Thèses sur le sujet « Dynamique asynchrone »
Créez une référence correcte selon les styles APA, MLA, Chicago, Harvard et plusieurs autres
Consultez les 46 meilleures thèses pour votre recherche sur le sujet « Dynamique asynchrone ».
À côté de chaque source dans la liste de références il y a un bouton « Ajouter à la bibliographie ». Cliquez sur ce bouton, et nous générerons automatiquement la référence bibliographique pour la source choisie selon votre style de citation préféré : APA, MLA, Harvard, Vancouver, Chicago, etc.
Vous pouvez aussi télécharger le texte intégral de la publication scolaire au format pdf et consulter son résumé en ligne lorsque ces informations sont inclues dans les métadonnées.
Parcourez les thèses sur diverses disciplines et organisez correctement votre bibliographie.
Van, Ham Philippe. « Dynamique des systèmes : une approche logique asynchrone ». Doctoral thesis, Universite Libre de Bruxelles, 1989. http://hdl.handle.net/2013/ULB-DIPOT:oai:dipot.ulb.ac.be:2013/213282.
Texte intégralHsu, Tiente. « Proposition d'une architecture de réseau d'interconnexion à reconfiguration dynamique et asynchrone ». Lille 1, 1993. http://www.theses.fr/1993LIL10052.
Texte intégralBudau, Victor. « Gestion dynamique des communications dans les systèmes distribués par commutation synchrone/asynchrone ». Evry, Institut national des télécommunications, 2003. http://www.theses.fr/2003TELE0002.
Texte intégralAs the programming of the distributed applications has become very complex, new programming models were created to facilitate the comprehension, the design and the implementation ot these applications. Amongst these models, the remote procedure call and the structures based on event notification imposed themselves as successful solutions for distributed programming. The middleware platforms have greatly eased the dialog between heterogenoeous parts of the systems and the component-oriented development helps to manage the complexity, the heterogeneity and the need for adaptability and reconfiguration of these systems. However, these platforms make an implicit, yet natural, association between these programming models and the communication models (synchronous and asynchronous) they support : remote calls are synchronous (simultaneous availability of the communication entities) and applications programmed in an event-based paradigm communicate asynchronously, with messages. The use of both types of communication, rarely addressed due to its programming difficulty, remains aprolific research subject. In this thesis we propose a mechanism of chanching dynamically (during the execution) and transparently 'without changing the application code) the communication model between the components of a distributed system. This mechanism is integrated in a architecture that adds an adaptative behavior to a communication platform. The utility of changing the communication type in the context of large scale or mobile applications, as well as the limits of this approach are studied. The mechanism was implemented on a J2EE open source application server in order to validate the working hypothesis and to perform the necessary benchmarks for performance gain and flexibility of programming comparisons
Canat, Sylvain. « Contribution à la modélisation dynamique d'ordre non entier de la machine asynchrone à cage ». Toulouse, INPT, 2005. http://ethesis.inp-toulouse.fr/archive/00000132/.
Texte intégralInduction machine is most widespread in industry. Its traditional modelling does not take into account the eddy currents in the rotor bars which however induce strong variations as well of the inductance as of the resistance of the rotor. This diffusive phenomenon, called “skin effect” could be modelled by a compact transfer function using the fractional derivative (non integer order). This report theoretically analyzes the electromagnetic phenomenon on a single rotor bar before approaching the rotor as a whole. This analysis is confirmed by the results of finite elements calculations of the magnetic field, exploited to identify a fractional order model of the induction machine (identification's method of Levenberg-Marquardt). Then, the model is confronted with an identification of experimental results. Lastly, an automatic method is carried out to approximate the dynamic model by integer order transfer function on a frequency band
Canat, Sylvain Faucher Jean. « Contribution à la modélisation dynamique d'ordre non entier de la machine asynchrone à cage ». Toulouse : INP Toulouse, 2005. http://ethesis.inp-toulouse.fr/archive/00000132.
Texte intégralPietrzak-David, Maria. « Algorithmes de commande d'un variateur électrique asynchrone : contrôle dynamique du flux : commande optimale de position ». Toulouse, INPT, 1988. http://www.theses.fr/1988INPT056H.
