Preliminary Study Of Two-layer Heterogeneous Fiber Sensor Network Based On Distributed Polarization Maintaining Fiber And Discrete Fiber Bragg Grating

The fiber sensor network consists of fiber sensors or sensing subsystems in a certain topology. With the intelligent feature, a new generation of fiber sensor network and its key components has become the significant orientation in the future development of sensing technique. It has great social needs and application prospects in the fields of national security, significant project, biomedicine and so on. Two-layer heterogeneous fiber sensor network is composed by distributed polarization maintaining fiber(PMF) and discrete fiber Bragg grating(FBG) layer. The control program is written in Labview with the same acquisition system to build a small heterogeneous sensor network. Further study has been done in the distributed PMF layer, discrete FBG layer and the sensor network experiments. In the distributed PMF layer, the carrier modulation method and the EMD-based method are presented to improve the detection sensitivity and the SNR. The attenuation phenomenon of coupling intensity is analyzed and a compensation method with reference point is proposed. The discrete FBG layer is composed by SLD light source, Fabry-Perot(F-P) type tunable optical filter(TOF) and FBG arrays with self-healing topology. An F-P calibration method is put forward to improve the sensing accuracy. A series of topology experiments are conducted in order to achieve the self-healing characteristics and the practicability of the triangle topology is verified.The major work of the thesis:1. Based on the review of fiber sensor network research at home and abroad, the overall design work of two-layer heterogeneous fiber sensor network is completed. It is composed of distributed PMF and discrete FBG layer. The control program is written in Labview with the same acquirsition system to realize the heterogeneous characteristic.2. In the distributed PMF layer, the carrier modulation method is proposed based on the signal feature. The gain of SNR is 8dB and the demodulation algorithm spends only 1.12s. An EMD-based method is presented and the detection sensitivity is improved to -70dB. A modified method with reference point is proposed to effectively compensate the intensity attenuation in the long PMF sensors and the absolute error of coupling intensity is less than 0.59 dB.3. In the descrete FBG layer, the influence of the light distribution is analyzed and a modified method is proposed. An F-P calibration method based on virtual instrument is presented to improve the sensitivity and the demodulation accuracy is decreased to less than 0.35με. Stain sensing experiments are done and several signals of common actions are extracted to prepare for the latter pattern recognition.4. A series of topology experiments are conducted in order to achieve the self-healing characteristics and the practicability of self-healing topology is verified. The remote panel access in Labview is described for initial networking. The sensing layers based on the distributed PMF and the discrete FBG preliminarily realize the heterogeneous and the self-healing of the fiber sensor network.