Search

The mathematical modelling and analysis of the collective behaviour of a cloud of N interacting particles or agents has attracted a lot of interest in the last years in several communities, such as biologists, physicists, mathematicians, computer scientists etc. This is motivated not only by fundamental reasons, such as the understanding of the natural phenomena occurring around us, but also by the wide applications of this field in several domains, such as collective robotics, unmanned areal vehicles, ...Several mathematical models appeared in literature in the last years, such as for ex. the Viscek model, the Kuramoto model, the Cucker-Smale model etc, each one being specifically adapted for a particular situation, and several mathematical and numerical studies have been performed, the literature being constantly growing. The basic models have been fully understood today, what is still open in our opinion is the design of more realistic models, permitting to get closer to reality, and the corresponding mathematical and numerical analysis. In fact, a truly good model must on one hand recreate the real-life behaviour one is investigating, and on the other hand it must be simple enough to enable a detailed mathematical and numerical study. So in our particular case of a drone swarm, all the specificities mentionned in Section 3.2 shall be step by step included in a realistic drone model and efficient, multi-scale numerical schemes designed to be proposed to the industrials

International audience; The envelope of the solidification front naturally develops a macroscopic curvature in a multitude of solidification processes. However, its effect on dendritic microstructure formation remains poorly understood. Here we exploit a microgravity environment where convection is suppressed to investigate quantitatively the effect of curvature on dendritic array growth during directional solidification of a transparent succinonitrile 0.46 wt% camphor alloy. In addition, we interpret the results using both theoretical analyses and phase-field simulations. In situ observations reveal that even a weak macroscopic interface curvature can have a major effect on both the array pattern evolution and grain structure. First, convex and concave interfaces lead to a continuous increase and decrease in time of the average primary spacing, respectively, which only attains a stationary value for a flat interface. We show that this results from the formation of drifting velocity gradients along the interface under the combined effect of the crystal misorientation and the interface misorientation. These grains form at sample boundaries and overgrow pre-existing interior grains of smaller misorientation to the detriment of well-oriented crystal growth. In addition, they induce a change of array structure when the directional normal to the sample boundary is parallel to one of axis of secondary branches. These results provide new insights for controlling and optimizing directionally solidified alloy microstructures.

The mathematical modelling and analysis of the collective behaviour of a cloud of N interacting particles or agents has attracted a lot of interest in the last years in several communities, such as biologists, physicists, mathematicians, computer scientists etc. This is motivated not only by fundamental reasons, such as the understanding of the natural phenomena occurring around us, but also by the wide applications of this field in several domains, such as collective robotics, unmanned areal vehicles, ...Several mathematical models appeared in literature in the last years, such as for ex. the Viscek model, the Kuramoto model, the Cucker-Smale model etc, each one being specifically adapted for a particular situation, and several mathematical and numerical studies have been performed, the literature being constantly growing. The basic models have been fully understood today, what is still open in our opinion is the design of more realistic models, permitting to get closer to reality, and the corresponding mathematical and numerical analysis. In fact, a truly good model must on one hand recreate the real-life behaviour one is investigating, and on the other hand it must be simple enough to enable a detailed mathematical and numerical study. So in our particular case of a drone swarm, all the specificities mentionned in Section 3.2 shall be step by step included in a realistic drone model and efficient, multi-scale numerical schemes designed to be proposed to the industrials

All blood cells originate from hematopoietic stem/progenitor cells (HSPCs). HSPCs are formed from endothelial cells (ECs) of the dorsal aorta (DA), via endothelial-to-hematopoietic transition (EHT). The zebrafish is a primary model organism to study the process in vivo. While the role of mechanical stress in controlling gene expression promoting cell differentiation is actively investigated, mechanisms driving shape changes of the DA and individual ECs remain poorly understood. We address this problem by developing a new DA micromechanical model and applying it to experimental data on zebrafish morphogenesis. The model considers the DA as an isotropic tubular membrane subjected to hydrostatic blood pressure and axial stress. The DA evolution is described as a movement in the dimensionless controlling parameters space: normalized hydrostatic pressure and axial stress. We argue that HSPC production is accompanied by two mechanical instabilities arising in the system due to the plane stress in the DA walls and show how a complex interplay between mechanical forces in the system drives the emerging morphological changes.

Introducing macromolecular micelles into a biocompatible hyaluronic acid (HA) hydrogel is a promising strategy to improve its mechanical properties for biomedical applications. However, it is still unclear whether the solvent nature has an influence on the structure and property of HA gels especially when they are used for those cases containing binary solvents because reversible hydrophobic association within micelles could be weakened or even dissociated by organic solvents. In this work, we demonstrated that a binary solvent consisting of water and low-toxic dimethyl sulfoxide (DMSO), a commonly used cryoprotectant agent in biomedicine, can enhance the mechanical properties of hydrophobic-associated methacrylated hyaluronate (MeHA) gels crosslinked by diacrylated PEO99-PPO65-PEO99 (F127DA) macromolecular micelles, namely FH gels. The resulting FH hydro/organo-gels showed a crystalline structure due to polymer/solvent interactions. The FH gels showed a low swelling degree and the maximum strength (10.12 MPa), modulus (106.8 kPa) and toughness (1540 J m-2) in DMSO with a volume fraction of around 0.6. Moreover, the FH gels displayed a rapid recoverability under cyclic loading-unloading stress particularly in the presence of DMSO within the network due to their dual-dynamic dissipation networks. Such novel hydrophobic associated polysaccharide gels with tunable mechanical properties in binary solvents would be attractive in a cryopreservation system for cell-based applications.

Molecular evolution is often conceptualised as adaptive walks on rugged fitness landscapes, driven by mutations and constrained by incremental fitness selection. It is well known that epistasis shapes the ruggedness of the landscape's surface, outlining their topography (with high-fitness peaks separated by valleys of lower fitness genotypes). However, within the strong selection weak mutation (SSWM) limit, once an adaptive walk reaches a local peak, natural selection restricts passage through downstream paths and hampers any possibility of reaching higher fitness values. Here, in addition to the widely used point mutations, we introduce a minimal model of sequence inversions to simulate adaptive walks. We use the well known NK model to instantiate rugged landscapes. We show that adaptive walks can reach higher fitness values through inversion mutations, which, compared to point mutations, allows the evolutionary process to escape local fitness peaks. To elucidate the effects of this chromosomal rearrangement, we use a graph-theoretical representation of accessible mutants and show how new evolutionary paths are uncovered. The present model suggests a simple mechanistic rationale to analyse escapes from local fitness peaks in molecular evolution driven by (intragenic) structural inversions and reveals some consequences of the limits of point mutations for simulations of molecular evolution., Comment: 35 pages, 10 figures, accepted for publication in PLoS Computational Biology

International audience; Spin–orbit photonic devices usually rely on 2D (transverse) material structuring and are designed for optimal coupling between the polarization state and the spatial degrees of freedom at a given wavelength. Exploiting the third dimension (longitudinal) provides ways to bypass monochromatic limitations. Within a singular optics framework, here we show that chiral liquid crystals endowed with non-singular 3D helix axis orientational distribution exhibit transmissive broadband spin–orbit optical vortex generation as well as an optical diode effect. These results are in stark contrast to the properties of spin–orbit optical elements fabricated from chiral liquid crystals with a uniform orientation of the helix axis, which are reflective devices that process forward and backward propagating waves equally. Moreover, the similarities between the proposed 3D chiral structure and that of the cuticle of some insects invites considering spin–orbit photonics from a biological perspective.

