Suchergebnisse

The continuously increasing of the intercontinental trade as well as the globalization lead to the need of higher productivity for ports and container terminals. In order to fulfill the need of competitiveness of terminal operators and in the same time respect the environmental legislation, container handling machine manufacturers are developping new solutions. Hybridization of energy sources is an interesting way but using more energy efficient actuation lines is also suitable. As a consequence, specifications are more and more complex, and can in particular have heterogeneous natures. In this thesis, the focus is made on two main topics linked to those problematics. On the one hand, a methodology concerning the sizing of over-actuated systems thanks to a coupling between inverse problems and dynamic optimization is exposed. On the other hand, the design, control strategies and an experimental validation of an hybrid actuation line applied to an container handling machine is presented. First of all, we propose a new procedure based on the bond graph language allowing the designer to take into account simultaneously two types of specifications, namely a cost function to minimize and functions of time specifying desired outputs of the system. This results in a coupled problem of state-space inversion and optimization. The procedure lead to a bond graph from which it is possible to directly derive the analytical system of the problem. The fundamental theory for proving the effectiveness of this procedure is carried out using the port hamiltonian systems. The bond graph representation of an optimal control problem is then extended to systems involving non linearities on dissipative R elements. In a second part, a new actuation architecture is proposed for a container handling machine, in order to improve the fuel efficiency. On the one hand, a transformer based system to recover the potential energy released during container lowering is exposed. This energy is stored into a hydropneumatic accumulator. Then, it is returned thanks to an extra hydraulic motor coupled to the engine shaft. On the other hand, a resizing of the main hydraulic pumps as well as a more suitable control law is proposed in order to make the engine work at better efficiency points. All those evolutions previously mentionned give now the possibility to perform kinetic energy recuperation during vehicle deceleration without adding any major component. The new architecture combined with more effetive control laws lead to a fuel consumption reduction of 16% up to 18%. Finally, the potential energy recovery system is validated on a test rig. The control laws are implemented and the dynamic and energetic performances are then analysed. ; L'accroissement permanent du commerce intercontinental et la mondialisation s'accompagnent d'une augmentation de la productivité des ports et des terminaux à conteneurs. Le besoin de compétitivité des exploitants auquel s'additionnent les législations environnementales, poussent les fabricants d'engins de manutention de conteneurs à développer de nouvelles solutions afin de produire des machines plus respectueuses de l'environnement à un coût de possession acceptable. Pour ce faire, l'hybridation des sources d'énergie et l'utilisation de chaînes d'actionnement plus efficaces d'un point de vue énergétique sont les solutions envisagées. En conséquence, les cahiers des charges sont de plus en plus complexes, et les spécifications qu'ils stipulent sont de nature hétérogène. Ce travail de thèse s'intéresse à deux aspects majeurs en lien avec ces problématiques. Il s'articule donc d'une part autour d'un développement méthodologique concernant le dimensionnement de systèmes suractionnés par couplage de problèmes d'inversion et d'optimisation, et d'autre part, autour d'un développement applicatif visant la conception, la commande et la validation expérimentale d'une chaîne d'actionnement hybride d'un engin portuaire. Premièrement, nous proposons une procédure de construction d'un modèle s'appuyant sur le langage bond graph permettant de prendre en compte des spécifications exprimées en termes de critère à minimiser et de fonctions du temps désirées en sortie du système. Ce problème prend alors la forme d'un problème couplé d'inversion-optimisation. Du modèle bond graph obtenu, il est possible de déterminer directement les conditions nécessaires d'optimalité, et donc le système d'équations algébrodifférentielles associé. Nous nous intéressons ensuite à l'extension de la procédure de représentation d'un problème d'optimisation seul, dans le cas où des non linéarités apparaissent au niveau des phénomènes de dissipation d'énergie. Dans un second temps, nous proposons une nouvelle architecture pour un engin de manutention de conteneurs afin d'améliorer son efficacité énergétique. Cette solution de récupération d'énergie à la descente du conteneur est basée sur un transformateur hydraulique. L'énergie est stockée dans un accumulateur hydropneumatique, puis, elle est restituée grâce à un moteur hydraulique complémentaire couplé au moteur diesel. Un redimensionnement de la chaîne d'actionnement hydraulique ainsi qu'une stratégie de commande plus adaptée sont également proposées. Ces modifications offrent la possibilité de réaliser une récupération d'énergie à la décélération du véhicule sans ajout de composant majeur. L'architecture globale obtenue associée à une stratégie de commande adéquate génèrent une réduction de la consommation de carburant de 16% à 18%. Finalement, une validation expérimentale du système de récupération d'énergie potentielle est effectuée et les performances dynamiques et énergétiques sont analysées avec la commande proposée.

