Search

International audience

International audience; Aims: Electroencephalography (EEG) provides insights for monitoring brain activity. Unfortunately, although scalp EEG can provide clinically useful information, its spatial resolution is low, making it difficult to determine the specific brain locations involved. However, noninvasive EEG-based source localization methods computationally estimate the locations, orientations, and strengths of the electric currents in the brain generating scalp potentials. Using anatomically faithful 3D head models, these approaches enable the accurate localization of epileptogenic foci, improve surgical efficacy, contribute to the improvement of critical care monitoring, and provide guidance for patient-tailored treatment. Our goal is to determine whether our method of source localization, called "Renormalization", is helpful for obtaining real-time feedback.Methods: Our approach is based on the lead field theory. Unlike other methods, ours computes the so-called “visibility field” of the recording system through a “3D Renormalisation” process. Then, to determine the spatial location and the strength of sources producing the potential scalp maps, this field is used to calculate a unique solution to the neuroelectric inverse problem.Results: We used the “Renormalisation” method with an “ANT Neuro eego sports” system and an anatomically realistic 3D head model. The visibility field characterizes the regions in which the recording system concentrates its source identification ability. We used it to provide an accurate real-time spatial mapping of electrical brain activity and characterization (localization, strength, orientation) of localized and distributed sources.Conclusions: By its specificities, we expect that our method will contribute to a better understanding of brain functions and improve the care of brain-injured patients.

International audience; Non-invasive EEG-based Brain-Computer Interfaces (BCIs) seek to translate the brain’s electrical activity into computer commands. To do so, one needs to interpret the measured electrode potentials by extracting the most relevant features and classifying the patterns of interest. To accomplish feature extraction, a variety of time-domain, frequency-domain, and time-frequency domain analysis methods are commonly used, followed by dimensionality reduction algorithms. Eventually, classification is generally carried out by applying a Machine Learning classifier. Alternatively, feature extraction and classification steps can be regarded as a whole within the Riemannian framework. This framework is based on the Riemannian geometry, allowing easy manipulation of covariance matrices and characterized by its simplicity, accuracy, robustness, and transfer learning capabilities.Even though EEG-based BCIs have excellent temporal resolution and usability, they suffer from poor spatial resolution. They are also more subject to noise from eye or muscle movements. Source reconstruction methods are known to improve the spatial resolution of scalp EEG. Consequently, they could be a way to improve classification accuracy.The Neuroelectromagnetic Inverse Problem aims to reconstruct the current distribution of electric sources within the brain. This involves potentials measured on the scalp and a source-receptor model to describe the electric propagation through the head tissues. To get a unique and stable solution to this ill-posed Source Term Estimation (STE) problem, we propose to use our ”Renormalization” method. This approach delivers a real-time solution based on both the lead field computed from the model and the innovative concept of ”visibility”.The purpose of the present work is to determine whether the ”Renormalization” approach can be helpful in the field of non-invasive EEG-based BCIs. Specifically, we investigate how far source reconstruction, as we propose, could complement existing BCI methodologies.

International audience; The statistics of concentration data, measured during unknown atmospheric dispersion events, are not fully understood although they are required in modeling, assessment, uncertainty analysis, and information fusion. The concentrations measured over a field of monitoring network is regarded as a vector which contains both magnitude and direction. Traditional statistics (mean, standard deviations, etc.) based on magnitude of concentration data summarize data properties but are limited in characterizing variations in data and their modeling quality. Comparatively, directions are efficient in providing valuable information to address these issues. Here we propose a statistical framework which transforms concentration measurements into directions projected on a hypersphere and analyzes their orientation and distribution. The directional data measured in identical conditions are expected to be rotationally symmetric around its principal axis and follow Watson distribution. The clustering parameter of Watson distribution measures tightness of directional data and, thus, can measure indirectly variations in observed data. It is shown that the clustering parameter is able to summarize an overall variation in data and modeling quality of data in a dispersion trial. The study analyzes real data taken from continuous release experiments, called “Fusion Field Trials,” conducted at Dugway Proving Ground, Utah, United States.

