Research on Ultrasonic Lamb Wave defect signal recognition method based on VMD

Abstract

In order to extract the defect from ultrasonic lamb wave signal accurately, the method based on variational mode decomposition (VMD) is proposed. The effect of decomposition produces different results due to the change of the decomposition layer - K. If the value of K is too large or too small, the intrinsic mode function (IMF) cannot be completely separated, resulting in mode aliasing phenomenon. The false components produced in the decomposition process will cause the summation of energy of the decomposed components to be larger than the original signal. In order to address this problem, the relationship between the mean of instantaneous frequency and the K value is used to determine the specific decomposition layer number of the VMD. The mutual information value of each IMF component and the original signal is calculated, and the IMF whose mutual information is lower than the threshold value will be eliminated to remove the false component. This method can decompose the fault signal well in the practical non-destructive testing and provide reference for the application of VMD in the faulted diagnosis of non-destructive testing.