| Distributed optical fiber vibration sensing systems as a novel sensing technology which can simultaneously transmit and sense information,and the vibration sensing information on the fiber can be obtained by detecting the light signal.This kind of technology has been successfully applied in various vibration sensing fields.Aiming at the challenges of long-distance precise position,distributed multi-point and wide frequency sensing,and multi-category disturbance mode detection that need to be urgently solved in practical applications,new schemes and new methods have been designed to significantly enhance the performances of the distributed optical fiber vibration sensing systems.Specifically,the schemes and methods are mainly based on the perspectives of dual wavelength based noise suppression,dual mechanism based integration sensing and intelligent sensing detection.The main research works are implemented in the aspects of theoretical model establishment,sensing system design,vibration signal demodulation and processing,and instrumentation design.The main research contents of this thesis are as follows:1.Aiming at the problems of low signal-to-noise ratio(SNR)and low positioning accuracy of dual Mach-Zehnder interferometer based optical fiber vibration sensing system in long distance sensing,an asymmetric Mach-Zehnder interferometer based vibration sensing structure with different central wavelength has been established.The influence of the different central wavelengths on the positioning accuracy of the asymmetric vibration sensing structure is firstly analyzed.Then a maximal overlap discrete wavelet transform based positioning scheme is proposed in this thesis.And experimental results show that a detection of more than 95% of positioning errors are distributed within the range of ± 40 m at a sensing length of ~101 km and a mean processing time of 135 ms is also achieved.2.Aiming at the detection problem that the optical fiber interferometer vibration sensing system cannot locate multiple points,a wavelength division multiplexing based distributed optical fiber vibration sensing system by integrating Φ-OTDR and Mach-Zehnder inteferometer is built.A demodulation scheme by using the phase information of the vibration signal is developed,so the dynamic linear demodulation range of the applied signals is expanded.A single fiber link based distributed optical fiber vibration sensing system is also proposed through integrating the Φ-OTDR and a Sagnac interferometer.Thus,the high positioning accuracy and the wide frequency response can be simultaneously detected in a single sensing fiber.Experimental results show that multiple vibrations with a spatial resolution of ~20 m have been achieved in a ~4 km fiber link.3.Aiming at the challenges of high false alarm rate in practical applications,a multi-dimensional eigenvector of the sensing signal is constructed.According to the effective features can be automatically extracted by the deep learning based neural network,a two-dimensional grayscale image based signal preprocessing method and a Mel time-frequency image based signal preprocessing method have been proposed.Therefore,the complexity of the signal preprocessing process can be significantly simplified.And the recognition efficiency and accuracy can be further improved.Field test results show that the proposed deep learning scheme can classify 8 types of sensing events,and a mean recognition rate of 98.6% has been achieved.4.Finally,based on industrial design requirements of equipment instrumentation and modular design idea,a engineering prototype based on the integrated distributed optical fiber vibration sensing system is designed.A distributed defense area module with multi-platform monitoring function is developed to realize partition prevention and accurate monitoring of intrusion events through using the UDP transmission protocol.A series of field tests have been implemented in the field of perimeter security.The results show that the integrated optical fiber vibration sensing system can satisfy the requirements of multi-point detection in practical applications. |