一种减小光纤传感系统频谱噪声的信号处理方法研究

Signal Processing Method Study of Reducing Spectral Noise in Optical Fiber Sensing System

针对基于短时傅里叶变换的布里渊光时域反射仪(STFT-BOTDR)系统噪声频谱泄漏导致的布里渊频移误差问题,提出利用布莱克曼窗函数解决频谱噪声的信号处理方法。首先搭建STFT-BOTDR系统,采集到包含系统变频模块产生低频噪声的布里渊信号,然后分别利用海明窗函数和布莱克曼窗函数解调的布里渊增益谱和布里渊频移分布。实验结果显示:利用布莱克曼窗函数比利用海明窗函数更能有效解决STFT-BOTDR系统中由于低频噪声发生频谱泄漏产生的谐波使布里渊增益谱失真的现象,减少布里渊频移的测量误差。实验表明利用窗函数解决系统频谱噪声的方法,在1.8 km的待测光纤上利用布莱克曼窗函数解调的布里渊频移分布测量精度达到1.4691 MHz,系统性能得到了提升,有利于加速低成本、中精度STFT-BOTDR系统的工程应用进程。

A signal processing method using Blackman window function was proposed to solve the issue of Brillouin frequency shift error caused by noise spectrum leakage in the STFT-BOTDR system based on short-time Fourier transform.Firstly,the STFT-BOTDR system was built,and the Brillouin signal including the low-frequency noise generated by the system frequency conversion module was collected.Then the Brillouin gain spectrum and Brillouin frequency shift distribution demodulated by Hamming window function and Blackman window function were used respectively.The experimental results show that using Blackman window function is more effective than using Hamming window function to solve the phenomenon of Brillouin gain spectrum distortion caused by harmonics caused by spectrum leakage of low-frequency noise in STFT-BOTDR system,and the measurement error of Brillouin frequency shift was reduced.The experimental results show that the Brillouin frequency shift distribution measurement accuracy using Blackman window function demodulation on the 1.8 km optical fiber to be measured can reach 1.4691 MHz.The overall performance of the system was improved,which is conducive to the engineering application of low-cost and medium precision STFT-BOTDR system.