摘要
随着电力供电可靠性要求的不断提高,电力巡线技术正在向自动化、智能化方向发展。无人机技术以其便捷的操作、精准拍照及图像识别等优点,正广泛应用于电力巡线技术领域。基于激光雷达实现无人机跟随导线飞行,进行线路通道巡检,提出一种基于改进卷积神经网络的可见光与热红外方法,通过采用可见光与热红外图像的融合,以识别出输电线路和障碍物信息,实现对输电线路故障的自动检测。通过对运行线路的实际巡视,输电线路故障的自动检测发现率达90%以上。结果验证了提出的无人机图像融合技术及故障自动检测技术的有效性及可行性。
At present,with the continuous improvement of reliability requirements of power supply,power line patrol technology is developing towards automation and intellectualization.Unmanned aerial vehicle(UAV)technology is widely used in the field of power line patrol technology because of its convenient operation,accurate photography and image recognition.Firstly,UAV wire following flight is realized based on lidar,and line channel patrol is carried out.Then a fusion method of visible and infrared images based on improved convolution neural network is proposed.By fusing visible light and infrared image,the information of transmission lines and obstacles can be identified to realize au・tomatic detection of transmission line faults.The effectiveness and feasibility of UAV image fusion technology and the automatic fault detection technology proposed in this paper are verified by the inspection results by selecting a practical operation line.
作者
汪勇
张英
廖如超
郭启迪
袁新星
康泰钟
WANG Yong;ZHANG Ying;LIAO Ruchao;GUO Qidi;YUAN Xinxing;KANG Taizhong(Patrol Operation Center of Guangdong Power Grid Co.,Ltd,Guangzhou 510145,China;Green Valley International Inc.,Beijing 100089,China)
出处
《激光杂志》
北大核心
2020年第2期141-145,共5页
Laser Journal
基金
广东电网有限责任公司科技项目(No.GDKJXM20184731)
关键词
无人机
热红外
激光雷达
输电线路
图像融合
unmanned aerial vehicle
infrared light
lidar
transmission line
image fusion
