基于点云数据的网架结构提升变形智能监测方法

Intelligent Deformation Monitoring for Lifting Space Frames Based on Point Cloud Data

变形监测是保障网架结构提升施工安全的重要手段之一,目前传统方法仅能实现对局部点的变形监测,难以实现对整体结构的变形监测。三维激光扫描技术可全覆盖地得到已完成结构的精准点云数据,这为解决上述难题提供了新思路。为此,以泸州高铁站为工程背景,开展基于点云数据的网架结构提升变形智能监测研究,包括点云数据预处理、点云数据非刚性配准以及提升变形可视化三个方面。针对网架结构提升前后的点云数据非刚性配准问题,基于聚类算法、随机采样一致性算法、图结构方法以及正交普氏分析等提出了集球心智能定位、球心粗匹配、球心非刚性配准于一体的非刚性配准算法。研究结果表明,基于点云数据的网架结构提升变形智能监测方法高效、全面且实用。

Deformation monitoring is one of the most essential means of ensuring construction safety for lifting space frames. Current traditional methods only enable deformation monitoring at local points, but not for the whole structure. Three-dimensional(3D) laser scanning technology can capture accurate point clouds of as-built structures through full-coverage scanning, thus providing a innovative solution to the above issue. To this end, the research on intelligent deformation monitoring for lifting space frames based on point cloud data, including point cloud data preprocessing, non-rigid matching of point clouds and lifting deformation visualization, was carried out based on a practical engineering project, namely Luzhou Railway Station. Based on clustering algorithms, random sample consensus, graph structural methods and orthogonal procrustes analysis, a non-rigid matching algorithm integrating intelligent sphere positioning, coarse sphere matching, non-rigid sphere matching was proposed for the non-rigid matching of point clouds of space frames before and after lifting. The results showed that the proposed intelligent deformation monitoring approach based on point cloud data was efficient, comprehensive and practical.