地面三维激光点云数据的滑坡变形分析与预测研究

发布时间：2017-12-28 01:20

本文关键词：地面三维激光点云数据的滑坡变形分析与预测研究　出处：《昆明理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：滑坡是一种结构复杂、不确定因素很强的自然灾害体,非接触式的地面三维激光扫描技术(TLS)能快速获取高精度、高密度的点云监测数据,对滑坡体的整体变形进行科学分析和合理预测。论文对TLS进行详细分类,对其应用领域进行概括论述;对地面三维激光点云数据的处理流程进行详细的理论研究;采用鲁甸地震灾区珍珠泉滑坡点云数据,利用Goemagic Studio 12进行数据处理、建模,对两期滑坡模型进行3D比较;利用Geomagic Qualify 12对滑坡山体两期局部区域的表面模型进行3D偏差分析,对三维位移变形分量建"变形量-颜色"偏差分布图,并提取特征点、断面线进行变形位移对比分析。用Sufer 7对原始点云进行规则格网差分比较和等高线重合度分析;用Cloud Compare对两期点云实现ICP配准和局部匹配,进行Haosdorff Distance比较;利用PolyWorks中的IMInspect模块分别对两期点云做体积计算并比较体积变化等;得出了滑坡表面整体变形不明显,滑坡体稳定的结论。最后利用GM(1,1)灰色模型和GM-BP灰色神经网络模型对滑坡区域的GNSS监测数据进行变形预测,通过实际变形量和预测变形量的对比分析,得出GM(1,1)灰色模型对滑坡变形预测效果较好,GM-BP灰色神经网络模型对滑坡变形预测精度更高的结论。
[Abstract]:Landslide is a natural disaster with complex structure and uncertain factors. Non-contact ground 3D laser scanning technology (TLS) can quickly acquire high-accuracy and high-density point cloud monitoring data, and make scientific analysis and reasonable prediction for the whole deformation of landslide. The detailed classification of TLS, the fields of its application are summarized; a detailed theoretical investigation on the process of 3D laser point cloud data; the Ludian earthquake stricken areas Zhenzhuquan landslide point cloud data, using Goemagic Studio 12 data processing, modeling, model of two landslide 3D compared to 3D; deviation analysis the surface model by using Geomagic Qualify 12 to two landslides in local area, the three-dimensional displacement component construction deformation - color deviation distribution map, and extract the feature points and section line analysis of displacement and deformation contrast. For regular differential comparison and contour coincidence analysis using Sufer 7 to the original point cloud; matching of two point cloud registration and local implementation of ICP Cloud Compare, Haosdorff Distance; using IMInspect module in PolyWorks of two point cloud volume do calculate and compare the volume change; the surface deformation of the whole landslide obviously, the stability of the landslide body conclusion. Finally, the use of GM (1,1) GNSS monitoring data of grey model and grey GM-BP neural network model of the area of the landslide deformation prediction, by comparing the actual deformation and deformation prediction analysis, obtains the GM (1,1) grey forecasting model is better for landslide deformation, more precise conclusion GM-BP grey neural network model to landslide deformation forecast.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P225.2;P642.22

【相似文献】