Journals Articles Conferences News About CLP
Submit a paper Email Alert
Search
Search
Articles
Journals
News
Advanced Search
Top Searches
2D MaterialsTransformation opticsQuantum PhotonicsSmall lasersSemiconductor UV PhotonicsSupersymmetric microring laser arrays

Chinese Journal of Lasers, Volume. 50, Issue 11, 1106003(2023)

Mode Recognition Method of Φ‑OTDR System Based on Mixed Input Neural Network

Xiao Li1, Yi Gao1、*, Hao Wu2, and Daoyu Wang1
Author Affiliations
  • 1Department of Information and Electronics, Wuhan Digital Engineering Institute, Wuhan 430202, Hubei, China
  • 2National Engineering Laboratory for Next Generation Internet Access System, Huazhong University of Science and Technology, Wuhan 430074, Hubei, China
    • show less
    AbstractGet PDF(in Chinese)
    Figures&Tables (7)
    Equations (0)
    References (15)
    Cited By (8)
    Tools
    Paper Information
    Topics
    Figures & Tables(7)
    Structure diagram of Φ-OTDR system

    Fig. 1. Structure diagram of Φ-OTDR system

    Download full size
    Phase change caused by coherent fading without fiber disturbance. (a) RBS intensity trace; (b) phase trace after demodulation

    Fig. 2. Phase change caused by coherent fading without fiber disturbance. (a) RBS intensity trace; (b) phase trace after demodulation

    Download full size
    Three-dimensional map of phase caused by beating fiber

    Fig. 3. Three-dimensional map of phase caused by beating fiber

    Download full size
    Structure of overall network

    Fig. 4. Structure of overall network

    Download full size
    Training process and verification results of the two models. (a) Cross entropy loss for the training processes; (b) training accuracy traces; (c) validation results of 1DCNN model; (d) validation results of HDNN model

    Fig. 5. Training process and verification results of the two models. (a) Cross entropy loss for the training processes; (b) training accuracy traces; (c) validation results of 1DCNN model; (d) validation results of HDNN model

    Download full size

    • Table 1. Type and quantity of the samples

      View table

      Table 1. Type and quantity of the samples

      Event typeSample sizeEvent label
      Background noise54000
      Human beating76201
      Machine excavation51302
      Human walking50703
      Human jumping50404

    • Table 2. F-score and accuracy of different types of events

      View table

      Table 2. F-score and accuracy of different types of events

      ModelEvent typeF-score /%Average accuracy /%
      1DCNNBackground noise93.1

      96.1

      Human beating94.4
      Machine excavation98.8
      Human walking96.9
      Human jumping97.8
      HDNNBackground noise99.4

      98.8

      Human beating99.4
      Machine excavation99.0
      Human walking97.6
      Human jumping98.2
    Tools

    Get Citation

    Copy Citation Text

    Xiao Li, Yi Gao, Hao Wu, Daoyu Wang. Mode Recognition Method of Φ‑OTDR System Based on Mixed Input Neural Network[J]. Chinese Journal of Lasers, 2023, 50(11): 1106003

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Set citation alerts for article
    Save article for my favorites
    Paper Information

    Category: Fiber optics and optical communication

    Received: Nov. 1, 2022

    Accepted: Dec. 15, 2022

    Published Online: May. 19, 2023

    The Author Email: Gao Yi (gaoyi709@163.com)

    DOI:10.3788/CJL221385

    Topics

    laser devices and laser physics
    Lasers and Laser Optics
    Laser physics
    laser manufacturing
    Instrumentation, Measurement and Metrology