Research On High-performance Rayleigh Scattering Distributed Optical Fiber Sensing Based On Compression In Time And Frequency Domain

Distributed optical fiber sensing(DOFS)based on Rayleigh scattering(RS)has be-come a research hotspot in recent years due to its advantages on fast response,high sensi-tivity,and long sensing distance.It has been initially applied to seismic wave monitoring,geological prospecting,intelligent transportation,large structure health monitoring,and other fields.There are three common types of RS-based DOFS systems: phase-sensitive optical time domain reflectometry(?-OTDR)for dynamic strain sensing,Coherent OTDR(COTDR)for dynamic/static temperature and strain sensing,and optical frequency do-main reflectometry(OFDR)for dynamic/static temperature and strain sensing.The per-formance parameters of the RS-based DOFS system are mutually restricted,and some of them such as frequency response bandwidth,spatial resolution,and measurement range are limited by the electronic bottleneck of traditional electrical devices.As a result,there is a huge challenge to improve the comprehensive performance of the sensing system.In this thesis,to achieving high-performance sensing,the limiting factors of the frequency response bandwidth,minimum detectable strain,and measurement slew rate of the sens-ing system have been systematically analyzed.Guided by those theoretical analysis,the positive and negative frequency multiplexing(PNM),pulse compression COTDR,sub-chirped-pulse extraction algorithm(SPEA)in the frequency domain,and SPEA in the time domain are proposed.Based on the time-domain compression of chirped pulses,the frequency response bandwidth,sensing distance,minimum detectable strain,measure-ment slew rate of the sensing system are improved.Based on the spectrum compression of the chirp signal,the limitation of the receiver bandwidth on the spatial resolution and measurement range of the COTDR sensing system has been broken.The specific research contents of this thesis are as follows:(1)To break the mutual limitation between frequency response bandwidth,sens-ing distance,and receiver bandwidth in phase-demodulation ?-OTDR,the PNM is pro-posed to improve the system spectrum utilization,and further improve the frequency re-sponse bandwidth without sacrificing other parameters or increasing the receiver band-width.From the basic theory of signal processing,the principle of the PNM is derived,and how to use the negative frequency band resources to multiply the maximum available bandwidth of the sensing system is clarified.By using the digital amplitude predistortion in single-sideband modulation,the effective output power and the sideband suppression ratio of the modulated signal are improved,paving the way for high-performance PNM.It is theoretically and experimentally demonstrated that the PNM does not affect the signal-to-noise ratio(SNR)of the sensing system.Finally,thanks to its good compatibility,the PNM was combined with time domain pulse compression,distributed Raman amplifica-tion,and frequency division multiplexing.As a result,a dynamic strain sensing with a frequency response bandwidth of 10.8 k Hz is realized on a 103 km sensing fiber.Com-pared with the traditional ?-OTDR,the frequency response bandwidth is increased by more than 20 times.(2)To break the trade-off between the spatial resolution and minimum detectable strain of the traditional COTDR sensing system and increase the frequency response band-width of it,the time-domain pulse compression and the frequency-domain SPEA are pro-posed.The Cramér-Rao Lower Bound(CRLB)of the detectable strain in RS pattern demodulation COTDR is derived.Guided by CRLB,a time-domain pulse compression COTDR sensing system is proposed,in which the minimum detectable strain can be im-proved without sacrificing the spatial resolution.In traditional COTDR sensing system,the time-consuming frequency-scanning process is necessary to form the complete RS pattern,which makes the frequency response bandwidth of this system is very low.By analyzing the frequency domain characteristics of the scattered signal,the SPEA is pro-posed to form an RS pattern with single-shot measurement.Thus,the frequency response bandwidth can be improved at least two orders of magnitude,and 237 times improve-ment has been realized experimentally.Finally,the system's capabilities in dynamic strain sensing,low-frequency response,anti-noise performance,long-distance sensing are fully explored.The obvious advantages are demonstrated compared with conventional phase demodulation ?-OTDR,traditional COTDR sensing systems,chirped pulse ?-OTDR.(3)To break the limitation of receiver bandwidth on the spatial resolution and mea-surement range of the COTDR sensing system,the time domain SPEA is proposed and combined with the de-chirping technique,which greatly reduces the receiver bandwidth requirement of the sensing system.Using de-chirp technology,the large-bandwidth RS signal is compressed in the frequency domain to form a small-bandwidth received signal.By analyzing the time-frequency characteristics of this signal,the time-domain SPEA is proposed,and realizes the acquisition of RS patterns with a large frequency range,thereby realizing high spatial resolution and large measurement range strain sensing.Static strain sensing experiments show that this method significantly reduces large errors compared with traditional demodulation methods of OFDR.Dynamic strain sensing with a mea-surement range of 60 ?? and a spatial resolution of 28.4 cm was achieved on a 920 m Rayleigh-enhanced fiber with a minimum detectable strain of 80.7 p?/(?).Compared with the COTDR based on the RS pattern,the receiver bandwidth requirement is reduced by about 9 times in the condition of the same spatial resolution and measurement range.In summary,this thesis focuses on improving the comprehensive performance of RS-based DOFS.By analyzing the main limiting factors of key performance indicators of the sensing system,the corresponding improvement method is proposed,which provides a new way for exploring the performance limit of RS-based DOFS.