HAL-INSA Toulouse
Not a member yet
34325 research outputs found
Sort by
Towards Real-Time Blood Pressure Measurement Using BallistoCardioGraphy and Machine Learning
No full textInternational audienceBlood pressure (BP) is one of the most challenging vital parameters to measure continuously without disrupting the patient. Developing a method for non-intrusive, continuous BP monitoring would be a significant medical advancement. Ballistocardiography (BCG) offers a non-invasive way to monitor cardiac signals by detecting small body deformations. Machine Learning (ML) can then use these signals to estimate blood pressure through regression. This work demonstrates a correlation between BP and BCG patterns, though it also identifies several key factors that must be addressed for effective relative BP measurement using BCG. These factors are thoroughly examined and discussed
Quantifying the sp3/sp2 ratio in functionalized graphene
No full textInternational audienceA major challenge in graphene nanoscience is quantifying the concentration of molecules or atoms grafted onto the graphene surface. Raman spectroscopy, a non-contact optical technique, is well suited to achieve this purpose. However, calibration becomes challenging at low coverage due to the weak signal, and at high coverage due to local fluctuations and large regions that yield no Raman response with visible light excitation. In this work, we propose a tailored continuous model that enables accurate quantification of the concentration of covalently bonded molecules or atoms and incorporates input parameters from quantum chemistry calculations that account for surface modifications. We analyze the intensity ratio of the defect-induced D band over the primary G band, along with their broadenings. This approach is validated through the comparison with both our experimental observations and previously published results on various adatom coverage rates. Good agreement is found for both low coverage rates, where the sizes of distorted and activated regions are derived from calculations, and high coverage rates, where sp 3 -hybridized, Raman-inactive regions are included. A detailed analysis of the electronic modifications induced by chemisorbed atomic hydrogen on the graphene surface enables a nonempirical approach to quantify the coverage rate
Analysis of the Compressive Buckling and Post-Buckling Behaviour of Wood-Based Sandwich Panels Used in Light Aviation
No full textInternational audienceThis work aims to investigate the buckling and post-buckling behaviour of wood-based sandwich structures with and without a manufacturing defect, under compressive loading. The specimens were made by gluing birch veneers to a balsa wood core. The defect consisted of a central zone where glue was lacking between the skin and the core. A compression load was applied to the plate using the VERTEX test rig, with the plate placed on the upper surface of a rectangular box and bolted at its borders. The upper surface of the plate was monitored using optical and infrared cameras. The stereo digital image correlation method was used to capture the in-plane and out-of-plane deformations of the specimen, and to calculate the strains and stresses. The infrared camera enabled the failure scenario to be identified. The buckling behaviour of pristine specimens showed small local debonding in the post-buckling range, which was not detrimental to overall performance. In the presence of a manufacturing defect, the decrease in buckling load was only about 15%, but final failure occurred at lower compressive loads
Dynamic structure of the cytoplasm
No full textThe cytoplasm is a dense and complex milieu in which a plethora of biochemical reactions occur. Its structure is not understood so far, albeit being central to cellular functioning. In this review, we highlight a novel perspective in which the physical properties of the cytoplasm are regulated in space and time and actively contribute to cellular function. Furthermore, we underscore recent findings that the dynamic formation of local assemblies within the cytoplasm, such as condensates and polysomes, serves as a key regulator of mesoscale cytoplasmic dynamics.</div
A new CUDA improved and optimised version announcement of the quantum dissipative dynamics package
No full textInternational audienceThe third release of the Quantum Dissipative Dynamics (QDD) package follows the second release [P. M. Dinh, et al., Comp. Phys. Comm. 295 (2024) 108947] in which the focus was mainly made on the optimization of the electronic dynamics, in particular with a CUDA fortran coding to allow the use of a GPU. In this new release, we pursued the exportation of other parts of QDD on CUDA-capable GPUs, as the ionic motion, the coupling of the electrons with a laser field and/or with the ions (via pseudopotentials), and all electronic observables, including the involved photo-electron spectra, energy- and/or angle-resolved. Several specific CUDA optimisations have also been implemented, to improve the performance and the memory usage while keeping the accuracy of the results
Flexibilité pour la durabilité des systèmes multiénergie : le cas d'une grande infrastructure de recherche
