乒乓球机器人视觉测量与控制

CORC > 自动化研究所 > 中国科学院自动化研究所 > 毕业生 > 博士学位论文

题名	乒乓球机器人视觉测量与控制
作者	张正涛
学位类别	工学博士
答辩日期	2010-09-10
授予单位	中国科学院研究生院
授予地点	中国科学院自动化研究所
导师	徐德
关键词	乒乓球机器人目标跟踪飞行模型碰撞模型轨迹预测三维视觉仿真 Table Tennis Robot target tracking flying model rebound model trajectory prediction 3D simulation
其他题名	Visual Measurement and Control for Table Tennis Robot
学位专业	控制理论与控制工程
中文摘要	机器人打乒乓时需要做到对快速运动的乒乓球进行可靠的跟踪、对球未来的轨迹做出准确的预测,以及控制球拍以合适的时间、位置、姿态和速度进行击球等，涉及机器人与外界信息的实时互动。因此，机器人打乒乓球涉及到视觉感知与智能控制的众多问题。通过对机器人打乒乓球的研究，可以对一系列感知和控制问题进行深入探索。本文围绕乒乓球机器人视觉测量与控制开展研究工作。在本文中，主要的工作和贡献有： 1.提出了一种基于智能相机的分布式并行处理的高速视觉系统。与传统的基于图像采集卡的视觉系统相比，本系统利用了智能相机具有的图像处理能力，将传统的串行处理的图像处理模式改为并行处理。将大量的运算分散到各个相机单元，大大提高了系统的实时性。 2.针对快速运动的乒乓球，提出了一套有效的乒乓球跟踪算法。算法将差帧法、动态窗口、游程编码等方法有机的组合在一起，使算法的实时性得到保证。并针对乒乓球差帧后可能出现的月牙形，提出了GSP算法，准确的定位乒乓球的中心坐标，保证了乒乓球三维坐标的精度。 3.利用乒乓球的飞行模型推导出了模型的解析解，可以有效的减少预测算法运行时间。提出了一种行之有效的乒乓球与球桌的碰撞模型，在水平方向利用一种线性模型建立碰撞前后的乒乓球速度变化关系，在竖直方向利用球桌的弹性系数，推导出反弹后的乒乓球速度变化。提出了三阶段的球拍运动规划方法及特殊情况下的直线击球策略。 4.针对飞行模型对初始点的速度敏感，而最小二乘拟合求导得到的初始点的速度精度较低的情况，提出了基于位置负反馈的初速度较正算法，大大提高了预测算法的鲁棒性与精度。 5.提出了三阶段的球拍轨迹规划算法，实现了人机对打50回合。同时为了逼真的再现实验场景及准确的标定模型参数，设计了基于OPENGL的三维仿真系统。最后，对本文的研究成果进行了总结，并指出了下一步的研究方向。
英文摘要	The robot should track the fast moving target and predict the future trajectory of the ball beforehand when the robot plays table tennis games. Also, it should strike the ball back in proper velocity, position, posture and right striking time. These mean that the robot should interact with the environment continuously. So the research of the table tennis robot contains visual perception and intelligent control and so on. By the study of the robot, a series of perceptional and control problem can be investigated. In this thesis, the visual measurement and control for the table tennis robot are investigated. The main contributions of this thesis include following issues: 1.A distributed parallel processing High-Speed Vision System based on smart camera is presented.Compared to traditional vision systems based on image acquisition card, this vision system makes good use of the powerful image processing performance. It transfers the serial mode to parallel and distributes the computation to each camera to improve the real-time performance. 2.A set of effectiveness table tennis tracking lgorithms to track the fast moving ball is proposed. The algorithms integrate adjacent frame difference, dynamic window, run length code together. They have good real-time performance. Also, the GSP algorithm is proposed to compute the center of the ball when the result of the adjacent frame difference may be a crescent to improve the accuracy of the 3D position of the ball. 3.The analytic solution of the flying model is deduced which is helpful to reduce the time-consuming of the prediction algorithm. An effective rebound model is proposed. In the horizontal direction a linear model is proposed to represent the velocity change before and after rebound on the table. In vertical direction, the rebound coefficient is employed to compute the velocity change of the ball. Three phases strike method is proposed with straight line shot in particular cases. 4.Because the flying model is very sensitive to the initial velocity, a negative feedback method based on position information is proposed. 5.A three-stage trajectory planning method is proposed. The 5-DOF table tennis robot can rally with human for 50 rounds based on this method. A 3D simulation vision system is designed to redisplay the experimental scene and adjust the model parameters. Finally, the research results are summarized and future work is discussed.
语种	中文
其他标识符	200718014628027
内容类型	学位论文
源URL	[http://ir.ia.ac.cn/handle/173211/6303]
专题	毕业生_博士学位论文
推荐引用方式 GB/T 7714	张正涛. 乒乓球机器人视觉测量与控制[D]. 中国科学院自动化研究所. 中国科学院研究生院. 2010.