Subspace-based optimization method for reconstructing 3-D scatterers in anisotropic laminates

Ping-Ping Ding (1) , Marc Lambert (2) , Dominique Lesselier (1)

1 L2S - Laboratoire des signaux et systèmes
2 GeePs - Laboratoire Génie électrique et électronique de Paris

This paper investigates the subspace-based optimization method (SOM) for reconstructing defects in the anisotropic laminates. Reconstruction of defects in such media, like planar composite panels applied in aeronautic and automotive industry, is greatly challenging to execute, due to the complexity in the anisotropy of materials and multi-layered structure. The main advantage of SOM is to split the space of induced currents into mathematical deterministic and ambiguous subspaces, as opposed to physical radiating and non-radiating subspaces in the noise-free scenario and mathematically measurable and non-measurable in the noisy scenario. The deterministic subspace is determined from the spectrum analysis, whereas the ambiguous subspace is calculated by an optimization method. This feature makes SOM fast convergent, robust against noise and the selection of the regularization parameter L that is used to split the space of induced currents. This work extends the SOM to multi-layered anisotropic inverse scattering problems involving 3-D complex defects.

Domaines: Electromagnétisme

Ping-Ping Ding, Marc Lambert, Dominique Lesselier. Subspace-based optimization method for reconstructing 3-D scatterers in anisotropic laminates. 20th International Workshop on Electromagnetic Nondestructive Evaluation (ENDE’2015), Sep 2015, Sendai, Japan. OS4-3 (2 p.). ⟨hal-01207371⟩