Prediction of fiber Rayleigh scattering responses based on deep learning  

作　　者：Yongxin LIANG Jianhui SUN Jialei ZHANG Yuyao WANG Anchi WAN Shibo ZHANG Zhenyu YE Shengtao LIN Zinan WANG 

机构地区：[1]Key Lab of Optical Fiber Sensing and Communications,University of Electronic Science and Technology of China,Chengdu 611731,China [2]Center for Information Geoscience,University of Electronic Science and Technology of China,Chengdu 611731,China

出　　处：《Science China(Information Sciences)》2023年第12期160-174,共15页中国科学（信息科学）（英文版）

基　　金：supported by National Natural Science Foundation of China (Grant No. 62075030);National Ten-Thousand Talent Program (Grant No. W030211001001);Sichuan Provincial Project for Outstanding Young Scholars in Science and Technology (Grant No. 2020JDJQ0024)。

摘　　要：Distributed acoustic sensing(DAS) is a fiber sensing technology based on Rayleigh scattering,which transforms optical fiber into a series of sensing units. It has become an indispensable part in the field of seismic monitoring, vehicle tracking, and pipeline monitoring. Fiber Rayleigh scattering responses lay at the core of DAS. However, there are few in-depth studies on the purpose of acquiring fiber Rayleigh scattering responses. In this paper, we establish a deep learning framework based on the bidirectional gated recurrent unit, which is the first time to predict the fiber Rayleigh scattering responses, to the best of our knowledge. The deep learning framework is trained with a numerical simulation dataset only, but it can process experimental data successfully. Moreover, since the responses could have a wider effective bandwidth than the experimental probing pulses, a finer spatial resolution could be obtained after demodulation. This work indicates that the deep learning framework can capture the characteristics of the fiber Rayleigh scattering responses effectively, which paves the way for intelligent DAS.

关 键 词：distributed acoustic sensing Rayleigh scattering deep learning bidirectional gated recurrent unit intelligent demodulation 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP212[自动化与计算机技术—控制科学与工程]