Portail HAL Ensta
Not a member yet
11080 research outputs found
Sort by
New Formulation and Experimental Validation of the Kinematic Guidance Principle on an Underactuated Uncrewed Surface Vessel
No full textInternational audienceThis work presents a new formulation and experimental validation of the recently introduced Kinematic Guidance Principle on a USV. While the original formulation of the controller was based on a matrix calculated with an algorithm, this new interpretation relies only on conventional control engineering techniques. It is shown in this work that the kinematic guidance principle provides a linear decoupled position error system which is its main advantage compared to other guidance principles. Additionally, this work presents the first experimental results of the controller applied to a US
Les spécialisations industrielles des PECO membres de l’Union européenne sur la période 1992-2018 : Une lecture structurale
No full textInternational audienceThe history of the transformation of the production systems of the former socialist countries that joined the European Union in the 2000s is now well known. After a transition period characterized by major macro- and meso-economic upheavals, which for some of them extended beyond the dates of their accession to the Union, these countries have stabilized these systems, which are now sometimes characterized by a strong high-tech industrial specialization. The aim of this article is to propose an empirical architecture, based on the economic dominance theory, to represent this transformation at the level of each of the eleven countries concerned. A typology of the observed historical trajectories is then established.L’histoire de la transformation des systèmes productifs des pays anciennement socialistes ayant rejoint l’Union européenne durant les années 2000 est aujourd’hui établie. Après une période de transition marquée par de fortes perturbations macro- et méso-économiques débordant pour certains d’entre eux leurs dates d’entrée dans l’Union, ces pays ont stabilisé ces systèmes quelquefois caractérisés désormais par des spécialisations industrielles high tech prononcées. L’objectif de l’article consiste à proposer une architecture empirique fondée sur la théorie de la dominance économique pour représenter cette transformation à l’échelle de chacun des onze pays concernés. Une typologie des trajectoires constatées est ensuite établie
Machine Learning for Scientific Computing and Numerical Analysis
No full textMasterThese notes are for third-year students at École Polytechnique (Palaiseau, France), who have completed two years of classes préparatoires, making the material equivalent to a graduate-level course in the UK or US. The development of a Scientific Machine Learning (SciML) course for third-year students bridges the gap between scientific computing and machine learning. SciML, an emerging field, enables efficient methods for solving complex problems in science and engineering, like partial differential equations (PDEs). This course offers a new, valuable specialization, helping students stand out in the evolving job market, where demand for ML experts with specialized skills is growing
Ionization-coherence-enhanced-vibration photoexcitation in nitrogen ions
No full textInternational audienceWe develop a vibronic coherence model for multiorbital ionization and excitation of nitrogen molecules, based on an up-to-date concept of coherent ionization into excited states [C. H. Yuen and C. D. Lin, Phys. Rev. A 108, 023123 (2023); 109, 033108 (2024)]. Compared to other models, where the coherence of excitation dynamics is described with fewer details, our model predicts larger populations at excited states, in particular, at levels corresponding to the transition wavelengths 391 or 428 nm relevant to air lasing in the atmospheric environment. The dependence of population inversion on the pump wavelength is investigated, showing that ionic coherence in ionization injection enhances the vibronic three-photon transition in the N2 postionization process
Steering Laser-Produced THz Radiation in Air with Superluminal Ionization Fronts
No full textInternational audienceWe demonstrate that pulsed THz radiation produced in air by a focused ultrashort laser pulse can be steered to large angles or even in the backward direction with respect to the laser propagation axis. The emission angle is adjusted by the flying focus technique, which determines the speed and direction of the ionization front created by the single-color laser pulse. This easily adjustable THz source, being well separated from the intense laser, opens exciting applications for remote THz spectroscopy
Stimulating Engineering Students' Potential for Sustainable Development
No full textInternational audienceEngineers can play a critical role in achieving sustainable development (SD). Despite this, there is a lack of specialized courses delivered in the undergraduate engineering curricula in Lebanon. The purpose of this study is to assess the outcomes of a newly developed core course on sustainability for engineering students integrated into the undergraduate engineering curriculum at one university in Lebanon. The study makes an original contribution as no similar course, in its current structure, has been found in the existing literature. A quantitative methodology was employed through a survey administered to the students both before and after course delivery. The survey aimed to assess the effectiveness of the course in changing students' awareness, perceptions, practices, and views on the engineering profession and how it relates to sustainability. The findings indicate that the course has significantly enhanced the understanding and awareness of the engineering students regarding sustainable development. This enhancement impacted positively on their perceptions, practices, and views regarding the importance of sustainable development in engineering education. This study presents an interdisciplinary course integrated into the engineering curriculum and augmented the students' awareness and knowledge on how to incorporate sustainable development into their design processes
Improved performances and motivation in intelligent tutoring systems: combining machine learning and learner choice
