基于复小波变换的超声信号分析技术研究

超声信号中标志实际测量目标的回波信号常常因受到由材料内部微缺陷(如裂纹、空穴、夹杂等)引起的众多回波干扰信号的干扰而难以识别。另外，由于超声波能量传递的各向异性特征，在同一位置不同方向测得的超声信号中，有效回波信号出现的位置常常会在一定范围内出现漂移，从而极大地影响超声检测结果的准确性。为此，本文利用复小波分析方法对实测超声信号进行了分析，并提出了小波包络检测及互相关分析技术。试验表明，该技术成功解决了上述难题，并充分提高了超声检测的准确性。

Application of Complex Wavelet Transform in Ultrasonic Detection Practice

The presence of numerous disturbing signals, which may be created when ultrasonic energy meets micro defects (e.g. micro cracks, voids or inclusions contained in material) during its propagation, makes it hard to distinguish the "true" feedback signal that indicates the object being inspected. Additionally, due to the anisotropic feature of the propagation of ultrasonic energy, the position of the 'true' feedback signal often fluctuates more or less. In order to tackle this difficult problem, the complex wavelet transform was adopted and two feasible methods--the complex wavelet based envelope detection method and the complex wavelet based correlation analysis method--were developed in this paper for analysing the ultrasonic signals collected by using the shear wave ultrasonic transducer. Experiments prove that, by using the proposed methods, the "true" feedback signal that indicates the object being inspected may be distinguished successfully and thereby its position is determined correctly. This is undoubtedly helpful for further improving the accuracy of ultrasonic detection significantly.