| Distributed optical fiber sensing technology has the advantages of anti-electromagnetic interference,corrosion preventive,long distance,and real-time monitoring of multiple physical quantities along the optical fiber,and is therefore of important value for the military defense,industrial production and aerospace fields.Brillouin optical time domain analysis(BOTDA)sensing systems have been widely used in major infrastructure safety monitoring such as the transportation,power industry,petrochemical industry and civil engineering due to the advantages of long sensing distance and high measurement accuracy.However,there are two key issues with traditional BOTDA distributed optical fiber sensing technology:the spatial resolution is difficult to exceed 1 m due to the limitation of 10 ns phonon lifetime in the fiber,the signal-to-noise ratio(SNR)deteriorates severely with increasing sensing distance due to the transmission loss and nonlinear effects of optical signals in the fiber.Aiming at the above scientific and technological issues,the differential pulse width pair BOTDA scheme based on gain switch modulated is proposed in this thesis,and develops an integrated Brillouin temperature/strain sensor based on this scheme.Furthermore,a pulse delay self-difference sensing scheme is proposed to achieve centimeter level spatial resolution.In addition,to enhance the SNR of the system in medium or long-distance sensing,a higher SNR sensing demodulation scheme based on time-frequency analysis local mean decomposition method,as well as a scheme combining complementary integrated empirical mode decomposition and particle swarm optimization are proposed respectively.The main innovation contribution completed in this dissertation are as follows:(1)A gain switched BOTDA with differential pulse-width pair(GS-DPP-BOTDA)scheme based on semiconductor optical amplifier(SOA)modulation is proposed.Firstly,the transmission process of probe and pump wave in three-wave coupled is investigated,and a BOTDA sensing model based on probe wave demodulation is established.Secondly,the demodulation and positioning principles of the DPP-BOTDA scheme are analyzed,and the effects of different differential pulse pairs,pump pulse rising/falling time,and pulse extinction ratio on the demodulation results are simulated.The experimental results show that compared to the DPP-BOTDA based on electro-optical modulator modulation,the pulse extinction ratio of GS-DPP-BOTDA system based on SOA modulation is improved by 11 dB,the SNR of differential signal is improved by 10 dB,and a spatial resolution of 10 cm is achieved over a sensing distance of 2 km.Finally,an integrated Brillouin temperature/strain sensor is designed and developed based on the above research scheme.(2)A centimeter-level spatial resolution BOTDA sensing scheme based on pulse delay self-difference(PDSD)is proposed.Firstly,the generation process of Brillouin scattering signal from a single-shot long-pulse in a traditional BOTDA system is analyzed,and a linear quantitative relationship between the falling edge of the BOTDA trace and the length of the event zone is established,small-scale event demodulated by performing delay self-difference of BOTDA trace in the time domain.Then,verified the demodulation principle of the scheme and its high spatial resolution demodulation ability for single and multiple events by the simulation.Finally,the distributed temperature measurement with 5 cm spatial resolution is realized based on a 40 ns pulse along a 2 km sensing fiber,and the spatial resolution at different sensing distances is measured,overcoming the limitation of phonon lifetime.(3)An adaptive high SNR demodulation scheme based on local mean decomposition(LMD)with time-frequency analysis characteristics is proposed.Firstly,the constraints of SNR on various indicators of the system are analyzed.And the distribution characteristics of BOTDA signal in the time and frequency domains are studied,then the signal is adaptively decomposed into a series of product function components using LMD,and the energy distribution of the signal at different spatial scales is obtained.Then,each component is classified and selected through energy calculation,and a Chebyshev low-pass filter is designed for noise filtering according to the low-frequency distribution characteristics of the effective information in each component.Finally,the distributed temperature sensing with an average SNR improvement of more than 10 dB over a sensing distance of about 23 km is achieved in experiment,the adaptive setting of parameters in noise processing is realized,which has strong engineering practicality.(4)A dual-adaptive high SNR demodulation scheme combining complementary ensemble empirical mode decomposition(CEEMD)and particle swarm optimization(PSO)is proposed.Firstly,the noisy BOTDA signal is efficiently decomposed using CEEMD with no mode mixing,the time-frequency distribution of each component is obtained.Then the mixed component of the signal and noise is obtained by calculating the centroid frequency.On this basis,an adaptive filtering method based on PSO is designed,which can adaptively search for the optimal filtering threshold according to the noise level of different components,overcoming the problem of excessive or insufficient denoising existing in traditional hard threshold/soft threshold filtering denoising methods and prior condition estimation.The results show that this scheme improves the SNR of the original signal more than 14 dB in a 23 km sensing fiber,which reduces the temperature measurement accuracy from 2.97℃ to 1.85℃,and has no deterioration in spatial resolution. |
PDF Full Text Request