【作者】 张奇；

【导师】 田威； 张新龙；

【作者基本信息】 南京航空航天大学 ， 工程硕士（专业学位）， 2021， 硕士

【摘要】 航空装配过程中对涂胶质量的要求越来越高,机器人自动化涂胶是实现涂胶高质量、高效率的发展趋势,采用视觉检测技术实现涂胶质量检测是保证涂胶质量的关键。本文结合航空涂胶的需求与特点,针对基于单目线结构光的涂胶胶体质量检测、饱和线结构光光条中心提取、视觉检测系统与机器人法兰盘的手眼标定等关键技术展开深入的研究。主要完成的工作有:(1)分析了基于视觉的机器人自动化涂胶质量检测的需求与难点,设计了涂胶质量视觉检测的总体方案,提出了视觉检测系统的标定与三维测量的数学模型,建立了涂胶胶体质量特征参数求解算法。(2)提出了基于高斯拟合迭代的线结构光光条中心提取方法,分析了线结构光光条形成机理与影响因素,提出了饱和线结构光光条灰度值估计算法,并通过仿真验证了该算法的有效性;提出了自适应阈值灰度重心法和B样条曲线平滑处理方法,提高了光条中心提取精度。(3)完成了机器人自动化涂胶检测系统的硬件通讯与软件设计,首先分析了机器人手眼标定原理,提出了基于中间测量系统的机器人手眼标定方法,然后研究了各坐标系建立与转换矩阵求解算法,最后得出手眼关系矩阵优化方法。(4)对本文线结构光光条中心提取方法、机器人手眼标定方法以及机器人自动化涂胶质量检测系统进行了实验验证,结果表明视觉检测系统宽度和高度测量误差均小于0.08 mm;机器人手眼标定平移误差小于0.232 mm,旋转误差小于0.165°,验证了本文光条中心提取方法、机器人手眼标定方法以及机器人自动化涂胶质量检测系统的可行性与有效性。更多还原

【Abstract】 In the process of aviation assembly,the requirement of gluing quality is higher and higher.Robot automatic gluing is the development trend to achieve high quality and high efficiency of gluing.Using visual inspection technology to achieve gluing quality detection is the key to ensure the gluing quality.In this paper,combined with the requirements and characteristics of aviation gelatinization,the key technologies such as quality detection of gelatinization gel based on monocular structured light,center extraction of saturation line structured light strip,vision detection system and hand eye calibration of robot flange are deeply studied.The main tasks are as follows:(1)This paper analyzes the requirements and difficulties of robot automatic gluing quality detection based on vision,designs the overall scheme of gluing quality visual detection,puts forward the mathematical model of visual detection system calibration and three-dimensional measurement,and establishes the algorithm for solving the characteristic parameters of gluing quality.(2)This paper studies the method of extracting the center of line structured light stripe based on Gaussian fitting iteration,analyzes the formation mechanism and influencing factors of line structured light stripe,proposes the gray value estimation algorithm of saturated line structured light stripe,and verifies the effectiveness of the algorithm through simulation;proposes the adaptive threshold gray gravity method and B-spline curve smoothing method,which improves the accuracy of light stripe center extraction.(3)The hardware communication and software design of robot automatic gluing detection system are completed.Firstly,the principle of robot hand eye calibration is analyzed,and a robot hand eye calibration method based on intermediate measurement system is proposed.Then,the establishment of each coordinate system and the algorithm of transformation matrix are studied.Finally,the optimization method of hand eye relationship matrix is obtained.(4)The experimental results show that the width and height measurement errors of the vision detection system are less than 0.08 mm,and the robot hand eye calibration translation error is less than0.232 mm The results show that the method of extracting the light stripe center,the method of robot hand eye calibration and the robot automatic gluing quality detection system are feasible and effective.更多还原