Proceedings of the

The 33rd European Safety and Reliability Conference (ESREL 2023)
3 – 8 September 2023, Southampton, UK

Complex-Valued-Autoencoder for Structural Health Monitoring with Frequency Modulated Continuous Wave Radar

Gaëtan Frusque1,a, Daniel Mitchell2,c, Jamie Blanche2,d, David Flynn2,e and Olga Fink1,b

1Intelligent Maintenance and Operations Systems Lab, EPFL, Lausanne.

2James Watt School of Engineering, University of Glasgow, Glasgow.

ABSTRACT

Frequency Modulated Continuous Wave (FMCW) radar is a low power, compact mechanism which can be used for non-destructive health monitoring and inspection of surface and subsurface materials. It enables the detection of defects that are internal to the analyzed structural element and not visible from the surface. The key benefits of this technology are that it offers a non-contact monitoring tool at reduced costs, risk and duration of inspection. Although a recent study proposed the use of FMCW radar sensing for composite material characterization, it has not yet been applied to the context of health monitoring. In this work, we propose to study the feasibility of FMCW radar to detect anomalies in monolithic surfaces. We propose to consider the analytic representation of the difference between the emitted signal and the incident wave to limit the interferences between the echo delay and gain difference between both signals. From a methodological perspective, we propose a complex-valued-AutoEncoder (AE) with a new activation function. We compare the proposed methodology to other state of the art methods.

Keywords: FMCW radar, Non-destructive evaluation, Complex-valued autoencoder, Anomaly detection, Analytical representation.



Download PDF