【作者】 王敏；

【导师】 童水光；

【作者基本信息】 浙江大学 ， 工程硕士（专业学位）， 2019， 硕士

【摘要】 随着机器人技术的发展和国家政策的大力支持,机器人在工业生产中的应用越来越广泛。在打磨加工作业中,机器人也有着比数控机床柔性灵活、比手工打磨高效安全的优点,但打磨机器人在实际应用中,还存在示教点繁多、接触力控制等诸多问题。本课题在调研今飞凯达轮毂制造工厂的基础上,以汽车轮毂的打磨加工为研究背景,在工业机器人的基础上引入工业相机和力传感器,开展基于视觉的加工轨迹在线获取和接触力反馈控制策略的研究,并搭建软硬件系统,完成试验验证。本文的研究内容如下:(1)提出了一种基于视觉的打磨加工轨迹在线获取方法,解决传统示教法记录机器人轨迹点效率低的问题。首先,采用图像预处理算法消除原始轮毂图像中的干扰因素;然后提取汽车轮毂的关键特征,生成离散的路径点坐标;最后标定图像处理坐标系和机器人工件坐标系,提出自定义间接标定法,使机器人末端完整复现图像处理得到的轨迹点,与实际加工路径相吻合。(2)提出了一种面向环境动力学参数未知的自适应阻抗控制方法。首先分析打磨过程中的受力特性,研究机器人力位控制中的阻抗控制策略;然后构建基于位置的阻抗控制模型,通过打磨法向分力的实时反馈控制末端位置;最后结合自适应模糊控制理论,在阻抗控制模型中加入模糊控制器,提高阻抗控制系统在环境动力学参数未知情况下的响应精度。(3)在视觉和力反馈的基础上规划与调整轨迹点,提出一种基于位姿分离和矩阵正交的机器人逆运动学求解算法。首先建立机器人运动学模型并分析机器人正逆运动学过程;然后改进逆运动学求解算法,分别根据位置子矩阵和旋转子矩阵求解各关节角,简化求解过程,减少运算时间;最后在视觉获取的轨迹点上完成直角坐标空间和关节空间的轨迹规划,基于力控策略调整初始轨迹,修正接触力异常点处的位姿。(4)搭建与开发了系统的软硬件部分并完成试验研究。首先安装完成了系统的硬件设备:ABB-IRB1600六自由度机械臂及其控制配件、工业相机及镜头、力传感器及数据采集卡;其次开发调试了图像采集与处理、力信息采集处理和机器人控制等软件系统;最后设计了轮廓轨迹跟踪试验和自适应阻抗控制试验,验证了提出方法的有效性。更多还原

【Abstract】 With the development of robotics technology and the strong support of national policy,robots are more and more widely used in industrial production.In deburring operation,robots also have the advantages of flexibility,flexibility,efficiency and safety compared with CNC machine tools,but in practical application,there are many problems such as teaching points,contact force control and so on.On the basis of investigating the present Jinfeikaida Hub Manufacturing Plant,taking the deburring of automobile hubs as the research background,introducing industrial cameras and force sensors on the basis of industrial robots,the research on online acquisition of processing trajectory based on vision and feedback control strategy of contact force is carried out,and the software and hardware systems are built to complete the test verification.The research contents of this paper are as follows:(1)Aiming at the inefficiency of traditional teaching method in recording robot trajectory points,a vision-based on-line deburring trajectory acquisition method is proposed.Firstly,image preprocessing algorithm is used to eliminate the interference factors in the original hub image;Secondly,the key features of automobile hub are extracted to generate discrete path point coordinates;Finally,image processing coordinate system and robot workpiece coordinate system are calibrated,and a self-defined indirect calibration method is proposed to make the end of the robot completely reproduce the trajectory points obtained from image processing,which is consistent with the actual processing path.(2)To solve the problem of contact force control in deburring process,an adaptive impedance control method for unknown environmental dynamics parameters is proposed.Firstly,the force characteristics in deburring process are analyzed,and the impedance control strategy in force position control of robot is studied.Then,a position-based impedance control model is constructed to control the end position by deburring the normal component of force with real-time feedback.Finally,a fuzzy controller is added to the impedance control model to improve the impedance control system in the absence of environmental dynamics parameters.Response accuracy under informed condition.(3)Planning and adjusting trajectory points on the basis of vision and force feedback,and an inverse kinematics algorithm based on pose separation and matrix orthogonality is proposed.Firstly,the kinematics model of the robot is established and the forward and inverse kinematics process of the robot is analyzed;Then the inverse kinematics algorithm is improved to solve the joint angles according to the position sub-matrix and the rotation sub-matrix respectively,which simplifies the solving process and reduces the computing time;Finally,the trajectory planning of the right-angle space and the joint space is completed at the trajectory points obtained by vision,and the image processing is adjusted based on the force control strategy.The initial trajectory is obtained and the position and attitude of the abnormal contact force point are corrected.(4)The software and hardware parts of the system are built and developed,and the experimental research is completed.Firstly,the hardware equipment of the system is installed and completed:ABB-IRB1600 six-degree-of-freedom manipulator and its control accessories,industrial camera and lens,force sensor and data acquisition card;Secondly,software systems such as image acquisition and processing,force information acquisition and processing and robot control are developed and debugged;Finally,the contour trajectory tracking test and adaptive impedance control test are designed to verify the effectiveness of the proposed method.更多还原