• Peer Review
  • Non-profit
  • Global Open Access
  • Green Channel for Rising Stars
Volume 10 Issue 6
Dec.  2023
Turn off MathJax
Article Contents

HUANG Junda, XU Rouna, TANG Siyao. Fast Analysis Method of Overhead Line Tree Barrier Based on Vectorized Power Line[J]. SOUTHERN ENERGY CONSTRUCTION, 2023, 10(6): 146-152. doi: 10.16516/j.gedi.issn2095-8676.2023.06.017
Citation: HUANG Junda, XU Rouna, TANG Siyao. Fast Analysis Method of Overhead Line Tree Barrier Based on Vectorized Power Line[J]. SOUTHERN ENERGY CONSTRUCTION, 2023, 10(6): 146-152. doi: 10.16516/j.gedi.issn2095-8676.2023.06.017

Fast Analysis Method of Overhead Line Tree Barrier Based on Vectorized Power Line

doi: 10.16516/j.gedi.issn2095-8676.2023.06.017
  • Received Date: 2022-10-22
  • Rev Recd Date: 2023-01-17
  • Available Online: 2023-12-26
  • Publish Date: 2023-11-10
  •   Introduction  With the continuous and rapid growth of the number of tree barrier inspections, visible light aerial photogrammetry has become the most important inspection method for hidden dangers of tree barriers in transmission lines of power supply bureaus in various cities. Since airborne laser radar cannot perform rapid iteration on a large number of point clouds, it is difficult to obtain high-precision point cloud power lines with visible light aerial photogrammetry. In the case of flourishing trees, it is difficult to comprehensively judge and analyze the relationship between tree tops, tree crowns and power lines, which results in problems such as misjudgment and omission of hidden dangers of tree barriers in the process of power line inspection. In order to solve this kind of problem, a solution of acquiring vectorized power line by laser radar is proposed. Through the intelligent analysis of vegetation information collected by multi-period visible light aerial photogrammetry, the solution of tree barrier is obtained.   Method  In view of the comprehensive situation that the image data collected by visible light aerial photogrammetry couldn't obtain high-precision point cloud power lines, it was difficult to simultaneously and comprehensively analyze the relationship between tree tops, tree crowns and power lines, and laser point cloud data couldn't be quickly iterated in large quantities, an overhead line tree barrier rapid analysis and processing system for multi-source spatio-temporal data of vectorized power lines was developed, which was suitable for the current conventional tree barrier inspection means, had comprehensive functions, saved resources, and was suitable for transmission and distribution networks.   Result  Through the three-dimensional real scene display of laser point cloud and visible light image data, the function of identifying and analyzing hidden dangers of tree barriers in power lines is realized. The vectorized power lines are complete and accurate. At the same time, combined with oblique photography technology, the data analysis of the hidden dangers of tree barriers results can identify 100% of the number of tree barrier hidden dangers.   Conclusion  The technical solution is accurate and effective, improves the efficiency and accuracy of the inspection of hidden dangers of tree barriers, and can provide guidance for practical application.
  • [1] 林火煅, 陈杰, 薛骅淳, 等. 基于LSD与统计分析的航拍图像电力线提取方法 [J]. 应用科技, 2019, 46(2): 30-34. DOI:  10.11991/yykj.201811001.

    LIN H D, CHEN J, XUE H C, et al. Segmenting power lines on the aerial photos based on line segment detector and statistical analysis [J]. Applied science and technology, 2019, 46(2): 30-34. DOI:  10.11991/yykj.201811001.
    [2] 陈驰, 麦晓明, 宋爽, 等. 机载激光点云数据中电力线自动提取方法 [J]. 武汉大学学报(信息科学版), 2015, 40(12): 1600-1605. DOI:  10.13203/j.whugis20130573.

    CHEN C, MAI X M, SONG S, et al. Automatic power lines extraction method from airborne LiDAR point cloud [J]. Geomatics and information science of Wuhan University, 2015, 40(12): 1600-1605. DOI:  10.13203/j.whugis20130573.
    [3] 吴争荣, 樊灵孟, 吴新桥, 等. 基于机载激光点云的输电线路树障隐患快速检测方法 [J]. 应用激光, 2022, 42(3): 128-134. DOI:  10.14128/j.cnki.al.20224203.128.

