Electromagnetic micro-structure non-destructive testing: sparsity-constrained and combined convolutional-recurrent neural networks methods - CentraleSupélec Accéder directement au contenu
Article Dans Une Revue Electronics Année : 2020

Electromagnetic micro-structure non-destructive testing: sparsity-constrained and combined convolutional-recurrent neural networks methods

Résumé

How to locate missing rods within a micro-structure composed of a grid-like, finite set of infinitely long circular cylindrical dielectric rods under sub-wavelength condition is investigated. Sub-wavelength distances between adjacent rods and sub-wavelength rod diameters require super-resolution, beyond the Rayleigh criterion. Two different methods are proposed to achieve it: One builds upon the multiple scattering expansion (MSM) and it enforces strong sparsity-prior information. The other is a data-driven method that combines convolutional neural networks (CNN) and recurrent neural networks (RNN), and it can be applied in effect with little knowledge of wavefield interactions involved, in much contrast with the previous one. Comprehensive numerical simulations are proposed in terms of missing rod number, shape, frequency of observation and configuration of tested structures. Both methods are shown to be achieving suitable detection, yet under more or less stringent conditions as discussed.

Dates et versions

hal-02905435 , version 1 (23-07-2020)

Identifiants

Citer

Peipei Ran, Dominique Lesselier, Mohammed Serhir. Electromagnetic micro-structure non-destructive testing: sparsity-constrained and combined convolutional-recurrent neural networks methods. Electronics, 2020, 9 (11), pp.1750. ⟨10.3390/electronics9111750⟩. ⟨hal-02905435⟩
96 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More