The increasing penetration of distributed energy resources, in particular Photovoltaic (PV) production units, and the ever-growing use of power electronics-based equipment has led to specific concern about Power Quality (PQ) in the Low-Voltage (LV) grid. These include high- and low-order current harmonics as well as voltage distortion at the point of common coupling. Solutions to overcome these issues, meeting international grid codes, are being proposed in the context of smart energy management schemes.This work proposes a novel three-phase topology for a PV system with enhanced PQ mitigation functionality, tackling the corresponding control challenges.First, a single-stage current-source inverter PV system with active filtering capability is preferred to the more common two-stage voltage-source inverter topology with additional voltage-step-up converter. The system also guarantees a nearly unitary displacement power factor in the connection to the grid and allows for Maximum Power Point Tracking (MPPT) with direct control of the PV array power. The grid-synchronised dq-axis grid current references are generated for the mitigation of nonlinear load low-order current harmonics, without the need for additional measurements. Active damping is used to minimise grid-side filter losses and reduce high-order harmonics resulting from the converter switching.Results on a 500W laboratory prototype confirm that active damping reduces the switching harmonics in the grid currents and active filtering properly mitigates the low-order current harmonics. The MPPT algorithm works effectively for various irradiance variations. Second, a PV system with a novel Indirect Matrix Converter (IMC)-based unified power quality conditioner topology is developed for enhanced current and voltage compensation capability, with compactness and reliability advantages. PQ issues such as current harmonics, and voltage sags, swells, undervoltage and overvoltage are mitigated by the shunt and series converters, respectively.The more common Space Vector Modulation (SVM) method used in IMCs is developed for this specific topology. In particular, a new shunt converter modulation method is proposed to additionally control the PV array current with zero switching vectors, resulting in a specific switching sequence.A direct sliding mode control method is also studied separately for the shunt and series converters, so that the zero-vector modulation method of the shunt converter can be used, with no sensitive synchronisation of the switching signals; this contrasts with the SVM method. A new dc link voltage modulation method with 12 voltage zones, instead of 6, is proposed to help overcome the limitation in the choice of shunt converter switching vectors due to the positive dc link voltage constraint.Results are obtained for the direct method on a 1 kW laboratory prototype with optimised IMC dc link connection and alternative shunt converter switching transitions to guarantee a positive dc link voltage. Current and voltage compensation capabilities are confirmed by tests in various operating conditions. / Doctorat en Sciences de l'ingénieur et technologie / info:eu-repo/semantics/nonPublished

Avec le renforcement des frontières extérieures de l'Union européenne, les îles en Méditerranée sont devenues des lieux clés de la politique de contrôle migratoire. Cette thèse s'intéresse au cas de Malte. Contrairement à Lampedusa ou Lesbos où les migrants qui y sont acheminés sont ensuite transférés vers une métropole, les migrants escortés sur le territoire maltais sont contraints, par la législation européenne, à demeurer sur l'île-Etat. Cette gestion des flux migratoires a impliqué une présence durable de populations subsahariennes sur l'île dont le traitement et l'expérience sont marqués par l'illégalité. En combinant l'analyse des politiques publiques avec l'anthropologie des camps d'enfermement et les governmentality studies, cette thèse pose les questions suivantes : comment les migrants deviennent-ils l'objet de stratégies gouvernementales depuis l'adhésion de l'île-Etat à l'Union européenne en 2004 et comment y réagissent-ils par leurs pratiques ? Dans une ambition théorique, comment conceptualiser ces modes de gouvernementalité ? La littérature sur le contrôle migratoire tend à étudier tantôt les concepteurs et les opérateurs de ces politiques tantôt les populations cibles. En cela, ce champ d'études a tendance à reproduire implicitement la déconnexion opérée par les politiques publiques entre gouvernés et gouvernants. Prenant le contre-pied de ces approches, ce travail se