finds a nearly optimal path subject to the non-holonomic constraints using an improved A* with a weighted heuristic function. By estimating the heuristic weight dynamically, the two-phase A* is able to adjust the optimality level of its path based on the hardness of the planning query. Therefore, the two-phase A* sacrifices little planning optimality, and its computation time is acceptable in most situations. The two-phase A* has been implemented and tested in the simulations and real-world experiments over various task environments. The results show that the two-phase A* can generate a nearly optimal global path dynamically, which satisfies the non-holonomic constraints of a car-like vehicle and reduces the total navigation time.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Two-phase A*: A real-time global motion planning method for non-holonomic unmanned ground vehicles


    Beteiligte:
    Zhang, Kai (Autor:in) / Yang, Yi (Autor:in) / Fu, Mengyin (Autor:in) / Wang, Meiling (Autor:in)


    Erscheinungsdatum :

    2021-03-01


    Format / Umfang :

    16 pages




    Medientyp :

    Aufsatz (Zeitschrift)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch








    Motion planning for parallel robots with non-holonomic joints

    Tchon, Krzysztof / Jakubiak, Janusz / Grosch Obregon, Patrick John et al. | BASE | 2012

    Freier Zugriff