Research On Dam Dynamic Three-dimensional Deformation Moinitoring Based On Laser Scanning Technology

Dam deformation monitoring(DDM) is an important monitoring item of dam safety monitoring, which has played an important role in flood protection and safety of reservoir. At present, there are some shortcomings in the means and methods of DDM.As the technology of Lidar developing rapidly, a new and advanced solution of DDM has been proposed. We can get rich and complete imformation(such as point cloud, image and intensity) by Lidar efficiently, precisely and actively. And this imformation can be used into DDM work. In this paper, the object of my research is DDM of Earth-Rock Dam. So, I do some work about obtaining and processing point cloud data of the dam and try to solve some key issues in using Lidar to do DDM work.DDM by point-by-point measure is outdated and inefficient, since Lidar can obtain the whole 3D data of the dam rapidly and accurately, which can provide more reliable scientific basis for safety assessment of the dam. In this paper, I have studied the techniques and methods of using point cloud data to solve DDM and imploved the accuracy and efficiency of Earth-Rock Dam deformation monitoring.In order to implove registration accuracy of mass point cloud data, a method of automatic registration for multi-station point cloud data based on sphere target is studied, researched the algorithms of fitting for sphere target and index of mass point cloud data, then showed these methods has high efficiency and precision by experiments. DDA(Dam Deformation Analysis) is the key and core issue of DDM. So this paper bring up two methods of DDA that bases on data of mass point cloud’s profile and based on precise 3D model of the dam. These algorithms solve the problems in hatch fitting and DDA. Also, I have improved building of precise 3D surface model of dam based on adaptive surface voxels, which solved the issues of inefficient triangulation, surface with holes, inconsistent normals and overlapping patches.After analyzing and comparing the methods of DDA that bases on DEM, on neighboring points and on C2M(Cloud to Mesh), I considered the uncertainty of mass point cloud data, then proposed a method based on multiscale mass point cloud roughness of the confidence interval which solved the problem of parameter setting in core point calculation, roughness calculation, confidence intervals at different spatial scales. Then I proofed its efficiency and precision by some experiments and validated it’s accuracy by comparing with the method based on precise 3D model of the dam.