Archive ouverte HAL-LAAS
Not a member yet
12189 research outputs found
Sort by
A laplace duality for integration
No full textInternational audienceWe consider the integral v(y) = Ky f (x)dx on a domain Ky = {x ∈ R d : g(x) ≤ y}, where g is nonnegative and Ky is compact for all y ∈ [0, +∞). Under some assumptions, we show that for every y ∈ (0, ∞) there exists a distinguished scalar λy ∈ (0, +∞) such thatwhich is the counterpart analogue for integration of Lagrangian duality for optimization. A crucial ingredient is the Laplace transform, the analogue for integration of Legendre-Fenchel transform in optimization. In particular, if both f and g are positively homogeneous then λy is a simple explicitly rational function of y. In addition if g is quadratic form then computing v(y) reduces to computing the integral of f with respect to a specific Gaussian measure for which exact and approximate numerical methods (e.g. cubatures) are available.</p
Absolute In situ measurement of Al x Ga 1-x As growth rate and composition by magnification inferred curvature method
No full textInternational audienceThe direct in situ measurement of III-element growth rates in epitaxy methods has been a continuous pursuit. Despitethe development of numerous in situ monitoring tools, the accurate measurement of growth rates remains challenging.The calibration of Group III element fluxes, and consequently the associated growth rate, is performed through variousmethods (such as RHEED oscillations, equivalent pressure measurement, optical methods, quartz crystal microbalance,etc.). However, these direct in situ measurements are often valid only under specific conditions and/or involve indirectmeasurements that require knowledge of additional parameters. In this study, we applied the magnification inferredcurvature method (MIC) to directly measure the growth rate of slightly strained AlAs layers, and thus enabling tocalibrate the Al effusion cells and controlling the thickness and composition of Alx Ga1-x As layers independently oftheir thickness, and the growth conditions. From MIC measurements, we were able to deduce the elastic parameters atgrowth temperature, enabling us to achieve a universal calibration of the Al incorporation rate. This calibration is alsoapplicable to other group-III or -V element cells, provided they induce strain relative to the substrate, and it is fullytransferable from one epitaxy chamber to another
Generalized Lyapunov conditions for k-contraction: analysis and feedback design
No full textInternational audienceRecently, the concept of k-contraction has been introduced as a promising generalization of contraction for dynamical systems. However, the study of k-contraction properties has faced significant challenges due to the reliance on complex mathematical objects called matrix compounds. As a result, related control design methodologies have yet to appear in the literature. In this paper, we overcome existing limitations and propose new sufficient conditions for k-contraction which do not require matrix compounds computation. Notably, these conditions are also necessary in the linear time-invariant framework. Leveraging on these findings, we propose a feedback design methodology for both the linear and the nonlinear scenarios which can be used to enforce k-contractivity properties on the closed-loop dynamics
Increasing Precision of Impedance Measurements for ESD Protection's Frequency Models
No full textInternational audienceAchieving system level ESD robustness by simulation required the use of precise models that should take into account the triggering behavior of the protections. This paper upgrades articles [1], [2], [3] dealing with frequency measurement methods to get the equivalent impedance of ESD protections exposed to strong pulses. In this paper, a two-port transmission measurement is used to increase the accuracy of measurement for very low or high impedance values of this type of protections. The pros and cons of one-port versus two-port measurements are presented, showing how the precision is increased. Some experimental results on commercial TVS are proposed and discussed. To conclude this article, different models also built from frequency measurements are introduced and compared
Adaptive Single Micro-Particle Detection and Segmentation in Self-Mixing Interferometry Signals
No full textMeilleur papier de la conférence 3rd Int. IEEE Applied Sensing Conference (APSCON 2025)International audienceThis paper introduces a signal processing algorithm for microparticle detection using the self-mixing interferometry (SMI) sensing scheme. Processing the laser output power modulation caused by the transit of single particles, the algorithm effectively segments the SMI signal associated with each particle. The algorithm was validated on the segmentation of different-size particles (2, 4, and 10 µm), and its performances are compared to the more established spectral amplitude detection method. The results highlight the SMI technique's potential for micro-particle analysis and facilitate the adoption of new signal processing algorithms to improve signal feature extraction
PROXDDP: Proximal Constrained Trajectory Optimization
No full textInternational audienceTrajectory optimization has been a popular choice for motion generation and control in robotics for at least a decade. Several numerical approaches have exhibited the required speed to enable online computation of trajectories for real-time of various systems, including complex robots. Many of these said are based on the differential dynamic programming (DDP) algorithm – initially designed for unconstrained trajectory optimization problems – and its variants, which are relatively easy to implement and provide good runtime performance. However, several problems in robot control call for using constrained formulations (e.g. torque limits, obstacle avoidance), from which several difficulties arise when trying to adapt DDP-type methods: numerical stability, computational efficiency, and constraint satisfaction.In this article, we leverage proximal methods for constrained optimization and introduce a DDP-type method for fast, constrained trajectory optimization suited for model-predictive control (MPC) applications with easy warm-starting.Compared to earlier solvers, our approach effectively manages hard constraints without warm-start limitations and exhibits good convergence behavior. We provide a complete implementation as part of an open-source and flexible C++ trajectory optimization library called ALIGATOR. These algorithmic contributions are validated through several trajectory planning scenarios from the robotics literature and the real-time whole-body MPC of a quadruped robot