Control of communication and information is a key asset in the strategies of power, whether military, political or commercial. Whoever is able to transmit information faster takes advantage of others. This is the engine of growth and progress in the field of telecommunications. The growing ubiquity of communicating devices demonstrates the exponential growth experienced by this area since the first wireless communications. At the time of all-in-one, multiple applications within a single device requires the use of ever more powerful and small components. At the heart of these systems, filters are becoming increasingly important. In a space environment, reliability constraints and space are particularly dramatic. The choice of materials is also limited, which reduces the possibilities of innovation. However, the improvement of accuracy and reliability in manufacturing technologies opens new opportunities for innovation and improved components. This work has been motivated by the desire to bring more performances and reliability, a smaller footprint by taking advantage of the potential offered by combining the concept SIW and LTCC. The results update serious technological dispersions. However, the potential association SIW / LTCC is shown, and the difficulties are surmountable. SIW filters in LTCC therefore have advantages for itself as a serious alternative to existing solutions. ; La maîtrise de la communication et de l'information est un atout primordial dans les stratégies de pouvoir, qu'elles soient militaires, politiques ou commerciales. Celui qui est capable de transmettre l'information plus vite prend l'avantage sur les autres. Tel est le moteur de la croissance et du progrès dans le domaine des télécommunications. L'omniprésence grandissante des dispositifs communicants témoigne de l'expansion exponentielle qu'a connu ce domaine depuis les premières communications sans fil. À l'époque du all-in-one, la multiplication des applications au sein d'un même appareil nécessite l'utilisation de composants toujours ...

Control of communication and information is a key asset in the strategies of power, whether military, political or commercial. Whoever is able to transmit information faster takes advantage of others. This is the engine of growth and progress in the field of telecommunications. The growing ubiquity of communicating devices demonstrates the exponential growth experienced by this area since the first wireless communications. At the time of all-in-one, multiple applications within a single device requires the use of ever more powerful and small components. At the heart of these systems, filters are becoming increasingly important. In a space environment, reliability constraints and space are particularly dramatic. The choice of materials is also limited, which reduces the possibilities of innovation. However, the improvement of accuracy and reliability in manufacturing technologies opens new opportunities for innovation and improved components. This work has been motivated by the desire to bring more performances and reliability, a smaller footprint by taking advantage of the potential offered by combining the concept SIW and LTCC. The results update serious technological dispersions. However, the potential association SIW / LTCC is shown, and the difficulties are surmountable. SIW filters in LTCC therefore have advantages for itself as a serious alternative to existing solutions. ; La maîtrise de la communication et de l'information est un atout primordial dans les stratégies de pouvoir, qu'elles soient militaires, politiques ou commerciales. Celui qui est capable de transmettre l'information plus vite prend l'avantage sur les autres. Tel est le moteur de la croissance et du progrès dans le domaine des télécommunications. L'omniprésence grandissante des dispositifs communicants témoigne de l'expansion exponentielle qu'a connu ce domaine depuis les premières communications sans fil. À l'époque du all-in-one, la multiplication des applications au sein d'un même appareil nécessite l'utilisation de composants toujours ...

Environmental concerns are increasingly impacting customers' behavior as well as companies' strategies. In this context, optimizing the supply chain with environmental considerations is becoming a critical issue. In this thesis we focus on this issue. An originality of our work is to show that supply chain design decisions (facility location, choice of production processes, transportation modes, etc.) impact the environmental quality of a product and therefore affect the level of its demand and / or its selling price. Demand and price are no longer considered as exogenous variables as often done in supply chain optimization existing works, but rather as endogenous variables. In this work we propose a set of supply chain optimization models that take into account the product environmental quality and customers' preferences towards product environmental quality. Based on a case study from the textile sector, we show the relevance of considering the products' environmental quality in supply chain optimization models. ; Les enjeux environnementaux impactent de plus en plus le comportement des clients ainsi que les stratégies des entreprises. Dans ce contexte, l'optimisation de la chaine logistique en tenant compte des questions environnementales devient un enjeu critique. Dans cette thèse nous nous intéressons à cette problématique. Une originalité de ce travail est de montrer que les décisions prises lors de la conception d'une chaine (localisation des sites, choix des processus de production, de transport,…) ont une influence sur la qualité environnementale d'un produit perçue par les clients, et influent donc sur le niveau de la demande et/ou prix de ces produits. La demande et le prix ne sont plus des variables exogènes comme pratiquement toujours considéré dans les travaux existants, mais bien des variables endogènes. Dans cette thèse nous proposons alors des modèles d'optimisation des activités logistiques en tenant compte de la qualité environnementale des produits élaborés et des préférences clients. Nous ...

