Cette thèse propose une technique innovante de diagnostic non invasive pour les systèmes piles à combustible. Cette technique s’appuie sur la mesure de la signature magnétique générée par ces systèmes. A l'aide de ces champs magnétiques externes, il est possible d'obtenir une cartographie de la densité de courant interne par résolution d'un problème inverse. Ce problème est néanmoins mal posé : la solution est non unique et est extrêmement sensible au bruit. Des techniques de régularisation ont ainsi été mises en place pour filtrer les erreurs de mesures et obtenir une solution physiquement acceptable. Afin d'augmenter la qualité de reconstruction des courants, nous avons conçu notre outil de diagnostic de manière à ce qu'il soit uniquement sensible aux défaillances de la pile (capteur de défauts). De plus, cette reconstruction se base sur un nombre extrêmement faible de mesures. Une telle approche facilite l'instrumentation du système et augmente la précision et la rapidité de celui-ci. La sensibilité de notre outil à certaines défaillances (assèchements, appauvrissement en réactifs, dégradations) est démontrée. / This thesis proposes a new non invasive technique for fuel cell diagnosis. This technique relies on the measurements of the magnetic field signature created by these systems. By solving an inverse problem, it is possible to get an internal current density map. However, the inverse problem is ill-posed: the solution is not unique and it is extremely sensitive to noise. Regularization techniques were used in order to filter out measurement errors and to obtain physical realistic solutions. In order to improve the quality of the current density estimators, a diagnostic tool was built which is only sensitive to faults occurring inside the fuel cell (fault sensor). More over, our approach is based on a very low number of measurements. Such technique simplifies the experimental setup and improves the accuracy and the speed of the diagnostic tool. The sensitivity of our tool to some faults (drying out, oxygen starvation and ageing) is demonstrated.

This research investigates on-blade partial span active plain trailing edge flaps (TEFs)with an aim to alleviate the helicopter vibrations. Among all the available smart materials, piezoelectric stack actuator(PEA)has shown its strong candidature for full scale rotor systems. Although, PEAs are quite robust in operation, however, they exhibit rate dependent hysteresis phenomenon and can generate only very small displacements. Dynamic hysteresis is a complex phenomenon which, if not modeled, can lead to drift in the vibration predictions. In this research, a comprehensive experimental analysis is performed on a commercially available piezostack actuator, APA-500L, which is well suited for full scale applications. Rate dependent hysteresis loops are obtained for helicopter operational frequencies. Nonlinear rate-dependent hysteresis loops are modeled using conic section approach and the results are validated with experimental data. Dynamic hysteresis exhibited by the PEA is further cascaded with the helicopter aeroelastic analysis and its effect on helicopter vibration predictions is investigated. PEAs generate high force but are limited by small translational motions. A linear to rotary motion amplification mechanism is required to actuate the TEF for vibration alleviation. A smart flap is designed and developed using computer-aided-design models. A rotor blade test section is fabricated and a lever-fulcrum mechanism (AM-1) is developed for a feasibility study. Smart flap actuation is demonstrated on the rotor blade test section. The conventional motion amplification devices contain several linkages, which are potential sites for structural failure. A novel pinned-pinned post-buckled beam linear-to-rotary motion amplifier (AM-2) is designed and developed to actuate the flaps. A new design of linear-to-linear amplification mechanism (LX-4) is developed and is employed in conjunction with AM-2 to increase the flap angles by an order of magnitude. An analytical model is developed using Mathieu-Hill type differential equations. Static and dynamic tests are conducted on a scaled flap model. Helicopter aeroelastic simulations show substantial reduction in hub loads using AM-2 mechanism. To further enhance the flap angles, an optimization study is performed and optimal beam dimensions are obtained. A new technique is also proposed to actively bias the flaps for both upward and downward motion. Critical flap design parameters, such as flap span, flap chord and flap location influences the flap power requirement and vibration objective function significantly. A comprehensive parametric investigation is performed to obtain the best design of TEFs at various advance ratios. Although, parametric study equips the designer with vital information about various critical system parameters, however, it is a computationally expensive exercise especially when used with large comprehensive helicopter aero elastic codes. A formal optimization procedure is employed to obtain the optimal flap design and location. Surrogate models are developed using design of experiments based on response surface methodology. Two new orthogonal arrays are proposed to construct the second order polynomial response surfaces. Pareto analysis is employed in conjunction with a newly developed computationally efficient evolutionary multi-objective bat algorithm. Optimal flap design and flap locations for dual trailing edge flaps are obtained for mutually conflicting objectives of minimum vibration levels and minimum power requirement to actuate the flaps.