    WU Z R, FAN L M, WU X Q, et al. Fast detection method for tree barrier of transmission line using airborne laser point cloud [J]. Applied laser, 2022, 42(3): 128-134. DOI:  10.14128/j.cnki.al.20224203.128.
    [4] 吴华, 刘海燕, 丁高峰, 等. 复杂环境中电力线激光点云的自动提取 [J]. 激光技术, 2020, 44(4): 509-514. DOI:  10.7510/jgjs.issn.1001-3806.2020.04.019.

    WU H, LIU H Y, DING G F, et al. Automatic extraction of power lines from laser point clouds in complex environments [J]. Laser technology, 2020, 44(4): 509-514. DOI:  10.7510/jgjs.issn.1001-3806.2020.04.019.
    [5] 吴建军, 陈灵, 李磊, 等. 机载LiDAR点云中电力线的提取和重建研究 [J]. 激光技术, 2019, 43(4): 500-505. DOI:  10.7510/jgjs.issn.1001-3806.2019.04.012.

    WU J J, CHEN L, LI L, et al. Power line extraction and reconstruction from airborne LiDAR point cloud [J]. Laser technology, 2019, 43(4): 500-505. DOI:  10.7510/jgjs.issn.1001-3806.2019.04.012.
    [6] 林祥国, 张继贤. 架空输电线路机载激光雷达点云电力线三维重建 [J]. 测绘学报, 2016, 45(3): 347-353. DOI:  10.11947/j.AGCS.2016.20150186.

    LIN X G, ZHANG J X. 3D power line reconstruction from airborne LiDAR point cloud of overhead electric power transmission corridors [J]. Acta geodaetica et cartographica sinica, 2016, 45(3): 347-353. DOI:  10.11947/j.AGCS.2016.20150186.
    [7] 张秋雁, 杨忠, 陈聪, 等. 一种复杂环境下的电力线检测方法 [J]. 应用科技, 2019, 46(5): 51-56. DOI:  10.11991/yykj.201901009.

    ZHANG Q Y, YANG Z, CHEN C, et al. A power line detection method in a complex background [J]. Applied science and technology, 2019, 46(5): 51-56. DOI:  10.11991/yykj.201901009.
    [8] 王诗超, 刘嘉畅, 刘展志. 基于高斯混合聚类的海上风电出力特征曲线提取方法及其在电量平衡计算中的应用 [J]. 南方能源建设, 2023, 10(1): 39-47. DOI:  10.16516/j.gedi.issn2095-8676.2023.01.005.

    WANG S C, LIU J C, LIU Z Z. Extraction method of offshore wind power output characteristic curve based on Gaussian mixture model and its application in electric quantity balance [J]. Southern energy construction, 2023, 10(1): 39-47. DOI:  10.16516/j.gedi.issn2095-8676.2023.01.005.
    [9] 张传明, 杨玲玲, 刘敏, 等. 面向海量智能终端的云定位系统设计与实现 [J]. 武汉大学学报(信息科学版), 2021, 46(12): 1872-1880. DOI:  10.13203/j.whugis20210249.

    ZHANG C M, YANG L L, LIU M, et al. Design and implementation of cloud positioning system for massive intelligent terminals [J]. Geomatics and information science of Wuhan University, 2021, 46(12): 1872-1880. DOI:  10.13203/j.whugis20210249.
    [10] 巫聪云, 刘斌, 沈梓正, 等. 基于边缘计算的故障录波主站信息快速智能处理方法 [J]. 南方能源建设, 2021, 8(2): 91-98. DOI:  10.16516/j.gedi.issn2095-8676.2021.02.014.

    WU C Y, LIU B, SHEN Z Z, et al. Fast and intelligent information processing method for fault recorder master station based on edge computing [J]. Southern energy construction, 2021, 8(2): 91-98. DOI:  10.16516/j.gedi.issn2095-8676.2021.02.014.
    [11] 麻卫峰, 王金亮, 王成, 等. 电力线点云精细提取与重建的模型残差实现 [J]. 测绘科学, 2020, 45(2): 60-66. DOI:  10.16251/j.cnki.1009-2307.2020.02.009.