No full textAlthough the benefits of personalized learning are now well documented, its generalization in schools is challenged by high class sizes. Educational technologies such as intelligent tutoring systems (ITS) could contribute to address this challenge and help teachers and students. Recently, an approach leveraging models of curiosity-driven learning was leveraged to build an ITS personalizing sequences of exercices. Based on the Learning Progress hypothesis (LPH), this approach consisted in proposing students exercices that maximize learning progress, and incrementally identified using multi-armed bandit machine learning techniques. The resulting algorithm (ZPDES) was previously shown in field studies to be more efficient in terms of learning performances for a diversity of student profiles when compared to a curriculum designed by a human expert. However, there were two limits. First, the motivational impact was not assessed. Second, the ZPDES algorithm did not leave students the ability to express choices. This limitation in agency was at odds with the LPH theory initially concerned with modeling curiosity-driven learning. Here, we introduce a system (ZCO) that combines adaptive exercise recommentations using LP, with the possibility for students to make choices. These choice possibilities concern dimensions that are orthogonal to exercice difficulty, and are an instance of gamification also used in many existing educational technologies. We present an extensive field study (265 7-8 years old children from 11 schools, RCT design) comparing an LP-based automatic curriculum generation system with a hand-designed curriculum, both with and without self-choice. We first show that LP-based personalization improves learning performance (reproducing and solidifying previous results) while producing a positive and motivating learning experience. We then show that the addition of self-choice as a playful feature triggers intrinsic motivation in the learner and reinforces the learning effectiveness of the LP-based personalizing. In doing so, it strengthens the links between intrinsic motivation and performance progress during the serious game. Conversely, deleterious effects of the playful feature are observed for hand-designed linear paths. Thus, the intrinsic motivation elicited by a playful feature is beneficial only if the curriculum personalization is effective for the learner. Such a result deserves great attention due to the increased use of playful features in non adaptive educational technologies available in the market.</div
Target-Soil Composite Scattering Investigation with Simulated SAR Images
No full textInternational audienceFor scenes of interest with a target located on a rough dielectric soil, the radar backscattering can be difficult to process due to the coupling field between the target and the rough surface. This paper aims to use simulated Synthetic Aperture Radar (SAR) images to investigate the EM signature of the composite scattering between the target and the soil. Simulation results are presented for different properties of the soil and a target considered as a perfect electric conductor (PEC). These results are analysed regarding the sensor parameters and the target configuration
The effect of microstructural inertia on plastic localization and void growth in porous solids
No full textInternational audienceThis paper investigates the impact of microinertia on plastic localization, void growth, and coalescence in ductile porous materials subjected to high strain rates. For that purpose, we have performed finite element calculations on a flat double-notched specimen subjected to dynamic plane strain tension. The simulations employ three distinct approaches to model the mechanical behavior of the porous aggregate: (1) discrete voids within a matrix material governed by von Mises plasticity; (2) homogenized porosity represented using standard quasi-static Gurson-Tvergaard plasticity; and (3) homogenized porosity described with Gurson-Tvergaard plasticity extended by Molinari and Mercier (2001) to account for microinertia effects. The porous microstructures used in the simulations are representative of additive manufactured metals, featuring initial void volume fractions varying between 0.5% and 4%, and pore diameters ranging from 30 μm to 150 μm (Marvi-Mashhadi et al., 2021, Nieto-Fuentes et al., 2023). The applied tensile velocities ranged from 100 m∕s to 1000 m∕s, producing strain rates between 10 5 s -1 and 10 6 s -1 , and stress triaxiality values spanning from 4 to 30. The simulations with discrete voids validate the calculations performed using homogenized porosity and microinertia effects, demonstrating that higher strain rates and larger pore sizes lead to slower void growth and a delayed, regularized plastic localization. Conversely, the standard Gurson-Tvergaard model shows notable mesh sensitivity and fails to describe the influence of the loading rate on plastic localization. Ultimately, the comparison between finite element models with discrete voids and those with homogenized porosity illustrates the stabilizing effects of porous microstructure and multiscale inertia on dynamic plastic flow, while also highlighting the strengths of the constitutive model introduced by Molinari and Mercier (2001) for simulating engineering problems involving porous ductile materials subjected to high-velocity impacts.</div
Algorithme bayésien hors-grille pour la localisation de sources sous-marines
No full textAcoustique sous-marine et navale, bioacoustique en milieu marin; GAPSUS - Acoustique Physique, Sous-Marine et Ultra-Sonore: GVB - Vibro acoustique et Contrôle du Bruit: GBIO - BioacoustiqueNational audienceEn acoustique sous-marine, les méthodes de localisation de sources peuvent être classées en deux catégories : les méthodes par scrutation (généralisant la formation de voie) et les méthodes globales. Les premières estiment indépendamment la contribution de chaque source, tandis que les secondes considèrent le problème d’estimation de la contribution des sources dans leur complexité combinatoire, i.e., selon un modèle génératif matriciel : y=Dz+n, où y représente les pressions acoustiques (en fréquentiel) reçues sur une antenne de M capteurs, D est une matrice modélisant l’incidence d’ondes provenant de N lieux différents, z contient les amplitudes complexes des sources et n représente un bruit additif complexe. Les méthodes globales résolvent le problème inverse d’estimation de z connaissant y et peuvent être d’intérêt notamment dans la prise en compte d’a priori supplémentaires sur la nature des sources (ex. parcimonie) par le moyen de termes de régularisation. Mais elles sont sensibles au bon conditionnement de la matrice D, particulièrement difficile à basses fréquences si l’on souhaite atteindre une bonne précision de localisation. Dans ce travail, nous proposons d’adapter un algorithme bayésien de la littérature, initialement proposé dans le cadre des décompositions parcimonieuses, à ce cas particulier. Pour s’affranchir des limitations liées à la discrétisation de la matrice, une recherche "hors-grille" des lieux des sources est proposé. En pratique, l’approche revient à modéliser les lieux des sources comme une variable continue paramétrant la matrice et à les estimer dans une étape d’optimisation supplémentaire introduite dans l’algorithme initial. Notre approche est validée sur données numériques, simulant un contexte de champ proche. Outre sa capacité à bien identifier des sources en dehors d’une grille prédéfinie, l’algorithme améliore le rapport signal à bruit en sortie, par rapport à une méthode classique de type formation de voie, le rendant particulièrement d’intérêt dans des environnements fortement bruités