针对轧件内部微小缺陷可能出现漏检的现象,提出一种基于对偶树复小波变换(DT-CWT)的超声信号检测方法;通过构造新的对偶树复小波,并利用对偶树复小波的平移不变性,改进现有的基于常规离散正交小波变换的超声回波信号处理方法。试验结果证实了该方法的优点和有效性。
Aiming to miss--detection of internal tiny flaw of workpiece for cross wedge rolling, a new ultrasonic signal testing method was proposed based on DT--CWT. By constructing new dual-- tree complex wavelet bases, and by taking advantage of near shift--invariance of DT--CWT, it can improve common wavelet analysis methods for ultrasonic signal testing based on real orthogonal discrete wavelet transform. Experiments indicate that the method can extract weak defect feature and its results demonstrate the validity of the proposed method.