Research On Optical Fiber Low-frequency Acoustic Wave Sensing Technologies And Signal Demodulation Algorithms

Low frequency acoustic waves have strong diffraction abilities due to its long wavelength,so that it's easy to steer clear of obstacles in the propagation.Besides,low frequency acoustic waves have relative low transmission loss in the media.Owing to these characteristics,low frequency acoustic waves have very long propagation distance,thus attracting a lot of interests in many applications such as underwater acoustic communications,sonar detection,anti-submarine early warning,and even infrasound weapons,etc.Besides,low frequency acoustic waves can be generated in many artificial activities and natural disasters such as debris flow,volcano eruption,pipeline leakage,and launching of rockets or missiles,etc.Therefore,low frequency acoustic wave sensing can be an effective tool for early warning and monitoring these events.Recently,optical fiber based sensing techniques of low frequency acoustic waves have attracted wide study interests.Compared with traditional electronic sensors,optical fiber acoustic sensors(OFAS)have inherent superiorities of compact size,light weight,resistance to harsh environments,and immune to eletro-magnetic interference(EMI).The principle of OFAS is the modulation of specific parameters(intensity,optical frequency,phase,polarization states,etc)of optical signals introduced by the acoustic waves.After this process,the output optical signals from the sensor will carry the informations of acoustic waves that received by the OFAS.It is easy to figure out from this working principle that the demodulation of the sensor signal also plays a significant role in the OFAS performance.This thesis mainly focuses on the research of OFAS techniques and signal demodulation algorithms.The main research contents and results of the thesis are summarized as follows:(1)Spectral modulation technology based on signal conversion from time to optical frequency domain assisted by optical wavelength scanning is proposed.The wavelength scanning process builds a mapping relationship between time and optical frequency domain.In the meantime,the sensor spectrum is receiving dynamic modulation from the time-varying acoustic pressure signals.Hence,the time-domain acoustic signals are converted to wavelength domain and the wavelength-scanned optical spectrum(WSOS)will carry the acoustic information.By analyzing the WSOS,acoustic signal is possible to be recovered.A long period grating(LPG)attached to a polyethylene terephthalate(PET)membrane is experimentally selected as a sensing element to verify the proposed technique.Acoustic signals below 1000Hz is well demodulated with minimum detectable pressure(MDP)of 604.6?Pa/Hz^1/2@1000Hz.It is also demonstrated that the proposed technique can isolate environmental perturbations to the recovered signals.(2)WSOS-based phase demodulation algorithm is proposed for interferometric sensors.According to the theory of optical interferometer,the mathmatical model of phase-modulated WSOS is derived.Based on this model,a phase demodulation algorithm is developed,which requires two channels of WSOS with a fixed phase difference(not integer multiple of?).Experimentally,a Michelson interferometer(MI)composed of a reference arm and a sensing arm is built up with a 3×3 fiber coupler.The 3×3 coupler can provide 2?/3 phase difference between two outputs of the MI sensor for the algorithm.Comparative experiment is also conducted on the same MI sensor by edge-flitering demodulation technique.Signals recovered by the proposed WSOS phase demodulation method exhibits excellent linear relationship with sound pressure,with R² fitting value of99.6%,while signals recovered by the edge-filtering technique shows fluctuations of more than 10d B under the same testing condition,due to the thermal fluctuation of the MI arms.Therefore,it is further proved that WSOS-based signal demodulation techniques are immune to slow-varying environmental instabilities.(3)A method for sound information extraction is proposed based on Fourier spatial frequency spectrum analysis of WSOS.Phase modulation loaded on the WSOS of the FP sensor will introduce a pair of dual-sideband signals in the spatial frequency domain.Sound information can be extracted by analyzing the spatial frequency and phase of the sideband components in the Fourier domain.This technique is also able to distinguish signals received by different sensors at the same time,which is demonstrated by two parallel connected FP sensors as a proof-of-concept experiment.This indicates that the proposed technique has great potential in applications for sensor array and multiplexing.(4)A white light interferometry(WLI)phase demodulation algorithm is proposed.By analyzing the interferential spectrum in the Fourier domain,the spatial frequencies(determined by optical path difference)of the sensor can be obtained.External signals received by the sensor can be demodulated by phase analysis of the sensor spatial frequencies.The proposed technique is firstly verified under static performance by an inline multimode interferometric sensing platform,which is made of single mode fiber offset splicing structure.By using WLI phase demodulation,dual-parameter sensing of temperature and strain is realized on this platform,with sensitivities of-1.47°/?and0.019°/??,respectively.Isolation ability of the WLI phase demodulation to spectral noise is also verified by the static experiment.The demodulation algorithm is further applied on dynamic acoustic wave sensing,with the help of silicon micromachined FP sensor.Low frequency acoustic wave sensing is realized in the range of 0.5-250Hz with a flat sensitivity response,with fluctuations of only 0.8d B.