Φ-OTDR光纤传感电缆防外破监测数据预处理方法

Preprocessing method of monitoring data for cable breaking prevention based onΦ-OTDR optical fiber sensor

针对Φ-OTDR光纤传感电缆外破隐患监测原始数据信噪比低的问题,文章介绍了一种空间域去趋势化与时间域改进谱减法去噪的数据预处理方法。提取光功率距离衰减趋势基线,去除原始数据空间域的衰减趋势,消除Φ-OTDR系统光功率因距离衰减对信号幅度造成的影响,然后采用改进谱减法对时域进行去噪。所提方法应用于现场实测的信号进行了检验,分析结果表明:去趋势化和改进谱减法可有效提高外破监测数据的信噪比,且两者具有叠加效果;现场破碎机、挖掘机、打桩机信号预处理后信噪比分别提高了2.25 dB、6.11 dB、4.72 dB。该方法可为外破监测后续的定位与识别、减少误报率提供参考。

In order to solve the problem of low signal-noise ratio(SNR)of raw data for optical fiber sensing cable external damage monitoring,in this paper,a hybrid signal-processing method is proposed based on an improved spectral subtraction denoising algorithm combined with detrending techniques from time-domain data by denoising in the time domain based on the improved spectral subtraction method.The method implement the extracted baseline of optical power attenuation to remove the attenuation trend in the spatial domain of the original data and to eliminate the influence of optical power on signal amplitude,which caused by distance attenuation.The effectiveness of proposed method was tested by the on-site construction signal.The results verified that both detrending and improved spectral subtraction can improve the signal-noise ratio(SNR),and both have a superimposing effect.And the signal-noise ratio(SNR)after signal-processing of the on-site crusher,excavator and pile driver was increased by 2.25 dB,6.11 dB and 4.72 dB respectively.This method has a certain reference value for follow-up research on the location and identification of monitoring potential external damage of optical fiber sensing cable.