HAL-Ecole des Ponts ParisTech
Not a member yet
42743 research outputs found
Sort by
A nonlinear reduced-order model for parametrized variational inequalities: application to crowd motion
No full textIn this work we adapt recent nonlinear model reduction approaches to predict the solutions of time-dependent parametrized variational inequalities. In our present context, we make a specific focus on discrete contact problems. A prototypical example is the agentbased model proposed in [MV11] to describe crowd motion in the presence of obstacles. In this model, in a discrete time setting, the set of velocities of each agent in the crowd is the solution at each time step to a constrained least-squares optimization statement. The parametric variations of the problem (associated with the geometric configuration of the domain where the agents evolve) have a very strong impact on the variability of the solution, both in terms of positions of the agents and of contact forces between them, the latter being given by the Lagrange multipliers associated to non-interpenetration constraints. More precisely, the Kolmogorov n-width of the solution set is very slowly decaying. Motivated by this observation, we investigate new developments and combinations of the reduced-basis method and supervised machine-learning techniques to provide more accurate approximations of the primal and dual solutions, inspired from the recent works [BFM23; Coh+23]. The proposed nonlinear compressive strategy then reads as a postprocessing step of the solution of a standard reduced basis reduced-order model, which yields an improvement of the accuracy of the approximation by an order of magnitude for a negligible extra computational cost. We see this work as a preliminary step before the investigation of more efficient nonlinear reduced order modeling approaches for this type of problems
Evidence from the Dead: New Estimates of Wealth Inequality based on the Distribution of Estates
No full textThis paper examines the estimation of the distribution of wealth using estates left at death. We establish formal conditions for implementing a simplified version of the classic estate multiplier method, relying solely on minimal information about estates and mortality. These conditions are empirically validated, and the simplified approach is applied to produce new long-run top wealth share series for Belgium, Japan, and South Africa, where estate data have previously been underutilized. This method holds potential for expanding the range of countries and years in which wealth concentration can be estimated, especially where estate data exist but the standard method with heterogeneous multipliers is inapplicable
Intelligent Task Offloading in Vehicular Networks: A Deep Reinforcement Learning Perspective
No full textInternational audienceThe proliferation of connected vehicles and the Internet of Things (IoT) has made access to high-quality services increasingly viable. However, the growing number of vehicular applications poses challenges for embedded systems that need to perform tasks efficiently despite network fluctuations. To solve this problem, we have developed a Vehicular Edge Computing (VEC) system with a task offloading algorithm adapted to the Internet of Vehicles (IoV). Our solution uses a four-stage Stackelberg game and a reinforcement learning model. The approach consists of analyzing the vehicle state in real time to determine computational requirements and cost functions, implementing communication methods and cost functions, and using multi-agent reinforcement learning to design an experience-based offloading strategy. Through simulations, our algorithm demonstrates a balanced optimization of time, expense, work vehicle utility and service vehicle utility, while improving the probability of task success under various constraints
Phase transition from turbulence to zonal flows in the Hasegawa–Wakatani system
No full textInternational audienceThe transition between two-dimensional hydrodynamic turbulence and quasi-one-dimensional zonostrophic turbulence is examined in the modified Hasegawa–Wakatani system, which is considered as a minimal model of β-plane-like drift-wave turbulence with an intrinsic instability. Extensive parameter scans were performed across a wide range of values for the adiabaticity parameter C describing the strength of coupling between the two equations. A sharp transition from 2D isotropic turbulence to a quasi-1D system, dominated by zonal flows, is observed using the fraction of the kinetic energy of the zonal modes as the order parameter, at C≈0.1. It is shown that this transition exhibits a hysteresis loop around the transition point, where the adiabaticity parameter plays the role of the control parameter of its nonlinear self-organization. It was also observed that the radial particle flux scales with the adiabaticity parameter following two different power law dependencies in the two regimes. A simple quasi-linear saturation rule which accounts for the presence of zonal flows is proposed, and is shown to agree very well with the observed nonlinear fluxes. Motivated by the phenomenon of quasi-one dimensionalisation of the system at high C, a number of reduction schemes based on a limited number of modes were investigated and the results were compared to direct numerical simulations. In particular, it was observed that a minimal reduced model consisting of 2 poloidal and 2 radial modes was able to replicate the phase transition behavior, while any further reduction failed to capture it
Étude des déchets plastiques et des fibres anthropiques lors d’événements transitoires : épisodes pluvieux en milieu urbain
