HAL-INSA Toulouse
Not a member yet
34325 research outputs found
Sort by
Steady State Analysis for Kalman Filters with Poisson-Sampled Observations
No full textWe study optimal filtering for continuous-time linear stochastic systems with Poisson-sampled observation processes. For each realization of the sampled observation process, the posterior distribution is a Gaussian process whose mean and covariance are described by continuous-discrete process. We are particularly interested in analyzing the expectation of the first and second moment of the estimation error with respect to the sampling process. Using the system-theoretic properties like observability and controllability, our results provide tractable conditions on the mean sampling rate for convergence of the expected error covariance, its boundedness and convergence of expected estimation error to zero. Some comparisons are also drawn with the solution of Riccati differential equation associated with the continuous-observation process
Non-Robustness of the Zero-Temperature-Limit Gibbs Measures to Perturbations of the Potential
No full text13 pages, 5 figuresThe robustness of properties of a statistical physics model to slight perturbations in the exact local interactions of the model is a very relevant philosophical question, considering real-life measurements on which we base some models can only ever reach a finite precision. In this article, we will discuss this topic in a formal mathematical setting, and notably exhibit a family of models for which the low-temperature behaviour is highly non-robust
A comprehensive review and benchmark of differential analysis tools for Hi-C data
No full textInternational audienceThe three-dimensional conformation of the genome has a key role in multiple biological processes such as gene expression regulation. Hi-C is a sequencing technique used to profile 3D chromosomal conformation. The data are summarized in a symmetric matrix, where each pixel (i,j) (or (j,i)) represents the interaction frequency between genomic positions i and j, estimating their spatial proximity. The objective of Hi-C data differential analysis is to identify genomic regions that display significant changes of interactions between two biological conditions. Several tools have been proposed to address this question and most propose to test each pixel of the matrix independently. Here, we present a review and a thorough statistical benchmark of these tools. We first focused on describing the tools on multiple aspects. First, we gave a technical description of the tools' implementations, specifically of their usability. Then, we focused on a statistical description, highlighting the different preprocessings, modelling choices (replicates and covariates, spatial dependency awareness, etc.) and multiple testing corrections proposed. In a second part, we performed an evaluation of the tools on simulations based on real data (mimicking a H0 and a H1 settings with no and controlled signal respectively), which led us to assess proper control of the Type-I error and power of the tools. Then, we evaluated tools to recover biological signal on a auxin/no auxin dataset validated through CTCF external data.Our experiments highlighted the strong impact of data preprocessings, the fact that tools did not control the FDR, and showed the superiority of diffHic
Detection of Dielectrically Heterogeneous 3D Multicellular Objects with Microwave Dielectric Spectroscopy
No full textInternational audienceMulticellular 3D biological models, also called spheroids or microtissues, are considered very promising for drug screening and cancer studies, as they integrate a structural heterogeneity and therefore exhibit a physiological complexity that closely mimics in vivo tissues. Such a model is usually characterized with optical techniques. These are very efficient but present however the implication of labels and high-cost equipment. To complete such methods, microwave dielectric spectroscopy presents attractive features, as it is non-invasive, label-free and rapid. To demonstrate its ability to evaluate 3D heterogeneous bio-models, a microwave biosensor suitable for the analysis of such bio-objects is evaluated. Three different types of elements with constitutional dielectric heterogeneity are used, spheroids with living HepG2 cells, polystyrene beads and the combination of these two models considered as hybrid spheroids. A clear discrimination of the different model categories is obtained, demonstrating the ability of the microwave sensing technique to detect a 3D bio-object with heterogeneous dielectric structuration
Charge State Tuning of Spin Defects in Hexagonal Boron Nitride
No full textInternational audienceBoron vacancies in hexagonal boron nitride (hBN) are among the most extensively studied optically active spin defects in van der Waals crystals, due to their promising potential to develop two-dimensional (2D) quantum sensors. In this letter, we demonstrate the tunability of the charge state of boron vacancies in ultrathin hBN layers, revealing a transition from the optically active singly negatively charged state to the optically inactive doubly negatively charged state when sandwiched between graphene electrodes. Notably, there is a photoluminescence quenching of a few percent upon the application of a bias voltage between the electrodes. Our findings emphasize the critical importance of considering the charge state of optically active defects in 2D materials, while also showing that the negatively charged boron vacancy remains robust against external perpendicular electric fields. This stability makes it a promising candidate for integration into various van der Waals heterostructures
