摘要
无人机巡检在220 kV输电线的日常巡检工作中占据越来越多的比重,然而无人机巡检存在定位不准、巡检目标识别与跟踪困难等系列问题。为此,基于惯性测量单元(inertial measurement unit,IMU)、全球导航卫星系统(global navigation satellite systems,GNSS)、视觉测程(visual odometry,VO)等多路传感器融合信息,提出了一种无人机输电线巡检定位与目标跟踪自适应方法。为解决GNSS信号中断以及视觉传感器在复杂场景中缺乏足够特征点导致的自主定位困难问题,提出一种基于联邦卡尔曼滤波的鲁棒自适应定位算法,该算法确保了无人机自主导航和定位的准确性;同时,为了实现输电线目标自动识别和跟踪,提出一种变化场景最佳阈值选择算法,实现预测性和连续的参数选择,以提高输电线跟踪的准确性。最后,以唐山某典型案例为试点,验证了方法的有效性,并进一步全面优化输电线路无人机巡检工作流程,为后续输电线巡检的智能巡检多维应用奠定较好的基础。
Unmanned aerial vehicle(UAV)inspection plays an increasingly important role in the daily inspection of 220kV transmission lines.However,UAV inspection has a series of problems,such as inaccurate positioning and the difficulty of identifying and tracking inspection targets.Therefore,an adaptive method for UAV transmission-line patrol positioning and target tracking based on the fusion of information from inertial measurement units,global navigation satellite systems(GNSS),visual odometry,and other multichannel sensors is proposed.To address the difficulty of autonomous localization caused by GNSS signal interruption and the lack of feature points for visual sensors in complex scenes,a robust adaptive localization algorithm based on federated Kalman filtering is proposed,which ensures the accuracy of autonomous navigation and localization for UAVs.Additionally,to achieve automatic target recognition and tracking of transmission lines,an optimal threshold selection algorithm for changing scenarios is proposed for predictive and continuous parameter selection,to increase the accuracy of transmission-line tracking.A typical case in Tangshan was used as a pilot to verify the effectiveness of the proposed method and optimize the workflow of UAV inspection for transmission lines,laying a foundation for multidimensional intelligent transmission-line inspection.This work is supported by the National Natural Science Foundation of China(No.52277118).
作者
朱长荣
吕文超
单超
王伟
范江涛
孙铭阳
范延赫
葛磊蛟
ZHU Changrong;LüWenchao;SHAN Chao;WANG Wei;FAN Jiangtao;SUN Mingyang;FAN Yanhe;GE Leijiao(Tangshan Power Supply Company of State Grid Jibei Electric Power Company Limited,Tangshan 063099,Hebei Province,China;School of Electrical and Information Engineering,Tianjin University,Tianjin 300072,China)
出处
《电力建设》
CSCD
北大核心
2023年第8期61-70,共10页
Electric Power Construction
基金
国家自然科学基金项目(52277118)。
关键词
视觉定位
精细化巡检
自适应滤波
目标追踪
霍夫变换
visual positioning
fine inspection
adaptive filtering
object tracking
Hough transformation
作者简介
朱长荣(1971),男,本科,高级工程师,主要研究方向为电网调度、电力系统运行和新能源系统分析;吕文超(1986),男,硕士,高级工程师,主要研究方向为新型电力系统及输电线路智能应用;单超(1988),男,本科,工程师,研究方向为无人机巡检技术和电力设施保护;王伟(1990),男,本科,工程师,研究方向为高压输电线路施工运行与维护;范江涛(1986),男,本科,高级工程师,研究方向为数字电网与输电线路智能化应用;孙铭阳(1995),男,硕士,助理工程师,研究方向为新型电力系统规划及电力设备运行状态监测;通信作者:范延赫(2000),男,硕士研究生,主要研究方向为智能配电网态势感知,综合能源优化,E-mail:asd456177300@126.com;葛磊蛟(1984),男,博士,副教授,主要研究方向为智能配电网态势感知、新能源并网优化控制和智能配用电大数据云计算技术等。