Environmental concerns are increasingly impacting customers' behavior as well as companies' strategies. In this context, optimizing the supply chain with environmental considerations is becoming a critical issue. In this thesis we focus on this issue. An originality of our work is to show that supply chain design decisions (facility location, choice of production processes, transportation modes, etc.) impact the environmental quality of a product and therefore affect the level of its demand and / or its selling price. Demand and price are no longer considered as exogenous variables as often done in supply chain optimization existing works, but rather as endogenous variables. In this work we propose a set of supply chain optimization models that take into account the product environmental quality and customers' preferences towards product environmental quality. Based on a case study from the textile sector, we show the relevance of considering the products' environmental quality in supply chain optimization models. ; Les enjeux environnementaux impactent de plus en plus le comportement des clients ainsi que les stratégies des entreprises. Dans ce contexte, l'optimisation de la chaine logistique en tenant compte des questions environnementales devient un enjeu critique. Dans cette thèse nous nous intéressons à cette problématique. Une originalité de ce travail est de montrer que les décisions prises lors de la conception d'une chaine (localisation des sites, choix des processus de production, de transport,…) ont une influence sur la qualité environnementale d'un produit perçue par les clients, et influent donc sur le niveau de la demande et/ou prix de ces produits. La demande et le prix ne sont plus des variables exogènes comme pratiquement toujours considéré dans les travaux existants, mais bien des variables endogènes. Dans cette thèse nous proposons alors des modèles d'optimisation des activités logistiques en tenant compte de la qualité environnementale des produits élaborés et des préférences clients. Nous montrons grâce à des expérimentations sur un cas d'étude du secteur de textile, la pertinence de la considération de la qualité environnementale des produits dans les modèles d'optimisation des activités logistiques.

This work focus on the design of wideband and high efficiency GaN MMIC high power amplifiers for military applications such as radar and electronic warfare. The main objectives consist in finding innovative power combining structures in order to decrease the overall size of amplifiers and increasing their efficiency at the same time. For these matters, an important part of this work consisted in the design and realization of ultra compact and low loss power combiners. Once the combiners realized and measured, they are integrated into power amplifiers to prove their functionality and the advantages they bring. Several kind of amplifiers have been realized whether regrind their architecture or their performances. ; Ces travaux de thèse se concentrent sur la conception d'amplificateur de puissance MMIC large-bande haut rendement en technologie GaN pour des applications militaires de type radar et guerre électronique. Les objectifs principaux sont de proposer des structures innovantes de combinaison de puissance notamment pour réduire la taille des amplificateurs actuels tout en essayant d'améliorer leur rendement dans le même temps. Pour cela, une partie importante de ces travaux consiste au développement de combineurs de puissance ultra compactes et faibles pertes. Une fois ces combineurs réalisés et mesurés, ils sont intégrés dans des amplificateurs de puissance afin de prouver leur fonctionnalité et les avantages qu'ils apportent. Différents types d'amplificateur tant au niveau de l'architecture que desperformances sont réalisés au cours de ces travaux.

