Characteristic spectrum research in ae signals based on wavelet analysis

Abstract

Acoustic emission (AE) signal is one kind of non-steady random signal, its frequency and the statistical nature change with the time variation. In traditional spectrum analysis, a spectrum can't be used to determine what the corresponding period of time domain signal is. According to Mallat decomposition algorithm, wavelet decomposition of each scale and structure is the convolution of a low-pass filter and a high-pass filter, acoustic emission signal is decomposed into different frequency range of time-domain signal components. The lower-scale decomposition component gives expression to high-frequency of the local information, and the higher-scale decomposition component gives expression to low-frequency of the local information. A measured AE signal was decomposed by 6-wavelet. The results showed that AE wave spectrum feature analysis based on wavelet analysis can be used to analyze the spectrum characteristics of the emission signal, and effectively extract useful information of AE.