【作者】 于涵伟；

【导师】 娄淑琴；

【作者基本信息】 北京交通大学 ， 光通信与移动通信， 2010， 硕士

【摘要】 光纤干涉仪具有体积小、重量轻、结构紧凑以及灵敏度高等特点,非常适合于光纤传感领域的应用。但是随着光纤传感应用范围的不断扩展,特别是远距离传感的需求,对光纤干涉型传感器提出了更高的要求。基于Sagnac环结构的光纤干涉仪具有较高的信噪比和检测灵敏度,并且成本低廉,易于构建和维护,在大范围入侵检测定位上具有明显优势。本课题在传感器结构以及信号处理算法的应用两方面进行了研究,提出了一种改进型Sagnac光纤干涉仪结构,并引入信号处理技术提高定位精度,进行信号特征提取。本课题首先研究了基于Sagnac光纤干涉仪的距离检测和信号识别系统,通过深入研究Sagnac环传感器的基本结构,提出了一种改进的Sagnac传感器结构,该结构克服了基本Sagnac传感结构中存在的一些缺陷,并且降低了制作成本。通过对该结构进行理论以及仿真分析,结果显示该结构能够同时进行距离检测并实现解调功能。同时,本文设计了具有入侵事件发生检测定位功能和入侵信号特征提取功能的传感器信号处理算法,并进一步进行了仿真分析以及实验验证,结果表明该算法具有较高的定位精度和入侵信号还原能力,在41.626km光纤中,实现了小于50m的检测精度。本文对采集的传感器信号首先进行小波去噪以及高通滤波处理,滤除信号中所夹杂的噪声,并且根据定位算法滤除一定频率段的频率。然后对经过去噪处理的信号进行FFT变换,基于Berry相位法,根据频谱上陷波点的位置进行定位。同时,本文对采集的数据分别进行过零点数和基于AR参数模型的功率谱估计,在时域和频域上进行特征提取。经过以上的过程,本系统精确地实现了定位,在150Ks/s的采样率设置下,41.626km的传感光纤上,实现了46.73m的定位精度,同时有效的提取了信号的特征。该研究成果可应用于检测第三方入侵,对需要监测的地域、管线进行监控、报警并提供精确定位。研究成果对于长距离分布式干涉型光纤传感器的实用化具有重要的理论意义和实际应用价值,并在工业和国防领域具有应用前景。更多还原

【Abstract】 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.更多还原