摘要
结合机器视觉和运动规划方法,对动态环境中移动物体的平滑抓取进行了研究。设计了一种运动物体跟踪算法,能够实现无障碍环境下移动物体的平滑抓取;针对环境中存在动态障碍物的情况,设计了一种基于排斥矢量的动态避障算法,与运动物体跟踪算法相结合,实现了机器人先避障后平滑抓取;基于机器人操作系统(Robotoperatingsystem,ROS)框架,在5自由度的KUKA Youbot机械臂平台上实现了无障碍环境下对传送带上物体的平滑抓取,以及动态障碍物环境中的平滑抓取,试验结果验证了所设计算法的优越性。
Using machine vision and motion planning technology, smooth catching of moving object in dynamic environment is studied. A tracking algorithm of moving object is designed to realize the smooth catching in obstacle-free environment. To avoid the moving obstacle in the environment, a collision avoidance algorithm is proposed by using repulsive vector, and then it is combined with the tracking algorithm to finish both obstacle avoidance and smooth pickup of moving object. Based on the robot operating system, several experiments are implemented on a 5-DOF KUKA Youbot robotic manipulator. The manipulator can catch moving object smoothly in the obstacle-free environment and the object in the environment with moving obstacle. The experimental results indicate that both the tracking algorithm and the collision avoidance algorithm are very effective to finish the object catching.
出处
《机械工程学报》
EI
CAS
CSCD
北大核心
2018年第19期10-17,共8页
Journal of Mechanical Engineering
基金
国家自然科学基金资助项目(51675501,51275500)
关键词
动态环境
平滑抓取
动态避障
物体抓取
dynamic environment
smooth pickup
moving obstacle avoidance
object grasping
