Highlights Proposed a Reasoning Graph framework to cooperate conflicting platoons with complex interactions. Introduced a situation-aware optimization to optimize the systemic utility in mixed traffic environments. Modeled the social interaction custom to guarantee the social compliance of the cooperation scheme. Tested the Reasoning Graph framework and did a comparative analysis with a reservation-based method and an optimization-based method.

    Abstract Cooperating unprotected turns in mixed connected and autonomous traffic environments is beneficial to improving the systemic utility of intersections. However, interactions among connected and automated vehicles (CAVs) and manually driving vehicles (MVs) are varied and interrelated. These factors make the cooperation a multimodal problem and bring many difficulties to interactions analyzing and the optimal plan solving. This paper proposes a situation-aware Reasoning Graph (RG) framework that consists of the Reasoning Process and the Solution Process. Complex interactions are first described by a graph, where the mapping from situations to the graph is constructed. Then the Reasoning Process divides future motion space into numerous situations and reasons potential situations using the graph search-based algorithm. Some rules for maneuver compatibility and social interaction customs are integrated into the graph search to ensure the maneuver compatibility and social compliance of these reasoned situations. The Solution Process develops a heuristic algorithm to quickly search for speed profiles that follow the reasoned situations. After comparative tests of the reservation-based and the optimization-based methods, the results show that the RG-based method helped decrease the average systemic cost by approximately 10%. However, it did not cause losses in individual performances. The statistical analysis of individual performance shows that the RG even increased the individual performances regarding safety, efficiency, and energy-saving.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Reasoning Graph: A Situation-aware framework for cooperating unprotected turns under mixed connected and autonomous traffic environments


    Beteiligte:
    Zhou, Donghao (Autor:in) / Ma, Zian (Autor:in) / Zhao, Xiaocong (Autor:in) / Sun, Jian (Autor:in)


    Erscheinungsdatum :

    2022-07-18




    Medientyp :

    Aufsatz (Zeitschrift)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch





    Early object detection for unprotected turns

    LUDERS BRANDON DOUGLAS | Europäisches Patentamt | 2022

    Freier Zugriff

    Early object detection for unprotected turns

    LUDERS BRANDON DOUGLAS | Europäisches Patentamt | 2019

    Freier Zugriff

    EARLY OBJECT DETECTION FOR UNPROTECTED TURNS

    LUDERS BRANDON | Europäisches Patentamt | 2019

    Freier Zugriff

    EARLY OBJECT DETECTION FOR UNPROTECTED TURNS

    LUDERS BRANDON DOUGLAS | Europäisches Patentamt | 2022

    Freier Zugriff