    MA W F, WANG J L, WANG C, et al. Precise extraction and reconstruction of power line by using model residual [J]. Science of surveying and mapping, 2020, 45(2): 60-66. DOI:  10.16251/j.cnki.1009-2307.2020.02.009.
    [12] 刘金生, 张磊, 赵振杰. 配电网网格可靠性目标制定方法研究 [J]. 南方能源建设, 2022, 9(1): 109-114. DOI:  10.16516/j.gedi.issn2095-8676.2022.01.016.

    LIU J S, ZHANG L, ZHAO Z J. Research on the methods of gridding reliability object setting for distribution network [J]. Southern energy construction, 2022, 9(1): 109-114. DOI:  10.16516/j.gedi.issn2095-8676.2022.01.016.
    [13] 周访滨, 邹联华, 刘学军, 等. 栅格DEM微地形分类的卷积神经网络法 [J]. 武汉大学学报(信息科学版), 2021, 46(8): 1186-1193. DOI:  10.13203/j.whugis20190311.

    ZHOU F B, ZOU L H, LIU X J, et al. Micro landform classification method of grid DEM based on convolutional neural network [J]. Geomatics and information science of Wuhan University, 2021, 46(8): 1186-1193. DOI:  10.13203/j.whugis20190311.
    [14] 刘焱雄, 郭锴, 何秀凤, 等. 机载激光测深技术及其研究进展 [J]. 武汉大学学报(信息科学版), 2017, 42(9): 1185-1194. DOI:  10.13203/j.whugis20150779.

    LIU Y X, GUO K, HE X F, et al. Research progress of airborne laser bathymetry technology [J]. Geomatics and information science of Wuhan University, 2017, 42(9): 1185-1194. DOI:  10.13203/j.whugis20150779.
    [15] 朱依民, 田林亚, 毕继鑫, 等. 基于无人机机载LiDAR的电力线点云提取与重建 [J]. 激光技术, 2021, 45(5): 554-560. DOI:  10.7510/jgjs.issn.1001-3806.2021.05.003.

    ZHU Y M, TIAN L Y, BI J X, et al. Power line point cloud extraction and reconstruction based on UAV-borne LiDAR [J]. Laser technology, 2021, 45(5): 554-560. DOI:  10.7510/jgjs.issn.1001-3806.2021.05.003.
    [16] 史洪云, 虢韬, 王迪, 等. 基于激光点云的电力线悬挂点定位方法 [J]. 激光技术, 2020, 44(3): 364-370. DOI:  10.7510/jgjs.issn.1001-3806.2020.03.017.

    SHI H Y, GUO T, WANG D, et al. Power line suspension point location method based on laser point cloud [J]. Laser technology, 2020, 44(3): 364-370. DOI:  10.7510/jgjs.issn.1001-3806.2020.03.017.
    [17] 时磊, 虢韬, 彭赤, 等. 电力线激光点云的分割及安全检测研究 [J]. 激光技术, 2019, 43(3): 341-346. DOI:  10.7510/jgjs.issn.1001-3806.2019.03.010.

    SHI L, GUO T, PENG C, et al. Segmentation of laser point cloud and safety detection of power lines [J]. Laser technology, 2019, 43(3): 341-346. DOI:  10.7510/jgjs.issn.1001-3806.2019.03.010.
    [18] 李靖, 钱建国, 王伟玺, 等. 基于机载激光点云的电力线自动提取方法 [J]. 激光技术, 2021, 45(3): 362-366. DOI:  10.7510/jgjs.issn.1001-3806.2021.03.017.

    LI J, QIAN J G, WANG W X, et al. Power line automatic extraction method based on airborne laser point cloud [J]. Laser technology, 2021, 45(3): 362-366. DOI:  10.7510/jgjs.issn.1001-3806.2021.03.017.
    [19] 曾旭, 陈伯建, 潘磊, 等. 基于机载激光点云的电网绝缘子识别方法 [J]. 激光技术, 2023, 47(1): 80-86. DOI:  10.7510/jgjs.issn.1001-3806.2023.01.012.