No full textInternational audienceFace aux enjeux de décarbonation et d'efficacité énergétique, cette étude examine les solutions de flexibilité énergétique et de valorisation de chaleur fatale appliquées à l'étude du Laboratoire National des Champs Magnétiques Intenses (LNCMI), une infrastructure de recherche consommant de l'ordre de 10 GWh annuel d'électricité. En s'appuyant sur une approche systémique, intégrant l'analyse des flux énergétiques et l'optimisation des émissions de gaz à effet de serre, nous évaluons les bénéfices environnementaux de ces solutions. Ce travail propose une méthodologie ouverte et reproductible, spécifiquement adaptée aux infrastructures électro-intensives
A gradient-based method for concurrent layout and topology optimization of modular lattice structures
No full textInternational audienceThe design of modular lattice structures, known for their adaptability and damage tolerance, has become increasingly significant with advancements in additive manufacturing and aerospace applications. This study introduces an innovative methodology to minimize structural weight by concurrently optimizing the layout of the structure and the topology of the repeating modules in such structures. The structure layout is parameterized using continuous design variables, enabling the use of a gradient descent optimizer. A two-step optimization strategy is implemented: an initial relaxed optimization phase addresses layout design and initial module topology, followed by a refinement phase that incorporates advanced mechanical constraints into the topology.Applications span 2D and 3D structures, including benchmark cases such as cantilever beams and modular truss bridges. The results highlight the framework’s capability to simultaneously handle stress and local buckling constraints, as well as its modularity. In the best case, it produces designs comparable in volume to monolithic structures, with a weight penalty of only up to 2.8 %. Furthermore, the algorithm demonstrates its efficiency for medium-scale applications, achieving optimized configurations within minutes on a standard laptop computer.<br /
Airplane panels inspection via 3D point cloud segmentation on low-volume training dataset
No full textInternational audienceThis paper presents an approach for inspecting airplane fuselage panels using deep learning techniques, specifically segmentation models applied to 3D point clouds. A common challenge in deep learning applications is the limited availability of annotated training data — a problem particularly pronounced in industrial settings due to the high cost, time-consuming nature of data collection, and confidentiality concerns. Focusing on the detection and classification of rivets in the point cloud, our method aims to optimize the identification of potential defects, such as missing rivets or damaged rivets, even with minimal training data. Additionally, we evaluate the impact of different scanners and scanning resolution
Effective Data Generation and Feature Selection in Learning for Planning
No full textInternational audiencePrevious studies have shown that leveraging data beyond optimal training plans improves the learning of search guidance for planning. Specifically, state ranking information can be extracted from states on optimal plan traces and their siblings. In this paper, we generalise this approach by extracting additional rankings from the A⋆ search tree for generating optimal training plans. As in the previous approach, we incur no additional search effort and negligible computational overhead for data extraction. However, extracting more data in this way may introduce many redundant features and states which slows down training. We formalise the problem of sound, redundant feature pruning and show that it is NP-complete to solve. Furthermore, we introduce several algorithms and approximations for redundant feature pruning. Experiments show that rankings learned by extracting more data from search trees for generating optimal training plans improve planner coverage. However, pairing with unsound pruning methods often results in diminishing performance, while our sound feature pruning methods provide consistent improvements across tested domains
Towards Canonical and Minimal Solutions in a Constraint-based Plan-Space Planner
No full textInternational audienceTemporal planning is the problem of selecting and scheduling a set of actions in order to achieve a distant objective from an initial state. Constraint Programming (CP) has historically played an important role in many automated temporal planners based on the Plan-Space Planning (PSP) algorithm. In a nutshell, PSP would drive the search for a plan by selecting which actions to add and how to resolve conflicts, with an embedded CP engine in charge of ensuring the coherence of the plan.In this paper, we consider an existing encoding of temporal planning into constraint programming that adheres to a plan-space approach, but encodes the decision of the PSP algorithm as decision variables of the CSP, thus letting the CP engine entirely drive the search for a solution plan. Analyzing the encoding, we show that the induced search space contains many symmetries that would be absent from the original PSP algorithm. Another caveat of the encoding is that it allows for non-minimal solutions, i.e., solutions in which some actions may be removed without invalidating it.We propose to extend the encoding with additional symmetry breaking constraints to enforce canonicity of the considered plans. The encoding is further improved with additional constraints to limit the presence of non-minimal plans. We show the extended encoding to very substantially improve the performance of the original version and to be competitive with state-of-the-art state-space temporal planners on the considered benchmarks