Cette thèse introduit différentes méthodes de vérification (ou model-checking) sur des modèles de systèmes à pile. En effet, les systèmes à pile (pushdown systems) modélisent naturellement les programmes séquentiels grâce à une pile infinie qui peut simuler la pile d'appel du logiciel. La première partie de cette thèse se concentre sur la vérification sur des systèmes à pile de la logique HyperLTL, qui enrichit la logique temporelle LTL de quantificateurs universels et existentiels sur des variables de chemin. Il a été prouvé que le problème de la vérification de la logique HyperLTL sur des systèmes d'états finis est décidable ; nous montrons que ce problème est en revanche indécidable pour les systèmes à pile ainsi que pour la sous-classe des systèmes à pile visibles (visibly pushdown systems). Nous introduisons donc des algorithmes d'approximation de ce problème, que nous appliquons ensuite à la vérification de politiques de sécurité. Dans la seconde partie de cette thèse, dans la mesure où la représentation de la pile d'appel par les systèmes à pile est approximative, nous introduisons les systèmes à surpile (pushdown systems with an upper stack) ; dans ce modèle, les symboles retirés de la pile d'appel persistent dans la zone mémoire au dessus du pointeur de pile, et peuvent être plus tard écrasés par des appels sur la pile. Nous montrons que les ensembles de successeurs post* et de prédécesseurs pre* d'un ensemble régulier de configurations ne sont pas réguliers pour ce modèle, mais que post* est toutefois contextuel (context-sensitive), et que l'on peut ainsi décider de l'accessibilité d'une configuration. Nous introduisons donc des algorithmes de sur-approximation de post* et de sous-approximation de pre*, que nous appliquons à la détection de débordements de pile et de manipulations nuisibles du pointeur de pile. Enfin, dans le but d'analyser des programmes avec plusieurs fils d'exécution, nous introduisons le modèle des réseaux à piles dynamiques synchronisés (synchronized dynamic pushdown networks), que l'on peut voir comme un réseau de systèmes à pile capables d'effectuer des changements d'états synchronisés, de créer de nouveaux systèmes à piles, et d'effectuer des actions internes sur leur pile. Le problème de l'accessibilité étant naturellement indécidable pour un tel modèle, nous calculons une abstraction des chemins d'exécutions entre deux ensembles réguliers de configurations. Nous appliquons ensuite cette méthode à un processus itératif de raffinement des abstractions. / In this thesis, we propose different model-checking techniques for pushdown system models. Pushdown systems (PDSs) are indeed known to be a natural model for sequential programs, as they feature an unbounded stack that can simulate the assembly stack of an actual program. Our first contribution consists in model-checking the logic HyperLTL that adds existential and universal quantifiers on path variables to LTL against pushdown systems (PDSs). The model-checking problem of HyperLTL has been shown to be decidable for finite state systems. We prove that this result does not hold for pushdown systems nor for the subclass of visibly pushdown systems. Therefore, we introduce approximation algorithms for the model-checking problem, and show how these can be used to check security policies. In the second part of this thesis, as pushdown systems can fail to accurately represent the way an assembly stack actually operates, we introduce pushdown systems with an upper stack (UPDSs), a model where symbols popped from the stack are not destroyed but instead remain just above its top, and may be overwritten by later push rules. We prove that the sets of successors post* and predecessors pre* of a regular set of configurations of such a system are not always regular, but that post* is context-sensitive, hence, we can decide whether a single configuration is forward reachable or not. We then present methods to overapproximate post* and under-approximate pre*. Finally, we show how these approximations can be used to detect stack overflows and stack pointer manipulations with malicious intent. Finally, in order to analyse multi-threaded programs, we introduce in this thesis a model called synchronized dynamic pushdown networks (SDPNs) that can be seen as a network of pushdown processes executing synchronized transitions, spawning new pushdown processes, and performing internal pushdown actions. The reachability problem for this model is obviously undecidable. Therefore, we compute an abstraction of the execution paths between two regular sets of configurations. We then apply this abstraction framework to a iterative abstraction refinement scheme.