The thesis focuses on the transition from laminar to turbulent flow in canonical wall-bounded shear flows presenting a linearly stable laminar flow. The subcritical transition in these shear flows is characterised by spatiotemporal intermittency in the transitional regime. This manifests as spatial localization of turbulence which self-organizes into large-scale coherent structures such as spots and bands. In this regard, these structures are investigated in different wall-bounded shear flows with high fidelity numerical simulations. The primary control parameter for the simulations is the Reynolds number Re. In the first part of the thesis, turbulent spots are investigated in four different shear flow cases: plane Couette, plane Poiseuille, Couette-Poiseuille and Waleffe flow. These flow scenarios present different symmetries and boundary conditions. Performing the simulations in large domains, the in-plane turbulent fluctuations are shown to decay algebraically away from the spot. The emergence of two distinct large-scale flow topologies dictates the spatial decay exponent. The large-scale flow structure and consequently the decay exponent are found to depend only on the symmetry of the flow and is independent of the Re. Arguments from 2D kinematics and flow symmetry are used to justify theoretically these observations. In the second part, the transition from turbulent to laminar in plane Poiseuille flow is investigated. The transitional regime is found to portray two distinct behaviours: (a) pattern formation with alternate regions of laminar and turbulent (b) spatially localised independent turbulent bands. Adopting the methodology of an impulse response with ensemble averaging, evidence for a linear instability of the turbulent flow leading to pattern formation is presented. The evolution of the pattern with Re is documented with its geometric properties as well as global observables such as the friction factor. High-order statistics of turbulent fluctuations reveal a continuous link between featureless turbulence and the transitional regime. In order to gather more insight into the formation and evolution of the pattern with Re, a low-order model of the shear flow is motivated. The third part explores a low-order model presented in the literature. The 1D PDE model presented by Manneville, featuring a Turing instability as an extension of the Waleffe model, is revisited. This was adapted with suitable parameters and extended with the introduction of nonlinear advection and stochastic noise. The model accurately captures the phenomenology of the transitional regime of plane Poiseuille flow. It features pattern formation with multistability, wavelength selection by noise and excitability at low Re. These results are extrapolated from the model and validated against the observations in the DNS simulations.; La thèse se concentre sur la transition laminaire/turbulent dans des écoulements de cisaillement canoniques délimités par des parois lorsque l'écoulement laminaire linéairement stable. La transition sous-critique dans ces écoulements cisaillés est caractérisée par l'intermittence spatio-temporelle du régime transitoire. Cela se manifeste par une localisation spatiale de la turbulence qui s'auto-organise en structures cohérentes à grande échelle telles que des bouffées turbulentes et des bandes. À cet égard, ces structures sont étudiées dans différents écoulements de cisaillement délimités par des parois, à l'aide de simulations numériques haute fidélité. Le paramètre de contrôle principal pour les simulations est le nombre de Reynolds Re. Dans la première partie de la thèse, les bouffées turbulentes sont étudiées dans quatre cas d'écoulement cisaillés différents : écoulement de Couette plan, de Poiseuille plan, de Couette-Poiseuille et de Waleffe. Ces écoulements présentent des symétries et des conditions aux limites différentes. En effectuant des simulations dans de grands domaines, on constate que les fluctuations turbulentes en vitesse dans le plan diminuent selon une loi de puissance à partir de l'origine. L'émergence de deux topologies d'écoulement à grande échelle distinctes déterminent l'exposant de décroissance spatiale. On constate que la structure de l'écoulement à grande échelle et, par conséquent, l'exposant de décroissance dépendent uniquement de la symétrie de l'écoulement et sont indépendants de Re. Des arguments cinématiques bidimensionnels et de symétrie sont utilisés pour justifier théoriquement ces observations. Dans la deuxième partie, la transition turbulent/laminaire dans un écoulement de Poiseuille plan est étudiée numériquement. On constate que le régime transitoire présente deux comportements distincts : (a) la formation de motifs avec des régions alternées de laminaire et de turbulent (b) des bandes turbulentes indépendantes localisées dans l'espace. En employant une méthodologie de réponse impulsionnelle avec moyenne d'ensemble, une instabilité linéaire de l'écoulement turbulent menant à la formation d'un modèle est mise en évidence. L'évolution du motif avec Re est documentée avec ses propriétés géométriques ainsi que des observables globaux tels que le coefficient de friction. Les statistiques d'ordre élevé des fluctuations turbulentes révèlent un lien continu entre la turbulence sans caractéristiques et le régime transitoire. Afin de mieux comprendre la formation et l'évolution du motif avec Re, un modèle dynamique réduit de l'écoulement de cisaillement est introduit, issu de la littérature. Ce modèle, introduit par P. Manneville, présente une instabilité de Turing comme une extension du modèle de Waleffe. Il est adapté avec des paramètres appropriés et étendu avec l'introduction de l'advection non linéaire et du bruit stochastique. Le modèle capture la phénoménologie du régime transitoire de l'écoulement de Poiseuille plan. Il présente la formation de motifs avec multistabilité, la sélection de la longueur d'onde par le bruit et l'excitabilité à faible Re. Ces résultats sont extrapolés à partir du modèle et validés par rapport aux observations dans les simulations DNS.

Densification and sparsification of temporal networks are attributed to two fundamental mechanisms: a change in the population in the system and/or a change in the chances that nodes in the system are connected. In theory, each of these mechanisms generates a distinctive type of densification scaling, but in reality both types are generally mixed. Here, we develop a Bayesian statistical method to identify the extent to which each of these mechanisms is at play at a given point in time, taking the mixed densification scaling as input. We apply the method to networks of face-to-face interactions of individuals and reveal that the main mechanism that causes densification and sparsification occasionally switches, the frequency of which depending on the social context. The proposed method uncovers an inherent regime-switching property of network dynamics, which will provide a new insight into the mechanics behind evolving social interactions., 12 pages, 6 figures + SI

We study the avalanche statistics observed in a minimal random growth model. The growth is governed by a reproduction rate obeying a probability distribution with finite mean a and variance va. These two control parameters determine if the avalanche size tends to a stationary distribution, (Finite Scale statistics with finite mean and variance or Power-Law tailed statistics with exponent in (1, 3]), or instead to a non-stationary regime with Log-Normal statistics. Numerical results and their statistical analysis are presented for a uniformly distributed growth rate, which are corroborated and generalized by analytical results. The latter show that the numerically observed avalanche regimes exist for a wide family of growth rate distributions and provide a precise definition of the boundaries between the three regimes., 5 pages, 3 figures

Modular robot network architectures are experiencing growing popularity. The problem of automatically reconfiguring a set of modular robots into a given target shape presents a real challenge to distributed computing.Many works on the subject restrict the nature of the constructed target forms. The bolder approaches focus on reducing the number of overall required movements. In this work, we propose a distributed asynchronous self-reconfiguration algorithm allowing to distribute the effort made by each robot to reach the final shape. This makes it possible to extend the life of the network of micro-robots.We compare our TBSR algorithm with C2SR self-reconfiguration algorithm using VisibleSim simulator. The obtained results show that globally TBSR outperforms C2SR except for rare cases. The TBSR algorithm allows reducing the number of required moves up to 17%. Besides, the ability of TBSR to balance the number of moves over the modular robots makes that the maximum number of moves per robot is reduced up to 40%.

