Research On Signal Denoising Algorithm Of Distributed Optical Fiber Vibration Sensing System

Phase-Sensitive Optical Time Domain Reflectometry(?-OTDR,Phase-Sensitive Optical Time Domain Reflectometry,?-OTDR)has the advantages of high measurement sensitivity,fast response,and simultaneous detection of multi-point disturbances.It has important application value in the fields of oil and gas pipeline leakage monitoring and underwater acoustic detection in the ocean.The sensing system will be affected by the system noise and the external environment,resulting in a low signal-to-noise ratio of the received sensing signal,affecting the correct demodulation of the disturbance signal by the system,reducing the accuracy of disturbance positioning and the accuracy of signal feature extraction.This article focuses on the denoising algorithm of ?-OTDR system sensor signal.The main research contents are as follows:(1)The structure and the detection principle of the ?-OTDR distributed optical fiber sensing system are introduced.The received sensing signal is converted into a time-distance two-dimensional matrix signal.Through analysis,it is concluded that the disturbance signal has high-frequency variation characteristics and sparsity characteristics on the distance axis.Combined with the characteristics of the disturbance signal,the basic principles of the moving average algorithm,wavelet transform algorithm and compressed sensing algorithm are analyzed.The ?-OTDR experimental system was built,and the single-frequency PZT vibration signal was used as the disturbance signal,and the received signals were processed by the above three methods.Experimental results show that,compared with the other two methods,the use of compressed sensing algorithm can more effectively improve the signal-to-noise ratio of the received signal in the sensing system,achieve accurate positioning of disturbance events,and accurately obtain the frequency spectrum of disturbance signals.(2)After denoising the received time-distance two-dimensional matrix signal in the direction of the distance axis,the signal strength has more obvious change characteristics near the disturbance point.Based on this feature,the paper uses the image processing algorithm to process the above-mentioned denoised two-dimensional signal to further improve the system signal-to-noise ratio,improve the accuracy of the sensor system for disturbance positioning,realize the correct demodulation of the disturbance signal and extract of signal features accurately.The thesis researches the uses of the combination method of the wavelet transform algorithm and the edge detection operator in image processing,as well as the combination method of the compressed sensing algorithm and the guided filtering algorithm to process and analyze the received signal of ?-OTDR system.The experimental results show that using a single-frequency signal as the disturbance signal,compared with the results of only using the compressed sensing algorithm,the combined algorithm processing of compressed sensing and guided filtering can significantly improve the signal-to-noise ratio of the system.Finally,simulating the two intrusion situations of shaking and cutting the protective net in practical applications,further experimental verification of the combined method of compressed sensing and guided filtering is done.