The presented work deals with the development of a monitoring system for monitoring the operation of the laboratory. The theoretical part of the work includes information on the safety of operation and work in biological laboratories, measured physical quantities and the principle of their measurement. The monitoring system is built on a compact ESP32-CAM camera and a WeMos D1 development board. The practical part consists of the design and implementation of a given system and the creation of a full-stack application for displaying live input from the laboratory for monitoring measured quantities in real time.

La gestion thermique des batteries Li-ion pour le véhicule électrique est essentielle, pour assurer une autonomie et une durée de vie optimales de ces batteries. Habituellement, des circuits d'air ou de liquide de refroidissement sont utilisés comme systèmes de gestion thermique. Cependant, ces systèmes sont coûteux en termes d'investissement et d'exploitation et doivent être dimensionnés sur la puissance maximale à extraire. L'utilisation de matériaux à changement de phase (MCP) pour l’absorption sous forme de chaleur latente de la chaleur à dissiper peut représenter une alternative moins coûteuse et plus facile à utiliser. En effet, les MCP peuvent stocker passivement la chaleur excédentaire produite et être utilisés en tant que systèmes passifs. Cependant, les MCP présentent de nombreux inconvénients comme la difficulté de décharger l’énergie thermique stockée, ce qui limite l’aptitude du système au cyclage, ou encore leur conductivité thermique peu élevée qui limite les capacités d’échange. Pour résoudre le problème de la régénération des MCP, un système actif supplémentaire peut être ajouté, dimensionné sur une puissance modérée; l'ensemble devient alors un système semi-passif. Dans cette étude, un système de gestion thermique composé d'un MCP et d’air en convection forcée est évalué. Ce système permet de coupler les avantages de ces deux techniques. Une modélisation du système est développée pour une cellule de batterie. Une comparaison avec de l’air uniquement, en convection forcée, montre l'utilité du MCP. Pour augmenter la capacité d’échange du MCP, un matériau à haute conductivité thermique peut être ajouté au MCP, ce qui permet d’obtenir un composite ayant une conductivité thermique plus élevée. Des composites basés sur les MCP étudiés et des nanostructures de carbone sont élaborés, leur conductivité thermique est mesurée. Ensuite, un système expérimental simulant la dissipation d’une cellule de batterie est construit et utilisé pour évaluer le MCP seul, le MCP inclus dans une mousse métallique et le meilleur composite obtenu. Enfin, pour se rapprocher des conditions réelles, un modèle représentant un stack entier de batterie est développé, des simulations sont produites et les résultats obtenus sont commentés. / Li-ion battery thermal management is essential for electric vehicles (EVs), to ensure an optimal autonomy and lifespan of those batteries. Usually, air or coolant circuits are employed as thermal management systems. However, those systems are expensive in terms of investment and operating costs and must be dimensioned on the maximal power to be extracted. The use of phase change materials (PCMs) as latent heat storage medium allowing the absorption of the heat to be dissipated as latent heat may represent an alternative cheaper and easier to operate. In fact, PCMs can passively store the excess heat produced by a device and be used as passive systems. However, PCMs have several drawbacks like the difficulty to discharge the stored thermal load which limits the system’s cyclability or their low thermal conductivity which limits their heat transfer capacity. To solve the problem of the PCM regeneration, an additional active system can be added, dimensioned on a moderate power; the whole becomes a semi-passive system. In this study, a thermal management system composed of a PCM and forced air convection is evaluated. This system permits to combine the respective advantages of the two techniques. A model of the system is developed for one battery cell. A comparison with forced air convection only points out the usefulness of the PCM. To overcome the PCM low thermal conductivity, a highly conductive material can be added to the PCM permitting to obtain a composite with a higher thermal conductivity. Composites based on the PCMs studied and carbon nanostructures are elaborated, and their thermal conductivity is measured. Then, an experimental system permitting to simulate the dissipation of a battery cell is build and used to evaluate the PCM alone, the PCM embedded in metal foam and the better obtained composite. Finally, to be closer to the real conditions, a model representing an entire battery stack is developed, simulations are produced and the obtained results are discussed.