    ZENG X, CHEN B J, PAN L, et al. Power grid insulator identification method based on airborne laser point cloud [J]. Laser technology, 2023, 47(1): 80-86. DOI:  10.7510/jgjs.issn.1001-3806.2023.01.012.
    [20] 谭永殿, 黎旭, 唐锋华, 等. 基于5G无人机在树障修剪过程中的研究 [J]. 电子测试, 2021(15): 73-75. DOI:  10.3969/j.issn.1000-8519.2021.15.024.

    TAN Y D, LI X, TANG F H, et al. Research on the application of 5G UAV in the process of tree barrier pruning [J]. Electronic test, 2021(15): 73-75. DOI:  10.3969/j.issn.1000-8519.2021.15.024.
    [21] 许强, 陆会燕, 李为乐, 等. 滑坡隐患类型与对应识别方法 [J]. 武汉大学学报(信息科学版), 2022, 47(3): 377-387. DOI:  10.13203/j.whugis20210618.

    XU Q, LU H Y, LI W L, et al. Types of potential landslide and corresponding identification technologies [J]. Geomatics and information science of Wuhan University, 2022, 47(3): 377-387. DOI:  10.13203/j.whugis20210618.
  • 通讯作者: 陈斌, bchen63@163.com
    • 1. 

      沈阳化工大学材料科学与工程学院 沈阳 110142

    1. 本站搜索
    2. 百度学术搜索
    3. 万方数据库搜索
    4. CNKI搜索

Figures(6)  / Tables(2)

Article Metrics

Article views(224) PDF downloads(59) Cited by()

Related

Fast Analysis Method of Overhead Line Tree Barrier Based on Vectorized Power Line

doi: 10.16516/j.gedi.issn2095-8676.2023.06.017

Abstract:   Introduction  With the continuous and rapid growth of the number of tree barrier inspections, visible light aerial photogrammetry has become the most important inspection method for hidden dangers of tree barriers in transmission lines of power supply bureaus in various cities. Since airborne laser radar cannot perform rapid iteration on a large number of point clouds, it is difficult to obtain high-precision point cloud power lines with visible light aerial photogrammetry. In the case of flourishing trees, it is difficult to comprehensively judge and analyze the relationship between tree tops, tree crowns and power lines, which results in problems such as misjudgment and omission of hidden dangers of tree barriers in the process of power line inspection. In order to solve this kind of problem, a solution of acquiring vectorized power line by laser radar is proposed. Through the intelligent analysis of vegetation information collected by multi-period visible light aerial photogrammetry, the solution of tree barrier is obtained.   Method  In view of the comprehensive situation that the image data collected by visible light aerial photogrammetry couldn't obtain high-precision point cloud power lines, it was difficult to simultaneously and comprehensively analyze the relationship between tree tops, tree crowns and power lines, and laser point cloud data couldn't be quickly iterated in large quantities, an overhead line tree barrier rapid analysis and processing system for multi-source spatio-temporal data of vectorized power lines was developed, which was suitable for the current conventional tree barrier inspection means, had comprehensive functions, saved resources, and was suitable for transmission and distribution networks.   Result  Through the three-dimensional real scene display of laser point cloud and visible light image data, the function of identifying and analyzing hidden dangers of tree barriers in power lines is realized. The vectorized power lines are complete and accurate. At the same time, combined with oblique photography technology, the data analysis of the hidden dangers of tree barriers results can identify 100% of the number of tree barrier hidden dangers.   Conclusion  The technical solution is accurate and effective, improves the efficiency and accuracy of the inspection of hidden dangers of tree barriers, and can provide guidance for practical application.

HUANG Junda, XU Rouna, TANG Siyao. Fast Analysis Method of Overhead Line Tree Barrier Based on Vectorized Power Line[J]. SOUTHERN ENERGY CONSTRUCTION, 2023, 10(6): 146-152. doi: 10.16516/j.gedi.issn2095-8676.2023.06.017
Citation: HUANG Junda, XU Rouna, TANG Siyao. Fast Analysis Method of Overhead Line Tree Barrier Based on Vectorized Power Line[J]. SOUTHERN ENERGY CONSTRUCTION, 2023, 10(6): 146-152. doi: 10.16516/j.gedi.issn2095-8676.2023.06.017
    • 随着多旋翼无人机的快速普及,可见光航测技术成为众多树障测量分析技术中应用最为广泛、综合性价比最高的技术。然而树障隐患分析中对点云的精度要求极高,难以直接使用多旋翼无人机的数据进行分析[1],特别是在树木繁盛的情况下,难以同时全方位观察树顶、树冠和电力线之间的情况下,经常造成误判、漏判。采用激光点云[2]数据进行分析数据的整合、分析工作量较大,通常采集时又伴随着耗费大量的人力、高成本的设备与处理效率低等问题。

