基于光纤传感技术的综合智能化光缆线路故障诊断  

Integrated intelligent fault diagnosis of the optical cable line based on optical fiber sensing technology

作　　者：方明 Fang Ming(Extra High Voltage Power Transmission Company Guiyang Branch,China Southern Power Grid Co., Ltd., Guizhou Guiyang, 550081, China)

机构地区：[1]中国南方电网有限责任公司超高压输电公司贵阳局,贵州贵阳550081

出　　处：《机械设计与制造工程》2021年第1期59-62,共4页Machine Design and Manufacturing Engineering

摘　　要：为提高光缆线路故障诊断的正确率,便于光缆线路的检修与维护,提出一种基于光纤传感技术的综合智能化光缆线路故障诊断方法。首先,通过光纤传感器监测、采集光缆线路故障信号数据,并利用小波包3层分解和重构光缆线路故障信号数据提取小波包能量特征、小波包标准差特征和Shannon熵特征;其次,将提取的特征数据划分成训练样本和测试样本,针对训练样本将光缆线路故障特征数据和故障类型(短路故障、漏电故障、钩挂故障以及锚砸故障)作为ELM模型的输入和输出,建立光缆线路故障的ELM识别模型。与SVM、RBFNN和BPNN相比,采用ELM进行光缆线路故障诊断具有更高的正确率,为光缆线路故障诊断提供新的方法。In order to improve the accuracy of fault diagnosis of optical fiber lines and facilitate the repair and maintenance of optical fiber lines,a comprehensive intelligent optical fiber line fault diagnosis method based on optical fiber sensing technology is proposed.Firstly,the optical fiber sensor is used to monitor and collect the fault signal data of optical fiber line,and the wavelet packet is used to decompose and reconstruct the fault signal data of optical fiber line to extract the energy characteristic,standard deviation characteristic and Shannon entropy characteristic of wavelet packet.Secondly,the extracted feature data is divided into training samples and test samples.For the training samples,the characteristic data and fault types(short circuit fault,leakage fault,hook fault and anchor fault)of optical cable line are taken as the input and output of ELM model,and the ELM identification model of optical cable line fault is established.Compared with SVM,RBFNN and BPNN,the ELM method presented in this paper has a higher accuracy rate for fault diagnosis of optical cable lines,which provides a new method for fault diagnosis of optical cable lines.

关 键 词：光缆线路故障 极限学习机 光纤传感器 短路故障 漏电故障 

分 类 号：TN248[电子电信—物理电子学] TP391.1[自动化与计算机技术—计算机应用技术]