Study On Modified Sagnac Interferometer's Position Detection Signal's Recognition And Processing Technology

Fiber interferometer is suitable for the field of fiber sensor application with a small size, light weight, compact structure and high sensitivity characteristics. However, because the application of fiber sensor is growing fast, particular the needs of large-scale intrusion detection, the requirement of interferometric fiber optic sensor is higher. Sagnac interferometer, which has high reliability noise ratio, high detection sensitivity, low cost and easiness to construct and maintain, is very suitable for large-scale intrusion detection. The paper researches the structure of Sagnac interferometer and signal processing technology, and proposes a modified Sagnac Interferometer, uses signal processing technology to increase the accuracy of position detection and features extraction.The research mainly focus on the distance and signal recognition systems based on Sagnac interferometer, through studying the basic structure of Sagnac loop sensor, a modified Sagnac sensor structure, which overcomes the basic Sagnac sensing some flaws exist in the structure and reduce production costs. By the theory and simulation analysis on this structure, results show that the structure can simultaneously detect the distance and achieve demodulation functions. Meanwhile, this event is designed with intrusion detection and intrusion positioning signal feature extraction capabilities of the sensor signal processing algorithms, which are further verified by simulations and experiments, show that the algorithm has higher accuracy and intrusion signal reduction capability, In the 41.626km optical fiber, the achieved detection accuracy of less than 50m.In this paper, the first sensor signal acquisition wavelet denoising and high-pass filter, filter out the noise signal in the mixture, and the positioning algorithm based on filter a certain frequency band. Then, apply FFT transform on the signal after denoising. With the Berry phase method, positioning based on the spectrum location of the point on the notch. Meanwhile, the number of zeros was conducted and the AR parameter model based on power spectrum estimation in the time domain and frequency domain feature extraction. After the above process, in 150Ks/s sampling rate settings,41.626km fiber sensing, positioning accuracy achieved 46.73m, while effectively extract the signal features.The results can be applied to third-party intrusion detection, the need for regional monitoring, pipeline monitoring, alarm and provide accurate positioning. Results for long-distance distributed fiber interferometer with practical application have theoretical significance and practical value, in addition, a good prospect in the field of industrial and defense.