西部黄土高原煤炭资源大规模开采导致严重的地面沉陷及生态环境损害。监测矿区地表沉陷可采用常规大地测量、合成孔径雷达干涉测量、摄影测量、激光扫描等多种技术手段，但这些方法目前都存在一定的适用条件和局限性。其中，利用无人机激光雷达点云构建沉陷区精细地形模型，通过模型叠加可高效地获取地表全盆地沉陷模型。然而，面对黄土高原矿区复杂地貌条件，无人机LiDAR扫描点云受扫描精度、地形坡度、植被覆盖、点云密度、下沉点水平位移等多因素影响，造成地面点云及所构建的地形模型和沉陷模型都存在显著的噪声和不确定性，制约着该技术在矿区沉陷监测中的实际应用。为此，本文针对西部黄土矿区地表沉陷无人机LiDAR扫描监测中点云的不确定性问题展开研究，通过现场实验、数据处理、不确定性分析和沉陷模型的修正，来提高基于无人机LiDAR点云数据构建矿区沉陷模型的精度。论文主要研究内容及结果如下：

（1）通过实验分析几种常用的点云滤波算法，确定了适用于黄土矿区不同地理环境的地面点云提取方法。基于四个实验区块的点云数据对比分析，发现在地形单一且平坦区域，高程阈值滤波算法效果较优，平均绝对误差最小，拟合优度达99.9%；在复杂地形区域，布料模拟滤波算法表现稳定，误差较小，拟合优度99.5%以上，滤波效果稳定。

（2）通过实验对比分析常用的数字高程模型（Digital Elevation Model, DEM）插值算法，确定了适用于黄土矿区不同地理环境的最佳插值算法。基于四个实验区块的点云数据对比分析了反距离权重插值、克里金插值、自然邻域插值、不规则三角网插值四种插值算法的地形建模误差，发现在复杂地形中反距离权重插值效果较佳，不规则三角网插值和自然邻域插值次之，克里金插值效果相对最差。

（3）通过实验研究分析了无人机LiDAR点云数据及DEM建模的不确定性误差及其影响因素。针对文中实验区块，结果显示点云数据的不确定性与地形坡度、栅格大小和点云密度相关。当坡度小于35º时，点云数据不确定性随坡度增加而增大；栅格大小在复杂地形区域影响较大，当栅格大小为0.5m时点云不确定性误差最小；点云密度越大时，不确定性误差越低，当点云密度超过30pts/m²时，其不确定性影响趋于稳定。

（4）针对开采沉陷过程中地表点本身水平位移所带来的沉陷模型不确定性问题，提出了基于概率积分法水平位移估算的沉陷模型不确定性修正方法。利用常规的概率积分法预计开采沉陷区内地表点在X、Y方向的水平位移量，在地形DEM建模中针对点云数据施加相应的水平向X、Y坐标修正，消除了DEM叠加生成的地表沉陷模型中由水平位移引起的不确定性误差，并通过实例验证了该方法在提高复杂地形区域地表沉陷模型精度方面的有效性。

研究结果对于复杂地貌矿区地表沉陷监测的无人机激光扫描数据处理具有一定的应用价值。

The large-scale exploitation of coal resources in the Loess Plateau in western China leads to serious ground subsidence and ecological environment damage. Conventional geodesy, synthetic aperture radar interferometry, photogrammetry, laser scanning and other technical means can be used to monitor surface subsidence in mining areas, but these methods have certain applicable conditions and limitations. Among them, the fine terrain model of subsidence area is constructed by using UAV LiDAR point cloud, and the subsidence model of the whole surface basin can be obtained efficiently by model superposition. However, in the face of complex geomorphic conditions in mining areas of the Loess Plateau, the UAV LiDAR scanning point cloud is affected by multiple factors such as scanning accuracy, terrain slope, vegetation cover, point cloud density, and horizontal displacement of subsidence point, resulting in significant noise and uncertainty in ground point cloud and the constructed terrain model and subsidence model, which restricts the practical application of this technology in mining subsidence monitoring. Therefore, this paper studies the uncertainty of midpoint cloud in surface subsidence monitoring by UAV LiDAR scanning in western loess mining area, and improves the accuracy of mining subsidence model construction based on UAV LiDAR point cloud data through field experiments, data processing, uncertainty analysis and correction of subsidence model. The main research contents and results of this paper are as follows:

