密集障碍物环境下基于凸包和微粒群优化的机器人路径规划

Robot path planning in environments with dense obstacles based on convex hull and particle swarm optimization

密集障碍物环境下,考虑机器人移动过程中的控制偏差进行路径规划,尚缺乏有效的方法.本文的方法是:首先根据障碍物之间的最小距离和机器人尺寸的大小关系,确定凸包形成的条件;然后,通过选择满足条件的顶点,形成密集障碍物的凸包;最后,基于凸包的关键点和稀疏障碍物的位置,采用微粒群优化规划机器人路径.仿真和实验结果验证了所提方法的可行性.

Effective methods are lacking for planning robot path while considering control deviation in an environment , ： with dense obstacles. In our approach, we first determine the prerequisites for forming a convex hull according to the relation between the robot size and the minimal distance between obstacles, and then, we form the convex hull of these dense obstacles by choosing vertices that satisfy certain conditions; finally, the particle swarm optimization is applied to planning the robot path based on the positions of the key vertices in the convex hull and the positions sparse obstacles. Simulation and experiment results validate the feasibility of our method.