Incommensurate Antiferromagnetism in UTe 2 under Pressure
No full textInternational audienceThe discovery of multiple superconducting phases in UTe 2 boosted research on correlated-electron physics. This heavy-fermion paramagnet was rapidly identified as a reference compound to study the interplay between magnetism and unconventional superconductivity with multiple degrees of freedom. The proximity to a ferromagnetic quantum phase transition was initially proposed as a driving force to triplet-pairing superconductivity. However, we find here that long-range incommensurate antiferromagnetic order is established under pressure. The propagation vector k m = ( 0.07 , 0.33 , 1 ) of the antiferromagnetic phase is close to a wave vector where antiferromagnetic fluctuations have previously been observed at ambient pressure. These elements support that UTe 2 is a nearly antiferromagnet at ambient pressure. Our work appeals for theories modeling the evolution of the magnetic interactions and electronic properties, driving a correlated paramagnetic regime at ambient pressure to a long-range antiferromagnetic order under pressure. A deeper understanding of itinerant- f -electron magnetism in UTe 2 will be a key for describing its unconventional superconducting phases
Growth driven phase transitions in Zinc Oxide nanoparticles through machine-learning assisted simulations
No full textInternational audienceThis study investigates the formation of zinc oxide (ZnO) nanoparticles, a material of significant technological interest with complex structural properties, through atom-by-atom deposition modeling a process common in bottom-up synthesis. Our findings demonstrate that, although the body-centered tetragonal (BCT) structure is thermodynamically stable at equilibrium for small particle sizes, the deposition process induces a crystal-to-crystal phase transition into the more stable wurtzite (WRZ) phase. This transformation is facilitated by a specific redistribution of the nanoparticle ions, which effectively compensates the emerging polar facets at the moment of transition. These insights offer a deeper understanding of oxide nanoparticle formation, which should ultimately help the design of materials with targeted structural features
SWOT-hydro2-learning: Regionalization of hydrological-hydraulic models and discharge laws over river networks with SWOT and multi-source data assimilation
No full textInternational audienc
Vers une meilleure compréhension des performances mécaniques et de la durabilité des bois pour la mobilité
No full textInternational audienceLes structures pour la mobilité telles que les véhicules autonomes terrestres, spatiaux ou encore aéronautiques (Castanié et al 2024) sont constamment soumises à des charges mécaniques et environnementales complexes. A titre d’exemple, la NASA et l'Agence aérospatiale japonaise (JAXA) ont lancé le premier satellite en bois (LignoSat) en orbite terrestre en été 2024 (Doi 2022). Ce satellite innovant est fabriqué en bois de magnolia, bois sélectionné après une première campagne expérimentale dans la station spatiale internationale, au cours de laquelle différentes essences ont été soumises aux conditions du vide spatial en orbite basse. Le choix du bois comme matériau se justifie par le fait qu’il se dégrade sans particules polluantes lorsqu'il rentre dans l'atmosphère. L'utilisation du bois permet à terme de contenir la pollution spatiale générée par les débris de matériaux non biodégradables autour de notre planète. Toutefois, ses propriétés intrinsèques – anisotropie, variabilité, sensibilité à l’humidité, aux UV ou aux attaques biologiques – ont historiquement limité son emploi à des composants secondaires. L’émergence de nouvelles technologies de modification du bois peut considérablement pallier les insuffisances du bois naturel. Des traitements thermiques, chimiques, ou encore la densification du bois, sont aujourd’hui des procédés utilisés afin d’améliorer les caractéristiques mécaniques, ainsi que la durabilité du bois (Hill 2006, Sandberg et al 2017). Parallèlement, l’essor de la fabrication additive de composites bois-polymères à base de PLA et de fibres lignocellulosiques ouvre de nouvelles voies pour l’impression 3D de structures légères, à géométries complexes. Par ailleurs, bien que la fabrication additive à base de bois représente une voie prometteuse pour produire des pièces légères et écologiques, potentiellement utilisables en mobilité, les travaux restent embryonnaires concernant la durabilité mécanique de ces matériaux imprimés (Zhaozhe et al 2021). Ce travail vise donc à étudier le comportement mécanique et à améliorer la durabilité du bois et des produits bois (bois massif, bois traité, matériaux d'impression 3D à base de bois) sous des charges THVM (Thermo-hydro-visco-mécaniques) couplées à la rupture bidimensionnelle (2D) et tridimensionnelle (3D) dans les structures pour la mobilité