Dynamic Strain Demodulation Techniques Of Distributed Optical Fiber Sensing Based On Φ-OTDR

As an emerging sensing technology,distributed optical fiber sensing has attracted wide attention due to its advantages of low loss,anti-electromagnetic interference,high temperature and high pressure resistance.At present,it has been used in energy,transportation,military,aerospace and other fields.This technology usually uses ordinary optical fiber as the sensing medium to realize the distributed continuous measurement of the whole link by the backscattering in the optical fiber.There are three main kinds of scattering in optical fiber:Rayleigh,Brillouin and Raman scattering.The phase-sensitive time domain reflectometer(Φ-OTDR)based on Rayleigh scattering has the characteristics of large scattering power,high sensitivity and fast response.It is more suitable for the field of vibration measurement,such as underwater monitoring,seismic monitoring,perimeter security,oil and gas pipeline condition monitoring,and high-speed railway condition monitoring.With the demand for quantitative disturbance monitoring in practical application,dynamic strain demodulation based on Φ-OTDR has become a research hotspot.In the past decade,researchers have proposed a variety of methods to retrieve dynamic strain from the perspective of phase demodulation and frequency demodulation.However,there are still some technical problems such as low measurement accuracy,poor real-time performance,and high demodulation cost.In view of the above problems,this thesis is mainly carried out from the following aspects:Firstly,the frequency demodulation technology based on traditional coherent detection Φ-OTDR is studied to reduce the system cost;Secondly,in order to improve the real-time performance of measurement,the corresponding low-sampling phase demodulation method is studied from the perspective of homodyne delay self-interference detection and heterodyne undesampling detection;Finally,the quantitative dependence of phase demodulation accuracy and stability on the SNR of the detected signal in heterodyne coherent Φ-OTDR is studied.Aiming at the problem of interference fading,two fading suppression algorithms are proposed.In detail,the main works of this thesis including:1.According to the wide spectrum characteristics of the narrow pulse response in Φ-OTDR and the characteristics of convolution operation,a pulse conversion algorithm is proposed to convert the signal detected by conventional narrow pulse into the signal detected by equivalent chirped pulse through chirped factor convolution operation.Therefore,the frequency demodulation once applied to chirped-pulse Φ-OTDR can be applied to the tradition coherentΦ-OTDR,which expands the applicable range and reduces the system cost of frequency demodulation.In the experiment,utilizing the natural anti-fading feature of frequency demodulation,the vibration measurement without dead zone is realized on 9-km fiber with dynamic strain of 56 pε/√Hz.2.In order to improve the real-time performance of quantitative measurement in Φ-OTDR,two methods for demodulating phase under the acquisition condition of single channel and low sampling rate.(1)A differential-cross-multiply(DCM)phase demodulation method for homodyne delay self-interference balanced detection system is proposed.After using the balanced detection to eliminate the common mode noise of the detected signal,the vibration waveform is obtained by the first-order DCM,and the phase information is accurately extracted by the second-order DCM compensation for the vibration amplitude.(2)A free undersampling phase demodulation method for heterodyne coherent detection system is proposed.The detected signal is acquired by undersampling and demodulated in the “time”direction to avoid the limitation of the undersampling sampling rate due to the spectral aliasing effect in the “distance” direction.3.Aiming at the problem that the phase demodulation is affected by interference fading in heterodyne coherent Φ-OTDR,we propose two fading suppression algorithm: multifrequency decomposition and moving rotation vector averaging.(1)In order to make full utilize the spectrum contents of the normal probe pulse,the detected signal is decomposed into a large number of signals detected by equivalent multi-frequency pulses by short-time Fourier transform.Interference fading is suppressed according to the characteristics of different fluctuations and fading point of the signals decomposed at different frequencies.(2)After rotating and aligning the initial phase angle of the detected signal along the “distance”direction,the amplitude fluctuation of the detected signal is effectively alleviated and the noise power is reduced by moving average,and then the interference fading is suppressed.Experimental results show that these two methods can effectively suppress fading,and improve the average SNR of the demodulated phase by at least 8d B.4.According to random signal analysis theory,the distribution functions of demodulated phase noise in heterodyne coherent Φ-OTDR are deduced.Based on this,the quantitative relationship between the phase demodulation accuracy and the SNR of the signal to be demodulated is completely studied.Aiming at the instability of phase demodulation caused by fading,the phase hopping rate(PHR)is defined as the evaluation index of demodulation stability,and the quantitative relationship between PHR and SNR is deduced.Ultimately,the experimental statistical results are consistent with the theoretical predictions.