propose de considérer cette déconnexion comme un objet d'étude à part entière. La fabrique du contrôle migratoire sera envisagée comme un processus auquel prennent part, tout en étant déconnectés, tant les récepteurs que les concepteurs. Ces derniers exercent leur agency dans une relation de pouvoir asymétrique. Dans cette perspective, je proposerai le concept de gouvernementalité transnationale à partir de l'étude de la construction des migrations subsahariennes à Malte comme problème européen, de l'enfermement des migrants sur l'île et des politiques de relocalisation et de réinstallation. A la différence des travaux centrés autour de la problématique de l'Etat, le concept de gouvernementalité transnationale permet de relativiser la centralité de la souveraineté. La gestion des migrations est envisagée comme le résultat de négociations politiques entre des acteurs nationaux, internationaux et non-gouvernementaux qui traversent les frontières étatiques. Partant l'enjeu n'est plus d'identifier quels acteurs détiennent le monopole du contrôle migratoire mais quelles sont les formes de contrainte qui s'exercent sur les migrants et quelles sont les marges de manœuvre dont ils disposent ? Sont-ils face à des techniques de pouvoir essentiellement restrictives et punitives ? Je montrerai que la gouvernementalité transnationale des migrations repose en partie sur l'exercice de la violence. L'enfermement systématique des migrants à l'arrivée peut-être prolongé mais il est transitoire ayant vocation à redistribuer les flux migratoires. Il s'apparente à une forme de biopolitique qui s'exerce sur des groupes plutôt que sur des individus. Les migrants sont catégorisés en fonction de critères juridiques et ethniques. Ces derniers sont classés et hiérarchisés en fonction de (non)statuts juridiques qui recoupent leur nationalité. Par ailleurs, les migrants sont présentés par les concepteurs de la politique migratoire de Malte comme étant immobilisés sur une île-frontière de l'Union européenne, ce qui justifie la mise en œuvre des politiques de relocalisation et de réinstallation. Néanmoins, si les migrants sont enfermés de manière temporaire sur l'île, ils sont en fait mobiles. En cela, ils constituent les véritables acteurs transnationaux de cette forme particulière de gouvernementalité. En revanche, le quotidien de l'attente vécu par les migrants à Malte ne se comprend comme un effet pervers et non prévus du dispositif européen de contrôle des frontières mais comme partie intégrante. / Due to the strengthening of the external borders of the European Union (EU), the islands of the Mediterranean became key sites of migration control policies. This doctoral research focuses on the case of Malta. Unlike Lampedusa or Lesbos from where migrants are transferred to mainland Europe, the island-state of Malta has to contain migrants who have been escorted to the Maltese territory, according to EU legislation. This management of migration flows has led to the settlement of Sub-Saharan populations on the island and whose treatment and experience are marked by illegality. This dissertation combines public policies analysis, anthropology of confinement with governmentality studies and asks the following questions: how have migrants become the object of governmental strategies following the accession of the island-state to the EU in 2004 and how have migrants reacted to them? How can we to conceptualise these modes of governmentality? The literature on migration control tends to study either policy-makers or target population. In doing so, it reproduces implicitly the disconnection between those who govern and those who are governed. To the contrary, this disconnection is at the centre of my research. I consider migration control as a process involving both policy-makers and beneficiaries, even if they are disconnected. They are subjected to asymmetrical power relations in which they both exercise agency. Following this perspective, I use the concept of transnational governmentality to study how Sub-Saharan migrants in Malta are constructed as an EU problem, how they are detained on the island and become the object of relocation and resettlement policies. Unlike studies that focuse on the State theory, the concept of governmentality goes beyond the centrality of sovereignty. The management of migration is the result of political negotiations between national, international and non-governmental actors. As such, my goal is not to identify who controls migration but rather what forms of coercion are exercised over migrants and how migrants circumvent the constraints