基于逐步逼近法的桥梁曲线轮廓检测方法

Detecting Bridge Curve Contour Based on Stepwise Approximation Method

目的为准确获取高耸曲线桥塔的外形,利用TLS与UAV的联合扫描获取数字点云模型,并提出基于逐步逼近法的曲线轮廓提取方法。方法首先,在保证测量精度的前提下,给出了TLS的最佳设站距离及标靶位置的优化方法;其次,针对某实际斜拉桥的高耸桥塔,利用基于空间相似变换公式的配准方法,将TLS与UAV扫描得到的点云数据融合成全桥点云模型。结果采用逐步逼近的方法,筛选得到了该高耸曲线桥塔轮廓边界处的点云数据,并通过数据拟合,得到了相应的实测曲线方程;通过与设计值的比较,得出实际曲线桥塔的整体安装误差小于6%。结论与传统单点测量方法相比,笔者所提方法不但丰富了测量内容,还可为后续桥梁运营状态评估提供基准数据。

To accurately obtain the shape of the tall curved bridge tower,the digital point cloud model is received by the joint scanning of TLS and UAV.A curve contour extraction method based on the stepwise approximation method is proposed.First,under the premise of ensuring the measurement accuracy,the optimal setting distance of TLS and the optimization method of target position are given.Secondly,for the high towers of an actual cable-stayed bridge,the point cloud data scanned by TLS and UAV is fused into a full bridge point cloud model using a registration method based on a spatial similarity transformation formula.Finally,the step-by-step approach is used to filter the point cloud data at the contour boundary of the tall curved bridge tower,and through data fitting,the corresponding measured curve equation is obtained.By comparing with the design value,the overall installation error of the curved tower is given.Compared with the traditional single-point measurement method,it enriches the measurement content and provides benchmark data for subsequent bridge operation status evaluation.