A Fast and Smooth Planning Framework for Autonomous Mobile Robot in Complex Environments
Kun Zhou, Liang Zhang, Shuping He · 2023
This technique note investigates the problem of path and motion planning for autonomous mobile robots in complex environments. A hybrid algorithm composed of a smoother A-star$(\mathrm{s}-\mathrm{A}^{*})$algorithm for global path planning and a faster & more accurate Dynamic Window Approach (fma-DWA) is proposed. Firstly, three technical tricks are designed based on the standard$\mathrm{A}^{*}$algorithm to ensure the smoothness and safety of the generated paths, which are the improvement of evaluation function, the redundant node deletion strategy and the B-splinebased path smoothing respectively. Then regarding the motion planning phase after path generation, we further advance the traditional DWA to choose a faster and more accurate motion command of the mobile robot. To this end, a composite evaluation function of the DWA is proposed by further considering the curvature of the generated path and the distance between the mobile robot and the target. Besides, the weight of the robot's velocity in the evaluation function is adjusted according to the closest distance to obstacles instead of keeping fixed in the traditional DWA method. Finally, extensive simulation results show the efficiency and accuracy of the proposed methods.