A novel feature representation method based on original waveforms for acoustic emission signals

One of the most important issues arising in the use of acoustic emission (AE) for nondestructive process monitoring is the accurate identification of potential process malfunctions to avoid premature failure. In some cases, the AE signals from malfunction sources are relatively weak, with high levels of background noises. Thus, these signals could easily be submerged, making it rather difficult to separate them. Therefore, it is of critically...