Liquefied natural gas (LNG) storage tanks are critical facilities for storing liquefied natural gas, making the monitoring of their outer walls essential for ensuring safety and structural integrity. Currently, deformation monitoring of LNG storage tanks primarily relies on total station measurements or terrestrial laser scanning, which suffer from low measurement efficiency and limited tank coverage. An analysis method was developed for LNG storage tank outer walls based on multi-source point cloud data fusion. By employing four data acquisition techniques—total station, terrestrial laser radar, airborne laser radar, and UAV oblique photogrammetry—and utilizing feature-based random sample consensus (RANSAC) and iterative closest point (ICP) algorithms, multi-source point clouds were accurately registered into a unified coordinate system to construct a 3D model of the tank outer wall. Quantitative analysis of tank inclination and deformation was conducted through segmented fitting and ovality calculations. Results indicate that the LNG storage tank exhibits a tilt variation of 2.58° and a deviation of 27 mm in the azimuth direction of 321.980°. The tank shows a slight tendency to tilt from the northwest. The average ovality of the LNG storage tank is 0.053%, and the average inclination and deviation of 27 mm both meet standard requirements. The recommended action is to conduct regular monitoring of the LNG storage tank. If measured values exceed standard limits, measures such as grouting reinforcement, additional support, hydraulic jacking, or gravity balancing can be implemented.
LI Bai-song, ZHANG Yun-wei, LEI Jiang-kai, LU Te, WU Feng-an.
Analysis Method of LNG Storage Tank Outer Wall Based on Multi-source Point Cloud Data[J].
Science Technology and Engineering, 2026, 26(3): 926-933 DOI:10.12404/j.issn.1671-1815.2501629
目前中外学者关于LiDAR和无人机摄影测量技术开展了大量的研究。Deng等[6]提出了一种3D激光雷达分阶段实时地面分割算法。Lee等[7]使用斜向图像实现了建筑模型详细的重建和提高定位精度。Petras等[8]研究了机载激光雷达点云的点密度变化对重建模型误差的影响规律,并提出了采用抽样和均匀化方式减少误差的方法。杨健等[9]提出了一种利用无人机激光雷达(LiDAR)技术与倾斜摄影技术完成测图的方法。刘慧等[10]提出基于Potree结构的激光点云与建筑信息模型(building information modeling,BIM)点云差异检测方法。徐郡彬等[11]提出激光与影像融合的古建筑建模方法,通过无人机航摄与三维激光扫描获取数据,采用迭代最近点算法(iterative cloest point,ICP)算法实现异源点云配准,重建精细化三维模型。朱旺煌等[12]提出融合空间密度检测的单体提取算法。首先结合过绿指数与法向量分类植被点,再通过密度聚类识别误分区域并重分类,最终分割出完整建筑物单体。
无人机倾斜摄影测量是使用运动恢复结构(structure from motion, SFM)技术从无人机拍摄二维图像中生成三维点云并提取有关三维结构和相机位置信息重建三维场景的技术。SFM算法主要步骤包括特征提取、匹配、相机姿势估计、三维重建和密集点云生成。首先,从图像中提取特征用于识别不同图像中的对应点,如角落或边缘。其次,在成对的图像之间匹配提取的特征建立最佳匹配关系。然后,通过求解摄像机位置和方向估计图像的相机姿态,由三角测量法计算匹配特征点的三维位置生成稀疏点云。最后,计算场景中所有点的三维位置生成密集点云。点云三维坐标与图像像素坐标之间的映射关系为
(2)基于ICP算法的精配准。为提高粗配准后点云配准精度,使用ICP进行精配准。使用奇异值分解(singular value decomposition,SVD)得到最优转换矩阵(旋转和平移)的解析解。首先计算最优旋转矩阵R,再计算最优平移矩阵T,为了去除平移的影响,先将源点云和目标点云都转换到质心坐标下,即
ZhangPuyang, BaiYu, LiuYang, et al. The experimental research on hypersonic flow around the double ellipsoid[J]. Industrial Construction, 2022, 52(8): 127-131.