This thesis deals with the design of distributed algorithms, network protocols, and cooperative applications with an object-oriented approach. This process requires the use of numerous system and network software components. We propose several tools and formalisms to carry out this activity. The originality of our approach is, first, to extend the local point of view of existing object-oriented methods in order to integrate the behaviors of groups of distributed objects, and second, to set up a distributed programming language. Thus, we propose a development process with three methodological levels: group, object, and method. This top-down process introduces, with successive refinement steps, more and more details in the behavioral models. These three levels deals, respectively, with distribution, concurrency, and sequentiality. In this thesis, we are interested in the first two levels. The group level deals with the inter-objects coordination of distributed objects. The interactions inside such groups are defined in terms of knowledge exchanges. They can be seen as the consequences of the run of some global actions performed by all the objects of the group. We put forward four global actions, called distributed control structures, frequently used in distributed applications: the phase, the distributed condition, the distributed iteration, and the distributed recursion. They can be seen as the extension, at a distributed level, of the four following basic algorithmic statements: the sequence, the \em case \em and the \em if \em statements, the \em while \em loop, and the backtrack traversal. We use an epistemic logic to describe the knowledge levels reached by the run of these structures. We propose a syntax, called group level knowledge based program, to describe the global actions and the epistemic predicates used by an application. The object level deals with intra-object coordination. According to the fact that inter-objects coordination is performed by objects whose concurrent activities need to be synchronized, the object level is a refinement of the group level. To carry out the description of intra-object coordination, we propose the CAOLAC language. We implemented it on top of the object language of the GUIDE distributed system. The CAOLAC language is a meta-object protocol that separates synchronization tasks from sequential ones. The former are defined in meta-classes, and the latter in classes. The originality of the CAOLAC language is to use both some state/transition models and some usual object-oriented code to write meta-classes. Each object is associated with a meta-object that traps method calls and that performs some coordination before delivering them to the object. The advantage of this approach is to clearly separate the different functionalities, and to facilitate the reuse of synchronization policies. The semantics of the CAOLAC language is partially defined with Lamport temporal logic of actions. We illustrate our approach with two examples. We present the design of a spanning tree construction distributed algorithm, and of a transactional two phases commit protocol. ; Dans cette thèse, nous nous intéressons à la conception d'algorithmes répartis, de protocoles réseaux et d'applications coopératives avec une approche objet. Cette activité, souvent complexe, nécessite la mise en \oe uvre de nombreux mécanismes systèmes et réseaux. Nous proposons donc différents outils et formalismes permettant de mener à bien cette tâche. L'originalité de notre approche est, d'une part, d'étendre le point de vue local des méthodologies de conception existantes afin d'intégrer les comportements de groupes d'objets distribués, et d'autre part, de proposer les premiers éléments d'une démarche systématique d'algorithmique répartie. Ainsi, nous proposons un processus de développement en trois niveaux méthodologiques : groupe, objet et méthode. Ce processus de type descendant permet d'introduire, par raffinements successifs, de plus en plus de détails dans les modèles de comportements. Ces trois niveaux traitent, respectivement, des aspects liés à la distribution, à la concurrence et aux traitements séquentiels. Dans cette thèse, nous nous intéressons aux deux premiers niveaux. Le niveau groupe concerne la coordination des comportements de groupes d'objets distribués. Nous modélisons les interactions au sein de tels groupes en terme d'échanges de connaissances. Ceux-ci peuvent être vus comme la conséquence d'actions globales entreprises par l'ensemble des objets du groupe. Nous avons mis en évidence quatre actions globales, ou structures de contrôle de groupe, qui apparaissent dans de nombreuses applications distribuées : le schéma de phasage, la conditionnelle distribuée, l'itération distribuée et la récursion distribuée. Elles peuvent être vues comme l'extension, à un niveau réparti, des structures algorithmiques de base que sont, respectivement, la séquence, les instructions de type \em case \em ou \em if\em , les boucles \em while \em et le parcours récursif. Nous utilisons une logique épistémique pour décrire les différents niveaux de connaissance atteints lors de l'exécution de ces structures. Nous proposons une notation appelée programme à base de connaissances de niveau groupe pour exprimer les actions globales et les prédicats épistémiques utilisés par une application. Le niveau objet s'intéresse à la coordination des comportements internes à un objet. C'est un raffinement du niveau précédent, au sens où la coordination inter-objets est implantée par des objets dont les activités concurrentes nécessitent une synchronisation. Pour mener à bien la description de cette coordination intra-objet, nous proposons le langage CAOLAC. Nous en avons réalisé une implantation au-dessus du langage objet du système distribué GUIDE. Le langage CAOLAC se présente sous la forme d'un protocole méta-objet et sépare les aspects de synchronisation des traitements effectifs. Les premiers sont définis dans des méta-classes, tandis que les seconds le sont dans des classes. L'originalité du langage CAOLAC est d'utiliser conjointement un modèle à base d'états et de transitions et du code objet habituel pour l'écriture des méta-classes. Ainsi, chaque objet est associé à un méta-objet qui intercepte les invocations de méthodes et les coordonne avant de les délivrer à l'objet. L'avantage de cette approche est de séparer clairement les différents fonctionnalités et de faciliter la réutilisation des politiques de synchronisation. La sémantique du langage CAOLAC est définie, partiellement, par une logique temporelle, la logique temporelle d'actions de Lamport. Enfin, nous illustrons notre propos par deux études de cas. Nous présentons la conception d'un algorithme réparti de calcul d'arbres couvrants et d'un protocole transactionnel de validation à deux phases.

