许逵, 李鑫卓, 张历, 张俊杰, 杨宁. 基于边缘AI的配网非定点作业现场实时智能安全监护成套装置研制[J]. 电力大数据, 2023, 26(6): 19-27. DOI: 10.19317/j.cnki.1008-083x.2023.06.003
引用本文: 许逵, 李鑫卓, 张历, 张俊杰, 杨宁. 基于边缘AI的配网非定点作业现场实时智能安全监护成套装置研制[J]. 电力大数据, 2023, 26(6): 19-27. DOI: 10.19317/j.cnki.1008-083x.2023.06.003
XU Kui, LI Xin-zhuo, ZHANG Li, ZHANG Jun-jie, YANG Ning. The Development of Real-Time Intelligent Safety Supervision Kit Based on Edge AI for Non-fixed-location Operations in Distribution Network[J]. Power Systems and Big Data, 2023, 26(6): 19-27. DOI: 10.19317/j.cnki.1008-083x.2023.06.003
Citation: XU Kui, LI Xin-zhuo, ZHANG Li, ZHANG Jun-jie, YANG Ning. The Development of Real-Time Intelligent Safety Supervision Kit Based on Edge AI for Non-fixed-location Operations in Distribution Network[J]. Power Systems and Big Data, 2023, 26(6): 19-27. DOI: 10.19317/j.cnki.1008-083x.2023.06.003

基于边缘AI的配网非定点作业现场实时智能安全监护成套装置研制

The Development of Real-Time Intelligent Safety Supervision Kit Based on Edge AI for Non-fixed-location Operations in Distribution Network

  • 摘要: 为满足带电作业、高处作业等配电网复杂作业场景的安全监护要求,解决配网非定点作业人工监护效率低、现场终端能力单一协同不足、通信安全较难保障等实际问题,文章提出了一种具有强边缘智能、主动监护、精准定位与辨识、多层通讯安全防护等特点的智能安全监护成套装置设计方案。该装置实现了视频监控设备的主动控制、作业人员活动的跟踪监测、监控视频的实时就地智能分析与违规行为预警;装置通过配套定位设备,实现了作业地点和作业范围的动态划定以及作业人员的精准定位;实现了可信通讯设备的自组网络和通讯内容的多层加密。实验结果表明,该装置在典型场景下的平均检测精度可以达到90%以上,可以极大提高配网户外非定点作业场景下的安全监护效率,提升作业现场安全的实时监管能力。

     

    Abstract: In order to meet the safety supervision requirements for complex operations in distribution networks such as live-line work and high-altitude work, and to solve practical problems such as the low efficiency of manual supervision for non-fixed-point operations in distribution networks, limited collaborative capabilities of on-site terminals, and difficulties in ensuring communication security, this article proposes a design scheme for an intelligent safety monitoring system with characteristics including strong edge intelligence, active monitoring, precise positioning and identification, and multi-layer communication security protection.The system achieves active control of video surveillance equipment, tracking and monitoring of operator activities, real-time intelligent analysis of monitored videos, and early warning of violations. Through the supporting positioning equipment, the system dynamically defines the operation location and range, as well as accurately locates the operators. It also establishes a self-organizing network for trusted communication devices and implements multi-layer encryption of communication content.Experimental results demonstrate that the system achieves an average detection accuracy of over 90% in typical scenarios. It significantly improves the efficiency of safety supervision in non-fixed-point outdoor operation scenarios in distribution networks, enhancing the real-time regulatory capabilities for on-site safety.

     

/

返回文章
返回