摘要
针对机械臂在工作空间内的避障问题,提出基于点云信息的七自由度冗余机械臂避障策略。首先利用D-H参数在机器人操作系统中建立七自由度冗余机械臂的运动学模型,再基于RGB-D相机引入的点云信息建立空间实时地图,使用基于点云信息的包围盒算法检测碰撞,最后在空间实时地图上通过快速扩展随机树进行避障路径规划,实现在机械臂运动过程中避障。实验结果表明,基于点云信息的七自由度冗余机械臂避障路径规划算法可以根据在工作空间内获得的障碍物点云信息进行路径规划,避开障碍物,达到目标位姿,在避障运动过程中与障碍物至少保持8 mm的安全距离,为后续的研究提供了参考。
In order to meet the requirements of the robot arm for obstacle avoidance in the workspace, a strategy for obstacle avoidance of 7-DOF redundant arm based on point cloud data was proposed.First,the kinematics model of a 7-DOF redundant arm was given in the Robot Operating System (ROS) by using D-H parameters.Then,space real-time map based on point cloud data introduced by RGB-D camera was built.Point cloud data based bounding box algorithm was used to detect collision.Finally,an obstacle avoidance strategy by Rapidly-exploring Random Trees (RRT) on space real-time map was provided,obstacle avoidance during the movement of the arm was achieved.The results of the physical model experiment show that the algorithm can perform obstacle avoidance movement based on point cloud data,and keep distance of at least 8 mm from obstacles during obstacle avoidance movement,which provide relevant references for follow up research.
作者
刘芳华
谢如坤
Liu Fanghua;Xie Rukun(School of Mechanical Engineering,Jiangsu University of Science and Technology,Zhenjiang 212003,Jiangsu,China)
出处
《现代制造工程》
CSCD
北大核心
2020年第12期6-13,35,共9页
Modern Manufacturing Engineering
基金
国家自然科学基金项目(51705217)。
关键词
点云信息
冗余机械臂
快速扩展随机树
机器人操作系统
point cloud data
redundant arm
Rapidly-exploring Random Trees(RRT)
Robot Operating System(ROS)
