The Research Of Novel Temperature Demodulation Method And Advance Prewarning Model For Fiber Raman Sensing System

Fiber sensing technology is a basic and strategic industry of the national economy.At present,fiber sensing technology based on Raman scattering has been applied to a certain extent,especially in the field of temperature monitoring of transportation infrastructure,smart grids,and gas pipelines.In recent years,with the development of distributed fiber Raman sensing technology,it has been paid attention by the scientific researchers because of its advantages of anti-interference ability,space applicability,high automation,convenient construction and low cost.Distributed fiber sensing technology based on Raman scattering has the problems of low measurement accuracy and long measurement time.And it cannot provide early warning.This paper firstly proposes a scattering signal strength difference algorithm of dual-ended calibration fiber,a dispersion compensation translation algorithm and a wavelet modulus maxima denoising algorithm to optimize the temperature measurement accuracy and system running time.Secondly,the early warning model based on the moving average forecasting method was developed to solve the problem that the current distributed fiber Raman temperature sensing technology cannot predict the temperature in the next period.We developed a novel long-distance distributed fiber Raman thermometer and a smart human-computer interaction upper computer system,which included comprehensive design of system calibration module,fiber length detection module,temperature demodulation module,wavelet denoising module,alarm module and temperature prediction module.The main research contents of this paper as follows:(1)A scattering signal strength difference algorithm of dual-ended calibration fiber is proposed.After anti-Stokes and Stokes backscattered light are corrected by using the dual-ended calibration optical fiber scattering signal intensity difference algorithm,the system can directly use the modified the light intensity ratio curve to demodulated the temperature information.It does not need to be calibrated before the temperature measurement,which reduces the system running time.(2)A dispersion compensation translation algorithm is proposed.In the process of demodulation,the Stokes signal needs to be translated so that the scattering position is consistent with the scattering position of the anti-Stokes signal,it can eliminate the dislocation error.The experimental results show that the temperature error of the system decreases from 5.5? to 0.87?,and the temperature fluctuation range decreases from 9.01? to 0.57? at the sensing distance of 5.8 km.(3)A wavelet maximum denoising algorithm is proposed to improve the temperature measurement accuracy,and its fluctuation range is reduced from 5.98? to 1.23? at the sensing distance of 11.5 km.(4)The data mining and fusion technology of real-time moving average forecasting method was developed.The historical temperature data of Raman thermometer was analyzed and a temperature prediction model was established.The accurate temperature prediction along the sensing fiber was achieved after 60 s.This solves the problem that the current distributed fiber Raman thermometer cannot predict the temperature in the next period.(5)An intelligent human-computer interaction interface was designed on the LabVIEW and MATLAB platform,it includes calibration module,fiber length detection module,temperature demodulation module,wavelet denoising module,alarm module and temperature prediction module.(6)A Novel distributed fiber Raman temperature detector was developed,and a sensing range of 30 km,temperature accuracy of ±1?,positioning accuracy of 2 m,and spatial resolution of 3.4 m were experimentally obtained.Finally,this novel system is applied to the temperature monitoring of large-scale buildings and natural gas transmission pipelines in Surabaya.