一种基于ICA的机械缺陷超声信号提取方法

针对强噪声背景下缺陷超声回波信号检测的问题,利用了基于独立分量分析的方法进行缺陷信号的提取。该方法首先对观测信号进行JADE分解,得出多导独立分量,再根据赫斯特指数,分离缺陷信号和噪声信号。通过对仿真和实测缺陷超声信号的去噪实验研究,结果表明,与小波去噪方法相比,ICA去噪方法能够得到很好的信噪比,有利于强噪声背景下缺陷的去噪处理及微弱信号的提取。

A method of mechanical defect ultrasonic signal extraction based on ICA

A method based on Independent Component Analysis to de-noise the ultrasonic signals is presented to overcome the disadvantage of ultrasonic testing signals that they are always strongly affected by the background noises. Firstly, the observed signals are decomposed of several independent components by JADE, then, set the noise component to be zero according to Hurst exponent. Finally, the denoised signals are figured out using the decompose matrix. The study on the flaw ultrasonic signals denoising of the simulation and experiment shows that the de-noising method based on ICA can get signal higher to noise ratio compared with wavelet deoising method. And ICA is beneficial to de-noise processing and extraction of characteristic signals for strong background noise.