QQ客服 官方微博 反馈留言

题名	机器视觉在工业机器人装配中的应用研究
作者	王帅
学位类别	硕士
答辩日期	2015-05-26
授予单位	中国科学院沈阳自动化研究所
授予地点	中国科学院沈阳自动化研究所
导师	徐方
关键词	机器视觉 工业机器人 底座 减速器 智能装配
其他题名	Industrial Robot Components Assembly Based on Machine Vision Technology
学位专业	控制工程
中文摘要	随着机器人与自动化技术不断向前发展，越来越多的工业机器人被应用于搬运、装配、分拣、码垛、喷涂等工业现场，代替人类完成那些危险、枯燥或者繁重的工作。然而，目前工厂实际应用的工业机器人绝大部分仍是以示教—再现的工作方式运行，缺乏对外部信息的了解，无法根据外部条件变化实时地调整其运动轨迹，缺乏灵活性和适应性。于是人们将视觉技术引入到机器人生产线，使得机器人具有类似人眼的功能，从而大大提升了生产线的柔性和自动化程度。本文针对工业机器人无人化自动生产线上的底座与一轴减速器装配问题，提出了用机器视觉识别并测量工件位姿，将得到的工件位姿信息通过离线编程的方式引导另一台工业机器人对底座与减速器分别进行抓取、移动、放置等操作，最终实现两者的精密装配，从而实现了“用工业机器人生产工业机器人”。本文首先给出了该视觉装配系统所需要的图像预处理算法，包括图像二值化以及连通域轮廓提取等，提出了使用8邻域轮廓跟踪方法找到连通区域轮廓，并利用边界轮廓扫描的方法提取连通域边界信息；随后针对待装配的底座与一轴减速器的位姿识别分别提出了基于改进Hough变换的圆检测以及基于类链码的减速器螺钉孔偏转角识别算法；由于视觉测量得到的工件位姿信息是在图像坐标系，要引导机器人运动就要将其变换到世界坐标系，故接着给出了摄像机成像模型以及摄像机标定方法；之后，本文给出了机器人装配过程中所需要的视觉引导指令；最后，搭建了视觉装配平台，详细介绍了该视觉装配系统的作业流程，并验证了本文提出的视觉算法的有效性。实验结果表明，对于工业机器人生产线上的底座与一轴减速器装配，本文提出的基于机器视觉的工件识别定位方法配合大负载工业机器人的使用，可以较好地完成两者的精密装配。此外，该方法具有较高实时性和鲁棒性，能够满足大规模、批量化生产的需求。
索取号	TP391.41/W36/2015
英文摘要	With the continuous improvement of the automatic level of manufacturing industry, the industrial application such as transportation, assembly, sorting, palletizing, buffing and spraying which using robotic technology is becoming more and more popular. Robot replace the human labor to accomplish the dangerous or burdensome work is an irreversible trend in the future. However, nowadays most robot work based on teaching playback or offline- programming , they could not adjust their moving trajectory according to the change of outside circumstance, so they lack flexibility and adaptability. In consequence, people apply the machine vision technology to robot production line, making robot possess the visual recognition ability like human eye. Aiming at the assembly work of robot base and first axis gear reducer in industrial robot body production line, this thesis proposed the visual recognition method. The proposed algorithm could accomplish the rapid recognition and positioning of the workpiece with high precision. Combined with the application of heavy-load industrial robot, the whole system realized the intelligent assembly of robot base and first axis gear reducer. That is “manufacturing robot with robot”. Finally, the feasibility and reliability of the algorithm is verified by experiment. The first part of this thesis is the introduction of some basic digital image processing algorithms which are used in the following chapter. A method in the use of eight neighborhood contour tracking is proposed to get a fast way to detect the outline of the connected domain. In the second part, an improved Hough Transform of Circle Detection algorithm and a chain code algorithm are developed to get the position and pose of the base and first axis gear reducer. As the pose information we have obtained are in the image coordinate system , in order to get their corresponding world coordinates, the third part is the description of camera model and monocular camera calibration. The following part is the description of visual guiding instruction. To verify the visual recognition and positioning algorithm this thesis developed, we build the experimental platform in the manufacturing workshop in SIASUN Robot & Automation Co, Ltd. and conduct the experimental verification. The platform includes 210Kg 6 DOF industrial robot, three claw gripper, camera, etc. Experimental result shows that the method is effective to recognize the pose information of the workpiece, combined with the heavy-load industrial robot, the whole system accomplish the automatic assembly of robot base and first axis gear reducer with high precision. What is more, the proposed assembly scheme has good stability and robustness and can satisfy the requirement of real-time demanding.
语种	中文
产权排序	1
页码	65页
内容类型	学位论文
源URL	[http://ir.sia.ac.cn/handle/173321/16736]
专题	沈阳自动化研究所_其他
推荐引用方式 GB/T 7714	王帅. 机器视觉在工业机器人装配中的应用研究[D]. 中国科学院沈阳自动化研究所. 中国科学院沈阳自动化研究所. 2015.

个性服务

查看访问统计

相关权益政策

暂无数据

收藏/分享

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。