Improve The Temperature Accuracy Of ROTDR System Based On WNNM Denoising Algorithm And Segmented Compressed Sensing OMP Algorithm

Raman Optical Time Domain Reflectometry(ROTDR)technology is a distributed optical fiber sensing technology based on Raman scattering,which can be used to detect temperature information along the optical fiber.The technology has the advantages of simple structure,corrosion resistance,anti-electromagnetic interference,long-distance real-time monitoring,etc.,so it is widely used in power industry,nuclear industry,chemical industry,rail transit,submarine detection and other fields.However,in some special application scenarios(such as aerospace),if the high accuracy of temperature cannot be achieved,it will cause unpredictable losses.Therefore,achieving high temperature accuracy is an urgent problem to be solved by the ROTDR system.Aiming at the above problems,this paper proposes two algorithms to improve the temperature accuracy of the system without increasing the hardware cost of the system.The main research work is as follows:(1)Build a ROTDR experimental platform based on the temperature measurement principle of the ROTDR system,and complete the prototype production.Based on Lab VIEW,the ROTDR system host computer software system was developed.(2)A denoising algorithm based on Weighted Nuclear Norm Minimization(WNNM)is proposed to improve the temperature accuracy of the system.WNNM algorithm not only achieves good denoising effect,but also effectively solves the problems existing in the denoising process of wavelet denoising algorithm and non-local mean algorithm.The experimental results show that under the condition of 500 Ms/s sampling rate and 13 d B SNR improvement,the temperature accuracy processed by the WNNM algorithm can reach 0.65 ℃.(3)A segmented compressed sensing OMP algorithm is proposed to improve the temperature accuracy of the ROTDR system.After a large number of experiments,it is found that the sampling rate is also an important factor affecting the temperature accuracy of the system.The lower the sampling rate,the worse the temperature accuracy of the system.Compressed sensing,as a new sampling theory,exploits the latent sparse nature of the signal to perfectly reconstruct the signal under the condition of much smaller than the Nyquist sampling rate.Segmented compressed sensing can solve the problems of long running time and poor reconstruction effect of overall compressed sensing.In addition,segmented compressed sensing can be combined with the WNNM algorithm to break through the limitation of sampling rate on system temperature accuracy.The experimental results show that the temperature accuracy of 0.98 ℃ can be achieved at the sampling rate of 150 Ms/s by using the two algorithms simultaneously.