| 基于主动立体视觉铁路道岔在位高精度测量方法 |
投稿时间：2025-12-10 修订日期：2025-12-17  点此下载全文 |
| 引用本文： |
| 摘要点击次数: 3 |
| 全文下载次数: 0 |
|
|
| 中文摘要:高速铁路道岔的在位几何参数对运行安全影响显著，为实现其在线精密监测，本文提出一种基于主动立体视觉的高精度测量方法。系统由双目工业相机与近红外激光条纹组成，可同步获取尖轨与基本轨的空间特征，并在轨向坐标系下精确定义密贴度与爬行量。依据一阶不确定度传播模型，构建了系统设计准则与精度保障方法；在算法方面，提出融合卡尔曼滤波与几何拓扑约束的对称拓扑中心线（STC）提取方法，有效提升振动与噪声条件下条纹中心的亚像素定位稳定性。实验表明，该方法长期运行可保持毫米级精度：标准工件最大误差 0.07 mm，现场密贴度与爬行量平均误差不超过 0.35 mm，零点漂移小于 0.06 mm。结果验证了系统的实时性与鲁棒性，可用于道岔状态的在线监测与预测性维护。 |
| 中文关键词:在位测量 主动立体视觉 条纹中心提取 卡尔曼滤波 高速铁路道岔 |
| |
| High-Precision In-Situ Measurement Method for Railway Turnouts Based on Active Stereo Vision |
|
|
| Abstract:The in-situ geometric parameters of high-speed railway turnouts are critical to operational safety. To achieve accurate online monitoring, this study proposes a high-precision measurement method based on active stereo vision. The system integrates a binocular industrial camera and a near-infrared laser stripe projector to simultaneously acquire the spatial features of the switch rail and the stock rail, and accurately define the rail–seat tightness and creeping displacement within a rail-aligned coordinate system. Based on first-order uncertainty propagation, system design criteria and precision assurance strategies are established. At the algorithmic level, a Symmetric Topological Centerline (STC) extraction method is developed by combining Kalman filtering with geometric–topological constraints, enabling robust sub-pixel localization of laser stripe centers under vibration and noise interference. Experimental results demonstrate that the proposed method maintains millimeter-level accuracy during long-term operation: the maximum deviation on standard test pieces is 0.07 mm, the average errors of tightness and creeping measurements in field tests are within 0.35 mm, and the zero-drift remains below 0.06 mm. These results verify the real-time performance and robustness of the system, indicating its suitability for continuous turnout monitoring and predictive maintenance. |
| keywords:in-situ measurement active stereo vision stripe center extraction Kalman filtering high-speed railway turnout |
| 查看全文 查看/发表评论 下载pdf阅读器 |
位访问者,今日访问
次