International audience; In case of a release of a hazardous material (e.g., a chemical or a biological agent) in the atmosphere, estimation of the source from concentration observations (provided by a network of sensors) is a challenging inverse problem known as the atmospheric source term estimation (STE) problem. This study emphasizes a method, known in the literature as the renormalization inversion technique, for addressing this problem. This method provides a solution that has been interpreted as a weighted minimal norm solution and can be computed in terms of a generalized inverse of the sensitivity matrix of the sensors. This inverse is constructed by using an appropriate diagonal weight matrix whose components fulfill the so-called renormalizing conditions. The main contribution of this paper is that it proposes a new compact algorithm (it requires less than 15 lines of MATLAB code) to obtain, in a fast and efficient way, those optimal weights. To show that the algorithm, based on the properties of the resolution matrix, matches the requirements of emergency situations, analysis of the computational complexity and memory requirements is included. Some numerical experiments are also reported to show the efficiency of the algorithm.; En cas de rejet de matières dangereuses (p. ex. un produit chimique ou un agent biologique) dans l'atmosphère, la caractérisation de la source d’émission à partir de mesures de concentration (fournies par un réseau de capteurs) nécessite la résolution d’un problème inverse, connu sous le nom de problème d'Estimation du Terme Source (STE). Cette étude se concentre sur une des méthodes de résolution de ce problème, connue dans la littérature sous le nom de « méthode de renormalisation ». Cette dernière fournit une solution de norme minimale pondérée qui peut être obtenue après avoir calculé un inverse généralisé de la matrice de sensibilité des capteurs. Cet inverse fait intervenir une matrice de pondération diagonale dont les composants doivent vérifier les conditions dites de renormalisation. La principale contribution de cet article est qu'il propose un nouvel algorithme, rapide, efficace et particulièrement compact (il comporte moins de 15 lignes de code MATLAB), qui permet d’obtenir cette matrice. Pour montrer que l'algorithme, basé sur la notion de matrice de résolution, est performant et peut être utilisé en situation opérationnelle, sa complexité informatique globale est étudiée. Enfin, son utilisation est illustrée sur des cas réels.

International audience; This article highlights an optimization-based approach to estimate the main characteristics (i.e., location and intensity) that describe the source of a release of a hazardous material (e.g., a chemical or biological agent) in the atmosphere. The method, which uses, as input data, concentration measurements provided by a network of sensors (and some additional meteorological data), is first formalized in the general framework of parametric source term estimation inverse problems. Some simple and new geometric interpretations of the method are proposed. Next, it is shown that the approach can also provide a realistic understanding of the uncertainties in the estimation of the characteristics of the source. Then, a strategy for an optimal deployment of mobile sensors in hazardous materials operations is proposed. To conclude, the source estimation approach and the network design strategy are evaluated with data from a wind tunnel experiment conducted at the Environmental Flow Research Center at the University of Surrey, UK.; Cet article présente une approche, basée sur l’optimisation, qui permet d'estimer les principales caractéristiques (c.-à-d. l'emplacement et l'intensité) d’une source de rejet de matières dangereuses (p. ex., un agent chimique ou biologique) dans l'atmosphère. Cette approche, qui utilise comme données d'entrée des mesures de concentration fournies par un réseau de capteurs (accompagnées de données météorologiques), est décrite dans le cadre général des problèmes inverses paramétriques. Plusieurs interprétations géométriques des résultats sont également présentées. Nous montrons ensuite que cette approche peut également être utilisée pour évaluer l’incertitude inhérente au processus d'estimation d’un terme source. Enfin, une stratégie pour le déploiement optimal, en situation opérationnelle, d’un réseau de capteurs mobiles est proposée. En conclusion, la technique d'estimation du terme source et la stratégie de déploiement d’un réseau mobile sont évaluées à l'aide de données issues d'une expérience réalisée en soufflerie à l’Environmental Flow Research Centre de l’Université de Surrey, UK.

International audience; The study highlights a theoretical comparison and various interpretations of a recent inversion technique, called renormalization, developed for the reconstruction of unknown tracer emissions from their measured concentrations. The comparative interpretations are presented in relation to the other inversion techniques based on principle of regularization, Bayesian, minimum norm, maximum entropy on mean, and model resolution optimization. It is shown that the renormalization technique can be interpreted in a similar manner to other techniques, with a practical choice of a priori information and error statistics, while eliminating the need of additional constraints. The study shows that the proposed weight matrix and weighted Gram matrix offer a suitable deterministic choice to the background error and measurement covariance matrices, respectively, in the absence of statistical knowledge about background and measurement errors. The technique is advantageous since it (i) utilizes weights representing a priori information apparent to the monitoring network, (ii) avoids dependence on background source estimates, (iii) improves on alternative choices for the error statistics, (iv) overcomes the colocalization problem in a natural manner, and (v) provides an optimally resolved source reconstruction. A comparative illustration of source retrieval is made by using the real measurements from a continuous point release conducted in Fusion Field Trials, Dugway Proving Ground, Utah.

