摘要： 环境建模和路径规划是轮式机器人视觉导航的关键技术。针对传统视觉导航中环境建模实时性较差的问题,提出利用边缘延伸原理对轮式机器人视觉环境进行感知的方法;针对人工势场法在路径规划中出现局部最优解、路径震荡的问题,提出改进的逆势场导向法,给出新的人工势场函数,得到广义上的轮式机器人安全通道;针对轮式机器人路径不能出现急转急停的问题,提出一种曲率映射法,将轮式机器人的运动轨迹映射成曲率,设计相应的平滑路径生成策略。最后给出路径生成的算法,通过仿真研究与实验验证了该算法的有效性和正确性。
关键词:
路径规划,
轮式移动机器人,
路径平滑,
环境建模,
逆势场导向法,
曲率映射法
Abstract: Environment modeling and path planning are two key problems for wheeled robot vision navigation.Aiming at the real-time shortage in the environment modeling,this paper proposes a method using the principle of extending edge;To overcome the oscillation and local trap problem in mobile robot path planning,this paper presents the contrarian potential field method to give a new function and the generalized wheeled robot safety channel;To solve the nonholonomic of the Wheeled Mobile Robot(WMR),this paper presents a curvature mapping method that the WMR path is mapped into a curvature to create smoothing path.Finally,the path generation algorithm is given,and some experiments are conducted to validate the algorithm.
Key words:
path planning,
Wheeled Mobile Robot(WMR),
smoothing of path,
environment modeling,
contrarian potential field method,
curvature mapping method
中图分类号: