Abstract
It is difficult to extract the acoustic emission characteristics of concrete signal accurately during the medium strain rate damage process due to the existence of noise, which can be solved by the wavelet threshold de-noising method. However, the determination of the optimal wavelet basis, the decomposition scale, the threshold strategy, threshold function and the evaluation index are complicated. Firstly, the above problems were analysed from the theoretical angle based on the acoustic emission characteristics of concrete signal. Secondly, the evaluation indexes were established based on an overall consideration of the above factors. Finally, we analysed S-R, and then the optimal de-noising method in the different damage stages were presented. In order to get the best denoising effect, the results show that the wavelet basis of coif5 and hard threshold function should be selected. And five decomposition scales and rigsure strategy should be selected before the 75% peak stress; four decomposition scales and rigsure or heusure strategy should be selected after the 75% peak stress. And at the same time, the S-R index is in its best position. It could clearly highlight the spectrum characteristics of acoustic emission signals.