International audience

International audience

International audience; This study describes a methodology combining a recently proposed renormalization inversion technique with a building-resolving computational fluid dynamics (CFD) approach for source retrieval in the geometrically complex urban regions. It presents the first application of the renormalization inversion approach to estimate an unknown continuous point release in real situations at an urban scale. The renormalization inversion approach is based on an adjoint source-receptor relationship and is purely deterministic in nature. The source parameters (i.e., source location and release rate) are reconstructed from a finite set of point measurements of concentration acquired from some sensors and the adjoint functions computed from a CFD model fluidyn-PANACHE that is able to represent the geometric and flow complexity inherent in the urban regions. The inversion procedure is evaluated for a point source reconstruction using measurements from the Mock Urban Setting Test (MUST) field experiment. Source reconstructions are performed for 20 trials of the MUST experiment of a continuous point release in an idealized urban geometry consisting of a regular array of shipping containers. The steady state flow fields are computed by solving the three-dimensional Reynolds-averaged Navier-Stokes equations by using a finite volume scheme. Then, in each MUST trial adjoint functions are obtained and used for the source identification. Inversion results are presented with both synthetic and real measurements in various atmospheric stabilities varying from neutral to stable and very stable conditions. With real concentration measurements, the point source is retrieved within an average Euclidean distance of 14.6 m from the actual source location. The estimated source intensity is overpredicted by an average factor of 1.37 of the true release rate. In a posterior uncertainty analysis with 10% random noise in measurements, it is demonstrated that standard deviation in the location error and release strength, respectively, varies by 5.22 m and ∼21% from their mean value for all 20 trials. A sensitivity analysis shows that the use of nonzero measurements helps in reducing the uncertainties involved in the source reconstruction. The source reconstruction results in various stability conditions exhibit the reliability of the renormalization inversion methodology coupled with the CFD approach in an urban area. The present methodology can be used by emergency regulators as a tool to detect the unknown accidental or deliberated releases in the complex urban environments.

International audience; In low-wind stable conditions, source reconstruction becomes sensitive to the height of the release and receptors measuring the concentrations. Therefore, an inversion technique, based on theory of renormalization, is proposed to identify the height of the release along with its location on the ground and emission rate in the atmosphere. The technique utilizes the information derived from the geometry of the monitoring networks. The retrieval algorithm is evaluated with real observations taken from Idaho diffusion experiment, conducted in low wind stable conditions.

International audience; Reconstruction of unknown atmospheric releases using measured concentrations is an ill-posed inverse problem. Due to insufficient measurements and dispersion model uncertainties, reliable interpretation of a retrieved source is limited by lack of resolution, nonuniqueness, and instability in the inverse solution. The study presents an optimality analysis, in terms of resolution, stability, and reliability, of an inverse solution given by a recently proposed inversion technique, called as renormalization. The inversion technique is based on an adjoint source-receptor framework and construction of a weight function which interprets a priori information about the unknown release apparent to the monitoring network. The properties of weight function provide a perfect data resolution, maximum model resolution, and minimum variance (or stability) for the retrieved source. The reliability of the retrieved source is interpreted in view of the information derived from the geometry of the monitoring network. The inversion technique and resolution features are evaluated for a point source reconstruction using measurements from a recent dispersion experiment (Fusion Field Trials 2007) conducted at Dugway Proving Ground, Utah. With the real measurements, the point release is reconstructed within an average distance of 23m from the true release where the average distance of the nearest receptor from the true source was 32m. In all the trials, the point release is retrieved within 3-60m Euclidean distance from their true location. The source strength is retrieved within a factor of 1.5 to the true release mass. The posterior uncertainty in the release parameters is observed to be within 20% of their mean value. The source localization features are resolved to its maximum extent feasible with the design of the monitoring network. The sensitivity studies are conducted to highlight the importance of receptors reporting zero concentration measurements and variations in the resolution features of the source retrieval with respect to the various arrangements of the receptors.