Texte intégralPietrzak, Maria. « Algorithmes de commande d'un variateur électrique asynchrone contrôle dynamique du flux, commande optimale de position / ». Grenoble 2 : ANRT, 1988. http://catalogue.bnf.fr/ark:/12148/cb37612966g.
Texte intégralNi, Zhenjiang. « Vision Asynchrone Événementielle : Algorithmes et Applications à la Microrobotique ». Phd thesis, Université Pierre et Marie Curie - Paris VI, 2013. http://tel.archives-ouvertes.fr/tel-00916995.
Texte intégralKérignard, Denis. « Contribution à la modélisation et à l'identification des paramètres électriques de la machine asynchrone en régime dynamique ». Poitiers, 1997. http://www.theses.fr/1997POIT2325.
Texte intégralMadani, Nadia. « Commande à structure variable d'une machine asynchrone double étoile alimentée par deux onduleurs MLI : modélisation dynamique, alimentation et validation expérimentale ». Nantes, 2004. http://www.theses.fr/2004NANT2105.
Texte intégralSupplying polyphase machine or double star machine by PWM voltage inverters causes currents peaks at each commutation. From a control point of view, these drives are multi-variable and strongly coupled. This thesis deals with the Double Star Asynchronous Machine (DSAM) supplied by two PWM inverters. It's divided in three parts : the first part concerns the elaboration of DSAM's dynamic modelling methodology. After presenting different modelling approaches, a vectorial formalism known concept, for polyphase machines modelling , is generalized to multi stars machines modelling. The second part of this work concerns the drives's behaviour analysis. The alimentation structures's influence (two or three levels inverters ) and their PWM command on the drive's temporal behaviour are investigated. The third part deals with the drive's variable structure control based on sliding mode technics. The objective of the control laws is to control machine flux and torque, and at the same time to minimize the amplitude of the currents circulating between the two stars. A methodology for synthesising a sliding mode control is elaborated for a three phase machine then extended to DSAM. Finally, in order to satisfy the defined objective, we propose a variable structure control algorithm based on the combination of six phase SVPWM control and sliding mode control. This study presents simulation and implementation results, obtained thanks to a benchmark built in the laboratory
Marino, Alban. « Caractérisation et modélisation des pertes magnétiques d'une machine asynchrone fonctionnant en charge ». Phd thesis, Grenoble INPG, 2003. http://tel.archives-ouvertes.fr/tel-00408127.
Texte intégralZhang, Yuan. « Comportement dynamique d'une chaine de traction asynchrone avec l'onduleur de tension MLI : analyse de la stabilité : commande scalaire et vectorielle ». Toulouse, INPT, 1988. http://www.theses.fr/1988INPT067H.
Texte intégralGlacet, Christian. « Algorithmes de routage : de la réduction des coûts de communication à la dynamique ». Phd thesis, Université Sciences et Technologies - Bordeaux I, 2013. http://tel.archives-ouvertes.fr/tel-00951393.
Texte intégralChevalier, Thierry. « Modélisation et mesure des pertes de fer dans les machines électriques : application à la machine asynchrone ». Phd thesis, Grenoble INPG, 1999. http://www.theses.fr/1999INPG0172.
Texte intégralChevalier, Thierry. « Modélisation et mesure des pertes de fer dans les machines électriques : application à la machine asynchrone ». Phd thesis, Grenoble INPG, 1999. http://tel.archives-ouvertes.fr/tel-00764023.
Texte intégralZhang, Yuan. « Comportement dynamique d'une chaîne de traction asynchrone avec un onduleur de tension M.L.I. analyse de la stabilité, commandes scalaire et vectorielle / ». Grenoble 2 : ANRT, 1988. http://catalogue.bnf.fr/ark:/12148/cb37619382r.
Texte intégralGara, Slim. « Allocation dynamique des ressources pour la transmission de la vidéo sur un réseau ATM ». Versailles-St Quentin en Yvelines, 1998. http://www.theses.fr/1998VERS0007.