imposed upon them. I demonstrate that transnational governmentality partly relies on the exercise of physical violence. Mandatory detention is implemented upon arrival and can last up to 18 months and it aims at redirecting migration flows. In this sense, detention is a form of biopolitics that is exercised over groups rather than over individuals. Migrants are categorised according to legal and ethnic criteria. They are classified and ranked by legal (non-)statuses that overlap with nationality. Although migrants are presented as immobilised populations on an island-border of the EU, which legitimises the implementation of resettlement and relocation policies, they are temporarily contained and they are actually mobile. As such, they are the real transnational actors of governmentality. However, the expectance of relocation, resettlement or departure that migrants experience on the island, is not understood as a side-effect of EU border control policies but as an actual part of them. / Doctorat en Sciences politiques et sociales / Thèse défendue le 25 janvier 2017 à Aix-en-Provence, Sciences Po Aix (CHERPA) / info:eu-repo/semantics/nonPublished

The main issue in swarm robotics is to design the behavior of the individual robots so that a desired collective behavior is achieved. A promising alternative to the classical trial-and-error design approach is to rely on automatic design methods. In an automatic design method, the problem of designing the control software for a robot swarm is cast into an optimization problem: the different design choices define a search space that is explored using an optimization algorithm. Most of the automatic design methods proposed so far belong to the framework of evolutionary robotics. Traditionally, in evolutionary robotics the control software is based on artificial neural networks and is optimized automatically via an evolutionary algorithm, following a process inspired by natural evolution. Evolutionary robotics has been successfully adopted to design robot swarms that perform various tasks. The results achieved show that automatic design is a viable and promising approach to designing the control software of robot swarms. Despite these successes, a widely recognized problem of evolutionary robotics is the difficulty to overcome the reality gap, that is, having a seamless transition from simulation to the real world. In this thesis, we aim at conceiving an effective automatic design approach that is able to deliver robot swarms that have high performance once deployed in the real world. To this, we consider the major problem in the automatic design of robot swarms: the reality gap problem. We analyze the reality gap problem from a machine learning perspective. We show that the reality gap problem bears a strong resemblance to the generalization problem encountered in supervised learning. By casting the reality gap problem into the bias-variance tradeoff, we show that the inability to overcome the reality gap experienced in evolutionary robotics could be explained by the excessive representational power of the control architecture adopted. Consequently, we propose AutoMoDe, a novel automatic design approach that adopts a control architecture with low representational power. AutoMoDe designs software in the form of a probabilistic finite state machine that is composed automatically starting from a number of pre-existing parametric modules. In the experimental analysis presented in this thesis, we show that adopting a control architecture that features a low representational power is beneficial: AutoMoDe performs better than an evolutionary approach. Moreover, AutoMoDe is able to design robot swarms that perform better that the ones designed by human designers. AutoMoDe is the first automatic design approach that it is shown to outperform human designers in a controlled experiment. / Doctorat en Sciences de l'ingénieur et technologie / info:eu-repo/semantics/nonPublished

Il s'agit de fonder une sémiotique du lien, c'est-à-dire une méthode d'établissement du sens par analyse des signifiants relationnels. Cette méthode est destinée à être utilisée en thérapie familiale systémique et plus particulièrement en thérapie de couple, mais l'auteur de cette recherche, pourtant master en psychologie et psychothérapeute, a choisi d'utiliser un corpus de pièces de Shakespeare autour de la question: qu'est-ce qui fait qu'une histoire contenant tous les ferments de la tragédie (par exemple "Beaucoup de bruit pour rien") finit en comédie et, inversement, qu'est-ce qui fait qu'une histoire contenant tous les ferments de la comédie (par exemple "Roméo et Juliette") se termine en tragédie? / Doctorat en Information et communication / info:eu-repo/semantics/nonPublished