(1) Through the experimental analysis of several commonly used point cloud filtering algorithms, the ground point cloud extraction method suitable for different geographical environments in the loess mining area is determined. Based on the comparative analysis of the point cloud data of four experimental blocks, it is found that in the single and flat terrain area, the elevation threshold filtering algorithm has a better effect, the average absolute error is the smallest, and the goodness of fit is 99.9%. In the complex terrain area, the cloth simulation filtering algorithm is stable, the error is small, the goodness of fit is more than 99.5%, and the filtering effect is stable.

(2) Through comparative analysis of the commonly used Digital Elevation Model (DEM) interpolation algorithm, the best interpolation algorithm suitable for different geographical environments in the loess mining area is determined. Based on the point cloud data of four experimental blocks, the terrain modeling errors of inverse distance weight interpolation, Kriging interpolation, natural neighborhood interpolation and triangulation IRregularity interpolation are compared and analyzed. It is found that the inverse distance weight interpolation has the best effect in complex terrain, followed by Triangulation IRregularity interpolation and natural neighborhood interpolation, and Kriging interpolation has the worst effect.

(3) The uncertainty error of UAV LiDAR point cloud data and DEM modeling and its influencing factors are analyzed through experimental research. For experimental blocks, the results show that the uncertainty of point cloud data is related to terrain slope, grid size and point cloud density. When the slope is less than 35º, the uncertainty of point cloud data increases with the increase of slope. The grid size has a great influence on the complex terrain area, and the point cloud uncertainty error is the smallest when the grid size is 0.5m. The higher the point cloud density, the lower the uncertainty error. When the point cloud density exceeds 30pts/m2, the uncertainty effect tends to be stable.

(4) Aiming at the uncertainty of the subsidence model caused by the horizontal displacement of the surface point itself in the process of mining subsidence, a method of revising the uncertainty of the subsidence model based on the probability integral method of horizontal displacement estimation is proposed. The conventional probability integral method is used to estimate the horizontal displacement of surface points in the X and Y directions in the mining subsidence area, and the corresponding horizontal X and Y coordinate correction is applied to the point cloud data in the topographic DEM modeling, which eliminates the uncertainty error caused by horizontal displacement in the surface subsidence model generated by the DEM superposition. The effectiveness of this method in improving the accuracy of surface subsidence model in complex terrain area is verified by an example.

The research results have certain application value for the processing of UAV laser scanning data for surface subsidence monitoring in complex landform mining areas.

王双明, 申艳军, 宋世杰, 等. “双碳”目标下煤炭能源地位变化与绿色低碳开发[J]. 煤炭学报, 2023, 48(7): 2599-2612.

[2] 徐智敏, 孙亚军, 高尚, 等. 干旱矿区采动顶板导水裂隙的演化规律及保水采煤意义[J]. 煤炭学报, 2019, 44(3): 767-776.

[3] 卞正富, 伍小杰, 周跃进, 等. 煤炭零碳开采技术[J]. 煤炭学报, 2023, 48(7): 2613-2625.

[4] 王燕. 煤田地质勘探与煤炭开发的关系[J]. 内蒙古煤炭经济, 2017(21): 33+77.

[5] 李帅. 基于SAR技术和Offset-Tracking技术的矿区形变监测研究[D]. 太原: 太原理工大学, 2021.

[6] 周健, 孙芸, 阮昊. 固态激光雷达探测器技术发展[J]. 红外, 2020, 41(05): 1-12.

[7] 何柯璐. 基于机载激光扫描特征点云的采煤沉陷信息提取[D]. 西安: 西安科技大学, 2021.

[8] Molebny V, Kamerman G, Steinvall O. Laser radar: from early history to new trends[C]//Electro-Optical Remote Sensing, Photonic Technologies, and Applications IV. SPIE, 2010, 7835: 9-38.

[9] Hyyppä J, Hyyppä H, Leckie D, et al. Review of methods of small‐footprint airborne laser scanning for extracting forest inventory data in boreal forests[J]. International Journal of Remote Sensing, 2008, 29(5): 1339-1366.