      本研究使用高精度点云进行电力线[3]的矢量化提取,结合可见光航测的低精度点云的植被进行树障隐患分析。该系统可以降低增量点云的采集成本、提高树障处理效率、增加可见光点云分析树障的精准度。以广东省区域为例,利用电力线矢量化和可见光点云文件,开展研究区域内便捷的树障隐患分析与工况分析并对所提取和制作的成果数据进行定位精度分析与评价。结果表明:电力线矢量化在树障分析隐患分析工程中具有可用性与适用性。该系统对树障隐患结果可以识别100%的树障隐患数,降低增量点云的采集成本,提高70%以上的树障处理效率。

    • 本次研究选择机载激光雷达扫描输电走廊获得高精度点云数据进行后续矢量化电力线提取[4-6]。电力线走廊激光点云数据采集的技术路线是:首先对测试区域环境进行考察[7],在考察过程中确定标靶和扫描仪的合适位置,利用这些工具保证测量数据的完整性、减少原始数据量及测试站数量、降低盲区的概率、提高航拍的工作效率。

    • 系统处理操作流程图如图1所示。

      Figure 1.  System processing operation flow chart

    • 依据相邻两座电塔连接的电力线俯视图呈直线且互相平行的特征,使用RANSAC电力线分条提取算法对分段电力线点云进行循环提取,得到各条电力线的点云。

      RANSAC算法[3]思路:(1)随机选择电力线俯视图点云中任意两点,通过这两个点确定一条直线L;(2)其距离阈值设置为d,计算电力线俯视图点云中与直线L的几何距离,其距离小于t的点构成局内点集S(L),否则为局外点(噪声点),统计局内点个数;(3)重复若干次随机选择,得到直线和相应的局内点集S( ),S( ),…,S( );(4)使用最小二乘法求最大局内点集的最佳拟合直线,作为电力线水平投影点云的最佳匹配直线。直线拟合的最小二乘法计算公式为:

      $$ \sum _{{i}=1}^{{n}}{({{Y}}_{{i}}-\widehat{{{Y}}_{{i}}})}^{2}={{\rm{min}}} $$ (1)

      其中表示最大局内点集点ii=1, 2 ,..., n)的坐标y值,表示点集内点ii=1, 2, ..., n)的坐标x值根据拟合直线方程计算得到的坐标y值。

      RANSAC电力线分条[3]提取算法中需要设置两个参数:采样次数和距离阈值,为了保证在随机抽样中至少有一组局内点构成的样本,距离阈值采用分段电力线投影直线最小间距的2/5,采样次数的数学关系公式为:

      $$ K=\frac{\mathrm{l}\mathrm{o}\mathrm{g}(1-z)}{\mathrm{l}\mathrm{o}\mathrm{g}(1-{w}^{n})} $$ (2)

      式中:

      K ——采样次数;

      z ——获取1组好样本的概率,一般设置为99%;

      w ——电力线点云中局内点的比例,一般在初始时设置为一个较小值,如0.1,然后迭代更新;

      n ——模型参数需要估计需要的最小局内点个数,电力线水平投影直线拟合中最少需要2个点。

      经过滤波的电力线走廊点云数据不可避免地存在植被点、地面点和电塔点组成的噪声点,因此RANSAC电力线分条提取算法每提取一条电力线后,就需要对电力线点云进行一次高程排查,以确保电力线点云提取的准确率,提高算法对噪声点的鲁棒性。高程排查的依据为电力线点在半径较小的邻域内高程变化不大且不存在突变,对提取的电力线点依次进行半径搜索,数量相对较少的噪声点与邻域内的电力线点在高程上存在突变,剔除噪声点后的点云即为单条电力线点云提取最终的结果。单档电力线分条提取结果展示图如图2所示。

