Proposition de thèse : Génération procédurale bio-inspirée de métamatériaux et métasurfaces


Bio-inspired procedural generation of metamaterials and metasurfaces

L’ED SISMI propose le sujet de thèse suivant :

Intitulé du sujet : Génération procédurale bio-inspirée de métamatériaux et métasurfaces

Ce projet serait sous la direction de Cyril Decroze du laboratoire XLIM à l’Université de Limoges

Co-directeurs renseignés : Thomas Fromentèze /

Les financement sont : MESR acquis, Projet région déposé

Le début de la thèse est prévu pour : 01/10/2021

Mots clés du sujet : Electromagnétisme, Métamatériaux, rayonnement, guidage, métasurfaces, imagerie, furtivitié, Experimentations, Programmation

Présentation du sujet : The proposed research project concerns the exploitation and improvement of a bio-inspired procedural generation technique, allowing the synthesis advanced electromagnetic structures such as functional metasurfaces and metamaterials. This research topic constitutes a real scientific advance and paves the way for numerous perspectives of applications and publications.

Objectifs : The objective of this work is to develop this new research topic by rapidly designing demonstrators allowing the radiation and the control interaction of incident electromagnetic waves. To this end, it will be necessary to understand the mathematical principles of the bio-inspired technique developed within our research team, to implement it using optimised algorithms and to fabricate the structures generated to design experimental demonstrations. The proposed work will be carried out in particular in K and W band.

Description du sujet : This thesis topic concerns a new research theme within our team focused on the development of bio-inspired procedural generation techniques for electromagnetic components. The design and optimization of passive components such as antennas, filters, guides and couplers is a crucial research and engineering activity for many practical applications. However, it is particularly difficult to program tools that can perform these designs in an automated way. Shape optimization techniques are available in the scientific literature and take advantage of the continuous increase in computing power, but they can hardly replace the talents of experienced designers at the present time. They are generally based on the guided exploration of very large solution spaces, analyzing the consequences of perturbations brought to the components until the expected criteria are satisfied.

New techniques are developed by our research team, allowing the automatic generation of electromagnetic structures capable of meeting pre-defined constraints. It is then no longer necessary to propose an initial structure that is then deformed until the expected properties are obtained.

The application fields anticipated for this work are particularly broad, ranging from the generation of antennas for various telecommunication and computational imaging applications to the development of guides, filters, couplers and materials for stealth. These techniques also seem to be particularly well adapted to more upstream research problems, linked in particular to the synthesis of electromagnetic materials with extraordinary properties.

It will be necessary for the good execution of this research project to master new mathematical models and to proceed to the experimental validation of the proposed principles using the means at our disposal in K and W band. The originality of this research theme will ensure to the highly motivated PhD student the publication of several articles in high level international journals, as well as the participation in international conferences.

Compétences acquises à l’issue de la thèse : The spectrum of skills required for the successful execution of this research project is very broad. This thesis will firstly involve the analysis and development of mathematical generative models. It will then be necessary to implement them numerically using scientific programming. The generated structures will then be evaluated in electromagnetic simulations using commercial or in-house developed softwares. The fabrication of these structures may involve the use of additive manufacturing technologies. Finally, the development of experimental beds for the validation of the proposed principles will require the mastering of RF tools for the generation and measurement of signals in time and frequency domains.

Présentation de l’équipe d’accueil : This work will be carried out within the Antennas and Signals team of Xlim’s RF System Axis. We are committed to start from new theoretical concepts and to go each time to the development of functional experimental validations, allowing us to train multi-skilled PhD and post-doctoral students. In recent years, our efforts have been focused on the development of new radar imaging techniques, taking advantage of our strong experience in the development of radiating systems and the associated numerical techniques.

Compétences souhaitées pour les candidats : The proper execution of this work will require high proficiency in areas of problem solving and numerical modelling. Our work is usually initiated by writing mathematical models, then studied through programming (Matlab preferred, Python accepted). The validation of these ideas then generally involves the use of commercial/in-house electromagnetic simulation software. Demonstrations finally require the development of experimental benches. Applicants will be required to demonstrate a certain level of proficiency on some parts of this chain and be motivated by a strong increase in skills in all these areas.

Pour plus d’informations et pour candidater, merci de contacter :

Date de dépôt : 03/31/2021 à  8 h 47 min




ED SISMI