[10] Colomina I, Molina P. Unmanned aerial systems for photogrammetry and remote sensing: A review[J]. ISPRS Journal of photogrammetry and remote sensing, 2014, 92: 79-97.

[11] Liu X. Airborne LiDAR for DEM generation: some critical issues[J]. Progress in physical geography, 2008, 32(1): 31-49.

[12] 靳克强, 龚志辉, 汤志强, 等. 机载LiDAR技术原理及其几点应用分析[J]. 测绘与空间地理信息, 2011, 34(1): 144-146+150.

[13] 张熠斌. 机载LiDAR点云数据处理理论及技术研究[D]. 西安: 长安大学, 2010.

[14] 于海洋, 罗玲, 杨强, 等. 三维激光扫描技术在河道测量中的应用[J]. 测绘学报, 2015, 44(S1): 49-53.

[15] Moudrý V, Klápště P, Fogl M, et al. Assessment of LiDAR ground filtering algorithms for determining ground surface of non-natural terrain overgrown with forest and steppe vegetation[J]. Measurement, 2020, 150: 107047.

[16] Hui Z, Hu Y, Yevenyo Y Z, et al. An improved morphological algorithm for filtering airborne LiDAR point cloud based on multi-level kriging interpolation[J]. Remote Sensing, 2016, 8(1): 35.

[17] Nurunnabi A, West G, Belton D. Robust locally weighted regression techniques for ground surface points filtering in mobile laser scanning three dimensional point cloud data[J]. IEEE Transactions on Geoscience and Remote Sensing, 2015, 54(4): 2181-2193.

[18] Hu K, Li M, Xia M, et al. Multi-scale feature aggregation network for water area segmentation[J]. Remote Sensing, 2022, 14(1): 206.

[19] Lian X, Liu X, Ge L, et al. Time-series unmanned aerial vehicle photogrammetry monitoring method without ground control points to measure mining subsidence[J]. Journal of Applied Remote Sensing, 2021, 15(2): 024505-024505.

[20] Pfeifer N, Mandlburger G. LiDAR data filtering and DTM generation[M]. Topographic Laser Ranging and Scanning. CRC Press, 2017: 307-334.

[21] Kraus M, Strengert M. Pyramid filters based on bilinear interpolation[C]//International Conference on Computer Graphics Theory and Applications. SCITEPRESS, 2007, 2: 21-28.

[22] Meng X, Wang L, Silván-Cárdenas J L, et al. A multi-directional ground filtering algorithm for airborne LIDAR[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2009, 64(1): 117-124.

[23] Zhang W, Qi J, Wan P, et al. An easy-to-use airborne LiDAR data filtering method based on cloth simulation[J]. Remote sensing, 2016, 8(6): 501.

[24] 张皓, 贾新梅, 张永生, 等. 基于虚拟网格与改进坡度滤波算法的机载LiDAR数据滤波[J]. 测绘科学技术学报, 2009, 26(3): 224-227+231.

[25] 汪文琪, 李宗春, 付永健, 等. 一种多尺度自适应点云坡度滤波算法[J]. 武汉大学学报(信息科学版), 2022, 47(3): 438-446.

[26] 何培培, 万幼川, 黄桂平, 等. 城区LiDAR点云自适应坡度的滤波算法[J]. 遥感信息, 2017, 32(5): 62-68.

[27] 丁少鹏, 刘如飞, 蔡永宁, 等. 一种顾及地形的点云自适应坡度滤波方法[J]. 遥感信息, 2019, 34(4): 108-113.

[28] 隋立春, 张熠斌, 柳艳, 等. 基于改进的数学形态学算法的LiDAR点云数据滤波[J]. 测绘学报, 2010, 39(4): 390-396.

[29] 徐旺, 官云兰, 张钊, 等. 结合薄板样条插值的机载LiDAR渐进形态学滤波算法[J]. 激光与光电子学进展, 2022, 59(10): 412-421.

[30] 张刚, 刘文彬, 张男. 基于区域特征分割的密集匹配点云渐进形态学滤波[J]. 地球信息科学学报, 2019, 21(4): 615-622.

