基于机器视觉的小型管道机器人研究

发布时间：2018-07-07 13:44

本文选题：管道机器人 + PIC单片机　；参考：《天津理工大学》2017年硕士论文

【摘要】：近年来,随着国家工农产业的飞速发展,管道作为水、石油、天然气等资源的重要运输方式,其需求量逐年提高。但由于管道通常长年处于恶劣的工作环境下,受地质灾害、化学腐蚀等外在因素的影响,会发生管道内部损坏、管道泄漏甚至导致爆炸等严重危险事故,给人们生活带来严重隐患。为了保证管道运输在安全可靠的前提下给人们生产生活带来便利,世界各国都制定了严格的法律法规,强制企业实施管道检测。因此管道检测机器人应运而生,它能代替检测人员进入管道内部检测,保证人员安全且提高检测效率。本文自主设计研发了一套具有远程监控功能的螺旋驱动式管道检测机器人。首先,在行走机构设计上,通过对比当前主流管道机器人行走方式的优缺点,结合被检测管道的实际情况,设计了带有弹簧压壁装置的螺旋轮式行走机构,测试结果表明该机构能实现在各种角度下的管道内稳定行走;在底层硬件系统设计上,根据管道机器人的技术需求,设计以PIC单片机为控制核心,辅以霍尔测速、超声波测距、倾角测量等传感器,并在前端带载CCD网络摄像头,实现对管道内部环境以及机器人自身状态的检测。其次,以Visual Studio2010作为开发环境,以C#作为主要编程语言,结合底层硬件系统,设计监控中心的上位机软件系统,该系统通过Wifi无线的方式与管道机器人进行通讯,能够实现远程控制管道机器人的运动、实时显示管道内状态信息、显示管道内部环境图像并对其进行处理等功能;作为管道机器人检测工作后的延伸,设计以ROS机器人操作系统为平台,以C++为编程语言的六轴机械臂控制系统。最后,根据实际管道内部缺陷图像特点,设计针对管内腐蚀图像的处理算法,由于摄像头成像的特点对采集的管道图像进行柱面展开处理,为了防止设备通道对图像造成影响采用LAB彩色空间,对图像进行分割处理采用分块+Otsu+K-means算法的形式,研究表明该算法在处理管道缺陷图像时达到的分割质量较其他算法更好,运算速度更快。
[Abstract]:In recent years, with the rapid development of the national industrial and agricultural industry, as an important transportation mode of water, oil, natural gas and other resources, the demand of pipeline is increasing year by year. However, the pipeline is damaged by the factors such as geological disasters, chemical corrosion and other factors, due to the pipeline usually in a bad working environment. Serious dangerous accidents, such as explosion, bring serious hidden danger to people's life. In order to ensure that pipeline transportation is convenient to people's production and life on the premise of safety and reliability, all countries in the world have formulated strict laws and regulations and forced enterprises to carry out pipeline inspection. Therefore, the pipeline inspectors have come into being, which can replace the inspectors to enter. The internal inspection of the pipeline ensures the safety of the personnel and improves the detection efficiency. This paper independently designs and develops a set of spiral driven pipeline detection robot with remote monitoring function. First, the design of the walking mechanism is designed to compare the current running mode of the mainstream pipeline robot, and the actual situation of the detected pipeline is designed. The test results show that the mechanism can achieve stable walking in various angles. Based on the design of the hardware system, the PIC microcontroller is designed based on the technical requirements of the pipeline robot, with Holzer velocity measurement, ultrasonic distance measurement, angle measurement and so on. With the CCD network camera on the front end, the internal environment of the pipeline and the self state of the robot are detected. Secondly, using the Visual Studio2010 as the development environment, using C# as the main programming language and the underlying hardware system, the upper computer software system of the monitoring center is designed. The system is entered into the pipeline robot by Wifi wireless mode. It can realize the movement of long-distance control pipe robot, display the state information in the pipeline in real time, display the internal environment image of the pipeline and process it. As the extension of the pipeline robot detection work, the six axis manipulator control system based on the ROS robot operating system and the programming language of C++ is designed. Then, according to the characteristics of the internal defect image of the actual pipeline, a processing algorithm for the internal corrosion image of the tube is designed. Because of the characteristics of the camera imaging, the cylinder image is processed by the cylinder image. In order to prevent the device channel from affecting the image, the LAB color space is used, and the block +Otsu+K-means algorithm is used for the image segmentation. The research shows that the segmentation quality of the algorithm is better than that of other algorithms in dealing with pipeline defect images, and the computation speed is faster.
【学位授予单位】：天津理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【相似文献】