La caractérisation des complexes de van der Waals est déterminante pour la compréhension et la modélisation des atmosphères planétaires et du milieu interstellaire. Cependant, ces agrégats moléculaires sont assez peu étudiés, particulièrement dans l’infrarouge proche. Cette thèse a pour but d’enrichir les connaissances sur les complexes d’eau et d’ammoniac dans cette région spectrale. Pour cela, nous avons utilisé un montage expérimental appellé FANTASIO+, couplant une expansion su- personique à un spectromètre à temps de déclin (CRDS). Plusieurs complexes de van der Waals ont pu être observés et analysés, à savoir 14/15NH3−Ar/Kr, (H2O)2, H2O−Ar/Kr et HDO−N2O.Ces complexes peuvent être décrits par deux régimes selon la hauteur de la barrière du potentiel intermoléculaire. Les complexes caractérisés par un potentiel inter- moléculaire élevé peuvent être traités par le régime semi-rigide, à savoir comme une molécule à part entière avec des constantes rotationelles bien définies [(H2O)2 et HDO−N2O]. Dans le régime du rotateur libre, au contraire, le monomère du complexe est soumis à une faible barrière de potentiel, et donc sa rotation est pra- tiquement libre (tous les complexes 14/15NH3− et H2O−gaz rare).Le déplacement et la séparation des niveaux d’énergie robivrationels du monomère dépend de l’effet de l’anisotropie du potentiel intermoléculaire par rapport au lien intermoléculaire. Les transitions entre ces sous-états sont appelées des sous-bandes.Les spectres observés dans ce travail de thèse correspondent à l’excitation vibra- tionnelle de 14/15NH3, H2O ou HDO. L’analyse rotationelle des sous-bandes a pu être réalisée pour toutes les espèces (y compris les 4 isotopes principaux du krypton dans H2O−Kr), excepté pour 15NH3−Kr car la sous-bande observée était à peine visible. Nous avons pu caractériser la structure rotationnelle de 1 (15NH3−Ar) à 14 (H2O−Ar) sous-états vibrationnellement excités et de l’état fondamental pour toutes les espèces sauf 14NH3−Kr, 15NH3−Ar/Kr et (H2O)2. Notre travail sur ce dernier, basé sur des calculs théoriques et sur une étude similaire en matrice de néon, est la première analyse rotationelle de ce complexe dans cette région spectrale. Les analyses des spectres de H2O−Ar et HDO−N2O s’appuient sur des calculs ab initio. Pour HDO−N2O, nous avons pu déterminer que le lien OD se trouve plus proche de N2O que le lien OH. Les spectres de 14/15NH3−Ar/Kr présentent des sous-bandes appartenant à des complexes de taille plus importante. Un séjour à l’Université d’Alberta a été organisé afin d’identifier la nature de ces complexes.Quand l’analyse rotationelle a pu être réalisée, nous avons déterminé les temps de vie de prédissociation des sous-états excités (sauf pour HDO−N2O car le signal-sur- bruit était trop faible). Ils sont de 20 ps pour (H2O)2 à 4 ns pour H2O−Ar. Nous avons observé que les temps de vie déterminés étaient inversement proportionnels à l’énergie de dissociation des complexes. Le déplacement observé entre l’origine des sous-bandes par rapport à la bande correspondante du monomère suit égale- ment la même tendance que l’énergie de dissociation du complexe. Nous avons pu déterminer les températures rotationnelles pour toutes les espèces dans le jet super- sonique. Celles-ci sont différentes uniquement dans le cas de 14NH3−Ar.Des spectres des monomères 14NH3 et 15NH3 ont également été enregistrés, notamment en utilisant un spectromètre à transformée de Fourier, enrichissant ainsi les données spectroscopiques de ces molécules dans l’infrarouge proche. / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

In this work, the general problematic of the transport phenomena in the lungs is addressed, with applications to lung diseases, in particular asthma and cystic fibrosis. Regarding these two major diseases, the dynamics of the nitric oxide and the mucus layer in the lungs, respectively, are of particular interest. In asthma, it has been shown that the measurement of the exhaled concentration of nitric oxide can be used as a proxy for monitoring the disease status. In cystic fibrosis, which is a disease of the mucus layer, pieces of evidence show that the water content of the mucus is linked to the severity of the symptoms of the disease. In both conditions, however, these links are far from being obvious. Regarding nitric oxide, its exhaled concentration is not straightforwardly related to its concentration in the lungs. Regarding