HAL Portal Cnam (French National Conservatory for Arts and Crafts)
Not a member yet
31010 research outputs found
Sort by
Projet Dolora : outil d'accompagnement et d'évaluation de la douleur avec et pour les femmes
No full textInternational audienc
Comparison between a Two-Wavelength Absolute Distance Meter and a GNSS-Based Distance Meter at CERN Geodetic Network
No full textInternational audienc
Lot-sizing under decision-dependent uncertainty: A probing-enhanced stochastic programming approach
No full textInternational audienceWe address the multi-item capacitated lot-sizing problem under decision-dependent uncertainty via a probing-enhanced stochastic programming framework. Demand is correlated with another random vector, and the decision-maker can acquire partial information by probing components of this vector, conditioning decisions on observed covariates. This generalizes classical models by embedding information acquisition into a three-stage framework. We propose a compact reformulation that removes non-anticipativity constraints, yielding stronger relaxations and better tractability. We extend classical inequalities and introduce value-function cuts that capture the link between probing and recourse costs. These are embedded in a branch-and-cut algorithm with a primal heuristic. Results show our method outperforms off-the-shelf solver, reducing optimality gaps by up to 85%, and achieving gaps below 1.5% on average. Results highlight the importance of structured reformulation, valid inequalities, and heuristics in solving decision-dependent stochastic programs
Comment appliquer la transposition didactique en anglais financier : le cas des indices boursiers. Épisode 2 : Suggestions de mise en œuvre n° 2 et 3
No full textInternational audienc
Scalable Hydrodynamics on multiple Field-Programmable Gate Arrays (FPGAs)
No full textInternational audienceThis paper presents scalable 2D and 3D Hydrodynamics solver implementations on FPGAs using Intel’s oneAPI framework, addressing challenges in porting stencil-based computations from CPU/GPU architectures on FPGAs.FPGAs offer customizable hardware with on-chip memory that can be custom-tailored for High Performance Computing (HPC) applications. Despite a much lower throughput, FPGAs achieve comparable energy efficiency with domain decomposition compared to NVIDIA A100 GPU. Optimizations leverage FPGA’s large cache to minimize data streaming and ensure data reuse, reducing memory bandwidth.Domain decomposition enabled pipeline duplication for scalable single FPGA workload, but also the distribution of the work on two Agilex 7 cards.This work demonstrates FPGA viability for computationally intensive simulations, contributing to scalable algorithms for heterogeneous HPC systems and offering insights for exascale computing
Dark-era : Dataflow Algorithm aRchitecture co-design of SKA pipeline for Exascale Radio Astronomy
No full textInternational audienc
Le professioni dell'educazione e della formazione a confronto con il lavoro : prospettive interdisciplinari e internazionali
No full textSecuring Offline CBDC Transactions: DigiVault Card and MarkoPayChain with Mobile Phone Integration
No full textInternational audienceResearch on offline payment systems for Central Bank Digital Currencies (CBDCs) faces challenges in convincing stakeholders of their practicality for widespread use. Implementing such systems entails overcoming significant technological hurdles while ensuring privacy and regulatory compliance. Although some solutions utilize mobile devices, few capitalize on the advantages of smart cards, which enhance security, flexibility, and portability. Moreover, integrating effective privacy and security measures in offline environments remains limited. This research proposes a novel solution with significant contributions:1) It employs a hybrid approach using smart cards for secure storage and cryptographic key management, coupled with smartphones for computational and authentication tasks. 2) It introduces proactive fraud protection utilizing Markov chains, enhancing security against double spending. 3) It employs two Privacy Enhancing Technologies (PETs): blind signatures and zk-SNARKs, ensuring robust privacy.4) It meets regulatory standards by enabling verifiable transactions without compromising user privacy. The evaluation of the offline payment protocol showcased strong reliability, achieving a success rate exceeding 88 % across all simulation
Data-Driven Optimisation Method for Tuning the HF Injection Parameters of an Extra Low Voltage Encoderless Synchronous Motor Drive
No full textInternational audienceThis paper discusses encoderless control of a permanent magnet synchronous machine that is fed by an extra-low DC bus voltage and operates within the low-speed range. The focus is on optimising the parameters of the high-frequency (HF) voltage injections necessary for estimating the electrical position of the rotor. Although the encoderless control method adopted is pulsating carrier voltage injection, the optimisation algorithm can be applied to other methods as well. The quality of the estimated position is influenced by the amplitude of the injected voltages, which also affect the disturbance of the machine torque. Therefore, this study aims to strike a balance between the quality of the estimated position and injection-related intrusion. The paper proposes an experimental approach to determining the optimal injection level, enabling accurate estimation of the electrical position while reducing torque disturbance. It also presents the experimental validation results of the optimal injection voltage implementation