基于三维环境模型的启发式路径规划算法

Heuristics Algorithm for Motion Planning Based on 3D Environment Model

针对多自由度机器人手臂在未知环境中实时避障的问题,提出了一种基于环境信息的连杆机器人实时路径规划方法。采用笛卡尔空间内的障碍物检测信息建立了障碍物的空间模型,并依据该模型设计一种基于启发式规则的机器人路径规划算法。该算法不断猜测和修正路径,通过模糊推理得到下一位姿点,通过曲线拟合得到到达该位姿点的路径。在Matlab下利用机器人工具箱建立了PUMA560型机器人的运动学模型,并在运动空间设置障碍物,对该算法进行仿真分析,分析结果说明所提出的路径规划算法可以在较短时间内完成避障运动,具有较好的实时性,同时运动关节的角度变化曲线比较平滑,运动中冲击力较小,这些特点使其便于在实际工程中使用。

To the MDOF robot arm motion planning problem in an unknown environment,a linking robot motion planning method based on the environment information is presented.By using the complete information about the workspace,an environmental model in the Cartesian space is presented and a heuristics algorithm for robot motion planning is developed.The presented algorithm continually surmises and amends motion path.The next posture is obtained through fuzzy inference and the path to the posture can be got through curve fitting.To test the presented algorithm,a motion planning control system based on PUMA560 is studied with robot toolbox of Matlab,and the simulation result shows that the presented algorithm can achieve motion planning in a less time,and the angle curve of motion joints is smooth,which improves the practicality of the algorithm.