Adaptive multisensor data fusion for acoustic emission source localization in noisy environment

Abstract

This article proposes an adaptive multisensor fusion algorithm for acoustic emission source location in isotropic plate-like structures in noisy environments. Overall, the approach consists of the following four main stages: (a) feature extraction, (b) sensor selection based on a binary hypothesis testing, (c) sensor weighting based on a well-defined reliability function, and (d) estimation of the acoustic emission source location based on the extended Kalman filter. The performance of the proposed algorithm is validated through pencil lead breaks performed on an aluminum plate instrumented with a sparse array of piezoelectric sensors. Two experimental setups have been used to simulate a highly noisy environment. In the first setup, the experimental signals have been artificially corrupted with different levels of noise generated by a Monte Carlo simulation. In the second setup, two piezoelectric transducers have been used to continuously generate high-power white noise during testing. The results show the capability of the proposed algorithm for inflight structural health monitoring of metallic aircraft panels in highly noisy operational situation.