Texte intégralDiop, Ousmane. « Analyse mathématique de la dynamique de réseaux de régulation biologique ». Electronic Thesis or Diss., université Paris-Saclay, 2020. http://www.theses.fr/2020UPASG013.
Texte intégralIn this thesis, we are interested in the qualitative analysis of the dynamics of two biological cycles that are central in eukaryotic cells, the cell division cycle and the circadian clock. For that purpose, we use asynchronous Boolean networks that provide an adapted qualitative framework. In these networks, cycles are captured by complex attractors containing hundreds of states. A new method for the analysis of such complex attractors is proposed. It is based on the construction of a summary graph of the attractor, enabling the comparison between the attractor's trajectories and qualitative properties of the biological cycle. The method is illustrated on a cell cycle model from the literature and of a circadian clock model we built from an existing continuous model. In both models our method proves to be efficient to visualize the attractor's structure and to compare it with the biological cycle. By combining the summary graph with a Markov chain, proportions of time spent in each phase are estimated. By combining it with a Boolean inference technique, we show how to locally adjust the asymptotic dynamics of the model in order to force specific dynamical properties. These two applications show the interest of our method in the modeling and analysis of cellular regulatory networks
Picard, Marchetto Aymeric. « Isomorphisme de réseaux d'automates et son influence sur le graphe d'interaction ». Electronic Thesis or Diss., Université Côte d'Azur, 2024. http://www.theses.fr/2024COAZ4034.
Texte intégralAn automata network with n components on a finite alphabet Q is a discrete dynamical system described by successives iterations of a function f : Qⁿ → Qⁿ. An important parameter in most applications is the interaction graph: a graph with vertices from 1 to n and with an arc from i to j if f_j(x) depends on x_i. In general, this interaction graph is easier to approximate than the network's actual dynamic. This raises an important question: what does the interaction graph tell us about the network's dynamic? In an attempt to find limits to this question, we study the inverse: what does the network's dynamic tell us about its interaction graph? To this end, we study networks up to isomorphism. Isomorphism preserves most studied properties, but does not preserve the interaction graph. We will thus study G(f), the set of interaction graphs of all networks isomorphic to f. Notably, we prove that K_n, the interaction graph with all arcs, is always in G(f), and it's the only graph with this property. This means if K_n is the interaction graph of f, then it gives no information on f up to isomorphism. Inversely, we show there are networks f that give no information on their interaction graph: G(f) contains all graphs, except the empty one. Finally, we also study the impact of isomorphism on asynchronous dynamics. We show that it preserves very little properties of those dynamics, except the number of fixed points
TARISIEN, Medhi. « Dynamique de la fragmentation de molécules simples induite par impacte d'ion multichargé ». Phd thesis, Université de Caen, 2003. http://tel.archives-ouvertes.fr/tel-00005217.
Texte intégralLe, Nôtre Yvan. « Etude de la réponse dynamique du Bloc Réacteur soumis à une sollicitation extrême : Co-simulation implicite/explicite multi-échelle en temps pour la dynamique du contact ». Electronic Thesis or Diss., Lyon, INSA, 2023. http://www.theses.fr/2023ISAL0055.
Texte intégralNuclear power plants are an important source of energy production in France. However, following the various accidents and risks associated with this technology, nuclear safety is a global concern. In France, standards are continually being imposed on existing nuclear facilities and on the next generations under development. Among the many aspects of nuclear safety, the dimensioning of mechanical structures is an important subject for industrial players. Framatome's core business is the design and justification of nuclear power plants. These must be designed to withstand extreme operating conditions, such as earthquakes, plane crashes or pipe ruptures. Numerical modeling of this type of loading requires dynamic temporal analyses to consider these multi-scale phenomena. However, such analyses are CPU and memory intensive. The aim of this thesis is to develop a new heterogeneous (different integration schemes) and asynchronous (different time steps) integrator, based on the GC coupling method, with improved energy conservation properties. The multi-scale phenomena present in the reactor block are simulated using an explicit integrator with small time steps for the contact zones and an implicit integrator with large time steps for the rest of the structure. A co-simulation demonstrator is developed between the Code Aster and Europlexus software packages in order to get closer to an industrial development and thus demonstrate the performance gain, for a three-dimensional reactor block model, provided by the multi-time-scale co-simulation approaches
Duboux, Thibault. « Régulation dynamique du partitionnement de données sur machines parallèles à mémoire distribuée ». Lyon, École normale supérieure (sciences), 1996. http://www.theses.fr/1996ENSL0009.
