科学研究
硕士论文

基于点云的既有桥梁信息模型构建方法与应用研究

来源:   作者:  发布时间:2017年10月18日  点击量:


摘要

建筑信息模型(BIM)越来越多地应用于桥梁工程的运维管理阶段,例如桥梁的结构监测、缺陷检测、更新改造与能量模拟等。桥梁运维管理系统是一个基于桥梁工程各阶段信息集成的复杂系统,组成该系统的基础是桥梁模型数据库,而获取桥梁信息模型的方式主要有全站仪测绘,图像摄影法,激光扫描法等。本文通过对比、选择点云建模的方法,即桥梁信息模型可以使用从激光扫描仪获取的密集三维测量点来构建,充分利用其快速捕获桥梁实际原貌的技术特点。

为创建有效和完整的桥梁信息模型,建模的关键问题主要包括:定义构件形状和几何尺寸,构件对象识别及描述构件之间的拓扑关系。激光扫描技术可以获取详细的几何表面信息,但在提取拓扑和语义方面却不足,但BIM技术作为数据信息的集成者,可以有效弥补该缺点。

本文的工作主要有以下两方面:(1)搜集并比较国内外常用的桥梁信息建模方法;(2)提出一套基于点云的既有桥梁信息模型构建方法,重点研究构件边界拟合算法和属性信息赋值的问题。通过收集原始数据历经数据缩减,边界检测和桥梁部件分类,使得桥梁部件被识别为不同的几何构件,最后赋予构件相关属性信息,从而构建完整的桥梁建筑信息模型。通过对华中科技大学校园内石拱桥收集的原始数据进行测试,以此验证技术的可行性并评估所提出方法的应用效果。结果表明,该方法可以快速获取既有桥梁的信息模型,并且集成桥梁运维过程中所需的数据信息。

关键词既有桥梁运维管理 三维激光扫描 BIM技术 建模方法

桥梁信息模型

ABSTRACT

Building information model(BIM) is increasingly being applied throughout a bridge's management stage for various applications, such as structure monitoring,defect detection,building renovation and energy simulation.The bridge management system is complexinvolvingvarious stages ofbuildinginformationand itsbasis is the bridge modelsdatabase. There are manymethodsto obtain bridge information models,includingthe total station mapping, image photography, laser scanning and so on. And this thesis choosesthemodeling method based on point clouddata. These as-isbridge BIM modelsare often constructed using dense 3D point measurements obtained from laser scanners using its rapid capture of original conditions.

In order to create valid and complete model,the modeling processinclude: defining the shape and geometry,object recognition and describing the topological relationships between the constituent components. Laser scanning technology can obtain detailed geometric surface information, but extraction of topology and semantics is insufficient. BIM can make up for it as an integrator of information data.

There are two aspects in the thesis: (1) Common BIM modeling methods arepresentedand evaluated; (2) Aparticularexisting bridge BIM modeling method based on point cloudsis proposed, the algorithm of boundary fitting and the information attribution are given. The collected raw data undergoesdata downsizing, boundary detection, and bridge component categorization, resulting in the building components being recognized as differentelements. Finally, the attribute information of the components is given, and the BIM model of the complete bridge is formed. The results of a test conducted on a bridge in HUST collected initial data to validate the technical feasibility and evaluate the performance of the proposed method,indicatingthat it can accelerate as-is bridge information modeling from point clouds,andalsointegrate the data needed for bridge management.

Key words:existing bridges operation and maintenance management

3D laser scanning BIM technology reconstruction method

bridges information model