基于声发射的输油管线破坏点定位方法研究

针对国内油田石油输送管线上频繁的钻孔盗油,且常规互相关检测法定位误差较大的情况,基于声发射信号的传播和改进小波包分解算法,提出了一种钻孔破坏点精确定位方法.根据管道两端声发射(AE)传感器的采样数据,采用改进小波包算法将信号分解到不同的频段,并结合快速傅立叶分析对高、低频的不同情况分别进行处理,再提取能量较丰富的小波包作为特征包进行重构,对特征包的重构信号进行相关分析来确定破坏信号到达两端传感器的延时,进而确定破坏点的位置.实验研究表明该方法可消除小波滤波器组造成的频谱混叠现象,提高了互相关分析定位的准确性,有助于实时检测埋地管线上的钻孔破坏信号.

Research on acoustic emission location method for oil pipeline breakage

Aimed at avoiding frequent drilling holes and stealing petroleum phenomena in pipeline of oil field and the direct cross-correlation method causing large errors, a novel method for accurate locating of pipeline leaks was proposed. Based on the acoustic emission (AE) sensor's signals of drilling hole in metallic pipe and a modified wavelet packet decomposition algorithm combined with fast Fourier transform, signals collected by AE sensors were decomposed into a series of time domain signals, each of which covered a specific frequency band. Wavelet packets containing abundant energy information were selected as feature packets to reconstruct the signals. Cross-correlation analysis was applied to detect the location of drilling hole through reconstruction signals. Industry experimental results show that the proposed method avoids the frequency aliasing caused by filter banks, and is effective to monitor the third party's damage activity in buried pipeline and improves the location accuracy.