摘要
主要研究了移动机器人在未知动态环境中的路径规划问题.提出一种将障碍预估与概率方向权值相结合的动态路径规划新方法.该方法将卡尔曼滤波引入到规划算法中,使得对障碍物运动状态的实时有效预估成为可能.同时,为实现移动机器人的实时路径规划,提出一种新的概率方向权值方法,基于周期规划将障碍物与目标信息进行融合,能够有效处理室内环境下对于障碍物的速度和运动轨迹均未知的动态路径规划问题.仿真结果以及基于SmartROB2移动机器人平台所进行的实验结果验证了该方法的有效性和实用性.
A novel dynamic path planning approach is presented to integrate obstacle prediction with probabilistic directional weighting and accomplish mobile robot path planning in an unknown and dynamic environment. In order to predict obstacles＇ states exactly, Kalman filter is employed in the path planning. Based on the concept of period planning, a new probabilistic directional weighting method is presented to complete the real-time path planning combining information of obstacles and goal. The method is applicable to the complex indoor environment without the prior information of moving obstacles＇ speeds and trajectories. Finally, simulation and experiment results implemented on SmartROB2 mobile robot show the method＇s validity and practicability.
出处
《控制理论与应用》
EI
CAS
CSCD
北大核心
2007年第3期337-342,共6页
Control Theory & Applications
基金
国家自然科学基金资助项目(60605023)
大连理工大学–中科院沈阳自动化研究所合作科研探索基金资助项目.
关键词
动态路径规划
卡尔曼滤波
周期规划
概率方向权值
障碍预估
dynamic path planning
Kalman filter
period planning
probabilistic directional weighting
obstacle prediction
