Multiple damage detection using sparse blind source separation in composite structures

Damage detection is a primary concern in composite structures. Multiple damage mechanisms have been identified for the aircraft composite structures. When the different damage mechanisms occur at numerous locations, the damage signals from acoustic emission (AE) sensors are mixed. This paper proposes an improved sparse component analysis (SCA) method to separate the mixing AE signals. Firstly, a bandpass filter is designed to remove the vibration interference, and pre-whitening is used to achieve the decorrelation. Secondly, implement SCA to reconstruct the source signals. Finally, the wavelet denoising (WD) is executed to the recovery signals after SCA. The simulation results show that the improved SCA can separate the composite structure damage AE simulation signals.