基于Kinect的七自由度空间机械臂体感控制方法
|
摘要
针对七自由度空间机械臂的体感控制问题,提出了一种基于操作周期和运动周期的增量式关节角人机运动映射方法。使用Kinect传感器采集人体关节点坐标数据并使用空间向量法计算关节角,根据关节角增量及手势命令计算得到机械臂关节角。为解决关节坐标数据的抖动与失真问题,分析了Kinect对不同关节点采集的数据精度的关系,并给出一种限幅平滑滤波算法。结合上述结果,提出对操作周期进行分段的方法,并给出了优化后的体感控制算法。最后,搭建了空间机械臂体感控制的仿真试验平台,在Unity中搭建仿真试验场景,经体感控制下的空间机械臂运动试验验证了所提出方法的可行性。
To solve the somatosensory control problem of 7-DOF space manipulator,an incremental joint angle man-machine motion mapping method was proposed based on action period and movement period. The Kinect sensor was used to collect the coordinate data of the human joints and the joint angle was calculated by the space-vector method. The joint angle of the space manipulator was calculated according to the joint angle increment and the gesture command. The accuracy of Kinect for different human joints was analyzed to solve the problem of jitter and distortion of data. A limitedwidth smooth filtering algorithm and a method for segmenting the operation cycle were proposed and the optimized somatosensory control algorithm was given. In the end,a simulation experiment platform for somatosensory control of the space manipulator was built and the simulation experiment scene was built in Unity. The feasibility of the proposed method was verified by space manipulator motion experiment under somatosensory control.
To solve the somatosensory control problem of 7-DOF space manipulator,an incremental joint angle man-machine motion mapping method was proposed based on action period and movement period. The Kinect sensor was used to collect the coordinate data of the human joints and the joint angle was calculated by the space-vector method. The joint angle of the space manipulator was calculated according to the joint angle increment and the gesture command. The accuracy of Kinect for different human joints was analyzed to solve the problem of jitter and distortion of data. A limitedwidth smooth filtering algorithm and a method for segmenting the operation cycle were proposed and the optimized somatosensory control algorithm was given. In the end,a simulation experiment platform for somatosensory control of the space manipulator was built and the simulation experiment scene was built in Unity. The feasibility of the proposed method was verified by space manipulator motion experiment under somatosensory control.
引文
[1]蔡自兴.机器人学[M].北京:清华大学出版社,2000:322-327.CAI Zixing. Robotics[M]. Beijing:Tsinghua UniversityPress,2000:322-327.(in Chinese)
[2] Stowers J,Hayes M,Bainbridge-Smith A. Altitude control ofa quadrotor helicopter using depth map from Microsoft Kinectsensor[C]//Mechatronics(ICM),2001 IEEE InternationalConference on Mechatronics,1994:205-221.
[3]张宝玉.机器人主从映射方法分析及实验研究[D].哈尔滨:哈尔滨工程大学,2015.ZHANG Baoyu. Master-slave Robot Mapping Method Analysisand Experimental Research[D]. Harbin:Harbin EngineeringUniversity,2015.(in Chinese)
[4] Nguyen V V,Lee J H. Full-body imitation of human motionswith Kinect and heterogeneous kinematic structure of human-oid robot[C]//IEEE/SICE International Symposium on Sys-tem Integration. Piscataway,USA:IEEE,2012:93-98.
[5]卫沅,赵京.基于运动捕捉下的7R机械臂的实时控制[J].机械工程学报,2013,49(23):68-73.WEI Yuan,ZHAO Jing. Real-time control of 7R manipulatorbased on motion capture[J]. Journal of Mechanical Engineer-ing,2013,49(23):68-73.(in Chinese)
[6]王朝阳.基于Kinect的类人机械臂关节运动直接示教[J].机器人,2017(9):699-700.WANG Chaoyang. Direct teaching in joint space for humanoidmanipulator using Kinect[J]. Robot,2017(9):699-700.(inChinese)
[7] Kinect开发Level 100 Guidelines[EB/OL]. Kinect for Win-dows.(2018)[2018].Kinect development level 100 guidelines[EB/OL]. Kinect forWindows.(2018)[2018].(in Chinese)http://www. k4w.cn/news/1.html.
[8]朱宇辉.基于Kinect的机械臂人机交互控制系统设计[D].绵阳:西南科技大学,2016.Zhu Yuhui. Design on Human-computer Interactive ControlSystem of Robot Arm Based on Kinect[D]. Mianyang:South-west University of Science and Technology,2016.(in Chi-nese)
[9]范勇涛.基于体感的空间机器臂人机交互技术[D].西安:西安电子科技大学,2015.Fan Yongtao. Somatosensory Technology of Human ComputerInteraction with Space Robot[D]. Xi'an:Xidian University,2015.(in Chinese)
[2] Stowers J,Hayes M,Bainbridge-Smith A. Altitude control ofa quadrotor helicopter using depth map from Microsoft Kinectsensor[C]//Mechatronics(ICM),2001 IEEE InternationalConference on Mechatronics,1994:205-221.
[3]张宝玉.机器人主从映射方法分析及实验研究[D].哈尔滨:哈尔滨工程大学,2015.ZHANG Baoyu. Master-slave Robot Mapping Method Analysisand Experimental Research[D]. Harbin:Harbin EngineeringUniversity,2015.(in Chinese)
[4] Nguyen V V,Lee J H. Full-body imitation of human motionswith Kinect and heterogeneous kinematic structure of human-oid robot[C]//IEEE/SICE International Symposium on Sys-tem Integration. Piscataway,USA:IEEE,2012:93-98.
[5]卫沅,赵京.基于运动捕捉下的7R机械臂的实时控制[J].机械工程学报,2013,49(23):68-73.WEI Yuan,ZHAO Jing. Real-time control of 7R manipulatorbased on motion capture[J]. Journal of Mechanical Engineer-ing,2013,49(23):68-73.(in Chinese)
[6]王朝阳.基于Kinect的类人机械臂关节运动直接示教[J].机器人,2017(9):699-700.WANG Chaoyang. Direct teaching in joint space for humanoidmanipulator using Kinect[J]. Robot,2017(9):699-700.(inChinese)
[7] Kinect开发Level 100 Guidelines[EB/OL]. Kinect for Win-dows.(2018)[2018].Kinect development level 100 guidelines[EB/OL]. Kinect forWindows.(2018)[2018].(in Chinese)http://www. k4w.cn/news/1.html.
[8]朱宇辉.基于Kinect的机械臂人机交互控制系统设计[D].绵阳:西南科技大学,2016.Zhu Yuhui. Design on Human-computer Interactive ControlSystem of Robot Arm Based on Kinect[D]. Mianyang:South-west University of Science and Technology,2016.(in Chi-nese)
[9]范勇涛.基于体感的空间机器臂人机交互技术[D].西安:西安电子科技大学,2015.Fan Yongtao. Somatosensory Technology of Human ComputerInteraction with Space Robot[D]. Xi'an:Xidian University,2015.(in Chinese)