列车轮对内侧距结构光视觉在线动态测量研究

Train wheel-set is the key running component that has a critical influence on the safety of train transport, especially heavy haul freight cars transport. The available measurement methods of the inboard distance of train wheel-set presently can't meet the needs of the railway development in our country. In this project, some systematic studies are carried out in the aspects of measurement theory and key technologies, centering on the on-line dynamic measuring the inboard distance of train wheel-set based on structured-light vision. The main research contents are shown explicitly as follows: firstly, a structured-light vision measurement model is constructed using the perspective projective geometry and homogeneous coordinate transformation, based on which the measurement model of inboard distance of wheel-set is proposed in accordance with the Railway Wheel-set Depot Repair Specifications; secondly, a set of field calibration methods of online dynamic measurement system of inboard distance of wheel-set are proposed, and the corresponding calibration equipments are developed; then, an extraction algorithm with high precision and speed for image features of laser stripe is put forward, and a structure of the measurement system with environmental adaptability are designed; lastly, a test bed for the inboard distance of train wheel-set measurement is established on field and dynamic testing is conducted. The key technologies of the on-line dynamic measurement of the inboard distance of train wheel-set are broken through by the study of this project. It provides effective technical approach and means for on-line dynamic measurement of the inboard distance of of the wheel-set, and contributes to the conversion from 'periodical maintain' to 'state maintain' for wheel-set and the safety assurance of train transport.

列车轮对作为铁路车辆重要的走行部件，对于铁路安全运输起着关键性的作用，尤其对于重载货车运输。目前，现有轮对内侧距检测方式已不能满足我国铁路快速发展的需要。本项目围绕列车轮对内侧距结构光视觉在线动态测量，从测量原理和关键技术等方面开展系统深入研究，主要研究内容为：采用透视投影变换和齐次坐标变换建立结构光视觉测量模型；按照轮对段修规程，建立轮对内侧距测量模型；提出轮对内侧距结构光视觉在线动态测量系统现场标定方法，并研制出标定装置；提出快速高精度激光光条图像特征提取算法；设计列车轮对内侧距结构光视觉在线动态测量系统环境适应性结构；构建列车轮对内侧距结构光视觉测量试验平台，并进行现场安装和动态测试。本项目的研究将突破列车轮对内侧距结构光视觉在线动态测量关键技术瓶颈，为轮对内侧距在线动态测量提供有效技术途径和手段，为实现轨道从"周期性修"向"状态修"转变和保证铁路运输安全作出贡献。

列车轮对作为铁路车辆重要的走形部件，对于铁路安全运输起着关键性的作用。而轮对内侧距是衡量列车运行平稳和安全的一项重要几何参数，对其在线动态检测可以提前预警，防止重大事故的发生。本项目围绕列车轮对内侧距结构光视觉在线动态测量，从测量原理和关键技术等方面开展系统深入研究，取得研究成果如下：. （1）根据透视投影原理建立了基于线阵相机的点结构光视觉测量精确模型，突破了传统三角测量模型存在非线性误差的技术瓶颈，从一维位移测量扩展到了空间点的三维测量。. （2）提出了内侧距测量新方法，采用多个点结构光视觉传感器获取轮辋内侧面上空间点坐标，以内侧面平行作为约束条件重构内侧面，进而完成内侧距测量。该方法大幅提高了内侧距测量的动态性，鲁棒性强,测量设备便于现场安装。. （3）提出了基于二维移动台带动平面靶标的点结构光视觉传感器标定方法，标定过程简单，易于实现；又提出基于自由移动立体靶标的点结构光传感器标定方法，不依赖高精度外部测量设备，靶标相对传感器自由移动，适合现场高精度标定。. （4）建立了球投影与绝对二次曲线投影关系方程（IS-IAC），提出了利用球投影模型的摄像机线型参数和畸变参数的标定方法和基于球心点定位的多相机全局标定方法，解决了复杂条件下视觉传感器现场标定受杂光干扰严重的技术瓶颈。. （5）为了满足轮对内侧距结构光视觉在线动态测量系统在铁路现场恶劣环境下应用的要求，项目组对结构光视觉传感器的抗杂光、防水除尘、抗高低温、防震等进行了设计，经现场试验，满足现场恶劣环境下长期运行要求。. （6）在实验室搭建了列车轮对内侧距静态视觉测量试验平台，完成静态测试，内侧距测量精度为0.25mm；在铁路现场安装了列车轮对内侧距在线动态视觉测量试验平台，完成货车轮对内侧距动态测试，在列车以80km/h速度通过时，内侧距测量精度为0.6mm。. 本项目突破了列车轮对内侧距结构光视觉测量在线动态测量关键技术瓶颈，为轮对内侧距在线动态测量提供了有效技术途径和手段。