Research And Realization Of The Distributed Optical Fiber Vibration Sensing System Technology

Distributed optical fiber sensing technology which has many advantages such as low cost, high concealment, high sensitivity, high temperature resistance, corrosion resistance and high resistance to electromagnetic interference, is a rapidly developed and widely used sensing technology in recent years. It has wide application prospect in border security, community security, long distance pipeline monitoring, earthquake monitoring and construction monitoring.At present, the research of distributed optical fiber sensing technology mainly focus on the improvement of the optical fiber sensing structure and less on the signal processing especially the extraction of feature parameters. But the quality of a distributed optical fiber sensing system not only depends on whether it can sense the external disturbance, but also to see whether it can identify the perturbation of the external environment disturbance or intrusion behavior accurately. In order to identify the invasion of vibration signals better, a recognition algorithm which is taking the speech signal processing method as a reference is introduced to the processing of optical fiber vibration signal analysis based on the characteristic of vibration signals.At first, it introduces the basic principle of distributed optical fiber sensing briefly and then present several kinds of optical fiber sensing structure which are used widely emphatically. In the third part, we analyze several kinds of signal processing algorithms from time domain, frequency domain and voice signal recognition respectively based on Mach-Zehnder structure, and choose one algorithm from the frequency domain and another from the voice signal recognition for comparison in the end. In the fourth part, the architecture of the system is introduced, mainly including front-end interface, data storage and the analysis of the algorithm processing module. At last we collected the signals of the rain and invasion through the experiment and used these two different algorithms to deal with respectively. It turns out that the signal processing algorithm based on speech can not only be applied to the optical fiber vibration signal analysis, but also get higher recognition accuracy.