钢轨轮廓匹配是钢轨廓形检测的关键步骤,决定轨道几何参数检测精度。在高速动态条件下实现检测钢轨轮廓与标准钢轨轮廓高精度自动匹配,是轨道几何参数高精度动态测量面临的重要问题。本文对钢轨轮廓高精度自动匹配方法进行了研究。首先,根据标准钢轨轮廓曲线曲率固有特征信息,对检测获取的轨腰和轨底廓形中不同圆弧、线段之间切点进行自动识别,实现不同圆弧、线段自动分割;然后,对获取的不同圆弧添加半径约束进行非线性拟合,准确提取各圆弧圆心坐标;最后,针对道岔处钢轨廓形复杂特点,提出采用Kalman滤波器对检测获取的标准轨腰和轨底特征点进行连续在线跟踪和预测,并根据结果构建道岔钢轨虚拟标准轨腰和轨底,以解决道岔处钢轨轮廓匹配问题。将该方法已在轨道检测车中进行实际应用,证明其是切实可行的。
Rail profile matching algorithm, which directly determines the accuracy of rail geometric parameters measurement results, is crucial in rail profile dynamic detection. It is an important and practical issue in the high precision dynamic measurement of rail geometric parameters to realize high precision automatic matching of the detection of rail profiles with standard rail profile under high speed dynamic conditions. This paper pres- ented a study on high-precision automatic matching algorithm for rail profile. First, based on the characteristic information on the curvature of standard rail profile curve, the tangent points between different arcs and line segments of rail waist and rail base profile that had been obtained through detection, were automatically identified, to achieve automatic segmentation of different arcs and line segments. Then, nonlinear fitting was conducted on different arcs that were subject to radius constraint to accurately extract the coordinates of the center of the circular arcs. Finally, in response to the complex characteristics of rail profile in the turnout area the Kalman filter was applied to perform continuous online tracking and prediction of the feature points of the standard rail waist and rail base and to establish virtual standard rail waist and rail base of the turnout rail based on the results in order to solve the matching problem of the rail profile in the turnout area. The proposed ap- proach proved to be practical and feasible after it was applied in the track inspection car.