基于机器视觉技术的工业机器人引导与抓取

发布时间：2018-03-31 11:51

本文选题：机器视觉　切入点：视觉引导　出处：《昆明理工大学》2017年硕士论文

【摘要】：当前制造业呈现出壁垒化、智能化、梯队化、多极化的趋势,在竞争激烈的大环境下,企业想在竞争中生存并保持长期发展,首先就要考虑生产制造成本、生产效率和产品质量等问题,只有成本降低了、效率提高了、产品质量高了,才能适应竞争激烈的市场。目前,随着人工生产成本的逐渐增加,导致对自动化的设备有很大的需求,自动化制造可以有效降低生产周期、提高生产的质量、并且可以取代复杂的和乏味的过程,减少劳动力成本,提高生产效率。机器视觉通过摄像机获取图像信息并进行处理,模仿人类的视觉图像分析功能,将分析的结果发送给机器人的控制中心,引导机器人完成指定的任务。本研究针对目前主要通过对工业机器人示教编程或离线编程来控制机器人进行定位抓取工件,高精度复杂模具件的难定位、机器人示教抓取定位、抓取误差较大等缺点,提出了一种单目视觉引导抓取技术,开发出新的机器视觉图像处理算法,研发出独立完整的一套基于库卡六轴工业机器人的机器视觉引导系统,利用KUKA六轴KR.180.R2500工业机器人为实验平台验证能较准确的定位与提取工件边缘特征,然后根据标定的视觉系统,将在像平面的特征点在世界坐标系中的坐标进行计算出来,与KUKA机器人进行通讯,进而使得能够准确的引导机器人进行精确的抓取工件。在图像处理方面,通过对工件进行系列不同算法的试验对比,得出适合特定目标的图像处理算法。在通信方面,详细研究KUKA六轴工业机器人控制器的通信接口协议,设计出了适合的工业机器人的数据通信协议,实时的与上下位机交互。针对机器视觉引导KUKA六轴工业机器人自适应抓取目标,设计了系统流程,并建立硬件和软件平台。选择的试验对象为四种不同形状、尺寸的卡盘工件,分别为五角星、正三角形、正方形卡盘工件和不规则扇形卡盘工件(均为合金钢材料),进行了位姿识别抓取实验,结果表明,本文提出的基于机器视觉技术的KUKA六轴工业机器人能够准确地抓取目标工件,位姿识别误差小,满足了本课题项目的要求,同时也可以满足工业生产中对机器人操作工件的预期要求。
[Abstract]:The current manufacturing industry is showing a trend of barrier, intelligence, echelon and multi-polarization. In the competitive environment, enterprises want to survive and maintain long-term development in the competition, first of all, they must consider the production cost. Only when the cost is reduced, the efficiency is improved and the product quality is high can the production efficiency and product quality be adapted to the fierce competition in the market. At present, with the increase of the cost of artificial production, Leading to a great demand for automated equipment, automated manufacturing can effectively reduce the production cycle, improve the quality of production, and can replace complex and tedious processes, reduce labor costs, Improve production efficiency. Machine vision acquires and processes the image information through the camera, imitates the human visual image analysis function, sends the analysis result to the robot control center, This research aims at controlling the robot to locate and grab the workpiece mainly by teaching or off-line programming to the industrial robot at present, which is difficult to locate the high precision and complex die parts, and the robot can teach the grasping localization. In this paper, a new machine vision image processing algorithm is developed, and an independent and complete machine vision guidance system based on Kuka six-axis industrial robot is developed. The KUKA six-axis KR.180.R2500 industrial robot is used as the experimental platform to verify the accuracy of locating and extracting the edge features of the workpiece. Then, according to the calibrated visual system, the coordinates of the feature points in the image plane in the world coordinate system are calculated. Communication with the KUKA robot, which makes it possible to accurately guide the robot to grab the workpiece accurately. In image processing, through a series of experiments of different algorithms for the workpiece, In the aspect of communication, the communication interface protocol of KUKA six-axis industrial robot controller is studied in detail, and the suitable data communication protocol of industrial robot is designed. Real time interaction with upper and lower computer. Aiming at machine vision guided KUKA six-axis industrial robot to grasp target adaptively, the system flow is designed, and the hardware and software platform is established. The test object is four kinds of chuck workpiece with different shape and size. As pentagram, equilateral triangle, square chuck workpiece and irregular sector chuck workpiece (all of which are alloy steel materials), the position and orientation recognition and grasping experiments are carried out. The results show that, The KUKA six-axis industrial robot based on machine vision technology is proposed in this paper, which can accurately grasp the target workpiece, and the position and pose recognition error is small, which meets the requirements of this project. At the same time, it can meet the expected requirements of robot operation in industrial production.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41;TP242

【参考文献】