浏览全部资源
扫码关注微信
基于分类器置信度的DAS信号定位和识别技术
An Localization and Classification Recognition Technology of DAS Signal based on Classifier Confidence
- 光通信研究 2025年第5期
- 作者机构：
1.国网湖北省电力有限公司武汉供电公司，武汉 430074
2.国网湖北省电力有限公司，武汉 430074
3.华中科技大学 光学与电子信息学院，武汉 430074
- 作者简介：
蔡超（1985-），男，江西九江人。博士，主要研究方向为配电网智能运维技术研究与应用。
秦知航，硕士。E-mail：qinzhihang@qq.com
- 基金信息：国网武汉供电公司城市电缆网监测资助项目(5215A0230006)
DOI：10.13756/j.gtxyj.2025.240119
中图分类号： TP273.5;TP212.9收稿：2024-06-11，
修回：2024-08-15，
纸质出版：2025-10-10
- 稿件说明：
移动端阅览
蔡超，秦知航，李璐，等. 基于分类器置信度的DAS信号定位和识别技术[J]. 光通信研究，2025(5): 240119.
Cai C, Qin Z H, Li L, et al. An Localization and Classification Recognition Technology of DAS Signal based on Classifier Confidence[J]. Study on Optical Communications, 2025(5): 240119.
蔡超，秦知航，李璐，等. 基于分类器置信度的DAS信号定位和识别技术[J]. 光通信研究，2025(5): 240119. DOI： 10.13756/j.gtxyj.2025.240119.
Cai C, Qin Z H, Li L, et al. An Localization and Classification Recognition Technology of DAS Signal based on Classifier Confidence[J]. Study on Optical Communications, 2025(5): 240119. DOI： 10.13756/j.gtxyj.2025.240119.
摘要
【目的】
2
为解决分布式光纤振动传感系统中目标事件模式识别率低和事件源定位准确性弱等问题。
【方法】
2
文章提出了基于相邻传感单元的分类器置信度分布联合分析方法，利用分布式声波传感（DAS）技术的光纤分布式特点，联立分布式光纤上多个相邻传感单元对外界目标信号数据进行分析，将每个传感单元所探测到的外界振动信号数据送入支持向量机（SVM）分类器进行计算，得到各个传感单元对应的预测事件置信度值。在事件模式识别端，根据各个传感单元对应的预测事件置信度值分布，结合反向传播（BP）神经网络实现目标事件的模式分类；在事件源定位端，提出了一种基于事件置信度空间距离定位算法，根据事件置信度空间距离加权公式对相邻传感单元对应的事件置信度空间分布进行计算，实现外界事件的定位。
【结果】
2
对文章所提算法进行了相应的实验测试，测试结果表明，该算法对外界4类扰动事件平均识别准确率高达94.85%，且事件源的定位精度高于90%。
【结论】
2
文章所提技术解决了DAS系统中目标事件类型识别准确率低、事件源精准定位困难的问题。
Abstract
【Objective】
2
In order to solve the problems of low events recognition rate and weak positioning accuracy in distributed fiber vibration sensor system
a pattern recognition algorithm based on classifier confidence distribution of adjacent sensor nodes is proposed.
【Methods】
2
According to the spatial distribution feature of Distributed Acoustic Sensing (DAS) technology
the target signals are analyzed by the combination of some adjacent sensor nodes. The external vibration signals received by sensor nodes are transmitted to Support Vector Machine (SVM) classifier for further process. And the probabilities prediction of target events for each sensor node can be calculated by SVM classifier. In the recognition stage
the target events can be identified by the combination of probabilities prediction of target events and Back Propagation (BP) neural network. In the events location stage
the spatial distance localization algorithm based on probabilities prediction distribution of target eventsis proposed. The external event locations are pinpointed through a distance-weighted probability distribution of predicted target events.
【Results】
2
Finally
the testing of this method is carried out. The experimental results demonstrated that a high average recognition accuracy rate of 94.85% for four disturbance events with a high positioning accuracy of over 90% can be achieved by the proposed scheme.
【Conclusion】
2
The proposed method solves the problems of low target events classification accuracy and imprecise event source positioning in DAS systems.
关键词
Keywords
references
Zhang X , Ding Z , Hong R , et al . Phase Sensitive Optical Time-domain Reflective Distributed Optical Fiber Sensing Technology [J ] . Acta Optica Sinica , 2021 , 41 ( 1 ): 0106004 .
