Research On Brillouin Optical Time Domain Analysis Employing Optical Pulse Coding

Brillouin Optical Time Domain Analysis(BOTDA) is a state-of-art full-distributed optical fiber sensor, which can receive strain and temperature information over a long fiber. Thus, BOTDA shows many superiorities in Giant Structures Health Monitoring such as Bridge, Dam, Oil and Gas pipelines, Railway and so on. In order to improve the performance of BOTDA, researchers have proposed a series of methods to make BOTDA gain high signal-noise-ratio(SNR). In this thesis, we mainly emphasize on Optical Pulse Coding.The impact of Optical Pulse Coding to BOTDA is theoretically analyzed through the research on the Stimulated Brillouin Amplification, whilst further study the improvement of SNR for coded-BOTDA through simulation and experiment. Meanwhile, a novel hybrid pulse-width code is proposed, which is theoretically and experimentally demonstrated.First of all, the principles of BOTDA is introduced in this paper. To get the modeling about BOTDA for simulation in MATLAB, Stimulated Brillouin Scattering is theoretically studied. Besides, a common method for advanced spatial resolution named differential pulse-width pair (DPP) is studied in simulation and experiment.Secondly, three common coding methods based on BOTDA, i.e. Simplex, Hadamard and Golay, are introduced, and these methods are theoretically analyzed and demonstrated in simulation.Meanwhile, the performance of each method is studied, and Golay code is experimentally demonstrated.Then, a method combinedbipolar codes with DPP is employed to further improve the performance of BOTDA. Simulation results indicate that the proposed hybrid pulse-width code shows 4 times enhancement of SNR. Besides, an experimental platform of BOTDA is set up, and the results show that the hybrid pulse-width code is an alternative method to improve SNR in BOTDA.Finally, digital image processing is adopted in BOTDA, and the experiment results are further optimized. Moreover, the comparison between two algorithms, i.e. Gaussian-filteringand Non-Local Mean filtering, is conducted, and it is proved that digital image processing improves the performance of BOTDA.