cystic fibrosis, the link between the disease and the alterations of the mucus content and dynamics is far from being understood. From these observations, several modeling approaches have emerged to complement the clinical measurements. In this work, the dynamics of the nitric oxide and of the mucus layer in the lungs is studied from a modeling approach, following a phenomenological point of view. Regarding the nitric oxide, a new model of its dynamics in the lungs is developed. When compared to previous ones, the model presents multiple new features that allow for a better description of this dynamics, especially in the case of asthma exacerbation, characterized by the presence of mucus obstructions and bronchoconstriction. From this model, the role of nitric oxide as a general marker of the bronchi caliber is suggested, in addition to its role in the asthma monitoring. Furthermore, based on clinical measurements of the exhaled nitric oxide concentration in cystic fibrosis patients and the use of the new model, the role of the measurement of nitric oxide in the understanding and control of other lung diseases, such as cystic fibrosis, is evaluated. Regarding the dynamics of the mucus layer in the lungs, a new analysis of the control of the mucus balance in the bronchial region of the lungs is presented in this work. Our approach is based on the combination of a balance equation for the mucus in an airway and a computational tool characterizing the evaporation of the mucus in the bronchial region. We show that this approach allows for new insights into the dynamics of the bronchial mucus and, more specifically, on the mechanisms controlling the amount of mucus in an airway. The results are analyzed in order to bring interesting new perspectives for the understanding and the treatment of mucus pulmonary diseases,such as cystic fibrosis. Altogether, this work demonstrates, applied to medical pathologies, the usefulness of modeling approaches in giving a mechanistic view of the encountered problematic. / Doctorat en Sciences de l'ingénieur et technologie / info:eu-repo/semantics/nonPublished

According to a conventional wisdom, politics is nowadays more personalized than before. The proposition of the personalization of politics is relevant for three specific areas: institutions, media and voting behavior. This dissertation deals with the latter, since it focuses on the influence of party leaders on vote choice. So far the empirical scholarship tested whether “leader effects” have increased overtime (longitudinal studies) or investigated which conditions can enhance or discourage the electoral influence of party leaders (conditionality literature). We argue that both approaches have their limits, the former being based on the customary assumption of linearity, the latter investigating the role of (micro, meso and macro) “characteristics” rather than overarching and dynamic condition(s). This dissertation puts the cursor on a specific overarching condition: the (in)stability of parties and party systems (supply-side complexity). Actually, we argue that the magnitude of leader effects on the vote is conditional to the quality of the electoral environment. We therefore attribute to the “leader variable” a heuristic value: the leader appeal is conceived as an electoral shortcut more likely to be activated in complex electoral environment than in clear and stable contexts. There are two ways of testing the link between (variations in) and supply-side complexity and (variations in) leader effect on the vote. The first requires the adoption of a synchronic perspective, which implies a cross-national and cross-partisan empirical posture. This perspective considers space (i.e. “horizontal”) variations of stability, by assuming that certain parties and party systems are stable, while others are less or not stable. Are leader effects on the vote greater in the latter than in the former? The second one focuses on time (i.e. “longitudinal”) variations of stability. Indeed, electoral contexts vary across time – and not only across spaces - thus modifying the perceptions that voters may have of their electoral environment. When parties and party systems get convulsed, voters lose their frame of references. As such, they become cue-takers and rely more easily on leader appeals. On the contrary, when the political environment becomes clarified, voters will be less encouraged to rely on leader heuristic.This dissertation gives voice to both dimensions (space variation vs time variation). In a first place, we dig into the synchronic perspective. The