[31] 王竞雪, 张雪洋, 洪绍轩, 等. 结合形态学和TIN三角网的机载LiDAR点云滤波算法[J]. 测绘科学, 2019, 44(5): 151-156, 183.

[32] 常兵涛, 陈传法, 郭娇娇, 等. 机载LiDAR点云分块插值滤波[J]. 红外与激光工程, 2021, 50(9): 209-217.

[33] 张凡, 李华山, 江涛. 基于布模拟算法在LiDAR点云中生成数字高程模型[J]. 激光与光电子学进展, 2020, 57(13): 130104.

[34] 王丽英, 赵元丁. 一种灰度体素结构分割模型下的机载LiDAR 3D滤波算法[J]. 地球信息科学学报, 2020, 22(11): 2118-2127.

[35] 宋晨洋, 王强, 高德涵, 等. 顾及地物邻域特征的LiDAR点云布料模拟滤波优化[J]. 激光与红外, 2023, 53(1): 51-56.

[36] Serifoglu C, Gungor O, Yilmaz V. Performance evaluation of different ground filtering algorithms for uav-based point clouds[J]. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2016, 41: 245-251.

[37] Montealegre A L, Lamelas M T, De La Riva J. Interpolation routines assessment in ALS-derived digital elevation models for forestry applications[J]. Remote sensing, 2015, 7(7): 8631-8654.

[38] Zeybek M, Şanlıoğlu İ. Point cloud filtering on UAV based point cloud[J]. Measurement, 2019, 133: 99-111.

[39] 胡永杰, 程朋根, 陈晓勇, 等. 机载激光雷达点云滤波算法分析与比较[J]. 测绘科学技术学报, 2015, 32(1): 72-77.

[40] 邹正, 邹进贵, 胡海洋. 不同机载LiDAR点云滤波算法对比分析[J]. 测绘地理信息, 2021, 46(5): 52-56.

[41] 李豆, 李朋飞, 穆兴民, 等. 机载LiDAR测量复杂地形中滤波算法的比较研究[J]. 水土保持研究, 2021, 28(4): 171-178.

[42] 王泽华, 甘淑, 吕杰, 等. 泥石流沟谷地形的机载激光雷达点云滤波算法对比[J]. 兰州大学学报(自然科学版), 2023, 59(6): 770-777.

[43] Bing S, Zheng N, Li D, et al. Reconstructing DEM using TLS point cloud data and NURBS surface[J]. Transactions of Nonferrous Metals Society of China, 2015, 25(9): 3165-3172.

[44] Cashman T J. NURBS-compatible subdivision surfaces[R]. University of Cambridge, Computer Laboratory, 2010.

[45] Yang X, Strahler A H, Schaaf C B, et al. Three-dimensional forest reconstruction and structural parameter retrievals using a terrestrial full-waveform lidar instrument (Echidna®)[J]. Remote sensing of environment, 2013, 135: 36-51.

[46] Gálvez A, Iglesias A. Particle swarm optimization for non-uniform rational B-spline surface reconstruction from clouds of 3D data points[J]. Information Sciences, 2012, 192: 174-192.

[47] Lu G Y, Wong D W. An adaptive inverse-distance weighting spatial interpolation technique[J]. Computers & geosciences, 2008, 34(9): 1044-1055.

[48] Abramov O, McEwen A. An evaluation of interpolation methods for Mars Orbiter Laser Altimeter (MOLA) data[J]. International Journal of Remote Sensing, 2004, 25(3): 669-676.

[49] Tran Q B, Nguyen T T. Assessment of the influence of interpolation techniques on the accuracy of digital elevation model[J]. VNU Journal of Science: Earth and Environmental Sciences, 2008, 24(4).

[50] 王娜, 刘江, 张世荣. 截面测量数据的B样条曲面拟合[J]. 机械设计与制造, 2007(8): 151-153.

[51] 罗尧治, 公晓莺. 基于双三次B样条插值的空间结构自由曲面[J]. 空间结构, 2004, 10(2): 30-34.

[52] 喻钧, 卢延新, 胡志毅, 等. 一种改进的双线性规则格网DEM内插算法[J]. 测绘科学技术学报, 2015, 32(5): 521-524,529.

