计算技术与自动化2024,Vol.43Issue(4) :46-52.DOI:10.16339/j.cnki.jsjsyzdh.202404008

基于北斗定位的便携式电缆局部放电检测系统

Portable Cable Partial Discharge Detection System Based on Beidou Positioning

于莉莉 吴庆 王志刚 马胜 张永欣
计算技术与自动化2024,Vol.43Issue(4) :46-52.DOI:10.16339/j.cnki.jsjsyzdh.202404008

基于北斗定位的便携式电缆局部放电检测系统

Portable Cable Partial Discharge Detection System Based on Beidou Positioning

于莉莉 1吴庆 2王志刚 2马胜 1张永欣1
扫码查看

作者信息

  • 1. 国网思极位置服务有限公司,北京 102200
  • 2. 国网信息通信产业集团有限公司,北京 100053
  • 折叠

摘要

我国电力发展较快,电缆使用时间较长,导致故障率和用电隐患较多,所以设计了一个基于北斗定位的便携式电缆局部放电检测系统.本系统使用长短期记忆网络算法(Long Short-term Memory,LSTM)来检测电缆的局部放电,平均功率越大,表示紫外光越强,就可推出局部放电的强度大小.在此基础上进一步加入算法反向向量,对算法进行改进,改进后的BiLSTM网络的隐含层更具特殊性.另外本系统首次引入北斗定位系统,实现对电缆非法放电区域进行精确定位的目标.采取OPC-UA技术实现对数据的统一传输和反馈,该研究缓解了系统主机的计算压力.通过实验,该系统技术核算的数据,误差率在可接受范围内,该研究为其他技术研究奠定基础.

Abstract

Due to the rapid development of electric power in China and the long service life of cables,there are many fail-ure rates and hidden dangers of power consumption.Therefore,this research designs a portable cable partial discharge detec-tion system based on Beidou positioning.The system uses long short-term memory(LSTM)algorithm to detect partial dis-charge of cables.The higher the average power is,the stronger the UV light is,and the strength of partial discharge can be deduced.On this basis,an algorithm reverse vector is further added to improve the algorithm.The improved BiLSTM net-work has a more specific hidden layer.In addition,this system introduces the Beidou positioning system for the first time,a-chieve precise positioning of illegal cable discharge areas.Adopting OPC-UA technology to achieve unified data transmission and feedback alleviates the computational pressure on the system host.Through experiments,it was found that the error rate of the system's technical accounting data is within an acceptable range,laying the foundation for other technical re-search.

关键词

放电检测/LSTM算法/OPC-UA技术/北斗定位系统

Key words

discharge detection/LSTM algorithm/OPC-UA technology/Beidou positioning system

引用本文复制引用

出版年

2024
计算技术与自动化
湖南大学

计算技术与自动化

CSTPCD
影响因子:0.295
ISSN:1003-6199
段落导航相关论文