Ce travail de thèse a pour but de construire et d’étudier une fibration de Hitchin pour les groupes qui apparaît naturellement lorsque l’on essaie de géométriser la formule des traces. On commence par construire une telle fibration en utilisant le semi-groupe de Vinberg. Sur ce semi-groupe de Vinberg, on montre qu’il existe un certain morphisme « polynôme caractéristique » muni d’une section naturelle, de même que dans le cas des algèbres de Lie. On montre également que l’on peut construire un centralisateur régulier au-dessus de cette base des polynômes caractéristiques qui est un schéma en groupes commutatif et lisse.On s’intéresse alors à des variantes pour les groupes des fibres de Springer affines pour lesquelles on remarque que l’introduction du semi-groupe de Vinberg permet d’obtenir une condition d’intégralité analogue à celle de Kazhdan-Lusztig. Ces fibres de Springer affines sont des analogues locaux des fibres de Hitchin. On obtient alors une formule de dimension pour ces fibres.Dans un troisième temps, on s’intéresse à l’aspect global de cette fibration pour laquelle on donne une interprétation modulaire et sur laquelle on construit l’action d’un champ de Picard, issu du centralisateur régulier. L’espace total de cette fibration étant en général singulier, nous étudions son complexe d’intersection. Cet espace de Hitchin s’obtient naturellement comme l’intersection du champ de Hecke avec la diagonale du champ des G-torseurs et on démontre que sur un ouvert suffisamment gros de la base de Hitchin, le complexe d’intersection de l’espace de Hitchin s’obtient par restriction de celui du champ de Hecke corrrespondant.Enfin, dans la dernière partie de cette thèse, on établit un théorème du support dans le cas où l’espace total est singulier analogue à celui de Ngô et l’on démontre que, dans le cas de la fibration de Hitchin, les supports qui interviennent sont reliés aux strates endoscopiques. / This main goal of this work is to construct and study the properties of Hitchin fibration for groups which appears naturally when we try to geometrize the trace formula. We begin by constructing this fibration using the Vinberg’s semigroup. On this semigroup, we show that there exists a characteristic polynomial morphism equipped with a natural section, analog at the Kostant’s one in the case of Lie algebras. We also show that there exists on the base of characteristic polynomials a regular centralizer scheme, which is a smooth commutative group scheme.Then, we are interested in some variant of affine Springer fibers, for which we see that the Vinberg’s semigroup appears naturally to obtain an integrality condition analog to Kazhdan-Lusztig’s one. These affine Springer fibers are local incarnation of Hitchin fibers.In a third time, we go back to the global case and give a modular interpretation of this new Hitchin fibration on which we construct an action of a Picard stack, coming from the regular centralizer.The total space of this fibration, even on the generically regular semisimple locus will be singular and we want to understand his intersection complex. This space can be obtained as the intersection of the Hecke stack with the diagonal of the stack of G-bundles and we show that on a sufficiently big open subset of the Hitchin base, the intersection complex of the Hitchin’s space is the restriction of the corresponding intersection complex on the Hecke stack.Finally, in the last part of this work, we establish a support theorem in the case of a singular total space, generalizing Ngo’s theorem et we show that in the case of Hitchin fibration, the supports that appear are related to the endoscopic strata.

No description available.

