摘要
针对智能配电网设备种类多样、日常安防巡检工作量大、隐患多等特点,且传统的人工巡检方式难以快速识别配电网设备多类型安防漏洞的难题,文章提出了一种基于四旋翼无人机的智能配电网设备巡检安防识别方法,应用无人机视频图像快速辨析技术,对配电网设备的安防问题进行快速巡检,从而保障系统的运行安全和稳定。针对高效率、低成本的配电网设备安防巡检需求,从坐标变换原理、摄像机成像模型、基于视觉的位姿估计原理和AprilTag位姿估计原理等多角度全方位设计了无人机的视觉定位算法。采用核相关滤波(kernel correlation filter, KCF)算法跟踪算法实现对目标的跟踪,并用卡尔曼滤波器建立目标物的运动模型对传统的KCF算法进行了改进,克服了边缘效应,提高了对目标物的跟踪可靠性。最后,实际案例验证了文章方法的可行性,以期为配电网设备精益化管理提供一定的借鉴。
In view of the characteristics of various types of distribution network equipment,heavy workload of daily security inspection and many potential safety hazards,moreover,the traditional manual inspection method is difficult to quickly identify a variety of security vulnerabilities of distribution network equipment,this paper proposes an identification method of security in patrol inspection for distribution network equipment based on four-rotor drone.The drone video image fast discrimination technology is applied to quickly inspect the security problems of distribution network equipment to ensure the safe and stable operation of the system.In view of the high-efficiency and low-cost security inspection requirements of distribution network equipment,the visual positioning algorithm of drone is designed from the principle of coordinate transformation,camera imaging model,vision based pose estimation principle and AprilTag pose estimation principle.The KCF tracking algorithm is used to track the target,and the motion model of the target established by Kalman filter is used to improve the traditional KCF algorithm,overcome the edge effect and improve the tracking reliability of the target.Finally,a practical case verifies the feasibility of the proposed method,which provides some reference for lean management of distribution network equipment.
作者
吴雪琼
于海平
范延赫
WU Xueqiong;YU Haiping;FAN Yanhe(State Grid Electric Power Research Institute Co.,Ltd.,Nanjing 211100,China;School of Electrical Automation and Information Engineering,Tianjin University,Tianjin 300072,China)
出处
《电测与仪表》
北大核心
2025年第4期208-216,共9页
Electrical Measurement & Instrumentation
基金
国网公司总部科技项目(5400-202112149A-0-0-00)。
关键词
配电网
无人机
智能巡检
图像识别
distribution network
drone
intelligent patrol inspection
image identification
作者简介
吴雪琼(1984-),女,高级工程师,研究方向为配电自动化、配网抢修、配网运维、信息集成。Email:wuxueqiong@sgepri.sgcc.com.cn;于海平(1985-),男,高级工程师,研究方向为配电自动化、配电物联网、配网运维、信息集成。Email:yuhaiping@sgepri.sgcc.com.cn;没得书范延赫(2000-),男,硕士研究生,研究方向为智能配电网态势感知,综合能源优化。Email:asd456177300@126.com。