We propose a minimal model to simulate long waiting times followed by evolutionary bursts on rugged landscapes. It combines point and inversions-like mutations as sources of genetic variation. The inversions are intended to simulate one of the main chromosomal rearrangements. Using the well-known family of NK fitness landscapes, we simulate random adaptive walks, i.e. successive mutational events constrained to incremental fitness selection. We report the emergence of different time scales: a short-term dynamics mainly driven by point mutations, followed by a long-term (stasis-like) waiting period until a new mutation arises. This new mutation is an inversion which can trigger a burst of successive point mutations, and then drives the system to new short-term increasing-fitness period. We analyse the effect of genes epistatic interactions on the evolutionary time scales. We suggest that the present model mimics the process of evolutionary innovation and punctuated equilibrium., 9 pages, 6 figures, The ALIFE conference 2021

Near- and far-field optical properties of silicon nanostructures under linear polarization (Gaussian beam) and azimuthally or radially focused cylindrical vector beams are investigated by finite-difference time-domain method (FDTD) in Meep open-source software. A python toolkit allowing FDTD simulations in Meep for using those excitation sources is provided. In addition to the preferential excitation of specific electric or magnetic resonance modes as a function of the excitation beam polarization, it is shown in the case of spheroids that shape anisotropy affects the resonance wavelength and the dipole orientation of the magnetic or electric dipole mode. Depending on the spheroid symmetry axis with respect to the electric field orientation, the electric dipole resonance can be split into two peaks, giving quasi-unidirectional scattering, separated by an anapole mode. The optical properties in both far-field (scattering pattern) and near-field (electric and magnetic field hot spots) can be tuned by changing the excitation polarization at a fixed wavelength and selecting properly the spheroid shape and dimensions. These numerical simulations are extended to top-down fabrication-friendly nanostructures such as nanocylinders with circular or elliptic sections.

International audience; Swarm robotics takes inspiration from natural selforganizing systems such as social insects, fish schools or bird flocks, deriving rules to build robotic systems that are more robust, fault-tolerant and flexible than single robots. Over the last two decades, swarm robotics has made significant progress, providing concrete demonstrations on how robot swarms could address complex problems, and also contributing to a better understanding of how complex behaviors emerge in nature. However, as of today, only few published experiments have been able to demonstrate collective behavior in a number of robots that can effectively be compared to the size of biological swarms, and further research is needed before swarm robotics is mature enough to be employed in real world applications. Here we summarize the main lessons learned during the first two decades of research in the field and the main open problems still to be solved, such as the lack of dedicated hardware and software platforms, the need for reliable methodologies for swarm design, or how to make the best out of biological inspiration. We then present what we expect to be the main avenues of research in the future, and discuss how application demands may drive swarm robotics research, transitioning from abstract studies towards an engineering practice. Specifically, we identify precision agriculture, infrastructure inspection and maintenance, defense, space missions, entertainment and precision medicine as the application fields that will most likely drive the science and technology of robot swarms in the next three decades.

International audience; A boost inverter is a versatile architecture able to supply DC or AC outputs from diverse alternative energy sources. The most relevant advantage of this multipurpose converter is to provide boosting and inversion in a single stage. The boost inverter has been studied from design and control perspectives with renewable energy sources. However, the real-time emulation of boost inverters is less widespread in literature. In this context, the main contribution of this paper is proposing an innovative methodology based on the Systems Modeling Language and Petri nets to emulate in real-time power converters using the boost inverter as a case of study. This approach develops real-time Hardware-in-the-Loop models using a graphical language and Petri nets. These graphical and Petri net features allow a formal validation of computational and time constraints before implementation in FPGA. The proposed methodology is also able to transform the developed models to the High Level Specification of Embedded Systems for automatic code generation. Comparison of real-time emulations and experimental results shows a suitable trade-off between the model accuracy and the computational time.; Un boost inverter est une architecture polyvalente capable de fournir des alimentations électriques CC ou CA à partir de diverses sources d'énergie alternatives. L’avantage le plus important de ce convertisseur multifonctionnel est de permettre l’élévation de tension et son ondulation en une seule étape. Le boost inverter a été étudié de sa conception jusqu'à son contrôle avec des sources d'énergie renouvelables. Cependant, l'émulation en temps réel des boost inverters est moins répandue dans la littérature. Dans ce contexte, la contribution principale de cet article propose une méthodologie innovante basée sur le langage de modélisation de systèmes et les réseaux de Petri pour émuler les convertisseurs de puissance en temps réel en utilisant le boost inverter comme étude. Cette approche développe des modèles de Hardware-in-the-Loop en temps réel utilisant un langage graphique et des réseaux de Petri. Ces fonctionnalités graphiques et le réseau de Petri associé permettent une validation formelle des contraintes de calcul et de temps avant la mise en œuvre dans le FPGA. La méthodologie proposée permet également de transformer les modèles développés en spécifications de haut niveau pour systèmes intégrés pour la génération automatique de code. La comparaison des émulations en temps réel et des résultats expérimentaux montre un compromis approprié entre la précision du modèle et le temps de calcul.

International audience; SOMA is a France-Switzerland collaborative project which aims to develop a computing machine with self-organizing properties inspired by the functioning of the brain. The SOMA project addresses this challenge by lying at the intersection of four main research fields, namely adaptive reconfigurable computing, cellular computing, computational neuroscience, and neuromorphic engineering. In the framework of SOMA, we designed the SCALP platform, a 3D array of FPGAs and processors permitting to prototype and evaluate self-organization mechanisms on physical cellular machines.

MAMSO is a tool to easily compare population-based stochastic black-box optimisers, and, more important, to combine strategies coming from them. By using it you can generate many other algorithms, and answer to some questions like 'What happens if I add an estimation of distribution algorithm to PSO?', or ' What happens if I alternate the strategies of Dierential Evolution and the ones of the Adaptive Population-based Simplex?'.

International audience; The dynamics of flow of a dielectric fluid in a vertical cylindrical annulus with a fixed temperature difference and an increasing alternating electric tension has been investigated using a direct numerical simulation (DNS). The temperature difference imposed on the cylindrical surfaces induces a radial temperature gradient perpendicular to the ground gravity which generates a baroclinic flow ascending near the hot surface and descending near the cold one. The electric field coupled with the permittivity gradient generates a dielectrophoretic buoyancy force which is a source of vorticity. Above a critical value of the electric tension, the flow bifurcates to a pattern of stationary columnar vortices. These columnar vortices which characterize the thermoelectric convection are not axially invariant, in contrast with classical Taylor columns. They bifurcate to regular wave patterns and then to spatio-temporal chaotic patterns when the electric field intensity is increased. The flow and temperature fields, the kinetic energy and the enstrophy of thermoelectric convective regimes are computed for different values of the electric tension.

International audience; Investigation of the natural convection in a dielectric fluid confined inside a vertical slot subject to a constant horizontal temperature gradient and a high-frequency electric field was revisited. The thermoelectric coupling produces a dielectrophoretic force which contains a term called electric buoyancy associated with an electric gravity. Similar to the Archimedean buoyancy, this force can induce thermo-electro-convective instability in a stable fluid flow. Including the feedback effect of the electric field on the flow, a three-dimensional linear stability analysis is performed to characterize the instability induced by this force. Different critical modes are found depending on the intensity of the electric field: stationary hydrodynamic modes, oscillatory thermal modes and stationary electric modes. The threshold of the electric modes is independent of the Prandtl number in contrast with the threshold of the hydrodynamic and thermal modes. The obtained results are compared with those of previous studies.

International audience; The termite nest is one of the architectural wonders of the living world, built by the collective action of workers in a colony. Each nest has several characteristic structural motifs that allow for efficient ventilation, cooling, and traversal. We use tomography to quantify the nest architecture of the African termite Apicotermes lamani , consisting of regularly spaced floors connected by scattered linear and helicoidal ramps. To understand how these elaborate structures are built and arranged, we formulate a minimal model for the spatiotemporal evolution of three hydrodynamic fields—mud, termites, and pheromones—linking environmental physics to collective building behavior using simple local rules based on experimental observations. We find that floors and ramps emerge as solutions of the governing equations, with statistics consistent with observations of A. lamani nests. Our study demonstrates how a local self-reinforcing biotectonic scheme is capable of generating an architecture that is simultaneously adaptable and functional, and likely to be relevant for a range of other animal-built structures.

