摘要
从变电站监控系统的应用概况出发,提出了一种基于深度学习技术的变电站视频监控隐患排查方法,通过训练深度学习网络模型以实现对监控视频中安全隐患的智能识别,具体包含视频质量检测、违章行为识别、安全风险识别、变电站设备故障识别、隔离开关设备状态识别等。另外,通过构建安全隐患数据集并进行实验初步验证了该方法的可行性,为智能图像识别技术在电力系统的应用提供参考。
Based on the application profile of the substation monitoring system,a method for detecting hidden dangers in substation video monitoring based on deep learning technology is presented.Through training in-depth learning network model to realize intelligent identification of security hidden dangers in monitoring video,including video quality detection,violation behavior identification,security risk identification,substation equipment failure identification,isolation switch equipment status identification,and so on.In addition,the feasibility of this method is preliminarily verified by building a security risk data set and experimentation,which provides reference for the application of smart image recognition technology in power system.
作者
李虎孬
陈富国
陈亮
LI Hunao;CHEN Fuguo;CHEN Liang(Pinggao Group Co.,Ltd.,Pingdingshan 467001,China;State Key Lab.of Electrical Insulation and Power Equipment,Xi′an Jiaotong University,Xi′an 710049,China)
出处
《电工技术》
2022年第16期37-41,122,共6页
Electric Engineering
基金
国家自然科学基金资助项目(编号51777154)。
关键词
图像识别
视频监控
深度学习
特征提取
随机森林
image recognition
video surveillance
deep learning
feature extraction
random forest