Since the 2010s, strategic local policies concerning urban resilience and biodiversity support are being promoted, focusing on ways to renew urban water management. Research points however, to the low effectiveness or efficiency of such urban environmental policies, and questions the need of renewing the relationship between the environment and the city. Other researches put forward that a major change occurs in urban rainwater management since the 1970s. We follow the idea of an ecological and climatic transformation of rainwater and urban rivers management since the 1970s. Three hypotheses underlie this thesis. First, based on ecological and climatic principles, a technical management doctrine transforms the urban water engineering. Secondly, urban hydrology has experienced a shift, carried out by new techniques. Thirdly, this technical doctrine has extended to urban planning and design. We tested the hypotheses by the study of a corpus of guides and technical documents produced between 1970 and 2015; of an inventory of techniques; and of five case-studies, completed with interviews with professionals. The study found that ecology is present, but as an auxiliary to sanitation principles. An ecological ethic is incorporated into technical doctrines : the use of living organisms in engineering is partially based their supposed superior effectiveness in spatial planning. The hydrology of the Paris metropolitan area seems to be marginally evolving. Urban neighborhoods are not transformed by water management : urban water spaces and facilities remain tenuous, discontinuous, underlying and preceded by other logics of spatial organization. ; Dans les années 2010, des politiques urbaines de résilience et de soutien de la biodiversité sont mises en avant, avec pour cœur une gestion renouvelée de l'eau en ville. Pourtant, les travaux de recherche font plutôt état d'une faible efficacité des politiques urbaines environnementales, et interrogent la désirabilité d'une réforme du rapport entre environnement et ville. Les ...

Since the 2010s, strategic local policies concerning urban resilience and biodiversity support are being promoted, focusing on ways to renew urban water management. Research points however, to the low effectiveness or efficiency of such urban environmental policies, and questions the need of renewing the relationship between the environment and the city. Other researches put forward that a major change occurs in urban rainwater management since the 1970s. We follow the idea of an ecological and climatic transformation of rainwater and urban rivers management since the 1970s. Three hypotheses underlie this thesis. First, based on ecological and climatic principles, a technical management doctrine transforms the urban water engineering. Secondly, urban hydrology has experienced a shift, carried out by new techniques. Thirdly, this technical doctrine has extended to urban planning and design. We tested the hypotheses by the study of a corpus of guides and technical documents produced between 1970 and 2015; of an inventory of techniques; and of five case-studies, completed with interviews with professionals. The study found that ecology is present, but as an auxiliary to sanitation principles. An ecological ethic is incorporated into technical doctrines : the use of living organisms in engineering is partially based their supposed superior effectiveness in spatial planning. The hydrology of the Paris metropolitan area seems to be marginally evolving. Urban neighborhoods are not transformed by water management : urban water spaces and facilities remain tenuous, discontinuous, underlying and preceded by other logics of spatial organization. ; Dans les années 2010, des politiques urbaines de résilience et de soutien de la biodiversité sont mises en avant, avec pour cœur une gestion renouvelée de l'eau en ville. Pourtant, les travaux de recherche font plutôt état d'une faible efficacité des politiques urbaines environnementales, et interrogent la désirabilité d'une réforme du rapport entre environnement et ville. Les ...