Dynamics underlying epileptic seizures span multiple scales in space and time, therefore, understanding seizure mechanisms requires identifying the relations between seizure components within and across these scales, together with the analysis of their dynamical repertoire. In this view, mathematical models have been developed, ranging from single neuron to neural population.In this study we consider a neural mass model able to exactly reproduce the dynamics of heterogeneous spiking neural networks. We combine the mathematical modelling with structural information from non-invasive brain imaging, thus building large-scale brain network models to explore emergent dynamics and test clinical hypothesis. We provide a comprehensive study on the effect of external drives on neuronal networks exhibiting multistability, in order to investigate the role played by the neuroanatomical connectivity matrices in shaping the emergent dynamics. In particular we systematically investigate the conditions under which the network displays a transition from a low activity regime to a high activity state, which we identify with a seizure-like event. This approach allows us to study the biophysical parameters and variables leading to multiple recruitment events at the network level. We further exploit topological network measures in order to explain the differences and the analogies among the subjects and their brain regions, in showing recruitment events at different parameter values.We demonstrate, along the example of diffusion-weighted magnetic resonance imaging (MRI) connectomes of 20 healthy subjects and 15 epileptic patients, that individual variations in structural connectivity, when linked with mathematical dynamic models, have the capacity to explain changes in spatiotemporal organization of brain dynamics, as observed in network-based brain disorders. In particular, for epileptic patients, by means of the integration of the clinical hypotheses on the epileptogenic zone (EZ), i.e. the local network where highly synchronous seizures originate, we have identified the sequence of recruitment events and discussed their links with the topological properties of the specific connectomes. The predictions made on the basis of the implemented set of exact mean-field equations turn out to be in line with the clinical pre-surgical evaluation on recruited secondary networks.

The intrinsic electrophysiological properties of single neurons can be described by a broad spectrum of models, from the most realistic Hodgkin-Huxley type models with numerous detailed mechanisms to the phenomeno-logical models. The Adaptive Exponential integrate-and-fire (AdEx) model has emerged as a convenient “middle-ground” model. With a low computational cost, but keeping biophysical interpretation of the parameters, it has been extensively used for simulations of large neural networks. However, because of its current-based adaptation, it can generate unrealistic behaviors. We show the limitations of the AdEx model and to avoid them, we introduce the Conductance-based Adaptive Exponential integrate-and-fire model (CAdEx). We give an analysis of the dynamics of the CAdEx model and show the variety of firing patterns it can produce. We propose the CAdEx model as a richer alternative to perform network simulations with simplified models reproducing neuronal intrinsic properties.

The architecture of coordination mechanisms is central to the performance and behaviour of (self-)integrated systems across natural, socio-technical and cyber–physical domains. Multi-scale coordination schemes are prevalent in large-scale systems with bounded performance requirements and limited resource constraints. However, theories to formalise how coordination can be implemented across multi-scale systems are often domain-specific, lacking generic, reusable principles. In these systems, feedback among system entities is a key component to coordination. Building on theories of hierarchies and complexity, in previous work we formalised Multi-Scale Abstraction Feedbacks (MSAF) as a design pattern to describe the architecture of feedback across system scales, highlighting the role played by micro-entities and macro-entities, as well as their interconnections. Focusing on exogenous coordination, this paper refines the MSAF pattern, describing a feedback cycle across scales as one where information flows bottom-up and top-down through five actions: state information communication, state information abstraction, information processing, control information communication, and adaptation from control information. Abstracted state information at each scale is processed with control input from the scale above and provides control input to the scale below. Using the example of distributed task allocation through exogenous coordination, NetLogo simulations are implemented to analyse the impact that different exogenous coordination strategies, and their internal timing configurations, have on resource consumption and on convergence performance. The experimental insights and refinement of the MSAF pattern contribute to a general theory of multi-scale feedback and adaptation. This architectural pattern and associated analysis and evaluation tools are still developing, but offer a concrete basis for further expansion, improvement, and implementation, while addressing questions that are at the core of the behaviour of multi-scale systems.

International audience

International audience

Geometry, calculus and in particular integrals, are too often seen by young students as technical tools with no link to the reality. This fact generates into the students a loss of interest with a consequent removal of motivation in the study of such topics and more widely in pursuing scientific curricula. With this note we put to the fore a simple example of practical interest where the above concepts prove central; our aim is thus to motivate students and to reverse the dropout trend by proposing an introduction to the theory starting from practical applications. More precisely, we will show how using a mixture of geometry, calculus and integrals one can easily share a watermelon into regular slices with equal volume., corrected version

Lorentzian distributions have been largely employed in statistical mechanics to obtain exact results for heterogeneous systems. Analytic continuation of these results is impossible even for slightly deformed Lorentzian distributions, due to the divergence of all the moments (cumulants). We have solved this problem by introducing a `pseudo-cumulants' expansion. This allows us to develop a reduction methodology for heterogeneous spiking neural networks subject to extrinsinc and endogenous noise sources, thus generalizing the mean-field formulation introduced in [E. Montbri\'o et al., Phys. Rev. X 5, 021028 (2015)]., Comment: 10 pages (with supplementary materials), 3 figures

Following up on a previous work we examine a model of transportation network in some source-sink flow paradigm subjected to growth and resource allocation. The model is inspired from plants, and we add rules and factors that are analogous to what plants are subjected to. We study how different resource allocation schemes affect the tree and how the schemes interact with additional factors such as embedding the network into a 3D space and applying gravity or shading. The different outcomes are discussed. PACS. 05.45.-a Nonlinear dynamics and chaos-05.65.+b Self-organized systems, Comment: European Physical Journal B: Condensed Matter and Complex Systems, Springer-Verlag, In press

International audience

The No Free Lunch does not hold for some sets of continuous functions or that are not closed under permutations (not-c.u.p.). On a computer everything is discrete nite so we consider more carefully not-c.u.p. sets of functions. It appears that except for very particular cases, not only the NFL does not hold, not only there is a best algorithm in average, but that an 'ultimate' algorithm does exist. However, when you compare some algorithms on a given benchmark, an algorithm that is the best can be the worst on another one.

