基于目标检测算法的电力资产识别研究  

Research on power asset recognition based on target detection algorithm

作　　者：管诗衡 皮梦婷 胡为民[1] 余杰[1] 朱佳 GUAN Shiheng;PI Mengting;HU Weimin;YU Jie;ZHU Jia(State Grid Yichang Power Supply Company,Yichang Hubei 443002)

机构地区：[1]国家电网宜昌供电公司,湖北宜昌443002

出　　处：《长江信息通信》2023年第10期82-84,共3页Changjiang Information & Communications

摘　　要：近年来,随着电力行业的不断深入发展,电力网络中的电力资产的发现和识别已成为了一个极其重要的研究热点。文章旨在探讨基于目标检测算法的电力资产识别方法,以提高电力资产设备的识别效率以及方便电力资产的管理。文章首先介绍了各种目标检测算法的工作原理,然后通过实验对比分析了不同的目标检测算法在电力资产图像数据集上的识别效果,实验结果表明,一阶段目标检测算法YOLOv5在电力资产数据集上的识别结果最好,精度达到97.4%。In recent years,with the continuous development of the power industry,the discovery and identification of power assets in the power network has become an extremely important research hotspot.The purpose of this paper is to explore the power asset identification method based on the target detection algorithm to improve the identification efficiency of power asset equipment and facilitate the management of power assets.First,this paper introduces the working principles of various target detection algorithms,and then compares and analyzes the recognition effects of different target detection algorithms on the power asset image dataset through experiments.The experimental results show that the first-stage target detection algorithm YOLOv5 has the best recognition results on the power asset dataset,with the accuracy reaching 97.4%.

关 键 词：电力资产 目标检测算法 识别效率 电力资产数据集 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]