配网带电作业机器人精准作业定位方法  

Precise Positioning for Live Working on Distribution Line

作　　者：张静 黄国方[1,2] 刘晓铭 单超[1,2] 王文政 童宇辉[1,2] 许茂洲 ZHANG Jing;HUANG Guofang;LIU Xiaoming;SHAN Chao;WANG Wenzheng;TONG Yuhui;XU Maozhou(NARI Group Corporation(State Grid Electric Power Research Institute),Nanjing 211106,Jiangsu Province,China;NARI Technology Co.,Ltd.,Nanjing 211106,Jiangsu Province,China)

机构地区：[1]南瑞集团(国网电力科学研究院)有限公司,江苏省南京市211106 [2]国电南瑞科技股份有限公司,江苏省南京市211106

出　　处：《电网技术》2022年第2期812-819,共8页Power System Technology

基　　金：国家电网有限公司科技项目《人工智能配网带电作业机器人关键技术及成套装备研究与应用》(SGTJBHOOYJJS1902138)。

摘　　要：针对户外光照、斗臂车振动、作业场景不固定等因素对配网带电作业机器人作业定位的影响,讨论了配网带电作业激光数据感知、设备对象识别和位置自动定位的问题。提出了基于场景语义的自动作业定位方法,解决了太阳强光直射对激光雷达致盲的影响,减少了机械臂与传感器的位置校准过程,通过整体特征量分析,减少离散点和抖动干扰,实现作业设备对象的自动识别,同时利用导线连通性的特征,实现整个导线的连通,为点云数据采集空缺处作业点选择提供可能,最后通过设备位置和长度语义的信息提供作业点自动选择输入,并实现作业点选择和安全校核。通过工程作业的应用分析,该方法提高了机器人对不同作业地点的环境适应性,通过目标对象识别和计算,将有效采集数据成功率提升了约20%,选择平均误差由手动离散3厘米降低到自动提取的1厘米,减少了人工选择和调整的作业过程,大大提升了配网带电作业机器人的作业效率。Since the positioning accuracy of a live working robot on the distribution line is significantly affected by the outdoor lighting,the bucket boom vibration,or the changing working spots,this article discusses the laser data perception,the equipment object recognition and the automatic location positioning.An automatic working positioning method based on the scenic semantics is proposed,which solves the blindness of radar caused by the strong light and reduces the position calibration process of the robot arms and the sensors.Through an analysis among the integral features,this method eliminates the interference of discrete points and achieves the automatic identification of the object working equipment.Meanwhile,using the semantic feature of wire connectivity the connection of the entire wire is realized,making it possible to select the operating point in the blank of the point cloud data collection.Finally,the automatic selection of operating points is provided through the information of equipment semantics,and the safety check of the operating point selection is realized.The comparative application analysis of this project shows that this method increases the successful rate of the effective point cloud data collection by about 20%and reduces the average error of selection from 3cm for manual discretization to 1cm for automatic extraction,which will reduce the manual adjustment process and improve the automation level of the distribution network live operation.

关 键 词：配网带电作业 机器人 立体感知 三维点云 精准定位 

分 类 号：TM721[电气工程—电力系统及自动化]