Texte intégralLi, Sijia. « Couches absorbantes pour la propagation d'ondes dans les sols non bornés à l’aide de la décomposition en sous-domaines et l'intégration temporelle hybride asynchrone ». Thesis, Lyon, 2019. http://www.theses.fr/2019LYSEI042.
Texte intégralThe purpose of this thesis is to propose novel designs and formulations of different absorbing layers at the boundaries of finite element meshes for the numerical simulation of wave propagation in unbounded 2D and 3D media, in order to deal with complex soil-structure interaction problems in transient dynamics. Three absorbing layers are developed: Rayleigh Absorbing Layer, Kosloff Absorbing Layer and Perfectly Matched Layer (PML). By studying analytically the strong form of elastic wave propagation in each medium, the frequency-independent absorbing ability of Kosloff damping and PML is derived, whereas the attenuation due to Rayleigh damping turns out be frequency-dependent. The interface problem between a non-dissipative elastic medium and a dissipative medium has been investigated by considering the wave propagation in the continuous setting, in order to establish optimal conditions to further reduce the spurious waves reflected at the interface. The three absorbing layers are discretized in space by the finite element method, while the time discretization is based on the powerful and flexible HATI framework (Heterogeneous Asynchronous Time Integrator), enabling adopt the most suitable time integration schemes and time steps, depending on the characteristics and physical phenomena occurring in the different subdomains. The efficiency of three asynchronous hybrid absorbing layers for modeling unbounded domains has been illustrated in various 2D and 3D numerical applications, such as Lamb’s test in stratified heterogeneous media or soil-structure interaction problems involving wave barriers. Finally, implicit and explicit versions of asynchronous PML are implemented and tested in problems with a large number of degrees of freedom. It is proved to be more accurate and efficient than the absorbing layers based on Rayleigh and Kosloff damping
Robert-Dehault, Étienne. « Modélisation dynamique, commande et conception de machines pentaphasées alimentées par des onduleurs MLI ». Nantes, 2005. http://www.theses.fr/2005NANT2119.
Texte intégralHamek, Saliha. « La coopération asynchrone dans les environnements dynamiques : le cas de la gestion des soins ». Paris 8, 2007. http://www.theses.fr/2007PA084191.
Texte intégralThis work aims at enlarging the research field of asynchronous cooperation. We adopt the perspective of working conditions in which the actors are placed in a time asynchronous situation implying sequential and delayed interdependences between them. Models and methods from the ergonomic psychology domain were used to carry out four studies. The work focuses particularly on the possibility of determining and maintaining a common frame of reference by nurses who take care of patients hospitalized at home (H. A. D. In French). The first one defines the organization of H. A. D activity and specifies the cognitive artefacts, which serve as a support for the functional structure of the horizontal cooperation. The second study describes structural cooperation this sequential and time delayed on a coordination implying a strong cognitive synchronization to maintain a common frame of reference. The third study analyses the representations expressed during the care providing process and shows that health care professionals can rebuild the coupling by adding the individuals representations, for which members they do not have a direct and synchronized access. The last study focuses on the “competence required to cooperate” by each coordination experience - priming and facilitation - in the cases of sequential and time delayed interactions for a collective adaptation reinforced progressively towards the maintenance of it he cognitive synchronization of the common frame of reference
Gouy-Pailler, Cédric. « Vers une modélisation dynamique de l'activité cérébrale pour la conception d'interfaces cerveau-machines asynchrones ». Grenoble 1, 2009. http://www.theses.fr/2009GRE10179.