relationship between supply-side complexity and leader effect on the vote is tested on a sample composed by 20 countries and 125 parties, included within the Comparative Studies of Electoral Systems. In a second place, we chart the magnitude of the leader variable according to longitudinal variations of supply-side complexity. Actually, our analytical effort will henceforth be centered on two distinguished case studies: Poland (1997-2011) and Italy (1996-2013), which have both experienced important variations in the stability of their electoral environment.We demonstrated that de facto differences (i.e. cross national and cross-partisan variations) in terms of stability do not automatically determine the magnitude of leader effects. On the contrary, we detect a link between the leader variable and longitudinal variations of the electoral environment. However, the quality of this link proves to be different in Poland and Italy. In the young democracy of Poland, leader effects and stability seem to be convergent rather than antagonist forces, while the Italian case properly fits our expectations: the convolution of the electoral environment makes voters more sensitive to the leader heuristic. / Doctorat en Sciences politiques et sociales / info:eu-repo/semantics/nonPublished

Afin de s’affranchir de contraintes liées à l’utilisation de milieux conventionnels, l’électroreduction du CO2 a été étudiée en milieux liquides ioniques. L’activité de l’interface électrode - électrolyte a été examinée par techniques électrochimiques et les produits de réactions formés après électrolyses sous contrôle potentiostatique, ont été analysés par techniques chromatographiques et spectroscopie RMN. L’ensemble des informations obtenu lors de ces analyses a permis de contribuer à une meilleure compréhension des mécanismes réactionnels de l’électroréduction du CO2 en milieux liquides ioniques. Au travers de ce travail, l’influence de la composition du liquide ionique sur l’électroréduction du CO2 conduite à une électrode d’or a pu être étudiée, en recourant à des liquides ioniques constitués de différents cations (imidazolium, pyrrolidinium, pipéridinium et tétraalkylammonium) et du même anion [NTf2]-. En milieu [BMIm][NTf2], la présence du proton acide H-2 du cation [BMIm]+ favorise la formation de CO, principal produit formé à l’électrode d’or polycristallin. Cependant, la réduction du cation [BMIm]+ apparaît comme une réaction compétitive, produisant non seulement du carbène, mais également des dimères hydrogénés et un monomère. La différence de sélectivité de ces produits observée à différents matériaux d’électrode a pu être interprétée sur base des capacités différentes des surfaces métalliques à adsorber l’hydrogène. En présence de ces produits générés électrochimiquement, le CO2 est indirectement converti produisant en partie un adduit. Dû fait de la faible chargée transféree durant électrolyse en milieu [N1,1,1,3][NTf2] et des grandes surtensions nécessaires à la réduction du CO2 en milieu [PMPip][NTf2], l’électroréduction du CO2 a été poursuivie en détail en milieu [BMPyrr][NTf2].En milieu [BMPyrr][NTf2], la réduction du cation n’apparaît plus comme une réaction compétitive, et CO peut être formé sélectivement avec d’excellentes efficacités faradiques, sans consommation des protons du liquide ionique. Dans ce milieu, la formation de CO est seulement inhibée à des potentiels plus négatifs par la formation d’oxalate, la dégradation de l’anion [NTf2]- et par la corrosion cathodique de l’électrode d’or. De plus, le présent travail a montré que la sélectivité pour la formation de CO et d’oxalate est affectée par la nature du matériau d’électrode, grâce à l’utilisation d’une surface de cuivre. A une électrode de cuivre polycristallin, la présence d’une source de proton, telle que H[NTf2], permet également la formation de produits à haute valeur ajoutée, telle que le méthane ou l’acide formique.A une électrode d’or polycristallin, la formation d’adduits en présence de CO2 n’a pas seulement été observée en milieu [BMIm][NTf2], mais également après ajout d’un co-réactant, telle que des bipyridines, en milieu [BMPyrr][NTf2]. Le mécanisme de réaction diffère cependant selon la structure de la bipyridine. / To avoid limitations encountered in conventional media, the electroreduction of CO2 has been studied in ionic liquids. The activity of the ionic liquid - electrode interface was examined by electrochemical techniques and the reaction products were analysed after controlled-potential electrolysis, by chromatographic techniques and NMR spectroscopy. The information