基于特征组合的机械臂视觉伺服协作抓取系统  

Visual Servo Collaborative Gripping System for Robotic Arms Based on Feature Combination

作　　者：李韵辰 胡晓兵[1,3] 杜玲羽 张哲源 陈海军 LI Yunchen;HU Xiaobing;DU Lingyu;ZHANG Zheyuan;CHEN Haijun(School of Mechanical Engineering,Sichuan University,Chengdu 610065,China;Sichuan Dawn Precision Technology Co.,Ltd.,Meishan 620460,China)

机构地区：[1]四川大学机械工程学院,成都610065 [2]四川德恩精工科技股份有限公司,眉山620460 [3]宜宾四川大学产业技术研究院,宜宾644600

出　　处：《组合机床与自动化加工技术》2025年第3期15-20,共6页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：四川省科技计划重点研发项目(2022YFG0072);川大-宜宾校市战略合作项目(2020CDYB-3)。

摘　　要：当前视觉伺服抓取策略大多注重工件自身的图像特征构建及复杂的控制模型设计,导致机械臂抓取系统设计困难,适用性和灵活性低。针对以上问题提出了基于图像特征组合的机械臂视觉伺服协作抓取系统。首先,利用姿态矫正台上装载的四边形平面光源的顶点特征,根据工件的质心位置和主方向,在工件位置处组合机械臂视觉伺服特征点,保证工件具有稳定的视觉特征;其次,搭建了机械臂和姿态矫正台的闭环协作控制回路,并设计姿态矫正位置补偿环节,用于减少运动复杂度,加快系统收敛速度。实验结果表明,所提出的视觉伺服协作抓取方案,能够快速稳定地完成工件的位姿矫正工作,视觉伺服最大像素误差为6个像素点,最大姿态角度误差为3.74°,协作运动均在40次迭代内完成,具有良好的抓取准确度和适用性。Most of the current visual servo grasping strategies focus on the construction of image features of the workpiece itself and the design of complex control models,which leads to the difficulty in the design of robotic arm grasping system,and low applicability and flexibility.In this paper,a visual servo collaborative grasping system for robotic arm based on image feature combination is proposed to address the above problems.Firstly,the vertex features of the quadrilateral planar light source loaded on the attitude correction table are used to combine the robotic arm visual servo feature points at the position of the workpiece according to the center of mass position and the main direction of the workpiece,so as to ensure that the workpiece has a stable visual feature;secondly,the closed-loop collaborative control loop of the robotic arm and the attitude correction table is built,and the position compensation link of the attitude correction is designed,which is used to reduce the complexity of the movement and accelerate the convergence of the system.speed.The experimental results show that the proposed visual servo collaborative grasping scheme can quickly and stably complete the position correction of the workpiece,the maximum pixel error of the visual servo is 6 pixels,the maximum attitude angle error is 3.74°,and the collaborative motion is completed within 40 iterations,which has good grasping accuracy and applicability.

关 键 词：特征组合 多机协作 上下料抓取 视觉伺服 机械臂 

分 类 号：TH165[机械工程—机械制造及自动化] TG659[金属学及工艺—金属切削加工及机床]