基于小波与相关性分析的木材声发射源直线定位算法改进

Linear positioning algorithm improvement of wood acoustic emission source based on wavelet and signal correlation analysis methods

为研究木材损伤过程中声发射(acoustic emission,AE)信号源定位问题,提出了一种基于小波和信号相关性分析的木材表面AE源直线定位方法。首先,采用NI高速采集设备和LabVIEW构建双通道的AE信号采集系统,并设定采样率为500 kHz。然后,通过5层小波分解的方法对原始AE信号进行降噪处理后重构AE波形,分别采用信号相关性和最大值分析方法确定AE信号在2个固定传感器之间的传播时差,并由此计算AE信号的传播速度。最后,在距原点150和200 mm处设置2个人工AE源,针对不同方法所确定的传播速度,采用基于时差的直线定位算法进行AE源定位。试验结果显示,对于原始AE信号采用信号相关性和最大值分析后,两处的定位相对误差分别为40%,32%和28%,15%。对原始AE信号进行小波降噪重构后,相应的相对误差分别为2%,26%和1%,11%。因此,对原始AE信号进行小波处理后,再基于相关性分析方法计算AE传播速度,能够明显提高AE源定位算法精度。

As an active dynamic non-destructive testing method,the acoustic emission(AE)technology has been widely used in the defect and damage monitoring of metals,rocks,natural organic polymer materials and composite materials.During the monitoring process,the damage or deformation of the measured material will generate an AE signal,and the positioning of the deformation or damage will be determined by the propagation rules and characteristics of the AE signal.In order to study the acoustic emission signal source positioning problem in the wood damage process,a linear positioning method based on wavelet and signal correlation analysis for wood surface AE source was proposed.Firstly,a two-channel AE signal acquisition system was built using the NI high-speed acquisition equipment and LabVIEW,and the sampling rate was set at 500 kHz.Then,the original AE signal was denoised by the 5-layer wavelet decomposition method to reconstruct the AE waveform,and the signal correlation and maximum value analysis methods were used to determine the propagation time difference between the two fixed sensors and calculate the propagation speed of the AE signal.Finally,two artificial AE sources were set at 150 and 200 mm from the origin.For the propagation speed determined by different methods,the AE source positioning was performed by the linear positioning algorithm based on the time difference.The test results showed that the relative errors of the two locations were 40%,32%and 28%,15%,respectively,after the signal correlation and maximum analysis were used for the original AE signal.After the wavelet denoising reconstruction of the original AE signal,the corresponding relative errors were 2%,26%and 1%,11%,respectively.The test results showed that the signal cross-correlation analysis method and the maximum value analysis method were used to process the original AE signal,and the time difference and rate between the arrival of the AE signals at the two sensors were estimated.After the noise on the original signal was reduced using the wavelet analysis method,the signal cross-correlation analysis and peak analysis method were used to estimate the time difference between the signals arriving at the two sensors to calculate the propagation rate of the AE signal on the wood surface.Finally,the linear positioning of the AE source using the time difference positioning method could improve the positioning accuracy of the AE source,especially after using the signal cross-correlation analysis method,resulting in the significantly improved positioning accuracy.