obtained by these analysis have contributed to a better understanding of the reaction pathways of the electroreduction of CO2 in ionic liquids media. In this work, the influence of the ionic liquid composition on the electroreduction of CO2 conducted at gold electrodes was studied by resorting to ionic liquids composed of different cations ( imidazolium, pyrrolidinium, piperidinium and tetraalkylammonium) and the same anion [NTf2]−. In [BMIm][NTf2], the presence of acidic H-2 proton on the [BMIm]+ cation favours the formation of CO, main product formed at Au (poly). However, the reduction of the [BMIm]+ cation appears as a competitive reaction, producing not only carbene, but also hydrogenated dimers and a monomer. The selectivity of these electrogenerated products observed at various electrode materials, has been interpreted taking into account the distinct H-adsorption abilities of the metallic surfaces. In the presence of these electrogenerated products, CO2 is indirectly converted into an adduct. In ammonium-based ionic liquids, the electroreduction has been investigated in detail only in [BMPyrr][NTf2], since inhibition of the charge transfer and a high overpotential for the reduction of CO2 were observed in [N1,1,1,3][NTf2] and [PMPip][NTf2], respectively.By contrast, in [BMPyrr][NTf2], presenting no reduction reaction of the cation in competition with the CO2 electroreduction at Au (poly), the formation of CO can be selectively formed with excellent faradaic efficiency and without consumption of the RTILs protons. The selectivity for the formation of CO only decreases at more negative potentials, since the formation of oxalate, the decomposition of the [NTf2]- anion and a cathodic corrosion of the gold electrode may also take place. It has been shown that the selectivity for the formation of CO and oxalate is affected by the nature of the electrode material, using a copper surface. At Cu (poly), the presence of an additional source of proton, such as H[NTf2], to [BMPyrr][NTf2] also provides the formation of methane or formic acid.At Au (poly), the formation of adducts in the presence of CO2 has not only been observed in [BMIm][NTf2], but also after the addition of co-reactant, such as bipyridines, in [BMPyrr][NTf2]. The reaction mechanism differs, however, according to the structure of bipyridine. / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

Bovine Leukemia Virus (BLV) latency is a viral strategy used to escape from the host immune system and contribute to tumor development. However, the recent discovery of a highly expressed miRNA cluster has suggested that BLV latency is partially true. In our PhD thesis, we studied the epigenetic and transcriptional regulations of this RNA polymerase III (RNAPIII)- dependent miRNA cluster and of a newly discovered RNA polymerase II (RNAPII)-dependent promoter (which drives an active antisense transcription). Moreover, our data suggested a putative collision phenomenon between RNAPII and RNAPIII convergent transcriptions. In the second part of our PhD thesis, we therefore provided new insights into this complex transcriptional network. In the third part of this work, we demonstrated the recruitment of CTCF, a multi-functional transcriptional regulator, along the BLV genome and provided data explaining its putative functions in BLV transcriptional and epigenetic regulations but also in viral-host genome long-range interactions. In conclusion, our PhD thesis provides a better understanding of the transcriptional network regulating BLV gene expression and new ideas to study BLV-induced leukemia development. La latence virale, principale caractéristique de l’infection par le virus de la leucémie bovine (BLV), permet au virus d’échapper au système immunitaire de l’hôte et contribue au développement tumoral. Cependant, la récente découverte d’une région codant pour des miRNAs viraux et activement transcrite par l’ARN polymérase III (RNAPIII), suggère que la latence virale n’est que partiellement vraie. Dans cette thèse, nous avons étudié les régulations transcriptionnelle et épigénétique du promoteur RNAPIII-dépendant contrôlant l’expression des miRNAs du BLV, ainsi que celles d’un nouveau promoteur ARN polymérase II (RNAPII)-dépendant, découvert au cours de notre travail de thèse. Suite à nos données, suggérant de l’interférence transcriptionnelle, nous avons ensuite étudié le fonctionnement de ce réseau transcriptionnel complexe composé de trois promoteurs distincts régulant l’expression du BLV. Finalement, nous avons démontré le recrutement de CTCF, un régulateur transcriptionnel multifonctionnel, le long du génome du BLV et nos résultats suggèrent des potentielles fonctions de CTCF dans les régulations transcriptionnelle et épigénétique du BLV, mais également dans des interactions longue-distances entre le virus et le génome de la cellule hôte. En conclusion, notre thèse de doctorat apporte une meilleure compréhension du réseau transcriptionnel régulant l’expression génique du BLV et offre de nouvelles pistes pour l’étude du développement tumoral induit par le BLV. / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