Made available in DSpace on 2016-04-04T18:31:39Z (GMT). No. of bitstreams: 1 Deivis Fernandes Pirani.pdf: 1504986 bytes, checksum: e4989c06d3fc9c76df085862e18609d0 (MD5) Previous issue date: 2014-02-07 / This paper shows a field-trial investigation comparing the use of IPv4 and IPv6 protocols on web pages access considering the real user perspective and the transition period from IPv4 to IPv6. To perform this study, it was developed a testbed in which three adressing configurations of Internet Protocol, IPv4, IPv6, and Duas Stack (IPv4+IPv6), were used to access the top 100 world most visited web pages with an IPv6 enabled address. In addiction, several software tools were developed, that enabled more than 600 simultaneous accesses between IPv4, IPv6 and Dual Stack, to each of these pages. Those accesses were performed every hour during the period of a month, enabling to obtain more than 20 GB of information on the web access. Results showed that lower loading times obtained by IPv6 and Dual Stack are related to the partial acquisition of the content. In fact, for IPv6, the number of web pages with faster loading times goes from 57 to 20% when the average number of bytes obtained by IPv6 and the IPv4 is similar. For Dual Stack the number goes from 25 to 5% given the same conditions. It was also noted that 18 of the 100 pages which were considered, no content would be presented to the user in more than 50% of the performed accesses. Regarding the preference for IPv6 instead of IPv4, on Dual Stack it was observed that 90% of the acquired content for 27 of 100 web pages was made by IPv6. / Este trabalho apresenta um estudo de campo, comparativo, sobre a utiliza??o dos protocolos IPv4 e IPv6 no acesso a p?ginas web. Foi considerada a perspectiva de um usu?rio real deste tipo de servi?o, no atual per?odo de transi??o do IPv4 para o IPv6. Para que tal estudo pudesse ser realizado, foi desenvolvido um arranjo no qual tr?s configura??es de endere?amento do Protocolo de Internet, IPv4, IPv6 e Pilha Dupla (IPv4+IPv6), foram utilizadas nos acessos ?s 100 p?ginas web, com suporte ao I v6, mundialmente mais visitadas. Al?m disso, foram desenvolvidas uma s?rie de shell scripts que possibilitaram mais de 600 acessos simult?neos do IPv4, IPv6 e Pilha Dupla, a cada uma dessas p?ginas. Tais acessos foram executados de hora em hora, durante o per?odo de um m?s, possibilitando a obten??o de mais de 20 GB de informa??es relativas ao acesso web. Os resultados do estudo mostraram que os tempos de carregamento menores obtidos pelo IPv6 e pela Pilha Dupla, est?o relacionados com a aquisi??o parcial do conte?do. De fato, para o IPv6, o n?mero de p?ginas web com tempos de carregamento menores diminui de 57 para 20% quando a m?dia do n?mero de bytes obtidos pelo IPv6 e pelo IPv4 ? similar. Para a Pilha Dupla, dadas as mesmas condi??es, o n?mero diminui de 25 para 5%. Tamb?m foi observado que para 18 das 100 p?ginas web consideradas, nenhum conte?do foi apresentado ao usu?rio em mais de 50% dos acessos efetuados. Com rela??o ? prefer?ncia pelo IPv6 ao IPv4, na Pilha Dupla, observou-se que 90% da aquisi??o do conte?do, de 27 das 100 p?ginas web, ocorreu pelo IPv6.

L'objectif de ce travail est de comprendre comment la forme intrinsèque des fibres individuelles contrôle les propriétés collectives des empilements, en particulier leurs propriétés mécaniques (élasticité) et électriques. Nous nous intéressons à des fibres longues, alignées selon une direction privilégiée et présentent un désordre de forme important. Notre étude est expérimentale et numérique. Nous proposons un modèle à deux dimensions, de champ moyen auto-cohérent, décrivant l'élasticité collective de l'empilement à partir des propriétés individuelles des fibres : leurs distribution de désordre ainsi que leurs module de courbure. Nous obtenons une équation d'état qui décrit avec un bon accord l'élasticité de nos empilements de fibres, sans paramètre ajustable, mais à un facteur multiplicatif près. Nous obtenons des résultats comparables entre les études expérimentale et numérique. / The purpose of this work is to understand how intrinsic shape of individual fibers controls the collective behavior of fiber stacks, in particular the mechanical (elasticity) and electrical properties. We consider long fibers, aligned towards one preferential direction with a significant disorder shape. Our study is experimental and numerical. We propose a two dimensions self consistent mean field model which describes the collective elasticity from the individual properties of fibers : the disorder distribution and the bending modulus. We obtain an equation of state which describes with a good agreement the stacks elasticity, without any fit parameters, however up to a multiplicative constant. We obtain similar results between experimental and numerical studies.

Modern embedded processors with dedicated address generation unit support memory accesses using indirect addressing mode with auto-increment and auto-decrement. The auto-increment/decrement mode, if properly utilized, can save address arithmetic instructions, reduce static and dynamic footprint of the program and speed up the execution as well. We propose an optimization framework for embedded processors based on the auto-increment and decrement addressing modes for address registers. Existing work on this class of optimizations focuses on using an access graph and finding the maximum weight path cover to find an optimized stack variables layout. We take this further by using coalescing, addressing mode selection and offset registers to find further opportunities for reducing the number of load-address instructions required. We also propose an algorithm for building the layout with considerations for memory accesses across basic blocks, because existing work mainly considers intra-basic-block information. We then use the available offset registers to try to further reduce the number of address arithmetic instructions after layout assignment.