动态环境下移动机器人增量式候选路径集生成及轨迹规划方法

Incremental candidate path set generation and trajectory planning method for mobile robots in dynamic environments

针对动态环境下移动机器人对轨迹高效规划的需求,提出一种基于增量式候选路径集的轨迹规划方法。首先,提出一种图搜索与凸优化相结合的初始候选路径集生成方法,优化了机器人的移动距离,提高了规划成功率;其次,基于初始候选路径集,在规划移动速度时添加绕开移动障碍物的增量候选路径,提高轨迹时长的最优性,构建增量式候选路径集;最后,设计了基于增量式候选路径集的启发式轨迹搜索算法,在不损失最优性的前提下提高了搜索效率。仿真实验表明,所提方法能够适应不同拥挤程度的规划场景,相比基于路图的规划方法,所提方法规划的轨迹时长更短,所需的规划时间也更短。

To address the demandforeffective mobile robot time-optimal trajectory planning in dynamic environments,a trajectory planning approachbasedon incremental candidate path set was proposed.An initial candidate path set generation method that used graph search and convex optimization was proposed for optimizing the moving distance of the robot and improving the success rate of planning.On this basis,the incremental candidate paths that avoided moving obstacles were added when the speed profile was under planning,which improved the optimality of trajectory duration and generated an incremental candidate path set.Based on the incremental candidate path set,a heuristic trajectory searching algorithm was developed,which improved searching efficiency without sacrificing optimality.The simulation experiments showed that the proposed method adjusted to planning scenarios with various levels of congestion,generated trajectories with shorter trajectory duration,and required less planning time than roadmap-based methods.