In this paper, we described the development of a hierarchical path-planning algorithm which accommodates changes in the driving environments for autonomous vehicles. The developed path-planning algorithm improves the existing local path-planning algorithm and enhances the computational efficiency with three functional components. The first component is the behaviour planner which performs the rule-based decision process based on perception and localization information. This decision process determines the overall vehicle behaviour related to the path-planning strategy. The second component is the map and path selector which preprocesses perception information and chooses the most appropriate local-path-planning algorithm based on the results of the behaviour planner. To use computational resources effectively, this component integrates only essential perception information according to the local path-planning strategy. The third component is the local-path planner, which provides a driveable and collision-free path. To find paths in various driving conditions, the proposed local-path planner is composed of two algorithms: a road-model-based path-planning algorithm and a graph-structure-based path-planning algorithm. In the first path-planning algorithm, the road model is derived as an arc-length-based cubic spline in curvilinear coordinates. Based on the road model, a set of path candidates is adaptively generated according to the driving conditions, and one of the candidates is selected through an optimization technique. This algorithm is mainly applied to structured road driving, such as lane keeping and changing. In contrast, the second path-planning algorithm is based on the graph data structure. For expansion of the graph data, a kinematic vehicle model and an obstacles grid map are used. The final path is formed as a series of vertices by using a cost function. In order to improve the real-time performance, the adaptive heuristic cost and a technique which reuses the previous path are applied. Since this algorithm is primarily dependent on perception information, it can be applied to unstructured road driving.


    Access

    Access via TIB

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Road-model-based and graph-structure-based hierarchical path-planning approach for autonomous vehicles


    Contributors:
    Kim, Junsoo (author) / Jo, Kichun (author) / Chu, Keonyup (author) / Sunwoo, Myoungho (author)


    Publication date :

    2014


    Size :

    20 Seiten, Bilder, Tabellen, 55 Quellen



    Type of media :

    Article (Journal)


    Type of material :

    Print


    Language :

    English