A synaptic theory of Working Memory (WM) has been developed in the last decade as a possible alternative to the persistent spiking paradigm. In this context, we have developed a neural mass model able to reproduce exactly the dynamics of heterogeneous spiking neural networks encompassing realistic cellular mechanisms for short-term synaptic plasticity. This population model reproduces the macroscopic dynamics of the network in terms of the firing rate and the mean membrane potential. The latter quantity allows us to gain insight of the Local Field Potential and electroencephalographic signals measured during WM tasks to characterize the brain activity. More specifically synaptic facilitation and depression integrate each other to efficiently mimic WM operations via either synaptic reactivation or persistent activity. Memory access and loading are related to stimulus-locked transient oscillations followed by a steady-state activity in the β-γ band, thus resembling what is observed in the cortex during vibrotactile stimuli in humans and object recognition in monkeys. Memory juggling and competition emerge already by loading only two items. However more items can be stored in WM by considering neural architectures composed of multiple excitatory populations and a common inhibitory pool. Memory capacity depends strongly on the presentation rate of the items and it maximizes for an optimal frequency range. In particular we provide an analytic expression for the maximal memory capacity. Furthermore, the mean membrane potential turns out to be a suitable proxy to measure the memory load, analogously to event driven potentials in experiments on humans. Finally we show that the γ power increases with the number of loaded items, as reported in many experiments, while θ and β power reveal non monotonic behaviours. In particular, β and γ rhythms are crucially sustained by the inhibitory activity, while the θ rhythm is controlled by excitatory synapses., Author summary Working Memory (WM) is the ability to temporarily store and manipulate stimuli representations that are no longer available to the senses. We have developed an innovative coarse-grained population model able to mimic several operations associated to WM. The novelty of the model consists in reproducing exactly the dynamics of spiking neural networks with realistic synaptic plasticity composed of hundreds of thousands of neurons in terms of a few macroscopic variables. These variables give access to experimentally measurable quantities such as local field potentials and electroencephalographic signals. Memory operations are joined to sustained or transient oscillations emerging in different frequency bands, in accordance with experimental results for primate and humans performing WM tasks. We have designed an architecture composed of many excitatory populations and a common inhibitory pool able to store and retain several memory items. The capacity of our multi-item architecture is around 3–5 items, a value similar to the WM capacities measured in many experiments. Furthermore, the maximal capacity is achievable only for presentation rates within an optimal frequency range. Finally, we have defined a measure of the memory load analogous to the event-related potentials employed to test humans’ WM capacity during visual memory tasks.

International audience; Actin is the major cytoskeletal protein of mammal cells that forms microfilaments organized into higher-order structures by a dynamic assembly-disassembly mechanism with crosslinkers. These networks provide the cells with mechanical support, and allow cells to change their shape, migrate, divide and develop a mechanical communication with their environment. The quick adaptation of these networks upon stretch or compression is important for cell survival in real situations. Using atomic force microscopy to poke living cells with sharp tips, we revealed that they respond to a local and quick shear through a cascade of random and abrupt ruptures of their cytoskeleton, suggesting that they behave as a quasi-rigid random network of intertwined filaments. Surprisingly, the distribution of the strength and the size of these rupture events did not follow power-law statistics but lognormal statistics, suggesting that the mechanics of living cells would not fit into selforganized critical systems. We propose a random Gilbert network to model a cell cytoskeleton, identifying the network nodes as the actin filaments, and its links as the actin cross-linkers. We study mainly two versions of avalanches. First, we do not include the fractional visco-elasticity of living cells, assuming that the ruptures are instantaneous, and we observe three avalanche regimes, 1) a regime where avalanches are rapidly interrupted, and their size follows a distribution decaying faster than a power-law; 2) an explosive regime with avalanches of large size where the whole network is damaged and 3) an intermediate regime where the avalanche distribution goes from a power-law, at the critical point, to a distribution containing both 1) and (ii). Then, we introduce a time varying breaking probability, to include the fractional visco-elasticity of living cells, and recover an approximated log-normal distribution of avalanche sizes, similar to those observed in experiments. Our simulations show that the log-normal statistics requires two simple ingredients: a random network without characteristic length scale, and a breaking rule capturing the broadly observed visco-elasticity of living cells. This work paves the way for future applications to large populations of non-linear individual elements (brain, heart, epidemics,. . .) where similar log-normal statistics have also been observed.

Problems to solve nowadays have never been so complex and are continuously increasing in complexity. In this context Systems of Systems (SoS) may be a solution but the study of such systems is far from over. An SoS is a complex system characterized by the particular nature of its components: the latter, which are systems, tend to be managerially and operationally independent as well as geographically distributed. This specific characterization led to re-think research fields of classic systems engineering, such as definition, taxonomy, modeling, architecting, etc. SoS architecting focuses on the way independent components of an SoS can be dynamically structured and can autonomously and efficiently modify their interactions in order to fulfill the goal of the SoS and to cope with the high dynamics of the environment. This paper contributes to the multi-agent and SoS research fields by proposing a new generic SoS model called SApHESIA which considers the SoS main characteristics found in the literature and extends the notion of environment and interactions between component systems. It also proposes a new SoS architecting procedure based on the Adaptive Multi-Agent System (AMAS) approach that advocates full cooperation between all the components of the SoS through the concept of criticality. This criticality is a measure, local to each component, expressing its difficulties to reach its local goals. In this procedure, the SoS architecture evolves to self-adapt to the dynamics of the environment in which it is plunged, while considering the respective local goals of its components. An instantiation of SApHESIA to two distinct cases studies from different domains (logistics and exploratory missions) is done to obtain two SoSs. These two SoS are then coupled to form a new SoS at an upper level. Evaluations of these SoSs show that their cooperation make each of them more efficient.

Many real-world social networks constantly change their global properties over time, such as the number of edges, size and density. While temporal and local properties of social networks have been extensively studied, the origin of their dynamical nature is not yet well understood. Networks may grow or shrink if a) the total population of nodes changes and/or b) the chance of two nodes being connected varies over time. Here, we develop a method that allows us to classify the source of time-varying nature of temporal networks. In doing so, we first show empirical evidence that real-world dynamical systems could be categorized into two classes, the difference of which is characterized by the way the number of edges grows with the number of active nodes, i.e., densification scaling. We develop a dynamic hidden-variable model to formally characterize the two dynamical classes. The model is fitted to the empirical data to identify whether the origin of scaling comes from a changing population in the system or shifts in the connecting probabilities., Comment: 12 pages, 6 figures (plus 7 figures in SI)

International audience; Pyrochlore (A 2 B 2 O 7) is an important, isometric structure-type because of its large variety of compositions and structural derivatives that are generally related to different disordering mechanisms at various spatial scales. The disordering is key to understanding variations in properties, such as magnetic behavior or ionic conduction. Neutron and X-ray total scattering methods were used to investigate the degree of structural disorder in the Ho 2 Ti 2-x Zr x O 7 (x = 0.0-2.0, dx =0.25) solid solution series as a function of the Zr-content, x. Ordered pyrochlores (Fd 3m) disorder to defect fluorite (Fm 3m) via cation and anion disordering. Total scattering experiments with sensitivity to the cation and anion sublattices provide unique insight into the underlying atomic processes. Using simultaneous Rietveld refinement (long-range structure) and small-box refinement PDF analysis (short-range structure), we show that the series undergoes a rapid transformation from pyrochlore to defect fluorite at x ~ 1.2, while the short-range structure exhibits a linear increase in a local weberite-type phase, C222_1 , over the entire composition range. Enthalpies of formation from the oxides DH f;ox determined using high temperature oxide melt solution calorimetry support the structural data and provide insight into the effect of local ordering on the energetics of disorder. The measured enthalpies of mixing are negative and are fit by a regular solution parameter of W =-31.8 +/- 3.7 kJ mol-1. However, the extensive short-range ordering determined from the structural analysis strongly suggests that the entropies of mixing must be far less positive than implied by the random mixing of a regular solution. We propose a local disordering scheme involving the pyrochlore 48f to 8a site oxygen Frenkel defect that creates 7-coordinated Zr sites contained within local weberite-type coherent nanodomains. Thus, the solid solution is best described as a mixture of two phases, with the weberite-type nanodomains triggering the long-range structural transformation to defect fluorite after accumulation above a critical concentration (50% Ti replaced by Zr).

