A Hierarchical Path Planning Approach Using Waypoint Visibility-Based Target Planner

Abstract

This paper presents a hierarchical path planning framework for efficient and practical robot navigation in environments with unknown obstacles not registered in a given map. While existing path planning approaches primarily focus on separately enhancing the performance of either the local planner or the global planner, we propose a novel framework that employs a target planner to integrate the two, enabling adaptive navigation toward the final goal position via reachable waypoints. The proposed framework integrates a visibility-based target planner to dynamically evaluate the distance and heading visibility of the current target waypoint along the global path computed by the global planner. By dynamically assessing visibility, the visibility-based target planner autonomously updates the robot’s target waypoint if the current one becomes unreachable due to obstacles, ensuring that the local planner efficiently navigates the robot to the updated target. Experiments were conducted to validate the proposed framework against a conventional sequential waypoint assignment approach. The results demonstrate that, compared to the conventional approach, the hierarchical framework using the visibility-based target planner reduces overall navigation time by 49.6 % and 52.6 % in two different experiments while requiring fewer control commands. These improvements indicate that the proposed framework effectively minimizes unnecessary detours and enables the robot to reach the final goal position more efficiently, even in the presence of unmapped obstacles.