International audience; Pour modéliser l'action du vent sur les structures, on génère des signaux représentatifs du vent à l'aide d'un modèle semi déterministe. Ce dernier est basé, d'une part, sur des réseaux de neurones artificiels, utilisant des données météo régionales, qui permettent de générer les fluctuations lentes, et d'autre part, sur un modèle classique de génération de processus stochastiques, utilisant des formules de densité spectrale empiriques, qui permet de générer les fluctuations turbulentes. Une procédure de vérification, basée sur la transformation en ondelettes continues, montre que seules les échelles de temps supérieures à 20000s (grandes échelles de temps) sont correctement reconstruites à l'aide des réseaux de neurones. Abstract : To compute wind action on structures, wind-speed time series are generated by using a semi-deterministic model. Artificial neural networks are used, with regional meteorological observations as input arguments, to generate the slowly varying wind speed component. Validation tests, using discrete wavelet transform, show that this model is only able to simulate timescales larger than 20000s. So the turbulent wind speed fluctuations are generated, as stochastic processes, by using a traditional method which uses analytical power spectrum formula.

International audience

International audience

International audience

International audience

International audience

International audience

The aim of this thesis is to compute atmospheric turbulent flows in order to study aerodynamic effects induced by the wind on civil engineering structures. A two-dimensional analysis has been carried out to provide a description of the phenomena involved when the wind acts ona bluff body elongated in the across wind direction.In the discussion, the wind has been assumed as an incompressible turbulent flow governed, in the atmospheric boundary layer, by the Navier-Stokes equations. The turbulent stresses have been determined by a first order turbulence model. The key parameters that have been taken into accounts are the surface roughness, the topogaphy, the obstructions and/or the development of large-scale organised vortices.The computational procedure has been performed with a finite-element software, CASTEM 2000, in which some uncommon methodology, like wall-functions based upon roughness length concept, have been incorporated. Validation tests have been performed for well-known pratical cases. The results show that the standard k-epsilon turbulence model is unable to simulate unsteady flows over changing terrains, with organised vortices. The more sophisticated RNG k-epsilon model appears to be more adapted to this kind of complex flows. Therefore, the latter model has been used to improve analytic formulations of buffeting forces. The existing quasi-steady formulation, commonly used in buffeting analysis, shows its limitations in the case of bluff bodies gust entry. A new approach is suggested which includes, through modified admittance functions, the turbulence induced by the structure itself. It has been found that the shape of these functions depends on both the creation of separation bubbles and the formation of vortices in the wake of the body.; L’objectif de notre travail de thèse est de modéliser numériquement la turbulence atmosphérique afin d’étudier l’action aérodynamique du vent sur les ouvrages de génie civil. L’analyse développée est bidimensionnelle et doit permettre, en vue d’une ´etude vibratoire, d’identifier les principaux mécanismes qui interviennent lorsque le vent aborde transversalement une structure élancée, non profilée.L’approche consiste à assimiler le vent à un écoulement de fluide incompressible régi, dans la couche limite atmosphérique, par les équations de Navier-Stokes et à représenter la turbulence à l’aide d’un modèle du premier ordre. Les principaux paramètres pris en compte sontla rugosité des sols, la topographie, la présence d’obstacles et/ou le développement de grosses structures tourbillonnaires.Le problème est traité avec le code de calcul par éléments finis CASTEM 2000, dans lequel des procédures spécifiques, telles des fonctions de paroi basées sur la notion de longueur de rugosité, ont été implantées. Des calculs de validation, portant sur des cas tests significatifs, montrent que le modèle k-epsilon standard est incapable de modéliser des écoulements instationnaires sur des sites non homogènes, en présence de tourbillons organisés, alors que le modèle RNG k-epsilon, plus élaboré, est adapté à ce type d’écoulements complexes.Ce dernier modèle est donc utilisé pour étudier les phénomènes rencontrés lorsqu’une rafale aborde une structure élancée. Pour des sections non profilées, les limitations de l’approche quasi-statique, généralement utilisée pour représenter les efforts aérodynamiques exercés par le vent turbulent, sont mis en évidence. Dans ce cas, on suggère que les effets de la turbulence, générée par la structure elle-même, doivent être pris en compte au travers de fonctions d’admittance modifiées. On montre que la forme de ces fonctions dépend explicitement de la création de poches de décollement et de la présence de grosses structures tourbillonnaires dans le sillage de l’obstacle.