分布式光纤拉曼测温系统信噪比优化研究

Research on Performance Optimization of SNR for Distributed Optical Fiber Raman Temperature Measurement System

针对分布式光纤拉曼测温系统(ROTDR)信噪比较低的问题,在传统累加平均算法的基础上,提出利用小波包去噪以及埃尔米特插值算法对后向拉曼散射信号进行处理来提升系统信噪比。该方法首先在数据采集端对散射信号进行累加平均去噪,然后使用埃尔米特插值算法进行色散补偿,最后选取sym6小波基对散射信号进行小波包去噪处理。实验表明,该方法将系统信噪比由27.6159 dB提升至32.6988 dB,而且有效补偿了系统的色散效应。选取8 km光纤进行实验,光纤全程温度波动范围从0.65℃~1.15℃降低至0.36℃~0.75℃,不同温度的测量误差中位数小于0.5℃。综上,本文提出的信号处理方案可有效提升系统信噪比及降低系统测温误差。

Aiming at the problem of low signal-to-noise ratio(SNR)in distributed fiber Raman temperature measurement system,a combined method which utilized wavelet packet de-noising and Hermite interpolation algorithm,is processed to improve SNR.The method firstly performs cumulative average de-noising on the data acquisition card,and then the Hermite interpolation algorithm is used to perform dispersion elimination.Finally,the sym6 wavelet base is selected to perform wavelet packet de-noising on the scattered signal.Experiments show that this method increases the SNR from 27.6159dB to 32.6988dB,and effectively eliminates the dispersion effect.The temperature fluctuation,along the 8Km multimode optical fiber,is reduced from 0.65℃~1.15℃to 0.36℃~0.75℃,and the temperature response median deviation is less than 0.5℃.In summary,the method mentioned above can be used to improve the SNR and reduce the system temperature error.