摘要
基于前端智能识别的架空输电线路在线监测模式能够有效解决现有在线监测识别实时性差和视频数据大规模传输导致的通信成本高昂等问题。文章首先提出基于前端识别的输电线路在线监测系统的总体架构。其次重点分析基于前端识别的输电线路在线监测目标检测、前端AI芯片应用选型以及AI模型压缩和加速3项关键技术,并综述相应技术的研究应用特点。最后探讨了前端识别在实践过程中存在的主要挑战,并对输电线路在线监测技术的发展方向进行了展望。
The on-line monitoring of overhead transmission line based on front-end intelligent recognition can effectively solve the problems of poor real-time performance and high communication cost caused by large-scale transmission of video data.Therefore,this paper firstly analyzes the overall architecture of on-line monitoring system based on front-end intelligent recognition is proposed.Then,three key technologies,which are on-line monitoring target detection based on front-end recognition,application selection of front-end AI chips and AI model compression and acceleration,are analyzed,and their research and application characteristics are summarized.Finally,the main challenges of front-end recognition in practice are discussed,and the development trend of on-line monitoring technology of transmission line is prospected.
作者
余盛达
叶保璇
冯毅强
易婷婷
韩旭君
施超
YU Shengda;YE Baoxuan;FENG Yiqiang;YI Tingting;HAN Xujun;SHI Chao(Wenchang Power Supply Bureau,Hainan Power Grid Co.,Ltd.,Wenchang 571300,China;Guangzhou Power Electrical Technology Co.,Ltd.,Guangzhou 510635,China)
出处
《通信电源技术》
2021年第3期44-46,共3页
Telecom Power Technology
基金
海南电网有限责任公司科技项目(070500Kk52190001)。
关键词
输电线路
前端智能识别
深度学习
AI芯片
transmission line
front-end intelligent recognition
deep learning
AI chip