[53] 汪燕麟, 黄现民. 一种改进的Delaunay三角网内插方法[J]. 测绘与空间地理信息, 2017, 40(5): 217-219.

[54] 韩富江, 刘学军, 潘胜玲. DEM内插方法与可视性分析结果的相似性研究[J]. 地理与地理信息科学, 2007, 23(1): 31-35.

[55] 贾旖旎, 汤国安, 刘学军. 高程内插方法对DEM所提取坡度、坡向精度的影响[J]. 地球信息科学学报, 2009, 11(1): 36-42.

[56] 和娴, 董有福, 王玉成. 径向基核函数在DEM内插中的适用性分析[J]. 测绘科学, 2017, 42(2): 5-10.

[57] 杨灿灿, 王春, 江岭, 等. DEM地形描述误差的概率分布特征研究[J]. 测绘工程, 2016, 25(8): 19-23,28.

[58] 曹庆安, 涂梨平, 宋佳, 等. 不同插值方法的栅格DEM质量分析与最佳分辨率的选取[J]. 江西农业学报, 2013(3): 115-117.

[59] 张永亮, 朱美正, 李欣, 等. 基于稠密与稀疏高程点的DEM插值算法[J]. 计算机工程与应用, 2014, 50(1): 167-174.

[60] 王猛, 田丰. 散乱数据内插平原地区DEM的方法[J]. 地理空间信息, 2011(4): 40-41+44+190.

[61] 陶本藻. GIS质量控制中不确定度理论[J]. 测绘科学技术学报, 2000, 17(4): 235-238.

[62] 杨必胜, 梁福逊, 黄荣刚. 三维激光扫描点云数据处理研究进展、挑战与趋势[J]. 测绘学报, 2017, 46(10): 1509-1516.

[63] 王家晖. 点云对象不确定性模型研究[D]. 淄博: 山东理工大学, 2022.

[64] 刘春, 史文中, 刘大杰. GIS空间数据面元与线元不确定性的关系[J]. 武汉大学学报(信息科学版), 2005, 30(1): 65-68.

[65] 于海洋, 杨礼, 张春芳, 等. 基于LiDAR DEM不确定性分析的矿区沉陷信息提取[J]. 金属矿山, 2017(10): 177-187.

[66] 陈西江, 章涛, 花向红, 等. 基于熵函数的点云变形可靠性评价指标确定[J]. 大地测量与地球动力学, 2018, 38(4): 402-406.

[67] 李斌兵, 黄磊, 冯林, 等. 基于点云数据的切沟泥沙负载量不确定性研究[J]. 农业工程学报, 2014, 30(17): 183-191.

[68] 玉院和, 王金亮. DEM不同分辨率对地形因子提取的不确定性[J]. 云南地理环境研究, 2018, 30(1): 40-45.

[69] 李朋飞, 李豆, 胡晋飞, 等. 机载LiDAR监测黄土高原土壤侵蚀的能力评估[J]. 测绘学报, 2023, 52(8): 1342-1354.

[70] Jung H C, Kim S W, Jung H S, et al. Satellite observation of coal mining subsidence by persistent scatterer analysis[J]. Engineering Geology, 2007, 92(1-2): 1-13.

[71] Pesci A, Teza G, Casula G, et al. Multitemporal laser scanner-based observation of the Mt. Vesuvius crater: Characterization of overall geometry and recognition of landslide events[J]. ISPRS journal of photogrammetry and remote sensing, 2011, 66(3): 327-336.

[72] Mukupa W, Roberts G W, Hancock C M, et al. A review of the use of terrestrial laser scanning application for change detection and deformation monitoring of structures[J]. Survey review, 2017, 49(353): 99-116.

[73] Schäfer T, Weber T, Kyrinovic P, et al. Deformation measurement using terrestrial laser scanning at the hydropower station of Gabcikovo[C]//INGEO 2004 and FIG Regional Central and Eastern European Conference on Engineering Surveying, Bratislava, Slovakia. 2004.

