Research On Key Algorithm Of Distributed Optical Fiber Sensor Signal Extraction Based On Brillouin Scattering Spectrum

The distributed optical fiber sensing technology which based on Brillouin scattering spectrum can collect the optical power spectrum of the Brillouin scattering spectrum in the sensing fiber.In order to achieve the purpose of sensing,it uses the changes of the center frequency shift to establish a mapping relationship with the changes of external physical parameter.It is a research focus of distributed optical fiber sensing technology based on Brillouin scattering spectrum to obtain the characteristic information of Brillouin scattering spectrum precisely.And it has a great academic value and application prospect.In this paper,the key algorithms of distributed fiber sensing signal extraction based on Brillouin scattering spectrum are studied.Firstly,this paper briefly introduces the classification of distributed optical fiber sensing technology,and introduces the research status of optical fiber sensing technology based on Brillouin scattering spectrum.The mechanism of Brillouin scattering,the fitting principle of Brillouin scattering spectrum and the relationship between Brillouin scattering spectrum and physical temperature and physical strain are analyzed.This paper also introduces different kinds of techniques for Brillouin signal detection.Secondly,a new hybrid optimization algorithm is presented,which is based on the genetic algorithm and QPSO algorithm.The hybrid optimization algorithm uses the crossover and mutation operators of GA to optimize the QPSO algorithm,improves its global search ability and overcomes the disadvantage that QPSO algorithm easily falls into local extremum.It is used to extract the character of the Pseudo-Voigt-shaped single peak Brillouin scattering spectrum.The parameters estimation and simulation analysis of the single peak Brillouin scattering spectrum are analyzed under different weight ratios,line widths and signal-to-noise ratios.Finally,this paper presents a hybrid optimization algorithm combining genetic algorithm with BP neural network.The hybrid optimization algorithm can provide better initial value for BP neural network by using GA's strong global search ability and improve the accuracy of BP neural network for global optimal solution.It is used to extract the character of the multi-peak Brillouin scattering spectrum.The parameters estimation and simulation analysis of multi-peak Brillouin scattering spectrum are analyzed under different conditions.