GuoMing, YanBingnan, ZhouTengfei, et al. Application of LiDAR technology in deformation analysis of Yingxian Wooden Pagoda[J]. Journal of Architecture and Civil Engineering, 2020, 37(2): 109-117.
WuLingxiao, DuanZhugeng, JiangXueliang. The experimental research on hypersonic flow around the double ellipsoid[J]. Science Technology and Engineering, 2023, 23(8): 3153-3161.
FanBing, HuXiaoning, MaLiang, et al. UAV LiDAR and tilt photogrammetry technology in complex river channels fusion research in engineering mapping[J]. Port and Waterway Engineering, 2023, 48(1): 11-17.
[11]
DengW, ChenX, JiangJ. A staged real-time ground segmentation algorithm of 3D LiDAR point cloud[J]. Electronics, 2024, 13(5): 841.
[12]
LeeE, ParkS, JangH, et al. Enhancement of low-cost UAV-based photogrammetric point cloud using MMS point cloud and oblique images for 3D urban reconstruction[J]. Measurement, 2024, 226(2): 114158.
[13]
PetrasV, PetrasovaA, McCarterJ B, et al. Point density variations in airborne lidar point clouds[J]. Sensors, 2023, 23(3): 1593.
YangJian, HuangXiaobing, LiKang, et al. Application of UAV airborne LiDAR technology and oblique photogrammetry in topographic mapping[J]. Beijing Surveying and Mapping, 2024, 38(7): 974-979.
LiuHui, LiuYuhang, ZhongChen. Potree based building change detection with BIM to point clouds of lidar[J]. Science Technology and Engineering, 2024, 24(3): 1176-1183.
XuJunbin, SongJianping, HuangXiaolin, et al. Method of optimizing regional ancient architectural models using laser and image fusion[J]. Science Technology and Engineering, 2025, 25(1): 37-43.
ZhuWanghuang, LiuRong, GongXunqiang, et al. Extraction of individual buildings based on point clouds from oblique photogrammetry[J]. Science Technology and Engineering, 2024, 24(30): 12843-12852.
[22]
GuoM, ZhuL, ZhaoY, et al. Intelligent extraction of surface cracks on LNG outer tanks based on close-range image point clouds and infrared imagery[J]. Journal of Nondestructive Evaluation, 2024, 43(3): 84.
[23]
FengY Y, LiH, LiC K, et al. 3D modelling method and application to a digital campus by fusing point cloud data and image data[J]. Heliyon, 2024, 10(17): e36529.
WangZonghui, WenShilin. Research on the 3D modeling of high-rise buildings based on multi-source data fusion[J]. Science and Technology Information, 2024, 22(5): 5-9.
DuShili, FuGui, TaoJia, et al. The experimental research on hypersonic flow around the double ellipsoid[J]. Journal of Huaiyin Institute of Technology, 2023, 32(3): 78-87.
HuangBeisheng, WangYunkai. 3D reconstruction of complex regional cultural relics based on multi-source data fusion[J]. Beijing Surveying and Mapping, 2023, 37(1): 43-47.
ChenLijia, WangYingtao, WangKai, et al. A 3D reconstruction method of virtual port with multi-source data fusion[J]. Journal of Wuhan University of Technology, 2024, 48(6): 1-12.
QuHonghui, YuJianli, WangZhengmao. Method of measuring tree diameter by fitting circle with total station[J]. Forest Inventory and Planning, 2024, 49(6): 6-9.
National Energy Administration.Code requirements for design and construction of concretes structures for the containment of refrigerated liquedied gases: SY/T 7304—2016[S]. Beijing: Petroleum Industry Press, 2016.
Ministry of Housing and Urban-Rural Development.Code for design of liquefied natural gas receiving terminal: GB 51156—2015[S]. Beijing: China Planning Press, 2015.