基于Φ-OTDR的振动事件识别分类器研究进展  

Research progress of vibration event recognition classifier based on Φ-OTDR

作　　者：赵丽娟[1] 魏迎健 徐志钮[1] ZHAO Lijuan;WEI Yingjian;XU Zhiniu(School of Electrical and Electronic Engineering,North China Electric Power University,Baoding Hebei 071003,China)

机构地区：[1]华北电力大学电气与电子工程学院,河北保定071003

出　　处：《光通信技术》2023年第2期1-5,共5页Optical Communication Technology

基　　金：国家自然科学基金项目(62171185、62273146)资助;河北省自然科学基金项目(E2019502177、E2020502010)资助。

摘　　要：相位敏感光时域反射计(Φ-OTDR)凭借着传感距离长、铺设简单、耐腐蚀和抗电磁干扰等特点被广泛应用于分布式振动监测领域。随着传感任务多样化及人工智能的广泛应用,对振动事件的类型识别成为研究的热点方向。为了使读者能更好理解识别分类器研究进展和发展趋势,先后介绍了传统识别分类器和基于深度学习的神经网络识别分类器,对不同分类器性能指标、优缺点和应用场合进行了比较,最后对Φ-OTDR振动事件识别研究方向进行了展望。Phase-sensitive optical time domain reflectometer (Φ-OTDR) is widely used in the field of distributed vibration monitoring due to its long sensing distance, simple laying, corrosion resistance and anti-electromagnetic interference. With the diversification of sensing tasks and the wide application of artificial intelligence, the type recognition of vibration events has become a hot research direction. In order to enable readers to better understand the research progress and development trend of recognition classifiers, this paper introduces the traditional recognition classifiers and the neural network recognition classifiers based on deep learning, compares the performance indexes, advantages and disadvantages of different classifiers and their applications, and finally prospects the research direction of Φ-OTDR vibration event recognition.

关 键 词：相位敏感光时域反射计 振动事件识别 深度学习 神经网络 

分 类 号：TN253[电子电信—物理电子学]