钢轨磨耗动态测量中轨廓自动配准方法研究

Study on Automatic Registration of Rail Profile during Dynamic Measurement of Rail Wear

高效准确的测量钢轨磨耗,对于确保铁路运营安全至关重要。提出一种基于结构光视觉的钢轨磨耗测量中轨廓自动配准方法。首先,基于RDP数据分割算法提取的特征点,对获取的钢轨轨腰和轨底廓形中不同半径的圆弧进行自动分割;然后,对分割圆弧轮廓进行基于半径约束的非线性拟合,针对圆心拟合误差,采用二次拟合减小噪声点对拟合精度的影响,将圆弧圆心坐标作为基准点,实现了钢轨测量轮廓和标准轮廓的初步配准;最后,采用基于KD树的改进ICP算法实现钢轨轮廓的高精度配准。实验结果表明,该方法配准精度高,重复性好,钢轨磨耗动态测量精度0.2 mm,满足铁路部门的标准要求。

Efficient and accurate measurement of rail wear is essential to ensure the railway operation safety. An automatic registration method for rail profile in rail wear measurement was proposed based on structured light vision. Firstly, the circular arcs with different radius in the measured rail waist and rail base profile were automatically segmented based on the feature points, which were extracted by the RDP data segmentation algorithm. Then, nonlinear fitting processed with radius constraint was conducted on segmented arcs. Considering the fitting error of the circle center, the quadratic fitting method was applied to reduce the influence of the noises on the fitting precision. The coordinates of the centers of the circular arcs were used as reference points to realize the preliminary alignment of the measured rail profile with the standard profile. Finally, the improved ICP algorithm based on KD tree was applied to achieve high-precision registration of rail profile. The experimental results indicate that the proposed method has high registration accuracy and good repeatability. The precision of dynamic measurement of rail wear reaches 0.2 mm, which meets the requirements of the railway departments.