Robust and Low-Complexity Time-Reversal Subspace Decomposition Methods for Acoustic Emission Imaging and Localization

This article focuses on the acoustic emission (AE) localization problem in the complex and unknown probed domain. The schemes based on time-reversal (TR) theorem are proposed to achieve accurate imaging and localization, even when there is no accurate priori knowledge of the Green's functions. In the proposed strategies, a signal enhancement method is proposed to de-noise sensor wavefields from single or multiple sources firstly. Then the wideband signal matrix is condensed into a single frequency matrix by coherent processing. The steering vectors are estimated based on minimizing noise variance. And by incorporating the steering vector estimation algorithm into the subspace decomposition-based methods, two novel robust imaging methods, robust coherent Decomposition of the time-reversal operator (R-DORT) and robust coherent time-reversal Multiple signal classification (R-TR-MUSIC), are presented. The performance of the proposed methods is evaluated using simulation studies that consider two background mismatches and different levels of signal-to-noise ratio (SNR) respectively. The simulation results demonstrate the effectiveness and the superiority of the presented methods in locating AE sources under the complex and uncertain circumstances. Moreover, the localization results based on experimental data show the potential of the proposed methods in practical applications.