Optimizing containerized edge service migration through file-level storage sharing
No full textInternational audienceEdge computing facilitates low-latency and high-bandwidth services by deploying edge servers close to data sources and end users. However, dynamic edge environments, characterized by fluctuating resources and user mobility, etc., necessitate efficient service migration to maintain consistently low latency. Traditional migration methods incur significant latency due to redundant data transfers, especially in containerized environments where shareable storage layers and redundant files across non-shareable layers in Docker containers exacerbate inefficiencies. Despite advances in layer sharing migration techniques, the challenges remain as follows: (i) shareable layers are limited due to coarse granularity, and (ii) approximately 97% of files across layers are duplicates. To solve this problem, we propose a framework for live migration of Docker containers, minimizing file system transfer size through file-level storage sharing and enhanced layer reuse. Our method reuses identical image layers and shares files both across layers and with those already present on the target server, transferring only the necessary files. Experimental results demonstrate that our file-level storage sharing strategy outperforms state-of-the-art techniques in terms of reducing the migration time of Docker containers
Machine learning pipeline for microparticle size classification in self-mixing interferometric signals for flow cytometry
No full textInternational audienceSelf-mixing interferometry (SMI) is an emerging optical sensing technique for detecting and classifying microparticles in non-contact and label-free flowmetry applications. High precision and reliability are essential for its integration into medical diagnostics, such as blood analysis, and quality control in chemical manufacturing processes. While theoretical models describe SMI-induced signal modulations caused by particle passage, challenges persist due to signal noise, variability, and interpretability under experimental conditions. This study enhances SMI-based particle size classification by integrating machine learning (ML) models to improve feature extraction and classification accuracy. Three ML pipelines are evaluated, achieving 98% classification accuracy in distinguishing particles of different sizes (2, 4, and 10 µm). The high classification accuracy demonstrates the scalability of our approach, ensuring its applicability across diverse particle analysis scenarios
Multi-Objective IoT Service Placement in Cloud-Fog-Edge Environments Using Deep Reinforcement Learning
No full textInternational audienceThe rapid proliferation of Internet of Things devices introduces significant challenges in determining optimal service placement across heterogeneous computing layers, spanning from centralized cloud servers to decentralized fog and edge nodes. Effective placement is essential to minimize latency, reduce energy consumption, and control costs, all while adhering to resource constraints such as limited memory and processing capacity. In this paper, we investigate several strategies for solving the IoT service placement problem. To overcome the limited adaptability of traditional optimization approaches, we introduce dynamic resource management in the simulation environment YAFS and propose a Deep Reinforcement Learning-based approach utilizing a Double Deep Q-Network (DDQN) architecture. The DRL agent autonomously learns placement policies through continuous interaction with the environment, optimizing a weighted multi-objective reward that balances execution time, energy efficiency, and cost. Experimental evaluations were conducted under two scenarios: memory-only constraints and combined memory plus CPU constraints. Results demonstrate that the DRL-based strategy consistently outperforms baseline approaches, including three meta-heuristic optimization methods-Genetic Algorithm, Simulated Annealing, and Tabu Search-as well as the cloud-only strategy, across all performance metrics
Are Diagnostic Concepts Within the Reach of LLMs?
No full textInternational audienceModel-based diagnosis is a cornerstone of system health monitoring, allowing for the identification of faulty components based on observed behavior and a formal system model. However, obtaining a useful and reliable model is often an expensive and manual task. While the generation of a formal model was the aim of previous work, in this paper, we propose a methodology to use large language models to generate Minimal Structurally Overdetermined sets (MSOs). MSOs are specific subsets of the model equations from which diagnosis tests can be obtained. We investigate two different directions: (i) the large-language-models' ability to generate MSO sets for hybrid systems, similar to those generated by the well-known Fault Diagnosis Toolbox (FDT) (ii) the automated generation of MSOs for Boolean circuits, as well as the computation of the diagnoses. We thus show how both dynamic and static systems can be analysed by large-language models and how their output can be used for effective fault diagnosis. We evaluate our approach on a set of arithmetic and logic circuits, using OpenAI’s LLMs 4o-mini, o1, and o3-mini