Since the 2010s, strategic local policies concerning urban resilience and biodiversity support are being promoted, focusing on ways to renew urban water management. Research points however, to the low effectiveness or efficiency of such urban environmental policies, and questions the need of renewing the relationship between the environment and the city. Other researches put forward that a major change occurs in urban rainwater management since the 1970s. We follow the idea of an ecological and climatic transformation of rainwater and urban rivers management since the 1970s. Three hypotheses underlie this thesis. First, based on ecological and climatic principles, a technical management doctrine transforms the urban water engineering. Secondly, urban hydrology has experienced a shift, carried out by new techniques. Thirdly, this technical doctrine has extended to urban planning and design. We tested the hypotheses by the study of a corpus of guides and technical documents produced between 1970 and 2015; of an inventory of techniques; and of five case-studies, completed with interviews with professionals. The study found that ecology is present, but as an auxiliary to sanitation principles. An ecological ethic is incorporated into technical doctrines : the use of living organisms in engineering is partially based their supposed superior effectiveness in spatial planning. The hydrology of the Paris metropolitan area seems to be marginally evolving. Urban neighborhoods are not transformed by water management : urban water spaces and facilities remain tenuous, discontinuous, underlying and preceded by other logics of spatial organization. ; Dans les années 2010, des politiques urbaines de résilience et de soutien de la biodiversité sont mises en avant, avec pour cœur une gestion renouvelée de l'eau en ville. Pourtant, les travaux de recherche font plutôt état d'une faible efficacité des politiques urbaines environnementales, et interrogent la désirabilité d'une réforme du rapport entre environnement et ville. Les ...

In urban and environmental planning, the private engineering actors hold a growing place in the implementation of public policies. At times considered as "experts", these invisible elements of projects are known for assisting local authorities on technical dimensions. Incidentally, they are not impervious to the contemporary mutations of capitalism, which affect local public action. In this context, they become actors of change and support renewed approaches to face a supposed complexification. Based on an ethnographic investigation, the thesis describes the work of engineering and consulting firms on local public action. The framework characterized by the pragmatic sociology of tests allows us to understand the work carried out to develop mechanisms, as well as to produce categories and critiques, which is the guiding principle of their activity. This research work documents a global normative change around the concept of project at the expense of a technicist rationality, which modifies local public action where private engineering represents a mirror. Firstly, in terms of organization and work, the expert's practices are downgraded in the face of the animation capabilities given to the project's leader. However, the latter is subjected to the contradictory orders stemming from the permanent discussion. Secondly, facing the client, engineering is increasingly known as a "consultant" to transform the executive relationship into a partnership. This coordination creates the negotiation of needs throughout the mission. Moreover, the global changes that divide the work into a planning work encourage a more local and uncertain coordination, at the expense of a technocratic State. Relying on private engineering is seen as a way to regain flexibility between an increasing number of actors. Lastly, the engineering work throughout the projects showcases, beyond technique, the important political work in the attribution of roles to carry out a project. Ultimately, this study shows the evolution of coordination in the ...

In urban and environmental planning, the private engineering actors hold a growing place in the implementation of public policies. At times considered as "experts", these invisible elements of projects are known for assisting local authorities on technical dimensions. Incidentally, they are not impervious to the contemporary mutations of capitalism, which affect local public action. In this context, they become actors of change and support renewed approaches to face a supposed complexification. Based on an ethnographic investigation, the thesis describes the work of engineering and consulting firms on local public action. The framework characterized by the pragmatic sociology of tests allows us to understand the work carried out to develop mechanisms, as well as to produce categories and critiques, which is the guiding principle of their activity. This research work documents a global normative change around the concept of project at the expense of a technicist rationality, which modifies local public action where private engineering represents a mirror. Firstly, in terms of organization and work, the expert's practices are downgraded in the face of the animation capabilities given to the project's leader. However, the latter is subjected to the contradictory orders stemming from the permanent discussion. Secondly, facing the client, engineering is increasingly known as a "consultant" to transform the executive relationship into a partnership. This coordination creates the negotiation of needs throughout the mission. Moreover, the global changes that divide the work into a planning work encourage a more local and uncertain coordination, at the expense of a technocratic State. Relying on private engineering is seen as a way to regain flexibility between an increasing number of actors. Lastly, the engineering work throughout the projects showcases, beyond technique, the important political work in the attribution of roles to carry out a project. Ultimately, this study shows the evolution of coordination in the ...

