【作者】 王振；

【导师】 王成福； 瞿寒冰；

【作者基本信息】 山东大学 ， 电气工程（专业学位）， 2020， 硕士

【摘要】 架空输电线路运行安全始终是考验电网安全的重要因素,尤其是在超、特高压输电线路实现大电网联网后,架空输电线路安全更是上升到国家能源安全的高度。然而,由于架空输电线路的分布呈现点多面广的特点,且分布地域环境复杂,加之近年来国网公司落实降本增效管理思路,使得一线运维缺员明显,从而导致架空输电线路的精益化管理面临诸多困难。与此同时,激光点云数据在测绘领域应用趋于成熟,可为输电线路运维困难问题提供一个新的解决思路。据此,笔者以解决实际运维中的工作困难为目的,针对输电线路激光点云数据挖掘与应用展开研究,并重点探讨适合线路实际数据处理的方法与应用效果。具体研究工作包括以下三点:(1)探索出了基于光滑分割的输电线路点云数据处理方法。首先,对点云数据进行预处理;然后,应用基于光滑分割的改进三角网渐进加密滤波算法完成点云滤波,并实现地面与非地面点云数据分类;最后,针对地上要素点云数据完成分类,基于经验针对居民地、山地等地形提出点云数据分类注意问题,并在济南地区的输电线路点云数据处理实验中检验了数据处理效果,数据整体处理效果较好,地表、植被数据处理效果优于Terrasolid。(2)提出了基于点云数据的输电线路建模方法。首先,运用点云数据VFH(Viewpoint Feature Histogram)特征描述因子,通过建立kd-tree结构的方法实现了点云模型数据库的构建;然后,通过模型驱动方法实现了对输电杆塔、输电导线的三维建模,最后,在济南地区输电线路开展建模实验以检验建模精度,实验结果证明所建模型的误差可以控制在3%以内。(3)开展了点云数据在输电线路自动巡检与交跨测量两方面的应用研究。首先,通过激光 SLAM(Simultaneous localization and mapping)方法实现 了航迹规划,针对输电线路自动巡检,确定了最佳数据抽稀比例,明确自动巡检作业拍摄规范、提出了典型输电杆塔航线规划方法,以及自动巡检安全距离标准,并在济南地区完成自动巡检精度检验;其次,提出了基于点云数据的分割逼近方法计算输电线路交跨距离,并对该方法测量精度进行了检验分析。更多还原

【Abstract】 The safety of overhead transmission lines is always an important factor to test the safety of power grids.Especially after the ultra high voltage transmission lines have realized the interconnection of large power grids,the safety of overhead transmission lines has risen to the height of national energy security.However,the distribution of overhead transmission lines presents a specific point with multiple areas and a complex geographical environment.In addition,in recent years,State Grid has implemented the management strategies of reducing cost and increasing efficiency,which leading an obvious shortage of front-line operators.In this case,it is difficult to realize lean management of overhead transmission lines.Meanwhile,the application of laser point cloud data in surveying and mapping field is on the way to maturity,which provides a new idea to solve the difficult problems of transmission lines.In order to solve the difficulties in actual operation and maintenance accordingly,the author focuses on the research on mining and application of laser-point cloud data for transmission lines,of which the key point is the method and effect of data processing which is more actual in the field of transmission lines.The main research work includes the following three points:(1)A method based on smooth segmentation of point cloud data processing for transmission lines is proposed.Firstly,point cloud data should be preprocessed,then the improved triangular network progressive encryption filtering algorithm based on smooth segmentation is applied to complete the point cloud filtering and realize the classification of ground and non-ground point data,finally,point cloud data of elements above ground are classified,attention in data classification of settlement places and mountains are raised based on experience,The data processing effect was tested in the point cloud data processing experiment of power transmission lines in Jinan.we get a good effect,especially the land and vegetation data effect is better than Terrasolid.(2)A modeling method of transmission line based on point cloud data is proposed.Firstly,the VFH feature description factor of point cloud data is used to construct the database of point cloud model by establishing kd-tree structure,then,the transmission tower and line are modeled through model-driven method,finally in the modeling experiment of transmission lines in Jinan,the modeling accuracy is verified.The error is proved less than 3%.(3)Research on two applications of point cloud data in transmission lines are proposed,which are automatic inspection and crossover survey.Firstly,route planning is realized by laser SLAM method,the best dilution ratio of data is for automatic inspection,the shooting specification is specified,the typical operation mode is put forward,as well as the safety distance standard.The precision of automatic inspection was inspected in Jinan.Secondly,a segmentation approximation method based on point cloud data is proposed in crossover survey,and the accuracy is tested and analyzed.更多还原