No full textInternational audienceThe presence of plastic debris, whether macro-, micro- or fibres, is now recognised worldwide as a major environmental issue. Due to its high activity and dense population, the urban environment is considered a major source of plastic pollution. This article presents new data concerning plastic waste and anthropogenic fibres in rainy weather discharges downstream of a peri-urban catchment. It also proposes extrapolations of flows transferred to receiving environments at the scale of Greater Paris and the Seine basin. The abundance and composition of macroplastics, microplastics and anthropogenic fibres in the stormwater of a peri-urban residential catchment (Sucy-en-Brie, France) of Greater Paris was studied over one year (macroplastics) and 4 rainfall events (fibres and microplastics). For this site, the concentrations of macroplastics and microplastics in the rainwater are of the same order of magnitude. Extrapolating to the scale of the Paris conurbation, the estimated quantity of macroplastic debris discharged into the environment by the separate networks varies from 8 to 33 metric tons per year, while the flows of microplastics and anthropogenic fibres vary respectively from 3 to 48 tons per year and from 0.3 to 0.8 tons per year. The comparison of the various data in order to establish flows at the scale of the conurbation but also at the scale of the Seine catchment area demonstrates the complexity of tackling this pollution in a systemic manner and their use in an operational manner on a large scale is premature.La présence de déchets plastiques, qu’il s’agisse de macrodéchets, de microplastiques ou de fibres, est reconnue aujourd’hui à l’échelle planétaire comme un enjeu environnemental majeur. Par les activités et les populations qu’il concentre, le milieu urbain est considéré comme une source majeure de pollution plastique. Cet article présente de nouvelles données concernant les déchets plastiques et les fibres anthropiques dans les eaux de ruissellement, à l’aval d’un bassin versant périurbain séparatif. Il propose aussi des extrapolations des flux transférés vers les milieux récepteurs à l’échelle du Grand Paris et du bassin de la Seine. L’abondance et la composition des macroplastiques, des microplastiques et des fibres anthropiques dans les eaux pluviales d’un bassin versant résidentiel périurbain (Sucy-en-Brie, France) du Grand Paris ont été étudiées sur un an, entre juin 2018 et mai 2019 (macroplastiques) et quatre événements pluvieux (fibres et microplastiques). Pour ce site, les concentrations en macroplastiques et microplastiques dans les eaux pluviales sont du même ordre de grandeur et atteignent quelques g/m3. En extrapolant à l’échelle de l’agglomération parisienne, la quantité estimée de déchets macroplastiques rejetés dans l’environnement par les réseaux séparatifs varie de 8 à 33 tonnes/an, tandis que les flux de microplastiques et de fibres anthropiques varient respectivement de 3 à 48 tonnes/an et de 0,3 à 0,8 tonne/an. La confrontation des différentes données en vue d’établir des flux à l’échelle de l’agglomération, mais aussi à l’échelle du bassin versant de la Seine, démontre toute la complexité d’aborder cette pollution de manière systémique. Leur utilisation de manière opérationnelle à grande échelle est prématurée
IA, MÉDECINE ET SCIENCES SOCIALES. Une mise en perspective
No full textInternational audienceHealthcare is an ideal field of application for devices based on Artificial Intelligence (AI). This introduction to the ‘AI and Medicine’ issue of Réseaux offers a social science perspective onthe intersection of these two fields. Drawing on the presentation and discussion of the six articles making up this issue, as well as other works – mainly from sociology and from Science and TechnologyStudies (STS) –, several issues and questions are highlighted: the unfinished, imperfect and even counter-productive nature of the automation generated by AI devices in work situations; the need fora fresh decoding of the notion of ‘explainability’ in relation to design processes and actual uses, which may not comply with legal requirements; and the relative absence of patients in the emergingpractices studied. The latter phenomenon is explained primarily by the fact that AI is used well beyond the framework of the clinical relationship, to equip various categories of professionals who are notdirectly involved in patient management and care.Le domaine de la santé constitue un champ d’application privilégié des dispositifs qui se revendiquent de l’intelligence artificielle (IA). Cette introduction au numéro « IA et médecine » propose une miseen perspective, du point de vue des sciences sociales, du croisement et de la rencontre de ces deux domaines. En s’appuyant sur la présentation et la discussion des six articles qui composent ce numéro,ainsi que sur d’autres travaux relevant principalement de la sociologie et des Science and Technology Studies (STS), plusieurs enjeux et questions sont mis en exergue : le caractère inachevé, imparfait,voire contre-productif, dans les situations de travail, de l’automatisation engendrée par des dispositifs d’IA ; l’acuité d’un décryptage à nouveaux frais de la notion d’« explicabilité » en rapport avec desprocessus de conception et des usages concrets, qui peuvent être en déphasage avec le droit en vigueur ; enfin, la relative absence des patients dans les pratiques émergentes étudiées, qui s’explique notamment par le fait que l’IA est mobilisée bien au-delà du cadre de la relation clinique afin d’équiper différentes catégories de professionnels qui n’interviennent pas directement dans la prise en charge des patients et dans les soins
Avec le télétravail et les achats en ligne, quelles évolutions des mobilités?
