Research On Key Technologies Of High Performance Raman Scattering Distributed Optical Fiber Temperature Sensing System

In recent years,the scale of China's large-scale infrastructure has grown rapidly.Large-scale infrastructure has a long service life,and it is very likely to cause structural damage during service,which may cause serious safety accidents and economic losses.Therefore,its structural health monitoring has received increasing attention.Temperature monitoring is an important part of structural health monitoring.Structural deformation,damage and aging caused by temperature changes have a significant impact on the structural reliability of infrastructure,and temperature information can be used to infer the health status of infrastructure.At the same time,temperature information can provide early warning of abnormal events such as fire or pipeline leaks.Distributed optical fiber sensing technology can sense the temperature,stress,vibration and other parameters along the fiber.It is especially suitable for the needs of large-scale,long-distance,continuous distribution in the health monitoring of infrastructure structures,and has unique advantages.As an important branch of distributed optical fiber sensing technology,a distributed optical fiber temperature sensing system based on Raman optical time-domain reflection uses spontaneous Raman scattering effect to obtain temperature information along the fiber,which can achieve a sensing distance of tens of kilometers.The spatial resolution of the order of magnitude provides a high-density,low-cost,long-distance distributed temperature monitoring solution for structural health monitoring,so it is very suitable for distributed temperature monitoring of large infrastructure.However,the current distributed optical fiber temperature sensing system based on Raman optical time domain reflection has some problems,which limits its promotion in engineering applications.First,the measurement accuracy of the system depends on the selection of the temperature reference point,and the existing temperature calibration method needs to be returned to the factory for multiple inspections,which cannot achieve real-time temperature calibration.Secondly,the improvement of system measurement accuracy will cause the deterioration of other performance parameters.At the same time,the improvement of measurement accuracy depends on complex hardware design,which increases the manufacturing cost of the system.Finally,the system has a single temperature warning method,which cannot fully assess the threat of temperature anomalies.The setting of the alarm threshold is highly dependent on manual participation.To this end,this article mainly carries out the following work: 1.In order to achieve the system temperature self-calibration function,an external self-calibration device and a dynamic temperature self-calibration scheme are designed,which eliminates the process of returning to the factory and improves the system's intelligence;2.Analyze the noise characteristics that limit the improvement of the system temperature measurement accuracy,and propose a corresponding noise suppression method.Using the high-speed data acquisition card in the embedded system,the cumulative average of the Raman scattered light signal is completed,and the system broadband noise is suppressed.A dynamic noise floor difference method is proposed to suppress the floor noise generated by the photoelectric detection module due to insufficient dynamic range during long-distance detection.The ratio of the anti-Stokes light and the Stokes light signal is subjected to attenuation compensation processing to eliminate the influence of the difference in attenuation coefficient of the anti-Stokes light and the Stokes light signal on temperature demodulation.Comparing different threshold selection schemes of wavelet transform,the Minimaxi scheme is preferred to perform wavelet threshold noise reduction on the demodulated temperature curve,which improves the temperature measurement accuracy of the system;3.The absolute temperature,temperature change,The multi-parameter comprehensive diagnosis of the temperature change rate and the temperature prediction alarm method based on the adaptive threshold solve the problem that the current distributed optical fiber temperature sensing system temperature alarm method cannot be applied to multiple scene applications,effectively shortening the temperature alarm Response time reduces the probability of false negatives and false positives.This paper has successfully developed a high-performance Raman scattering distributed optical fiber temperature sensing system with self-calibration capability.The system achieves a spatial resolution of 3m,a temperature measurement accuracy of ±1.5? and a temperature resolution of 0.11? at a measurement distance of 10 km.Through signal processing,without changing the hardware structure,the system measurement temperature fluctuation is reduced by 4.02?,the signal-to-noise ratio is increased by 13.87 d B,and the temperature measurement accuracy is increased by 35.77%.The simulation experiment of pipeline leakage was set up under the external field conditions,which proved that the system can adapt to the external field environment and provided a reliable solution for distributed temperature monitoring in the health monitoring of large infrastructure structures.