轨道车辆静态限界三维视觉测量系统
文献类型：学位论文
| 作者 | 郭庆耀
|
| 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 |
| 导师 | 李论 |
| 关键词 | 静态限界 视觉测量 全局标定 光顺算法 |
| 其他题名 | 3D vision system for the measurement of the rail vehicles's static gauge |
| 学位专业 | 控制工程 |
| 中文摘要 | 静态限界是指停靠在平直轨道上的空载机车车辆的截面投影轮廓，是保证机车车辆安全运行的最基本要求。机车车辆在实际部署运行之前，需要先进行静态限界检测。传统的静态限界测量通常采用限界规完成，存在操作复杂、测量精度低、效率不高、依赖人工等缺点。近年来，随着对静态限界测量精度和测量效率的要求不断提高，传统的检测方法已不能满足这种新的要求。 针对这种情况，本文在归纳分析现有的机车车辆静态限界检测方法的基础上，设计了一种基于工业相机和线结构光传感器的三维视觉测量系统。该限界测量系统主要由硬件和软件两部分组成。其中，硬件部分主要包括工业控制计算机、龙门架和13个测量单元等。13个测量单元按照设计的间隔分布在龙门架上，是测量系统中执行测量任务的关键组成部分。每个测量单元均由一个工业相机和一个线结构光传感器组成，并且只负责整个截面中部分截面的测量工作。通过整合13个测量单元的测量数据，便可获得整个截面的测量数据。 本文研究了基于激光跟踪仪的全局标定方法，推导了测量单元相机坐标系到全局坐标系的数学转换公式，得到了每个测量单元相机坐标系到全局坐标系的齐次变换矩阵。 本文还研究了从光条图像反求空间点三维坐标的方法。针对测量数据中存在的数据重叠情况，提出了一种基于向量的重叠部分剔除算法。然后对处理后的数据进行了曲线光顺处理，直观地显示了测量结果。最后研究了截面超限判断算法，并提出了一种改进的叉乘判别法。 实践证明，该系统可以适应多个不同型号机车车辆的静态限界测量工作，测量精度小于0.5mm，测量单个截面所需时间小于100ms，整车测量时间小于20分钟，很好地满足了机车车辆静态限界测量项目的技术要求。目前，该测量系统已投入实际使用，大大提高了机车车辆静态限界测量工作的效率，为企业带来了很好的经济效益。 |
| 索取号 | TP391.41/G95/2015 |
| 英文摘要 | Static gauge is the projection profile of the the rail vehicle, which rests on the straight track without load, and is the basic requirement to make sure that the rail vehicle can run safely. Before practically deployed, the the rail vehicle need to go through the static gauge detection.The traditional method of static gauge commonly use limit gauge. This method which is complicated to operate has some obvious shortcomings that its precision and efficiency are both lowly, along with high dependence on manpower. In recent years, with increasing requirement for accuracy and efficiency of measurement, the traditional detection method cannot meets the new requirements. Due to the circumstance, this paper designs a kind of non-contact system, which is based on industrial camera and structured light stripe vision sensor, to measure the static gauge of the rail vehicle, based on the analysis of existing static gauge detection method. The system is composed of software and hardware. The hardware mainly includes industrial control computer, portal frame and 13 measuring unit, etc. The 13 measuring unit, which is the key part of measurement system to perform the measurement function, is distributed in the portal frame according to a certain interval. Each measuring unit is composed of an industrial camera and a structured light stripe vision sensor, and is only responsible for surveying a part of the whole section. By integrating the measurement data of the 13 measuring unit, We can get the measurement data of the whole section. We studied the method of the global calibration, which is based on laser tracker, and deduced the transform formula from the measuring unit’s camera coordinate to the global coordinate, gained the homogeneous transformation matrix for each measurement unit. In this paper, we also study the method of deduce the 3D coordinates of the space points from the image of the light stripe. Due to the overlap of the measurement data, we proposed an algorithm based on vector to eliminate the overlapping part of the mesurement data. After that we did a curve faring of the processed data and got an viusal result of the measurement data. At the last, we studied the algorithm to judge whether the section is over gauge, and proposed an improved discrimination method used multiplication cross. Practice has proved that the system can adapt to the static gauge measurement of a number of different types of rail vehicle, with the measuring accuracy less than 0.5mm and the time of measuring a single section less than 100ms, as well as the time of measuring the section of a whole rail vehicle less than 20 minutes. It well meet the technical requirement of the project of the rail vehicle’s static gauge measurement. |
| 语种 | 中文 |
| 公开日期 | 2015-08-31 |
| 产权排序 | 1 |
| 源URL | [http://ir.sia.ac.cn/handle/173321/16763]  |
| 专题 | 沈阳自动化研究所_装备制造技术研究室 |
| 推荐引用方式 GB/T 7714 | 郭庆耀. 轨道车辆静态限界三维视觉测量系统[D]. 中国科学院沈阳自动化研究所. 2015. |
入库方式： OAI收割
来源：沈阳自动化研究所
浏览0
下载0
收藏0
其他版本
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。