管道泄漏声发射信号降噪中最优小波基的选取

Selection of optimal wavelet basis in noise reduction of acoustic emission signals in pipeline leakage

针对输油或输气管道泄露源精确定位问题,介绍了使用小波变换对管道泄漏声发射信号进行降噪过程中选择最优小波基函数的方法。首先在相对较长的时间间隔内监测管道泄漏的声发射信号;然后应用短时傅立叶变换来捕获时域中的泄漏信号特征,使用捕获的信号来寻找在时域和频域中最佳匹配的小波基函数;最后根据泄漏信号小波降噪后的信噪比和均方根误差作为准则对小波基函数的适用性进行定量分析。实验结果表明,筛选后的小波基函数对信号进行降噪分析处理时,能够很好地滤除其中掺杂的干扰噪声,并减少误差。

Aiming at the problem of accurate positioning of leaking sources in oil pipelines or gas pipelines,in this paper,the reasonable selection of optimal wavelet basis function was introduced in the process of signal denoising by using wavelet transform for acoustic emission(AE)signal in pipeline leakage.First,the AE signals of pipe leakage was monitored in a relatively long time interval.Then the leakage signal characteristics in time domain were extracted by using the Short-time Fourier Transform(STFT),and the captured signals were used to find the best matching wavelet basis function in the time and frequency domains.Finally,after the wavelet noise reduction of the leakage signal,the signal-to-noise ratio and the Root Mean Square Error were used as the criterion to analyze the applicability of wavelet basis function.The experimental results show that the filtrated wavelet basis function can effectively filter out the doping noise and reduce the error when the leakage signal was denoised by wavelet transform.