基于激光点云的滑坡监测数据处理方法研究

发布时间：2018-03-01 19:06

本文关键词： 滑坡点云三角网数据处理模糊计算　出处：《江西理工大学》2015年硕士论文　论文类型：学位论文

【摘要】：滑坡是我国最为常见且多发的一种地质灾害。由于滑坡监测预防的时效性差,尽管国家投入大量的资金进行滑坡治理,但我国滑坡地质灾害依然较为严重,据统计每年因滑坡灾害造成的直接经济损失高达20亿元,因此滑坡监测与预防迫在眉睫。目前,三维激光扫描仪技术已经成熟地应用于滑坡监测,为滑坡的预防奠定了数据基础,则滑坡监测预防的核心在于数据后处理,然而数据后处理存在两方面问题：一是由于现用的曲面重构方法应用于滑坡表面重构中曲面边界适应性较差等缺点,导致影响滑坡表面模型的构建精度；二是数据分析方面,现今的数据分析方法虽然能通过建立滑坡表面的整体模型,来对滑坡做出整体变形监测,但未能考虑滑坡变形与影响因素之间的联系,这对于探索滑坡变形规律与滑坡稳定性分析还远远不够。这些不足直接导致滑坡监测数据处理分析效率低和准确性差。为此,通过选用曲面重建方法中更加灵活的三角网模型,并结合逐点插入算法和分块算法进行三角网的构建,建立滑坡表面DEM进行点云数据的特征信息提取,进而为滑坡变形分析提供资料；基于模糊计算原理,确定滑坡变形与各影响因子之间的作用大小。同时采用C#语言研制基于点云的滑坡数据处理原型系统,该原型系统具有界面操作友好、快速获取滑坡特征信息、滑坡变形与各影响因子关系可定量计算等特点。将其运用于某滑坡监测实例当中,得到了较为形象直观的滑坡表面模型,并基于此获得了该滑坡的变形等特征信息,并通过影响因子的量化,得出引起该滑坡变形的影响因子作用力最大的是人类工程活动。实验表明,原型系统功能运行可靠,通过引用三角网建模方法和滑坡模糊计算综合评估模型,实现了滑坡特征信息快速获取以及滑坡影响因子的定量计算,使得滑坡监测预防准确可靠,满足基于点云的滑坡数据处理与分析的研究工作需求,同时也为滑坡数据处理提供全新的思路。
[Abstract]:Landslide is one of the most common and frequent geological disasters in China. Because of the poor timeliness of landslide monitoring and prevention, although the state has invested a lot of funds for landslide control, the landslide geological hazard in China is still relatively serious. According to statistics, the direct economic loss caused by landslide disasters is as high as 2 billion yuan per year, so landslide monitoring and prevention are urgent. At present, 3D laser scanner technology has been used in landslide monitoring. It lays a data base for landslide prevention, so the core of landslide monitoring and prevention lies in data post-processing. However, there are two problems in data post-processing: one is that the surface boundary adaptability of the current surface reconstruction method is poor, which results in the influence of the landslide surface model construction accuracy; the other is the data analysis. Although the present data analysis method can monitor the whole deformation of the landslide by establishing the whole model of the landslide surface, it fails to consider the relationship between the landslide deformation and the influencing factors. This is far from enough to explore the law of landslide deformation and the analysis of landslide stability, which directly leads to the low efficiency and poor accuracy of landslide monitoring data processing and analysis. Combined with point-by-point insertion algorithm and block algorithm to construct triangle network, establish the landslide surface DEM to extract the feature information of point cloud data, and then provide data for landslide deformation analysis, based on the principle of fuzzy calculation, At the same time, using C # language, a landslide data processing prototype system based on point cloud is developed. The prototype system has friendly interface and can obtain landslide characteristic information quickly. The relationship between landslide deformation and various influencing factors can be calculated quantitatively. By applying it to a landslide monitoring example, a more visual landslide surface model is obtained, and the characteristic information such as deformation of the landslide is obtained. Through the quantification of the influencing factors, it is concluded that the most influential force of the factors causing the landslide deformation is the human engineering activities. The experimental results show that the function of the prototype system is reliable. By using the triangle network modeling method and the comprehensive evaluation model of landslide fuzzy calculation, the landslide characteristic information is obtained quickly and the quantitative calculation of landslide influence factors is realized, which makes landslide monitoring and prevention accurate and reliable. It meets the research requirements of landslide data processing and analysis based on point cloud, and also provides a new way of thinking for landslide data processing.
【学位授予单位】：江西理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P642.22

【参考文献】