基于Alexnet网络的绝缘子自爆无人机巡检技术研究  

Insulator Self-explosive Inspection Technology Based on Alexnet Network in UAV Grid Inspection

作　　者：李映国 杨宏 徐郁 周杰 赵家乐 LI Yingguo;YANG Hong;XU Yu;ZHOU Jie;ZHAO Jiale(State Grid Chongqing Yongchuan Power Supply Company,Chongqing 402160,China;School of Computer Information and Engineering,Jiangxi Normal University,Nanchang 330022,China)

机构地区：[1]国网永川供电分公司,重庆402160 [2]江西师范大学计算机信息工程学院,江西南昌330022

出　　处：《智慧电力》2021年第8期104-109,共6页Smart Power

基　　金：国家自然科学基金资助项目(61977038)。

摘　　要：绝缘子是输电系统中与安全相关的关键部件,绝缘子自爆问题的高效快速识别对电力系统的保护具有重要的意义。随着无人机(UAV)相关产业的不断发展,可以采用无人机技术对输电线路进行巡检拍摄。以此为背景提出了一种基于Alexnet网络的绝缘子自爆无人机巡检技术。首先,应用无人机巡检这一先进技术得到绝缘子的清晰实时图片。然后,采用Alexnet网络对绝缘子自爆图片进行学习和识别。与传统的识别方法相比,Alexnet网络模型不但结构上有所加深,对卷积的功能也进行了强化,对无人机巡检过程中拍摄的复杂图像进行识别和检测有很好的效果。Insulators are the key components in the safe operation of power transmission systems.Efficient and rapid identification of insulator self-explosion is of great significance to the protection of power systems.With the continuous development of unmanned aerial vehicle(UAV)related industries,UAV technology can be used to inspect and photograph power transmission lines.An insulator self explosive UAV inspection technology is proposed based on Alexnet network.Firstly,the advanced technology of UAV inspection is applied to obtain clear real-time images of insulators.Then,Alexnet network is used to learn and identify the self-explosion pictures of insulators.Compared with the traditional recognition method,Alexnet network model not only deepens the structure and strengthens the convolution function,but also shows good effect in the recognition and detection of complex images taken during UAV inspection.

关 键 词：绝缘子自爆 Alexnet网络 无人机巡检 识别 

分 类 号：TM712[电气工程—电力系统及自动化]