杨斌 , 皋魏 , 席刚 . Φ-OTDR分布式光纤传感系统的关键技术研究 [J ] . 光通信研究 , 2012 ( 4 ): 19 - 22 .
Yang B , Gao W , Xi G . Key Technologies for Φ-OTDR-based Distributed Fiber-Optic Sensing Systems [J ] . Study on Optical Communications , 2012 ( 4 ): 19 - 22 .
Muanenda Y . Recent Advances in Distributed Acoustic Sensing based on Phase-Sensitive Optical Time Domain Reflectometry [J ] . Journal of Sensors , 2018 ( 1 ): 3897873 .
He T , Sun Q , Zhang S , et al . A Dual-Stage-Recognition Network for Distributed Optical Fiber Sensing Perimeter Security System [J ] . Journal of Lightwave Technology , 2023 , 41 ( 13 ): 4331 - 4340 .
Tejedor J , Macias-Guarasa J , Martins H F , et al . Real Field Deployment of a Smart Fiber-Optic Surveillance System for Pipeline Integrity Threat Detection: Architectural Issues and Blind Field Test Results [J ] . Journal of Lightwave Technology , 2018 , 36 ( 4 ): 1052 - 1062 .
Li S , Peng R , Liu Z . A Surveillance System for Urban Buried Pipeline Subject to Third-Party Threats based on Fiber Optic Sensing and Convolutional Neural Network [J ] . Structural Health Monitoring , 2021 , 20 ( 4 ): 1704 - 1715 .
张均伟 , 蓝波 , 黄嘉庚 , 等 . 基于分布式光纤传感的光缆防外破监测研究 [J ] . 光通信研究 , 2022 ( 5 ): 53 - 57 .
Zhang J W , Lan B , Huang J G , et al . Monitoring and Early Warning System for Anti-Breakage of Optical Cable based on Distributed Optical Fiber Vibration Sensing [J ] . Study on Optical Communications , 2022 ( 5 ): 53 - 57 .
张俊楠 , 娄淑琴 , 梁生 . 基于SVM算法的Φ-OTDR分布式光纤扰动传感系统模式识别研究 [J ] . 红外与激光工程 , 2017 , 46 ( 4 ): 422003 .
Zhang J N , Lou S Q , Liang S . Study of Pattern Recognition based on SVM Algorithm for Φ-OTDR Distributed Optical Fiber Disturbance Sensing System [J ] . Infrared and Laser Engineering , 2017 , 46 ( 4 ): 422003 .
钱君霞 , 郭家兴 . 光纤振动信号的特征提取与识别方法综述 [J ] . 光通信研究 , 2024 ( 6 ): 230116 .
Qian J X , Guo J X . Overview of Feature Extraction and Recognition Methods for Fiber Optic Vibration Signals [J ] . Study on Optical Communications , 2024 ( 6 ): 230116 .
Ma P , Liu K , Jiang J , et al . Probabilistic Event Discrimination Algorithm for Fiber Optic Perimeter Security Systems [J ] . Journal of Lightwave Technology , 2018 , 36 ( 11 ): 2069 - 2075 .
Shi Y , Wang Y , Zhao L , et al . An Event Recognition Method for Φ-OTDR Sensing System based on Deep Learning [J ] . Sensors , 2019 , 19 ( 15 ): 3421 .
Wu H , Yang M , Yang S , et al . A Novel DAS Signal Recognition Method based on Spatiotemporal Information Extraction with 1DCNNs-BiLSTM Network [J ] . IEEE Access , 2020 , 8 : 119448 - 119457 .
0
浏览量
20
下载量
0
CSCD
0
CNKI被引量
- 网络PDF下载
- Meta-XML下载
- BibTeX下载
- Endnote下载
- RefMan 下载
- NoteExpress下载
- NoteFirst下载
关联资源
相关文章
相关作者
相关机构
- 地址：武汉市江夏区高新四路6号烽火科技园4号楼3层 邮编：430205
- 技术支持由北京北大方正电子有限公司提供 鄂ICP备2022010197
鄂公网安备 42011202002092号 - 本系统建议在Chrome、 IE9+ 以上版本浏览器阅读本站内容，360浏览器请切换至极速模式
- Cookies帮助我们提供服务并提供个性化体验。使用本网站，即表示您同意我们使用Cookies
- 总访问量：304943 今日访问量：552