Beschreibung:
<jats:p>This paper presents an algorithm for isolating a useful acoustic signal (corresponding to damage accumulation) against the background of a signal used to model the performance of an industrial rotary equipment. Acoustic emission signals induced by deformation and fracture were studied using a uniaxial tensile test on woven laminate samples cut along the fiber and weft directions. The background signal is a random composition of acoustic pulses used to model the performance of an industrial rotary equipment. A comparison of useful and noise signals enables us to develop two algorithms based on frequency filtering of a signal and its decomposition into empirical modes. These algorithms can be used to isolate useful AE pulses against the background of all signal intensities under consideration.</jats:p>