The Key Signal Processing Algorithm Research Of Distributed Sensing Based On Brillouin Effect

Distributed optical fiber sensing technology can be used to monitor the change of the the measured information along the fiber in continuous time and space at the same time. It has great academic value and application prospect. With long sensing distance and high measurement precision, the distributed optical fiber sensing technique based on Brillouin scattering become the mainstream technology in optical fiber sensor field. This paper mainly research the Brillouin distributed optical fiber sensing system in sensing accuracy and sensing distance problem. The specific contents are as follows:Firstly, the classification and the domestic and international development of distributed optical fiber sensing technology are introduced. Then this paper expound the mechanism of spontaneous and stimulated Brillouin scattering, mainly analysis its sensing detection technology and the quantitative relation between the Brillouin scattering spectrum shift, strength and temperature, strain. The performance index of BOTDR sensing system and the technology of the three current problems are explored.Secondly, in view of high precision extraction requirement of Brillouin scattering spectrum frequency shift characteristic, this paper proposes a new Brillouin scattering spectral fitting algorithm CNS algorithm. On this basis for further improvement of the algorithm, FOAAM-GRNN algorithm which used to identify the multiple frequency shift is proposed. The experiment and simulation analysis show that the proposed algorithms can effectively improve the accuracy of measurement.Thirdly, in view of the extraction and recognition problems of the sensing invasion signal information, this paper first proposes an EMD- AWPP denoising algorithm. On this basis, the high order spectrum analysis is effectively used to extract the information characteristics of the sensor signals. Finally through the support vector machine based on particle swarm optimization algorithm, the signal types can be accurately identify. Simulation analysis show that the technology can effectively extract and recognize the information of invasion sensing signal.Finally, in order to improve the signal strength and the sensing distance, this paper proposed a G-S coding technology and use it combining with the average superposition.This method not only improves the sensing distance with guarantee of the space resolution but also enhances the intensity of impulse response on the basis of reducing the superposition times.