Swarm robotics is a young multidisciplinary research field at the<p>intersection of disciplines such as distributed systems, robotics,<p>artificial intelligence, and complex systems. Considerable research<p>effort has been dedicated to the study of algorithms targeted to<p>specific problems. Nonetheless, implementation and comparison remain difficult due to the lack of shared tools and benchmarks. Among the tools necessary to enable experimentation, the most fundamental is a simulator that offers an adequate level of accuracy and flexibility to suit the diverse needs of the swarm robotics<p>community. The very nature of swarm robotics, in which systems may comprise large numbers of robots, forces the design to provide<p>runtimes that increase gracefully with increasing swarm sizes.<p><p>In this thesis, I argue that none of the existing simulators offers<p>satisfactory levels of accuracy, flexibility, and efficiency, due to<p>fundamental limitations of their design. To overcome these<p>limitations, I present ARGoS---a general, multi-robot simulator that<p>currently benchmarks as the fastest in the literature.<p><p>In the design of ARGoS, I faced a number of unsolved issues. First, in existing simulators, accuracy is an intrinsic feature of the<p>design. For single-robot applications this choice is reasonable, but<p>for the large number of robots typically involved in a swarm, it<p>results in an unacceptable trade-off between accuracy and<p>efficiency. Second, the prospect of swarm robotics spans diverse<p>potential applications, such as space exploration, ocean restoration,<p>deep-underground mining, and construction of large structures. These applications differ in terms of physics (e.g. motion dynamics) and available communication means. The existing general-purpose simulators are not suitable to simulate such diverse environments accurately and efficiently.<p><p>To design ARGoS I introduced novel concepts. First, in ARGoS accuracy is framed as a property of the experimental setup, and is tunable to the requirements of the experiment. To achieve this, I designed the architecture of ARGoS to offer unprecedented levels of modularity. The user can provide customized versions of individual modules, thus assigning computational resources to the relevant aspects. This feature enhances efficiency, since the user can lower the computational cost of unnecessary aspects of a simulation.<p><p>To further decrease runtimes, the architecture of ARGoS exploits the computational resources of modern multi-core systems. In contrast to existing designs with comparable features, ARGoS allows the user to define both the granularity and the scheduling strategy of the parallel tasks, attaining unmatched levels of scalability and efficiency in resource usage.<p><p>A further unique feature of ARGoS is the possibility to partition the<p>simulated space in regions managed by dedicated physics engines<p>running in parallel. This feature, besides enhancing parallelism,<p>enables experiments in which multiple regions with different features are simulated. For instance, ARGoS can perform accurate and efficient simulations of scenarios in which amphibian robots act both underwater and on sandy shores.<p><p>ARGoS is listed among the major results of the Swarmanoid<p>project. It is currently<p>the official simulator of 4 European projects<p>(ASCENS, H2SWARM, E-SWARM, Swarmix) and is used by 15<p>universities worldwide. While the core architecture of ARGoS is<p>complete, extensions are continually added by a community of<p>contributors. In particular, ARGoS was the first robot simulator to be<p>integrated with the ns3 network simulator, yielding a software<p>able to simulate both the physics and the network aspects of a<p>swarm. Further extensions under development include support for<p>large-scale modular robots, construction of 3D structures with<p>deformable material, and integration with advanced statistical<p>analysis tools such as MultiVeStA. / Doctorat en Sciences de l'ingénieur / info:eu-repo/semantics/nonPublished