基于相关分析的声发射储罐罐底检测降噪方法

Denoising method for acoustic emission testing of tank bottom based on cross-correlation analysis

基于声发射原理的大型储罐罐底腐蚀和泄漏检测方法近些年被业界普遍认同。作为一种高效率、低成本的先进方法取得了越来越多的应用。然而由于该方法本身的特点,在检测过程中不可避免地会采集到大量的噪声信号。如不能有效地消除这些噪声信号,就会严重影响对罐底评估的准确性。因此,人们提出了多种去噪方法,然而这些方法却不能很好地消除"虚定位"事件。针对该问题,提出了一种基于相关分析的声发射罐底检测降噪方法以消除"虚定位"事件。通过对现场实际检测数据的处理表明,该方法能够有效消除"虚定位"事件,使得罐底评估的结果与开罐实测的结果更加相符。

During the process of acoustic emission(AE) testing on tank bottom,due to the characteristics of the method,a lot of noise signals would be collected,which would seriously affect the evaluation accuracy of the tank bottom's corrosion degree.Therefore,various approaches have been developed to reduce the noise signals.However,it is hard for them to reduce the virtually located AE events.To solve this problem,a denoising method was proposed.The treatment result of the actual data by using the method shows that it could reduce the virtually located AE events effectively so as to make the evaluation result be more in line with the real situation.