Smart multi-sensor systems for gas detection (or electronic noses), are already deployed in areas as diverse as cosmetics and food industry, environment monitoring and military and medical purposes. The ongoing collaboration within the European SNIFFER Project, focused on the development of an innovative electronic nose-like system based on MEMS sensors combined with olfactory bioreceptors for the detection of illicit substances. In another context and, within an internal approach to ESIEE-Paris, we have also chosen to expand the topic to design a multi-sensor system for application in volatile organic compounds detection.Thus, in view of improving these technologies through the use of MEMS sensors, this manuscript comprehensively investigates the design, fabrication and characterization of silicon and diamond resonant micro-cantilevers sensors, and also presents a proof of concept of a multi-sensor system for gas detection. To date, silicon has been used as a building block of micro sensors, whose features are very well known and argued extensively in the literature. On the other hand, diamond as a unique material in terms of its superior physical and chemical properties, has been received attention in microelectronics. Although the realization of diamond MEMS sensors has been presented before, it has been never compared to silicon MEMS gas sensors. Therefore, to establish elements of comparison for the two technologies, this thesis aimed to use these two competitive materials as a building block of micro-cantilever based MEMS gas sensors.Firstly, we set up specific micro-machining processes for the realization of diamond patterns that have been previously developed in our laboratory in order to optimize them for a fully clean room compatible manufacturing, independent of changes in diamond synthesis conditions. We have, for the first time, realized diamond micro-cantilevers with integrated polysilicon gauges. The Young's modulus of diamond was characterized using two different methods resulting in the best ...

The UAVs (Unmanned Aerial Vehicle) have long been a military tool. In recent years, the advent of mini-UAVs, smaller, lighter and less expensive than their counterpart, extended the use of UAVs in civil domain. They offer the solution to many dangerous operations such as search and rescue of survivors after natural disasters, cartography, communication relaying between several areas of intervention of firefighters. UAVs replace piloted airplanes and helicopters which may avoid possible human casualties. In order to enhance the performance of UAVs, Researches is being carried out to make cooperative aircraft. A swarm of cooperative UAVs sharing different tasks between drones, will be able to perform complex operation faster. Such a mission requires a high level of coordination between UAVs made possible by a frequent exchange of information between them and their control station. The ad hoc network is a promising solution to have the UAV communicate with each other (inside the swarm) and with the control station (outside the swam). It allows UAVs to perform in difficult situations such as mountains and after natural disasters. During a mission, a swarm of cooperative UAVs, having different role, exchanges a variety of message type between the aircraft and their control station. These messages may have different needs in term of quality of service (QoS) which varies with the task given to the transmitter. The needs of each UAV may vary depending on the content of the message and over time with the evolution of its task. Therefore, QoS must be managed by a module aware of these evolutions and able to provide each track requirements throughout the operation. This research work, financed by the European project D3CoS 1, proposes a communication system, called DAN (DCoS (Distributed Cooperative System) Ad hoc Network), for a swarm of cooperative UAVs (Unmanned Aerial Vehicle). This system provides a service differentiation according to a variety of criteria chosen by applications such as the role or the task assigned to the node, the message, etc. DAN architecture is a combination of several QoS mechanisms created to be used in real environment with Paparazzi UAVs. Therefore, it was evaluated using two evaluation means: simulation and real experiments. Simulations were realized in realistic environment using a specific mobility model, called PPRZM (PaPaRaZzi Mobility). PPRZM is a stochastic mobility model that imitates Paparazzi UAVs movement designed and implemented for this study. It was validated using a comparative study between PPRZM, the well known mobility model RWP (Random Way-Point) and real Paparazzi UAV movement traces. In addition, DAN was evaluated by real experiments. It was implemented on embedded Linux on Raspberry Pi cards in order to be integrated to Paparazzi UAVs. Several tests were performed to validate the operation of the new Paparazzi communication system and the simulation results. This work creates a new communication system for Paparazzi system that can provide the suitable medium to operate UAVs over an ad hoc network and to respect the need of each agent in term of QoS. ; Les drones, engins volants sans pilotes généralement appelés UAV pour Unmanned Aerial Vehicle, ont longtemps été un outil militaire. Ces dernières années, l'avènement de micro-drones, plus petits, légers et moins coûteux que leurs homologues a étendu l'utilisation des UAV au domaine civil.Afin d'améliorer les performances des drones sur de telles missions, des recherches sont menées afin de rendre les drones coopératifs. Une flotte de drones coopératifs serait alors capable de réaliser plus rapidement des missions plus complexes en partageant les didifférentes taches entre les UAV. Ce genre d'opération nécessite un niveau élevé de coordination entre les drones rendu possible par un échange continu d'information entre eux et avec leur station de contrôle. Cet échange permet aux contrôleurs d'ajuster la mission selon les circonstances. La communication est donc vue comme un enjeu majeur dans l'évolution des opérations de ce genre de systèmes. Le réseau ad hoc est une solution prometteuse pour faire communiquer les drones entre eux et avec la station sol. Ses capacités d'auto-organisation lui permettent de fonctionner dans des situations difficiles comme des canyons urbains ou après les catastrophes naturelles.Au cours d'une mission, une flotte de drones coopératifs, ayant chacun un rôle différent des autres, échange une variété de types de messages envoyés entre les UAV sur la liaison descendante (des drones vers la station sol) et sur la liaison montante. Ces messages ont des priorités et des besoins en termes de la qualité de service (QdS) qui différent selon la tâche accordée a l'émetteur.Ces besoins peuvent varier selon le contenu du message et au cours du temps avec l'évolution et le changement de la tache. Par conséquent, la QdS doit être gérée par un module conscient de ces évolutions et capable de s'adapter a la variation des besoins et de fournir a chaque trafic sa demande durant toute l'opération.Cette thèse, financée par le projet européen D3CoS, propose un système nomme DAN « DCoS (Distributed Cooperative Systems) Ad hoc Network », offrant des garanties en termes de différenciation de services selon une variété de critères choisis par les applications comme la tache allouée a l'émetteur et le contenu du message. Ce système permet de proposer aux agents collaboratifs différents niveaux de QdS en fonction du type d'application utilisée qui peut être variable au cours du temps. DAN est composé d'un ensemble de modules permettant à une application d'exprimer ces besoins en termes de QdS. Il permet, par la suite, de classifier les paquets émis selon la demande et le changement des besoins de l'application dans le but de garantir la QdS demandée et de respecter les priorités de chaque classe. Le système DAN devant être utilise en environnement réel, il a été évalué a l'aide de deux moyens : la simulation et l'expérimentation réelle. Les simulations ont été réalisées dans un environnement réaliste en utilisant un modèle de mobilité reproduisant le mouvement des drones Paparazzi et nommé PPRZM pour (PaPaRazzi Mobility). PPRZM a _été implémenté et validé par une étude comparative avec le modèle de mobilité (Random Way-Point) et des traces de mouvements réels de drones Paparazzi. Par la suite, DAN a été évalué par des expérimentations réelles. Il a été implémenté sur Linux embarqué sur des cartes Raspberry Pi intégrées aux drones Paparazzi. Plusieurs tests ont été réalisée permettant de valider les résultats obtenus par simulations. L'intégration de ce de système aux drones Paparazzi permettra de faire _évoluer les missions impliquant une flotte de drones puisque avant cette réalisation, des fonctionnalités comme l'évitement de collisions ou l'allocation dynamique de tâches impliquaient directement la station sol, comme point de relai ou de prise de décision. Le système proposé permet de faciliter les échanges dans ce genre de missions tout en garantissant la QdS pour chaque drone.

