轨道曲率预瞄算法及空间几何滞后机理分析

Pre-scanning Algorithm of Railway Track Curvature and Analysis of Geometric Lag Principle

实时获取轨道线路的曲率可为诸多轨道车辆主动控制技术提供关键输入参量。通过相机获取前方轨道线路的图像从而预瞄曲率,是解决信号滞后问题的有效途径。为此,首先基于既有的机器视觉技术进行合理假设,进而提出一种通过轨道线路中心线横向偏移量映射线路曲率的预瞄算法。随后,重点阐述探测曲率相对于预瞄点处曲率的滞后机理,研究结果表明由预瞄点形成的虚拟车体引入了几何惯性。在此基础上,给出探测曲率与车辆所处位置实际曲率之间的空间位置修正公式。最后,通过Simpack和Matlab仿真验证了上述结论。上述研究结果可为视觉技术应用于轨道曲率测量领域提供一定的理论参考。

Obtaining the railway track curvature in real time can provide key input parameters for many active control technologies of railway vehicles.It is an effective way to completely solve the problem of signal lag to obtain the image of the front track line through camera and pre-scanning the curvature.Therefore,some reasonable assumptions based on the existing machine vision technology is made and a pre-scanning algorithm is proposed to map the track curvature through the lateral offset of the track center line.Then,the lag principle of the measured curvature relative to the curvature at the pre-scanning point is further analyzed.The results show that the virtual vehicle body formed by the pre-scanning point introduces geometric inertia.On this basis,the spatial correction formula between the measured curvature and the actual curvature of the vehicle position is given.Finally,the above conclusions are verified by Simpack and Matlab simulation.The above research results can provide some theoretical reference for the application of vision technology in the field of railway vehicle active control.