[74] 郭文兵, 白二虎, 陈俊杰. 三维激光扫描监测开采沉陷的精度分析[J]. 煤炭科学技术, 2014(11): 85-89.

[75] 曾凯, 姜岩. 三维激光扫描技术在地表沉陷监测中的应用[J]. 地矿测绘, 2015, 31(2): 28-30.

[76] 郭超, 李建委, 冯华俊, 等. 三维激光扫描技术在矿区大坝沉陷监测中的应用[J]. 矿山测量, 2014(6): 70-72+7.

[77] 石鹏卿, 胡向德, 魏洁, 等. 三维激光扫描技术在华亭东峡煤矿塌陷区监测中的应用[J]. 矿山测量, 2019, 47(3): 97-102.

[78] 胡大贺, 吴侃, 陈冉丽. 三维激光扫描用于开采沉陷监测研究[J]. 煤矿开采, 2013, 18(1): 20-22+35.

[79] 敖建锋, 吴侃. 急倾斜煤层开采沉陷预计研究[J]. 煤矿开采, 2009, 14(5): 10-12.

[80] 刘杰, 阎跃观, 李军, 等. 基于激光点云的沉陷区水平位移求取方法研究[J]. 煤矿开采, 2016, 21(2): 99-102.

[81] 阎跃观, 戴华阳, 葛林林, 等. 基于DInSAR沉陷区超前影响角计算修正方法[J]. 深圳大学学报(理工版), 2012, 29(4): 290-294.

[82] 陈优良, 王之玉, 栗正刚, 等. 基于D-InSAR的矿山地表沉陷监测研究[J]. 化工矿物与加工, 2019, 48(8): 10-14.

[83] 王宗良, 汤伏全. 基于D-InSAR监测的黄土矿区地表沉陷特征分析[J]. 技术与创新管理, 2019, 40(3): 394-397.

[84] 汤伏全, 芦家欣, 韦书平, 等. 基于无人机LiDAR的榆神矿区采煤沉陷建模方法改进[J]. 煤炭学报, 2020, 45(7): 2655-2666.

[85] 谷延超. 机载激光雷达点云数据滤波方法研究[D]. 成都: 西南交通大学, 2014.

[86] 马东洋, 尚大帅, 保永强. 一种基于高程差粗糙度的建筑物激光脚点提取方法[J]. 测绘科学技术学报, 2013, 30(2): 158-161.

[87] 安葳鹏, 程小博, 刘雨. Fleiss' Kappa系数在贝叶斯决策树算法中的应用[J]. 计算机工程与应用, 2020, 56(7): 137-140.

[88] 梁波, 梁晓鹏, 王健, 等. 利用激光点云数据进行露天矿采剥工程量的计算方法[J]. 北京测绘, 2022, 36(1): 50-54.

[89] 丁海勇, 杨延珍, 王新志, 等. 基于移动曲面拟合方法的GPS高程转换[J]. 大地测量与地球动力学, 2010, 30(6): 82-85.

[90] 董志南, 郑拴宁, 赵会兵, 等. 基于空间插值的风场模拟方法比较分析[J]. 地球信息科学学报, 2015, 17(1): 37-44.

[91] 刘强, 陶钧, 刘旭, 等. 基于克里金插值的磁法数据格网化研究[J]. 河南科学, 2013(7): 1039-1044.

[92] 赵祥鸿, 陈欣, 刘聚. 基于二维经验模态多波束测深插值算法比较[J]. 海洋测绘, 2018, 38(1): 18-22.

[93] 党磊, 王子佳, 王光磊. 基于ArcGis不同插值方法制作水文要素等值线[J]. 东北水利水电, 2020, 38(1): 35-36+41.

[94] 刘党卫, 曹叔巍, 岳长青, 等. 海底数字高程模型在OBN勘探中的应用[J]. 物探装备, 2021, 31(2): 104-107.

[95] 刘峰, 武金辉, 张孝荣, 等. 机载激光点云构建黄土高原复杂地貌区精细DEM实验研究[J]. 测绘工程, 2023, 32(03): 44-53.

[96] 李培现, 万昊明, 许月, 等. 基于地表移动矢量的概率积分法参数反演方法[J]. 岩土工程学报, 2018, 40(4): 767-776.