基于机器视觉的物料分拣工业机器人关键技术研究

发布时间：2018-06-11 15:44

本文选题：库卡工业机器人 + 机器视觉　；参考：《深圳大学》2017年硕士论文

【摘要】：随着工业机器人技术在生产制造领域的广泛应用,越来越多的工业机器人被应用到制造领域,代替人工完成对生产线上物料的分拣。虽然通过对工业机器人示教编程可以完成分拣任务,但是往往对周边设备要求较高,并且成本也较高。对于多目标,工件的形状、尺寸和摆放位置不确定的分拣任务,如何让机器人完成工件的识别和定位,并准确的抓取目标是使分拣机器人更加智能化的关键。将机器视觉技术运用到工业机器人当中,使机器人具有人眼的功能,通过视觉引导使工业机器人完成对目标的抓取具有重要意义。1本文对库卡KR 6 R700 sixx工业机器人进行了视觉分拣技术研究。单个工业机器人不能定位分拣工作平面上物料的位置,为了使机器人能确定物料的位置,基于图像传感技术设计了视觉识别、定位方法,实现了机器视觉定位。将机器视觉技术运用于库卡工业机器人使机器人确定目标物料的位置。建立了工业机器人的连杆坐标系,根据连杆坐标系确定了D-H参数表,建立了机器人的运动学方程,并求解了库卡工业机器人的正逆解。采用五次多项式曲线对库卡机器人进行了关节空间的轨迹规划。利用正运动学解,采用随机数的方法分析了机器人的工作空间。工业摄像机作为分拣机器人的图像传感器。分析了视觉系统中的四个坐标系,并分析了工业摄像机的线性成像模型和非线性成像模型。研究了基于平面靶标的工业摄像机标定方法,构建了工业摄像机的标定实验,获得了工业摄像机的内外参数矩阵。在分拣工作平面上的18cm×18cm区域进行了视觉定位实验,分析了视觉定位误差。最大定位误差小于0.9mm。工业摄像机获取工件的图像后,采用双边滤波器对图像进行滤波,采用了大津法阈值分割方法提取工件图像。提取了工件的轮廓,通过判断轮廓的面积来完成对工件的识别。通过提取轮廓的中心获得了工件的像素位置。设计了角度计算方法获取工件的角度信息。分析PC与库卡工业机器人的位姿数据通信,确定了通信数据的字符串格式。通过搭建PC端数据发送服务器实现了PC与机器人的数据通信。最后搭建了视觉分拣系统并进行分拣实验。实验结果表明,库卡KR 6 R700 sixx工业机器人能够完成对目标工件的视觉分拣,提高了分拣机器人的自适应性。
[Abstract]:With the wide application of industrial robot technology in the field of production and manufacture, more and more industrial robots have been applied to the field of manufacturing, instead of manually completing the sorting of materials on the production line. Although the sorting task can be accomplished by instructing the industrial robot, it often requires high peripheral equipment and high cost. How to make the robot complete the identification and localization of the work piece and grasp the target accurately is the key to make the sorting robot more intelligent for the task of sorting the multi-target, the shape, the size and the position of the work piece which is uncertain. The application of machine vision technology to industrial robots, so that the robot has the function of the human eye, It is of great significance to make the industrial robot complete the target capture through visual guidance. In this paper, the visual sorting technology of KR6 R700 sixx industrial robot is studied. A single industrial robot can not locate the position of the material in the sorting work plane. In order to make the robot determine the position of the material, a vision recognition and location method is designed based on the image sensing technology, and the machine vision positioning is realized. The machine vision technology is applied to the Kuka industrial robot to determine the position of the target material. The connecting rod coordinate system of the industrial robot is established, the D-H parameter table is determined according to the connecting rod coordinate system, the kinematics equation of the robot is established, and the forward and inverse solutions of the Kuka industrial robot are solved. The joint space trajectory planning of Kuka robot is carried out by using the polynomial curve of the fifth degree. Using the positive kinematics solution, the workspace of the robot is analyzed by the method of random number. Industrial cameras are used as image sensors for sorting robots. Four coordinate systems in the vision system are analyzed, and the linear and nonlinear imaging models of industrial cameras are analyzed. The calibration method of industrial camera based on planar target is studied. The calibration experiment of industrial camera is constructed and the internal and external parameter matrix of industrial camera is obtained. The visual localization experiment was carried out in the 18cm 脳 18cm area of the sorting work plane, and the visual positioning error was analyzed. The maximum positioning error is less than 0.9mm. After the industrial camera acquires the image of the workpiece, the two-sided filter is used to filter the image, and the threshold segmentation method is used to extract the image of the workpiece. The contour of the workpiece is extracted, and the recognition of the workpiece is completed by judging the area of the contour. The pixel position of the workpiece is obtained by extracting the center of the contour. The angle calculation method is designed to obtain the angle information of the workpiece. The data communication between PC and Kuka industrial robot is analyzed, and the string format of communication data is determined. The data communication between PC and robot is realized by building PC data sending server. Finally, the visual sorting system is built and the sorting experiment is carried out. The experimental results show that the KR6 R700 sixx industrial robot can achieve the visual sorting of the target workpiece and improve the self-adaptability of the sorting robot.
【学位授予单位】：深圳大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41;TP242.2

【参考文献】