International audience; One contribution of 14 to a theme issue 'Multi-scale analysis and modelling of collective migration in biological systems'. Collective migration has become a paradigm for emergent behaviour in systems of moving and interacting individual units resulting in coherent motion. In biology, these units are cells or organisms. Collective cell migration is important in embryonic development, where it underlies tissue and organ formation, as well as pathological processes, such as cancer invasion and metastasis. In animal groups, collective movements may enhance individuals' decisions and facilitate navigation through complex environments and access to food resources. Mathematical models can extract unifying principles behind the diverse manifestations of collective migration. In biology, with a few exceptions, collective migration typically occurs at a 'mesoscopic scale' where the number of units ranges from only a few dozen to a few thousands, in contrast to the large systems treated by statistical mechanics. Recent developments in multi-scale analysis have allowed linkage of mesoscopic to micro-and macroscopic scales, and for different biological systems. The articles in this theme issue on 'Multi-scale analysis and modelling of collective migration' compile a range of mathematical modelling ideas and multi-scale methods for the analysis of collective migration. These approaches (i) uncover new unifying organization principles of collective behaviour, (ii) shed light on the transition from single to collective migration, and (iii) allow us to define similarities and differences of collective behaviour in groups of cells and organisms. As a common theme, self-organized collective migration is the result of ecological and evolutionary constraints both at the cell and organismic levels. Thereby, the rules governing physiological collective behaviours also underlie pathological processes, albeit with different upstream inputs and consequences for the group. This article is part of the theme issue 'Multi-scale analysis and modelling of collective migration in biological systems'.

Most studies of dewetting fronts in 3D with a “corner formation”, as happens behind a drop sliding down an incline are based on a generalisation of Voinov theory, with (at least implicitly) a slip length at small scale. I here first examine what happens, if instead of considering a free slip at small scale, one admits a non-zero additional line friction of microscopic origin. Concerning the selection of cone angles, I show that most features of the model are unchanged, except that the “slip length” must be replaced in the equations with an “effective” cut off that can become apparently unphysically small. I suggest that these results could explain problematical cut-offs in the hydrodynamical modelling observed recently by Winkels et al. on water drops [K.G. Winkels, I.R. Peters, J. Snoeijer, F. Evangelista, M. Riepen, A. Daerr, L. Limat, Eur. Phys. J. Special Topics 192, 195 (2011)]. The sole difficulty with this interpretation is the law ruling the radius of curvature of the corner tip at small scale, which remains unsatisfactory. I suggest that evaporation of the liquid should also be considered at these very small scales and propose a preliminary “toy model” to take this effect into account. The orders of magnitude are better recovered without changing the structure of the equations developed initially for “classical” wetting dynamics with silicon oil drops.

Group-living organisms that collectively migrate range from cells and bacteria to human crowds, and include swarms of insects, schools of fish, and flocks of birds or ungulates. Unveiling the behavioural and cognitive mechanisms by which these groups coordinate their movements is a challenging task. These mechanisms take place at the individual scale and can be described as a combination of interactions between individuals and interactions between these individuals and the physical obstacles in the environment. Thanks to the development of novel tracking techniques that provide large and accurate datasets, the main characteristics of individual and collective behavioural patterns can be quantified with an unprecedented level of precision. However, in a large number of studies, social interactions are usually described by force map methods that only have a limited capacity of explanation and prediction, being rarely suitable for a direct implementation in a concise and explicit mathematical model. Here, we present a general method to extract the interactions between individuals that are involved in the coordination of collective movements in groups of organisms. We then apply this method to characterize social interactions in two species of shoaling fish, the rummy-nose tetra ( Hemigrammus rhodostomus ) and the zebrafish ( Danio rerio ), which both present a burst-and-coast motion. From the detailed quantitative description of individual-level interactions, it is thus possible to develop a quantitative model of the emergent dynamics observed at the group level, whose predictions can be checked against experimental results. This method can be applied to a wide range of biological and social systems. This article is part of the theme issue ‘Multi-scale analysis and modelling of collective migration in biological systems’.

With the advent of deep learning supported by substantial technological advances, the Artificial Intelligence has taken a decisive step towards the automation of large dimension tasks. Reinforcement learning has been revolutionized thanks to new representation concepts introduced by deep learning. However, the extension of this paradigm application to the real world has triggered new challenges of generalization and optimization associated with higher level of tasks non-stationarity. In this thesis, we are interested in the recent methodological evolution of machine learning towards meta-learning in order to remedy the deep learning limits. The proposed approach is built on the basis of a Markovian formulation gradually evolving along 2 axes of improvement. In terms of learning robustness, we integrate in the policy gradient expression, the advantage function estimated by a generalized version of temporal difference learning. Regarding the generalization capacity, we implement a gradient meta-optimizer, learning the evaluations of the base-level algorithm across tasks. The resulting generalizable model showed a rapid adaptation to new contexts of urban autonomous driving.; Avec l'avènement de l'apprentissage profond, l'intelligence artificielle a franchi un pas décisif vers l'automatisation des tâches de grande dimensionnalité. L'apprentissage par renforcement a été révolutionné grâce aux nouveaux concepts de représentation profonde. Toutefois, l'extension de l'application de ce paradigme vers la sphère du monde réel a engendré des nouveaux défis de généralisation et d'optimisation face à la non-stationnarité des tâches. Dans cette thèse, nous nous intéressons à l'évolution méthodologique récente de l'apprentissage automatique vers le méta-apprentissage afin de remédier aux limites de l'apprentissage profond. L'approche proposée est construite sur la base d'une formulation Markovienne évoluant graduellement selon 2 axes d'amélioration. Au niveau de la robustesse de l'apprentissage, nous intégrons dans l'expression du gradient de la politique la fonction avantage estimée par une version généralisée de l'apprentissage par différence temporelle. Concernant la capacité de généralisation, nous implémentons un méta-optimiseur de type gradient apprenant les évaluations de l'algorithme de base à travers les tâches. Le modèle généralisable obtenu a montré une adaptation rapide aux nouveaux contextes de conduite autonome en milieu urbain.

The well-known Vicsek model describes the dynamics of a flock of self-propelled particles (SPPs). Surprisingly, there is no direct measure of the chaotic behavior of such systems. Here, we discuss the dynamical phase transition present in Vicsek systems in light of the largest Lyapunov exponent (LLE), which is numerically computed by following the dynamical evolution in tangent space for up to two million SPPs. As discontinuities in the neighbor weighting factor hinder the computations, we propose a smooth form of the Vicsek model. We find a chaotic regime for the collective behavior of the SPPs based on the LLE. The dependence of LLE with the applied noise, used as a control parameter, changes sensibly in the vicinity of the well-known transition points of the Vicsek model., 7 pages, 16 equations, 6 figures

Autonomic systems manage resources to achieve high-level goals with minimum human intervention, by continuously adapting based on a model of themselves and of their environment. Adaptation requires managing runtime uncertainties produced by changes within a dynamic environment, by modeling unexpected situations that were not considered at design time. This challenge becomes even more critical in a self-integration context where autonomic systems, traditionally engineered systems, and human organizations, influence each other in a shared environment. In this paper, we propose a cognitive control system that aims to manage runtime uncertainty in self-integrating autonomic systems. The cognitive control system autonomously extends the existing modeling language, by generating knowledge to represent unexpected situations as they are encountered at runtime, and reasons on the resulting model, in order to determine the required adaptation knowledge. As proof-of-concept validation, we present a scenario supported by an implementation of our proposal, illustrating the capabilities of the cognitive control system in the smart home domain. The obtained results set the basis for further studying the modeling language for describing the adaptation knowledge, with the objective of enriching the collaboration capabilities of autonomic systems in a self-integration context as future work.