The UAVs (Unmanned Aerial Vehicle) have long been a military tool. In recent years, the advent of mini-UAVs, smaller, lighter and less expensive than their counterpart, extended the use of UAVs in civil domain. They offer the solution to many dangerous operations such as search and rescue of survivors after natural disasters, cartography, communication relaying between several areas of intervention of firefighters. UAVs replace piloted airplanes and helicopters which may avoid possible human casualties. In order to enhance the performance of UAVs, Researches is being carried out to make cooperative aircraft. A swarm of cooperative UAVs sharing different tasks between drones, will be able to perform complex operation faster. Such a mission requires a high level of coordination between UAVs made possible by a frequent exchange of information between them and their control station. The ad hoc network is a promising solution to have the UAV communicate with each other (inside the swarm) and with the control station (outside the swam). It allows UAVs to perform in difficult situations such as mountains and after natural disasters. During a mission, a swarm of cooperative UAVs, having different role, exchanges a variety of message type between the aircraft and their control station. These messages may have different needs in term of quality of service (QoS) which varies with the task given to the transmitter. The needs of each UAV may vary depending on the content of the message and over time with the evolution of its task. Therefore, QoS must be managed by a module aware of these evolutions and able to provide each track requirements throughout the operation. This research work, financed by the European project D3CoS 1, proposes a communication system, called DAN (DCoS (Distributed Cooperative System) Ad hoc Network), for a swarm of cooperative UAVs (Unmanned Aerial Vehicle). This system provides a service differentiation according to a variety of criteria chosen by applications such as the role or the task assigned ...