Acoustic Emission Signal Processing Based On Wavelet Analysis And Its Application In Damage Diagnosis

Damage location and damage assessment which have been in the internal structure of components or materials are two important aspects of non-destructive testing. The traditional signal processing methods (Fourier transform) can only be used to handle the transient nature of stationary signal, but lack of local transient non-stationary signal analysis function; However, wavelet transform, which due to a good time-frequency localization analysis of performance, is currently the most effective tool to deal with the acoustic emission signals. When we use wavelet transform for signal processing, usually, a beam structure is the research object: wavelet transform modulus maxima are the singular points, that is, damage locations; while for the assessment of degree of injury, often establish the relationship between modulus maxima and Lipschitzα. However, it has large amount of calculation, and the computer slow to deal, which has a long testing process and not easy in practical application. In this paper, based on the predecessors, it has studied the Daubechies wavelet system and proposed a 'node-peak' positioning method and testing procedure. The acoustic emission signals collected and recorded from SWAES full-waveform acoustic emission detector are used as the detect objects, as well as combine parameters analysis and waveform analysis. Wavelet transform signal decomposition, wavelet compression treatment are both able to more accurately identify the damage location. And on this basis, this paper analyzes and discusses the extent of beam damage.In Chapter 1, we briefly review the development of acoustic emission at home and abroad, and summarize the study of this topic status and the main contents of this paper. In Chapter 2, the basic theory of acoustic emission and acoustic emission with the fundamental principles are discussed.In Chapter 3, it focuses on continuous wavelet and discrete wavelet nature, as well as time-frequency analysis and multi-resolution analysis.In Chapter 4, the relationship between singularity and wavelet transform is studied, that is, the relationship between Lipschitzαand the maximum modulus; In addition, the principle of selection of wavelet bases is discussed.In Chapter 5, we have studied the signal processing based on wavelet transform, including the threshold de-noising, damage location and extent of damage location.In Chapter 6, we sum up this work and look into the future of the wavelet transform in signal detection.