      Figure 2.  Single-file power line strip extraction result display diagram

      单条电力线的曲线拟合模型:电力线理想情况下拟合时,在理论力学中其描述的空间形态为数学模型中的悬链线模型,依据已有研究结果表明抛物线方程近似等于悬链线方程的表达,相较悬链线方程更适合电力线三维重建,其模型的重建效率和精度。在曲线拟合前,优先获取求取电力线的方向,其次将某条电力线投影到获取的电力线方向和Z轴所在平面中,最后采用最小二乘原理计算得到电力线曲线模型的最佳参数[8]

    • 将矢量化后的电力线结果信息记录在数据库中,成果文件提交至存储空间并反馈[9-10]。电力线模板可以通过本系统直接进行本次以及后续的树障隐患分析,用户也可以将电力线模型通过不同坐标系的转换与数据形式获取自己所需的数据应用于其他平台系统。系统所有的电力线模型通过一张图的形式进行展示,精细化模型后续可应用于建筑施工预警、区域工程监管、生态工程监管,山体滑坡危害等情况对经济社会发展提供可靠的支撑和保障。

    • 本次研究低精度点云使用倾斜摄影技术制作成三维模型的点云。点云重叠度从80%降低到30%,飞行方式使用从以往蛇型环绕方式改为杆塔区间U型。

      进行低精度点云采集时,存在电力线的点云大量缺失的情况。为避免点云存在电力线影响后续分析效果,将低精度点云中的电力线进行抽取并去除。

      通过比较两个杆塔之间的杆塔点位位置信息,将矢量化电力线与低精度点云进行位置的校准与修改[11]。低精度点云结合矢量化电力线展示图如图3所示。

      Figure 3.  Low-precision point cloud combined with vectorized power line display

    • 根据电力行业标准和规范对安全距离的规定要求,利用分类后点云数据[3],自动进行输电导线与地面、树木等之间间距自动计算,完成输电线路树障隐患检测[6]。树障隐患自动分析技术具体流程是:根据可见光点云导线自动提取技术提取出完整准确的导线点云,以及基于可见光的植被提取方法提取用于架空线路树障分析的植被点云,使用KdTree遍历导线点云,将导线搜索点添加到植被点云中,根据国家规定的安全距离设置垂直距离阈值R,将阈值R作为KdTree[12]半径搜索的半径值,位于该半径内的植被点视为树障隐患部位,相应植被点被标记,从植被点云中删除导线搜索点,直到所有导线点完成搜索。输出被标记的植被点,即为该输电线路的树障分析隐患点,为后续核实修复工作提供原始资料。某段35 kV输电线路可见光点云的树障隐患分析结果如下图所示,该段导线到植被最短距离为4 m,不满足国家规定的安全距离,因此该段输电线路存在树障隐患[13-14]

      判断为隐患时将该区域的点全部标记为红色树障隐患,再通过获取距离隐患最近电力线点坐标、区域隐患点中最近电力线一点坐标进行连线与净空距离计算形成隐患信息。隐患缺陷等级包括严重、异常、注意。安全距离设置与缺陷等级相互关联,不同的隐患缺陷等级通过电力行业标准和规范设置不同的安全距离。

    • 本研究生成的矢量化电力线[14-18]还可以对输电线路中的实际状况进行模拟,导线在考虑自身应力的内在因素以及外界的风速、覆冰、温度等不同环境中引起矢量化电力线的弧垂变化,计算此时电力线与植被的距离,分析生成工况树障隐患报告。隐患分析实验统计结果如表1所示。

      实验点云隐患总
      数/个
      隐患1/m隐患2/m隐患3/m隐患4/m隐患5/m隐患6/m隐患7/m
      电力线分析72.853.483.230.441.133.823.50
      矢量化分析73.023.573.320.771.423.923.67
      偏差00.170.090.090.330.290.100.17

      Table 1.  Statistical results of hidden danger analysis experiment

      本研究将在温度为50摄氏度的情况下对矢量化电力线的工况隐患分析与点云电力工况隐患分析进行对比,隐患分析实验统计结果如表1 所示,模型分析最大误差为0.38 m,隐患总数误差为0。可得出在极端天气中利用矢量化电力线进行工况模拟隐患分析的结果与激光点云进行工况模拟隐患分析结果差异较小且准确可靠。

