Location Algorithm Of Distributed Disturbance Sensing System Based On ?-OTDR

Distributed sensing system based on phase-sensitive optical time domain reflectometer(?-OTDR)is widely used in the field of perimeter security because of its unique advantages such as high sensitivity,long-distance distributed sensing,strong anti-electromagnetic interference ability,small size and light weight.Aiming at the problems of existing distributed sensing system based on ?-OTDR,a wavelet threshold de-noising algorithm based on simulated annealing algorithm(SA)is proposed.Combining with a two-dimensional edge detection method,the signal-to-noise ratio of far-end disturbance signals in the sensing system can be efficiently improved and thus location of multi-point disturbance events can be recognized.The main achievement of this thesis are described as the following.1.The sensing principle of distributed sensing system based on ?-OTDR is analyzed,and the sources of optical background noise of ?-OTDR technology are studied.The results show that the noise of light-source caused by laser frequency drift and polarization-related noise caused by fiber birefringence are the main optical background noises of ?-OTDR technology.The theoretical analysis of the system and noise sources lay a foundation for the subsequent signal de-noising.2.The effect of wavelet threshold de-noising(WTD)algorithm on the noise reduction of vibration signal is studied in order to improve the signal-to-noise ratio of vibration signal through signal processing.The vibration signal received by distributed sensing system based on ?-OTDR is analyzed by wavelet transform and time-frequency analysis.The distribution of disturbance information and noise is summarized by time-domain and frequency-domain analysis.The noise of vibration signal is reduced by WTD algorithm.The experimental results show that it is impossible to effectively de-noise the far-end disturbance signals of sensing system only by using WTD algorithm.3.In view of the shortcomings of the traditional WTD algorithm,a wavelet threshold de-noising algorithm based on SA is proposed,which can search the global optimal threshold adaptively and effectively improve the signal-to-noise ratio for the far-end disturbance signals of the sensing system.The experimental results show that the signal-to-noise ratios of the four disturbance events at 21.5 km are 7.3 dB(watering),8.8 dB(pressing),6.2 dB(climbing)and 7.9 dB(knocking).On this basis,aiming at the vibration signal after noise reduction,the two-dimensional edge detection method in image processing is applied to effectively locate the multi-point disturbance events.Four disturbance events are detected and located at four different locations simultaneously,which is the best localization result reported at present.4.The distributed disturbance sensing system based on ?-OTDR is built and the data is sampled.The ability of noise reduction and location of the proposed algorithm is verified by experiments.The experimental results show that the multi-point disturbance events location with an average positioning accuracy of more than 90%can be achieved by combining the wavelet threshold denoising algorithm based on SA with two-dimensional edge detection method.