Potential and limitations of NARX for defect detection in guided wave signals

Abstract

Previously, a nonlinear autoregressive network with exogenous input (NARX) demonstrated an excellent performance, far outperforming an established method in optimal baseline subtraction, for defect detection in guided wave signals. The principle is to train a NARX network on defect-free guided wave signals to obtain a filter that predicts the next point from the previous points in the signal. The trained network is then applied to new measurement and the output subtracted from the measurement to reveal the presence of defect responses. However, as shown in this paper, the performance of the previous NARX implementation lacks robustness; it is highly dependent on the initialisation of the network and detection performance sometimes improves and then worsens over the course of training. It is shown that this is due to the previous NARX implementation only making predictions one point ahead. Subsequently, it is shown that multi-step prediction using a newly proposed NARX structure creates a more robust training procedure, by enhancing the correlation between the training loss metric and the defect detection performance. The physical significance of the network structure is explored, allowing a simple hyperparameter tuning strategy to be used for determining the optimal structure. The overall detection performance of NARX is also improved by multi-step prediction, and this is demonstrated on defect responses at different times as well as on data from different sensor pairs, revealing the generalisability of this method.

A comparison of ultrasonic temperature monitoring using machine learning and physics-based methods for high-cycle thermal fatigue monitoring

“…Machine learning is a technique that is being increasingly implemented in non-destructive evaluation scenarios, including: Ultrasonic flaw classification 25 Deconvolution of ultrasonic signals 26 Artefact identification and suppression in ultrasonic images 27 Defect detection in guided wave signals 28 Noise quantification in ultrasonic images 29 Ultrasonic crack characterisation 30 Machine learning has several beneficial characteristics including:…”

Active sensing ultrasonic guided wave-based damage diagnosis via stochastic stationary time-series models

“…Recently, nonlinear autoregressive with exogenous excitation (NARX) models were introduced in the context of ultrasonic guided wave-based damage diagnosis. 47,48 It was shown that instead of a single-step ahead prediction, a multi-step ahead prediction scheme with a suitable training procedure may improve the damage detection performance. Moreover, Da Silva et al 49 have used a Gaussian process NARX model for performing damage detection in composite structures.…”

“…The majority of the literature on guided wave-based SHM employs user-defined, and thus arbitrary thresholds, for detecting, localizing, and quantifying damage. In addition, existing probabilistic methods are based on complex nonlinear model representations that exhibit unnecessary complexity, 47,48 and Bayesian approaches that solve computationally costly inverse problems and/or require the use of multi-physics-based numerical models and various sampling approaches to determine the statistical distributions of the quantities of interest. (iii) The method currently postulated is relatively simple, computationally efficient, and can be easily automated without the need for user intervention.…”