Recognition of Acoustic Emission Signal Based on MAE and Propagation Theory

Because of containing several model waveforms, the source rub-impact acoustic emission (AE) signal transmitting in complicated body would cause waveform distortion for varying transmission speed of each model and dispersion effect. It is difficult to distinguish useful AE from complicated mixture singals by traditional detect methods. To avoid the deviation caused by these factors,. A transmission model of AE was established based on modal acoustic emission (MAE) and stochastic process theory, the mathematical expression of multi-modal AE signal was given, a mixed parameter composed of logarithm cepstral coefficients and fractal dimension was presented as the characteristic coefficients of rub-impact AE signal, and the AE recognition system based on improvement Gaussian mixture model was also established. The theory analysis and experiment result show that the system not only can be applied to recognize bursting AE signal, but also continuous signal.