    • 结合曲线多项式方程得到电力线的分条提取结果,将获取的电力线投影到电力线方向与Z轴构成的平面上,再利用最小二乘原理对抛物线进行处理,得出最佳拟合曲线方程沿X轴每0.01 m生成一个电力线三维矢量节点,近似表达电力线矢量。如图4所示,该图为某档电力线在原始点云中的三维重建结果。在实验中电力线的重建精度采用原始电力线点云与矢量化模型的最佳拟合点进行最小距离、平均距离、最大距离等多项指标进行对比。其中该档的矢量化精度统计表如表2所示,在实验结果中的平均值0.042 m为最大误差。由该结果得出实验误差在可接受范围内,矢量化导线方法可行。单档电力线三维重建实验统计结果如表2所示。单档电力线矢量化结果如图4所示。

      Figure 4.  Single-file power line vectorization result

      实验数据$ {\mathrm{d}\mathrm{i}\mathrm{s}}_{\mathrm{m}\mathrm{e}\mathrm{a}\mathrm{n}} $/cm$ {\mathrm{d}\mathrm{i}\mathrm{s}}_{\mathrm{m}\mathrm{a}\mathrm{x}} $/cm$ {\mathrm{d}\mathrm{i}\mathrm{s}}_{\mathrm{m}\mathrm{i}\mathrm{n}} $/m电力线长度/m
      电力线13.592 6913.363 77.324 22 × 10−4281.5
      电力线23.928 6212.311 02. 082 42 × 10−3280.3
      电力线34.061 8822.912 67.000 78 × 10−4281.4
      电力线44.144 4512.882 39.153 05 × 10−4280.2
      电力线54.167 6714.958 33.222 69 × 10−3282.1
      电力线63.966 5516.971 81.119 44 × 10−3282.0
      电力线73.946 3911.320 49.085 07 × 10−4280.8
      电力线83.945 8511.459 71.178 41 × 10−3280.4

      Table 2.  Statistical results of single-file power line 3D reconstruction experiment

    • 本研究将矢量化电力线的隐患分析与点云电力隐患分析进行对比,隐患分析结果统计结果如表2所示,模型拟合最大误差为0.33 m,隐患总数误差为0。得出低精度点云加上矢量化电力线在隐患分析结果精度较在可接受范围内、隐患识别率100%,准确度较好,矢量化电力线的树障隐患分析结果准确可靠。激光点云与矢量化点树障分析结果如图5图6所示。

      Figure 5.  Analysis results of laser point cloud tree barrier

      Figure 6.  Analysis results of vectorized point cloud tree barrier

    • 研究表明通过利用高精度点云矢量化电力线与低精度点云相结合的方式,可以保证电力线三维重建精度和树障分析准确度的情况下大量降低树障巡检的成本且减少飞行的危险性。该情况还能有效发挥数据整合的优势,提供跨系统跨平台复用。

      本研究可实现对点云树障监测预报预警,提供及时准确全面的点云情报预报信息。该模型后续可应用于建筑施工预警、区域工程监管、生态工程监管,山体滑坡危害等情况,对经济社会发展提供可靠的支撑和保障。

      本研究的后续展望有如下几点:

      1)绝缘子等电塔附件对导线提取算法的精度产生影响,因此激光点云导线提取算法在电塔附近的导线精度有待提高,提高复杂输电线路场景下的导线提取的鲁棒性;提高激光点云导线数据矢量化的三维重建精度[19-21]

      2)研究激光点云和可见光点云的地面点提取算法,提高可见光点云的植被提取结果精度,优化算法技术方案,提升算法效率。

      3)提高激光点云和可见光点云的杆塔提取精度,能更加准确地提取出不同类型的电杆或者电塔,使点云分类结果更优。

      4)研究激光点云和可见光点云在没有电塔位置情况下的电塔点云提取算法,改进技术流程,提高算法的自动化程度、鲁棒性,改善算法效率。研究大场景下的激光点云和可见光点云的点云分类提取算法。

      5)输电线路的地面点可以用于制作描述输电线路地形的DEM等方面,因此地面点提取算法的研究可以作为下一步的研究内容。

Reference (21)

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return