基于几何约束的种子取样机器人示教及自标定方法研究

发布时间：2018-05-07 22:20

本文选题：种子切削 + 几何约束　；参考：《中国科学技术大学》2017年硕士论文

【摘要】：操作对象的自动定位和机器人示教是机器人应用基础研究中的关键问题,对此开展研究工作具有重要的应用价值和理论意义。本文在中科院先导课题支持下,针对种子取样过程中操作对象的自动定位和机器人示教问题,设计并实现了种子切削取样机器人系统,解决了种子取样过程的自定位和示教问题,并通过实验证明了所提方法的正确性。论文的主要工作如下:1.依据种子切削取样课题任务的要求,结合国内外育种专家对种子取样自动化装备的需求,设计并实现了一种作物种子切削取样机器人系统,编写了上位机程序,解决了取样过程的自定位和直接示教问题。2.针对种子切削取样机器人系统中取样操作的自定位问题,分析了种子取样盒的几何结构特点,提出了基于几何约束的机器人自标定方法。该自标定方法基于种子取样盒坐标点的几何约束关系,利用取样盒坐标系与机器人基坐标系的齐次坐标变换方法,解决了取样操作中的自定位问题,实现了取样点坐标位置的自动获取。3.为了实现取样过程中机器人直接示教功能,研究了机器人直接示教问题,采用基于位置调整的主动柔顺控制方法,在ER6C60上开发了一种基于拖曳的机器人直接示教系统。该直接示教系统采用了基于力传感器的主动柔顺控制算法,利用多维力传感器获得外部的示教力/力矩信息,将经力坐标变换后的力信号矢量转换为机器人的位置调整量,达到了顺应性控制的目的。上位机采用基于MFC对话框程序,从而实现了机器人的直接牵引示教功能。4.实验表明,基于几何约束的自标定方法计算的种子坐标位置与实际位置坐标误差小于0.8 mm,满足课题需求。此方法测量过程简单,成本低,定位精度高,对解决工件坐标的自定位问题具有借鉴意义。基于位置控制策略的直接示教方法使得机器人很好地跟踪示教作用力,直接示教系统的可靠性得到了实验的验证。该示教方法适用于封闭的机器人控制系统,通用性较好。
[Abstract]:The automatic location of operating objects and the demonstration of robot teaching are the key problems in the basic research of robot application. It has important application value and theoretical significance to carry out the research work. In this paper, a seed cutting sampling robot system is designed and implemented under the support of the pilot project of the Chinese Academy of Sciences, aiming at the problem of automatic positioning of the operation object and robot teaching during the seed sampling process. The problem of self-localization and teaching in seed sampling process is solved, and the correctness of the proposed method is proved by experiments. The main work of the thesis is as follows: 1: 1. According to the requirements of seed cutting sampling task and the requirement of breeding experts at home and abroad for seed sampling automation equipment, a crop seed cutting sampling robot system is designed and implemented, and the upper computer program is compiled. The problem of self location and direct teaching in the sampling process is solved. In order to solve the problem of self-localization of sampling operation in seed cutting sampling robot system, the geometric structure of seed sampling box is analyzed, and a self-calibration method based on geometric constraints is proposed. The self-calibration method is based on the geometric constraint relation of the coordinate points of the seed sampling box. By using the homogeneous coordinate transformation method between the sampling box coordinate system and the robot basic coordinate system, the self-positioning problem in the sampling operation is solved. The automatic acquisition of coordinate position of sampling point. 3. In order to realize the direct teaching function of robot in the sampling process, the direct teaching problem of robot is studied. An active compliance control method based on position adjustment is adopted to develop a robot direct teaching system based on towing on ER6C60. The direct teaching system adopts the active compliance control algorithm based on the force sensor. The multi-dimension force sensor is used to obtain the external teaching force / torque information, and the force signal vector after the force coordinate transformation is converted into the position adjustment quantity of the robot. The purpose of compliance control is achieved. The upper computer adopts the program based on MFC dialog box, thus realizing the robot's direct traction teaching function. 4. The experimental results show that the error between the seed coordinate and the actual position calculated by the self-calibration method based on geometric constraints is less than 0.8 mm, which meets the requirements of the project. This method is simple in measurement, low in cost and high in positioning accuracy. It is useful for solving the problem of self-localization of workpiece coordinates. The direct teaching method based on the position control strategy makes the robot track the teaching force well, and the reliability of the direct teaching system is verified by experiments. This teaching method is suitable for closed robot control system and has good generality.
【学位授予单位】：中国科学技术大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【参考文献】