International audience

For a N nodes graph, the complexity of the proposed method, called NK-method, is O (N 2) in number of qubits and O (N 4) in number of gates. It works for any K, but needs N 2 − N K + 1 qubits (and N K + 1 classical bits). As this number is quite high, we also propose a Q-method that requires significantly less qubits, but whose quantum circuit is far more complicated to build when K increases. For the NK-method we give a general Qiskit code that builds and simulates the quantum circuit. For the Q-method we just give a code for the 3-coloring. In these codes we use only gates whose unitary matrices are real. These codes can find all solutions that require at most K colours.

Indirect, or dipolar excitons are bosonic quasi-particles in semiconductors composed of spatially separated but still Coulomb-bound electron and hole. They have long lifetime and can travel over large distances before recombination, offering a unique system that can be both optically active and electrically controllable. It is suitable for studies of fundamental properties of light and matter and for the development of conceptually new excitonic devices. Excitons in polar GaN quantum wells can be considered as naturally indirect excitons, because of the strong built-in electric field in the growth direction. This dissertation describes an experimental realization and investigation of indirect excitons engineered in GaN/(Al,Ga)N heterostructures, and the collective states that these can form. The main results of this work are (i) the demonstration of the in-plane confinement and cooling of indirect excitons, when trapped in the electrostatic potential created by semitransparent electrodes of various shapes carefully designed and deposited on the sample surface, this is a prerequisite for studies of the complex phase diagram of these dipolar bosons at low temperatures ; (ii) The proof-of-principle for electrical control of the indirect exciton densities and fluxes in the plane of the heterostructure, which opens attractive prospects for realization of excitonic devices ; (iii) the first points on the dipolar boson phase diagram, providing first evidence of the density-induced correlated state (dipolar liquid) and dissociation (Mott transition) of the indirect excitons in GaN/(Al,Ga)N heterostructures.; Les excitons indirects dipolaires, sont des quasi-particules bosoniques dans les semi-conducteurs, composés d'un électron et d'un trou spatialement séparés mais toujours liés par interaction coulombienne. Leur grande durée de vie radiative et leur capacité à se déplacer sur de grandes distances avant leur recombinaison en font un système unique qui peut être à la fois optiquement actif mais également électriquement contrôlable. Ce système permet l'étude des propriétés fondamentales de la lumière et de la matière, mais aussi le développement de dispositifs excitoniques conceptuellement nouveaux. Les excitons dans les puits quantiques polaires en nitrure de gallium (GaN) peuvent être considérés comme des excitons naturellement indirects, en raison du fort champ électrique intrinsèque existant dans la direction de croissance cristalline. Cette thèse est consacrée à l'étude expérimentale et une étude des excitons indirects dans des hétérostructures GaN/(Al,Ga)N, ainsi que les états collectifs, depuis la conception et la fabrication jusqu'à la spectroscopie optique de leurs états collectifs.Les principaux résultats de ce travail sont (i) la démonstration du confinement spatial dans le plan et du refroidissement des excitons indirects, lorsqu'ils sont piégés dans le potentiel électrostatique créé par des électrodes semi-transparentes de géométries diverses soigneusement conçues et déposées sur la surface de l'échantillon, qui est une condition préalable à l'étude du diagramme de phase complexe de ces bosons dipolaires à basse température ; (ii) la preuve de principe du contrôle électrique des densités et des flux d'excitons indirects dans le plan du puits quantique. Cela ouvre des perspectives intéressantes pour la réalisation de dispositifs excitoniques ; (iii) les premiers points sur le diagramme de phase des bosons dipolaires, fournissant une première mise en évidence non seulement de l'existence d'un phase fortement corrélée résultant des corrélations induites à forte densité (phase de liquide dipolaire) mais aussi la dissociation (transition de Mott) de ces excitons indirects dans les hétérostructures GaN/(Al,Ga)N.

Indirect, or dipolar excitons are bosonic quasi-particles in semiconductors composed of spatially separated but still Coulomb-bound electron and hole. They have long lifetime and can travel over large distances before recombination, offering a unique system that can be both optically active and electrically controllable. It is suitable for studies offundamental properties of light and matter and for the development of conceptually new excitonic devices. Excitons in polar GaN quantum wells can be considered as naturally indirect excitons, because of the strong built-in electric field in the growth direction. This dissertation describes an experimental realization and investigation of indirect excitons engineered in GaN/(Al,Ga)N heterostructures, and the collective states that these can form. The main results of this work are (i) the demonstration of the in-plane confinement and cooling of indirect excitons, when trapped in the electrostaticpotential created by semitransparent electrodes of various shapes carefully designed and deposited on the sample surface, this is a prerequisite for studies of the complex phase diagram of these dipolar bosons at low temperatures ; (ii) The proof-of-principle for electrical control of the indirect exciton densities and fluxes in the plane of theheterostructure, which opens attractive prospects for realization of excitonic devices ; (iii) the first points on the dipolar boson phase diagram, providing first evidence of the density-induced correlated state (dipolar liquid) and dissociation (Mott transition) of the indirect excitons in GaN/(Al,Ga)N heterostructures.; Les excitons indirects dipolaires, sont des quasi-particules bosoniques dans les semiconducteurs, composés d’un électron et d’un trou spatialement séparés mais toujours liés par interaction coulombienne. Leur grande durée de vie radiative et leur capacité à se déplacer sur de grandes distances avant leur recombinaison en font un système uniquequi peut être à la fois optiquement actif mais également électriquement contrôlable. Ce système permet l’étude des propriétés fondamentales de la lumière et de la matière, mais aussi le développement de dispositifs excitoniques conceptuellement nouveaux. Les excitons dans les puits quantiques polaires en nitrure de gallium (GaN) peuvent êtreconsidérés comme des excitons naturellement indirects, en raison du fort champ électrique intrinsèque existant dans la direction de croissance cristalline. Cette thèse est consacrée à l’étude expérimentale et une étude des excitons indirects dans des hétérostructures GaN/(Al,Ga)N, ainsi que les états collectifs, depuis la conception et la fabricationjusqu’à la spectroscopie optique de leurs états collectifs. Les principaux résultats de ce travail sont (i) la démonstration du confinement spatial dans le plan et du refroidissement des excitons indirects, lorsqu’ils sont piégés dans le potentiel électrostatique créé par des électrodes semi-transparentes de géométries diverses soigneusement conçues et déposées sur la surface de l’échantillon, qui est une condition préalable à l’étude du diagramme dephase complexe de ces bosons dipolaires à basse température ; (ii) la preuve de principe du contrôle électrique des densités et des flux d’excitons indirects dans le plan du puits quantique. Cela ouvre des perspectives intéressantes pour la réalisation de dispositifs excitoniques ; (iii) les premiers points sur le diagramme de phase des bosons dipolaires,fournissant une première mise en évidence non seulement de l’existence d’un phase fortement corrélée résultant des corrélations induites à forte densité (phase de liquide dipolaire) mais aussi la dissociation (transition de Mott) de ces excitons indirects dans les hétérostructures GaN/(Al,Ga)N.

Controlled self-assembly of polymeric systems behave according to their composition and physical-chemical properties. Inclusion of a guest molecule may influence interactions and response to the media and to other surfaces and bodies, for example, cells. Hence, loading the polymeric system with curcumin (CUR), a natural and anticancer drug, confer interesting features to the carrier material. Micelle formation was achieved, in aqueous solution of carbohydrate-based block copolymer consisting in maltoheptaose-block-polyisoprene (MH-b-PI3.8kDa), by adding a large amount of water (a selective solvent for maltoheptaose, MH) into a solution of well-dissolved MH-b-PI3.8kDa (THF/H2O 9:1 (% w/w)]. Morphology and size (ca. 89 nm) of these glyco-nanoparticles were characterized using light scattering (static and dynamic modes – SLS and DLS) complemented by nanoparticle tracking analysis (NTA), transmission electron microscopy (TEM) and atomic force microscopy (AFM). The lyophilization efficiency was confirmed by measuring the relation between final and initial diameters (Sf/Si). CUR was successfully loaded into the nanoparticles with an entrapment efficiency of ca. 70%. CUR-charged nanoparticles showed stability into simulated gastric and intestinal fluids. Such micelles served as drug delivery system to carry CUR and presented a great role in wiping out unhealthy cells from many sorts.

