Hierarchical Trajectory Planning for AGVs with a Local Adaptive Search Space
Chun Liu, Jieying Lu, Junhui Li · 2021
This paper presents a real-time and responsive trajectory planner for Autonomous Ground Vehicles (AGVs). We improve a state-based sampling search space method to generate well separated kinematically feasible trajectories according to the obstacle distribution ahead in a finer-scale range of vehicle. Meanwhile, the planner utilizes a cost function, which introduces a global guidance with the nonholonomic nature of AGVs, to ensure the optimality and responsivity for vehicles navigation in the dynamic environments. The experimental results demonstrate that the proposed planner has an improvement in generating superior and smoother trajectories, compared with the Global Adaptive Hierarchical Trajectory Planner (GAHTP). Simulation results show that the proposed method can provide safe and smooth trajectories in highly constrained semi-structured or unstructured environments.