Research On Signal Detection And Processing Of Metal Plate Cracks Based On Electromagnetically Induced Acoustic Emission

Electromagnetically induced acoustic emission (EMAE) is a new method for nondestructive testing based on acoustic emission. It did nondestructive detection with the effect of dynamic electromagnetic loading to generate a stress field stimulating stress waves from the defects. This thesis mainly focuses on the application of the EMAE in the nonmagnetic metal plate crack detection. It adopts the study methods of combining theory with practice. It deeply analyses the EMAE signal and builds a recognition system based on neural network technology to identify the EMAE signal.At first, based on the analysis of the generate principle of EMAE, EMAE experiment of defect detection is achieved. The pulse-current loading unit is designed and applied in the detection experiment of Aluminum crack. It analyses the influence of detection conditions, loading conditions and specimen type on the EMAE signal in detail.Secondly, a full research on the EMAE signal processing has been done in this thesis. It adopts Joint Time-Frequency Analysis methods (JTFA) to process the EMAE signal. For extracting the signal feature effectively, the time-frequency reassignment method is introduced. It deeply discusses the application of wavelet packet transform in the signal processing and feature extracting of EMAE. Meanwhile the program of preprocessing and feature extracting of the EMAE signal is proposed, and a criterion of the metal plate with the crack based on the wavelet packet transform and energy partition is given.Finally, the recognition issue of EMAE signal is studied. It builds BP neural network and wavelet neural network to identify the EMAE signals, and an adaptive optimization method of input feature vector of neural network is presented for achieving good performance of the recognition system.