Currently, energy management within buildings is essentialto participate in the green transition. To this end, buildingsare increasingly equipped with sensors to assist the buil-ding manager. But the heterogeneity and the large amountof data generated makes this task quite difficult. The SAND-MAN multi-agent system, described in this paper, aims toassist in the automatic detection, in real time, of seve-ral types of anomalies using raw and heterogeneous data.SANDMAN features a semi-supervised learning by consi-dering some feedbacks from an expert in the field. The re-sults show that SANDMAN, after a learning phase, detectsthe different types of anomalies, is resistant to noise and isscalable., Actuellement, la gestion de l'énergie au sein des bâtiments est essentielle pour participer à la transition écologique. Pour cela, les bâtiments sont de plus en plus équipés de capteurs pour aider le gestionnaire du bâtiment. Mais l'hétérogénéité et la grande quantité de données générées rend sa tâche assez ardue. Le système multi-agent SANDMAN, décrit dans cet article, a pour objectif d'aider à la détection automatique en temps réel, de plusieurs types d'anomalies en utilisant des données brutes et hétérogènes. SANDMAN fait un apprentissage semi-supervisé en considérant quelques avis d'un expert du domaine. Les résultats montrent que SANDMAN, après une phase d'apprentissage, détecte les différents types d'anomalies, est résistant au bruit et passe l'échelle. Mots Clef Système multi-agent auto-adaptatif, Détection d'anomalies, Bâtiments intelligents

International audience; Effective upon acceptance for publication, copyright (including all rights hereunder and the right to obtain copyright registration in the name of the copyright owner), the above article is hereby transferred throughout the world and for the full term and all extensions and renewals thereof, to EUROPEAN CONTROL ASSOCIATION (EUCA) This transfer includes the right to adapt the article for use in conjunction with computer systems and programs, including reproduction or publication in machine-readable form and incorporation in retrieval systems. Some papers accepted for ECC are considered for subsequent publication in the European Journal of Control (EJC). If the signing author has not been informed in writing within three months of the closing date of the conference that the paper is being so considered, he and his co-authors are free to submit the paper elsewhere, provided that full acknowledgement is given in any subsequent publication to the original ECC Proceedings CD-ROM and to the conference. Note: All copies, paper or electronic, or other use of the information must include an indication of the EUCA copyright and a full citation of the ECC Proceedings source.

Theta-nested gamma oscillations have been reported in many areas of the brain and are believed to represent a fundamental mechanism to transfer information across spatial and temporal scales. In a series of recent experiments in vitro it has been possible to replicate with an optogenetic theta frequency stimulation several features of cross-frequency coupling among theta and gamma rhythms observed in behaving animals. In order to reproduce the main findings of these experiments we have considered a new class of neural mass models able to reproduce exactly the macroscopic dynamics of spiking neural networks. In this framework, we have examined two set-ups able to support collective gamma oscillations: the pyramidal interneuronal network gamma (PING) and the interneuronal network gamma (ING). In both set-ups we observe the emergence of theta-nested gamma oscillations by driving the system with a sinusoidal theta-forcing in proximity of a Hopf bifurcation. These mixed rhythms display always phase amplitude coupling. However 2 different types of nested oscillations can be identified: one characterized by a perfect phase locking between theta and gamma rhythms, corresponding to an overall periodic behaviour; another one where the locking is imperfect and the dynamics is quasi-periodic or even chaotic. From our analysis it emerges that the locked states are more frequent in the ING set-up. In agreement with the experiments, we find theta-nested gamma oscillations for forcing frequencies in the range [1:10] Hz, whose amplitudes grow proportionally to the forcing one and which are clearly modulated by the theta phase. At variance with experimental findings, the gamma-power peak does not shift to higher frequencies by increasing the theta frequency. This effect can be obtained, in or model, only by incrementing, at the same time, also the noise or the forcing amplitude., 28 pages, 14 figures

A version in English is available here: http://cormas.cirad.fr/pdf/CormasforIsaga2015.pdfThis paper presents the new functionalities of Cormas, a generic agent- based modeling (ABM) platform dedicated to common-pool resource management. As free software, Cormas is used by an international community of researchers willing to understand the relationships between societies and their environment. It is intended to facilitate the design of ABM as well as the monitoring and analysis of simulation scenarios. Cormas has taken an innovative direction oriented toward the collective design of models and interactive simulation. In accordance with the principles of participatory methods and serious games, many experiments devel- oped with Cormas combine two layers of complexity: the natural dynamic of the system, simulated by the computer, and the stakeholders’ interactions and decisions played by the actors. Between these two extremes, a range of intermediate situa- tions exists where some decisions are human and others are computer specified. The term hybrid agent simulation covers these intermediary situations. The main idea is to enable the stakeholders to interact with the execution of a simulation by modi- fying the behavior of the agents and the way they use the resources. Thus, it is possible to collectively explore scenarios to better understand how a desired situation may be reached. This may feed back into the collective design of the model. As our intention is to involve more deeply the stakeholders into the modeling process, it is necessary to have an easily changeable tool to act on the simulation and to modify the conceptual model on the fly. After having explained the purpose and the philosophy of the companion modeling, this chatper presents how the Cormas functionalities (asymmetry of information, agent manipulation, modification of behavior, stepping back, and distributed simulation) are put into practice through three experiments with stakeholders facing actual environmental challenges.; Resumen: Esta nota de investigación presenta las nuevas características de Cormas, una plataforma de modelización multiagente dedicada a la gestión de recursos renovables. Cormas es un software libre que tiene como objetivo diseñar fácilmente un Sistema Multi-Agente (SMA) y analizar escenarios. Recientemente Cormas ha adoptado una dirección innovadora orientada al diseño colectivo de modelos y simulación interactiva. Estas simulaciones híbridas mezclan las decisiones tomadas por las partes interesadas y otras por el modelo. Esto le permite interactuar con una simulación cambiando el comportamiento de los agentes y cómo ellos utilizan los recursos. Por lo tanto, es posible explorar colectivamente escenarios de mediano y largo plazo para comprender mejor cómo lograr una situación deseada. A su vez, esto permite una revisión colectiva del modelo conceptual.; Une version en français est disponible ici : https://hal.archives-ouvertes.fr/hal-01385722Ce document présente les fonctionnalités de Cormas, une plate-forme de modélisation multi-agent dédiée à la gestion des ressources renouvelables. Logiciel libre, Cormas vise à concevoir facilement un SMA et analyser des scénarios. Aujourd’hui, Cormas a pris une direction novatrice orientée vers la conception collective de modèles et la simulation interactive. Ces simulations hybrides mixent des décisions prises par les joueurs et d'autres par le modèle. Cela permet d'interagir avec une simulation en modifiant le comportement des agents et la façon dont ils utilisent les ressources. Ainsi, il est possible d'explorer collectivement des scénarios à moyen et long terme afin de mieux comprendre comment atteindre une situation souhaitée. En retour, cela permet de réviser collectivement le modèle conceptuel. Après avoir expliqué la philosophie de la modélisation d'accompagnement, ce document présente la façon dont les fonctionnalités de Cormas sont appliquées à travers trois expériences.