Research And Implementation Of The Distributed Optical Fiber Sensing Perimeter System

Along with the rapid socio-economic development and the continuous progress of science and technology, optical fiber communication technology to the optical fiber for the transmission medium and the optical fiber sensing technology based on the principle of guided wave fiber in people's life in various fields have been widely applied. Due to the wide monitoring range, high sensitivity, low energy consumption and adapt to the harsh environment advantages, perimeter security system based on distributed optical fiber sensing technology obtained social recognition and favor. But how to improve the ability of detection and reduce the false alarm rate of the system is a difficult problem.We can solve this problem from two aspects: one is to optimize the structure of the sensor, so that the system can better perceive and record the external disturbance, the other is to optimize the signal processing algorithms, so that the system can better grasp the characteristics of vibration signals and improve the recognition ability. The current research on the sensing structure has been relatively mature, while the vibration signal processing algorithm is relatively lagging behind. By studying of the distributed sensing structure and signal processing algorithms, in this thesis we introduce the time-frequency analysis technology to process optical fiber vibration signals.Firstly, this thesis introduces the principles of the distributed optical fiber sensing technology briefly and then presents several interferometric distributed sensors. According to the comparison of their advantages and defects we choose Mach-Zehnder structure as the sensing structure for system development. Secondly, we analyze several commonly used signal processing algorithms, and choose one algorithm from time domain and another from time-frequency domain for further study. Thirdly, we propose the architecture of the whole system, especially describe in detail the module of signal processing, data storage, the communication between the host computer and the slave computer. Finally, we use the two algorithms to deal with intrusion signals in the strong interference environment respective. The result shows that the time-frequency analysis algorithm gets higher recognition accuracy and can be extended to the optical fiber sensing vibration signal recognition.