No full textInternational audienc
How Manipulable Are Prediction Markets ?
No full textIn this paper, we conduct a large-scale field experiment to investigate the manipulability of prediction markets. The main experiment involves randomly shocking prices across 817 separate markets; we then collect hourly price data to examine whether the effects of these shocks persist over time. We find that prediction markets can be manipulated: the effects of our trades are visible even 60 days after they have occurred. However, as predicted by our model, the effects of the manipulations somewhat fade over time. Markets with more traders, greater trading volume, and an external source of probability estimates are harder to manipulate
Topological Properties of the Effective Reproduction Number in an Heterogeneous SIS Model
No full textThis present results lay the foundations for the study of the optimal allocation of vaccine in the simple epidemiological SIS model where one consider a very general heterogeneous population. In the present setting each individual has a type x belonging to a general space, and a vaccination strategy is a function η where η(x) ∈ [0, 1] represents the proportion of non-vaccinated among individuals of type x. We shall consider two loss functions associated to a vaccination strategy η: either the effective reproduction number, a classical quantity appearing in many models in epidemiology, and which is given here by the spectral radius of a compact operator that depends on η; or the overall proportion of infected individuals after vaccination in the maximal endemic state.By considering the weak-* topology on the set ∆ of vaccination strategies, so that it is a compact set, we can prove that those two loss functions are continuous using the notion of collective compactness for a family of operators. We also prove their stability with respect to the parameters of the SIS model. Eventually, we consider their monotonicity and related properties in particular when the model is “almost” irreducible
Global Carbon Budget 2024
No full textInternational audienceAccurate assessment of anthropogenic carbon dioxide (CO2) emissions and their redistribution among the atmosphere, ocean, and terrestrial biosphere in a changing climate is critical to better understand the global carbon cycle, support the development of climate policies, and project future climate change. Here we describe and synthesize datasets and methodologies to quantify the five major components of the global carbon budget and their uncertainties. Fossil CO2 emissions (EFOS) are based on energy statistics and cement production data, while emissions from land-use change (ELUC) are based on land-use and land-use change data and bookkeeping models. Atmospheric CO2 concentration is measured directly, and its growth rate (GATM) is computed from the annual changes in concentration. The global net uptake of CO2 by the ocean (SOCEAN, called the ocean sink) is estimated with global ocean biogeochemistry models and observation-based fCO2 products (fCO2 is the fugacity of CO2). The global net uptake of CO2 by the land (SLAND, called the land sink) is estimated with dynamic global vegetation models. Additional lines of evidence on land and ocean sinks are provided by atmospheric inversions, atmospheric oxygen measurements, and Earth system models. The sum of all sources and sinks results in the carbon budget imbalance (BIM), a measure of imperfect data and incomplete understanding of the contemporary carbon cycle. All uncertainties are reported as ±1σ. For the year 2023, EFOS increased by 1.3 % relative to 2022, with fossil emissions at 10.1 ± 0.5 GtC yr−1 (10.3 ± 0.5 GtC yr−1 when the cement carbonation sink is not included), and ELUC was 1.0 ± 0.7 GtC yr−1, for a total anthropogenic CO2 emission (including the cement carbonation sink) of 11.1 ± 0.9 GtC yr−1 (40.6 ± 3.2 GtCO2 yr−1). Also, for 2023, GATM was 5.9 ± 0.2 GtC yr−1 (2.79 ± 0.1 ppm yr−1; ppm denotes parts per million), SOCEAN was 2.9 ± 0.4 GtC yr−1, and SLAND was 2.3 ± 1.0 GtC yr−1, with a near-zero BIM (−0.02 GtC yr−1). The global atmospheric CO2 concentration averaged over 2023 reached 419.31 ± 0.1 ppm. Preliminary data for 2024 suggest an increase in EFOS relative to 2023 of +0.8 % (−0.2 % to 1.7 %) globally and an atmospheric CO2 concentration increase by 2.87 ppm, reaching 422.45 ppm, 52 % above the pre-industrial level (around 278 ppm in 1750). Overall, the mean of and trend in the components of the global carbon budget are consistently estimated over the period 1959–2023, with a near-zero overall budget imbalance, although discrepancies of up to around 1 GtC yr−1 persist for the representation of annual to semi-decadal variability in CO2 fluxes. Comparison of estimates from multiple approaches and observations shows the following: (1) a persistent large uncertainty in the estimate of land-use change emissions, (2) low agreement between the different methods on the magnitude of the land CO2 flux in the northern extra-tropics, and (3) a discrepancy between the different methods on the mean ocean sink. This living-data update documents changes in methods and datasets applied to this most recent global carbon budget as well as evolving community understanding of the global carbon cycle. The data presented in this work are available at https://doi.org/10.18160/GCP-2024 (Friedlingstein et al., 2024)