Texte intégralBrain-Computer Interfaces (BCIs) aim at establishing a direct communication pathway between the human brain and an electronic device. In the past twenty years, many breakthroughs have been made in the field, but current systems remain slow and difficult to use. They indeed often constrain users to send their command at precise moments (synchronous systems). Our research work is devoted to asynchronous systems based on electroencephalography (EEG): subjects are allowed to send their commands whenever they want. This entails brain activity of the subject to be analyzed in real time to detect the task-related brain activity among the whole brain activity. Our research work focuses on three main issues raised by asynchronous BCIs: 1. To allow users to send their command whenever they want; 2. To improve the ability of the system to correctly recognize a particular mental task; 3. To increase the robustness of the system when coping with possible variations in the way the subject performs mental tasks. The proposed methods aim at jointly using the spatial (cerebral areas implied), frequential (frequency bands involved) and temporal (succession of activations) features of the different mental tasks in order to compensate the low signal to interference ratio. The approaches are evaluated and compared to state-of-the-art results using datasets from the successive BCI Competitions
Lacressonnière, Fabien. « Conception de structures de conversion d'énergie à double fonctionnalité pour chariots électriques de manutention ». Artois, 2005. http://www.theses.fr/2005ARTO0205.
Texte intégralNowadays, in electric industrial trucks, two distinct power converters are used : on the one hand a speed controller and on the other hand a battery charger. The aim of the study is to use the existing components (speed controller) to realize a battery charger in order to reduce the cost and improve performances. To undertake this study, two industrial trucks have been selected : a transpallet with DC motorization and a fork lift truck with AC motorization. After a brief reminder on the technologies and standards to respect (harmonic pollution, galvanic isolation), the choice of the topology was made on the CUK converter with a galvanic isolation for the transpallet and on a P. W. M. Inverter/rectifier for the fork lift truck. A mechanical study for the two industrials trucks has allowed to conclude on the effectiveness from the regenerative braking. In order to analyse the influence of a charging current with a sinusoidal perturbation on the performances of the lead-acid battery, the electrical model has been introduced. For the transpallet, the full analyse (dimensioning, dynamical model, snubber) of the CUK converter has been done. For the fork lift truck, the induction motor is directly coupled at the main wheel and it is used like as a transformer during the charge of the battery. A dimensioning study of the motor has been presented and its interaction with the P. W. M. Inverter/rectifier has been analysed
Abboud, Mohssen. « Tolérance aux défaillances dans les réseaux dynamiques ». Paris 7, 2008. http://www.theses.fr/2008PA077176.
Texte intégralWe study how some fault-tolerant algorithms for classical Systems can be extended to t}e used in larger scale networks. In the first part of this thesis we address the problems of reliable broadcast and consensus in sensor networks communicating with radio-broadcast. Communication is prone to collision when several sensors broadcast simultaneously. Moreover sensors may crash and stop sending. In this framework, reliable broadcast and consensus are not possible to solve. Sensors are equipped with collision detectors. We propose some specifications of collision detectors that enable us to achieve reliable broadcast, consensus and we give some algorithms for this. In the second part we consider a dynamic network of processes communicating by sending messages. The network is dynamic in the sense that the processes are created dynamically and each process does not know either the number or the set of created processes, but it has a unique identity. Created processes are prone to failure. We study three classical problems of fault-tolerance in the case where the set of processes is unknown. The three problems are: the consensus problem, the implementation of atomic registers and the eventual leader election. For this we consider different models in respect of their degree of synchrony (asynchronous, partially synchronous, synchronous), and we prove how to solve these problems in every considered models
Peschanski, Frédéric. « Composition et adaptation dynamiques de systèmes distribués : une approche à base de composants asynchrones typés ». Paris 6, 2002. http://www.theses.fr/2002PA066293.
Texte intégralHugues, Jean-Guy. « Contribution à l’étude dynamique, lors du démarrage, de chaînes cinématiques a engrenages, entraînées par moteur électrique : Sous-titre eng ». Lyon, INSA, 1993. http://www.theses.fr/1993ISAL0108.
Texte intégralThe purpose of this thesis is the study of the dynamic torsional behaviour, during startup, of gear transmissions having a large flexibility extremal to the gearbox and driven by electrical motor. In order to achieve this, a transmission including an induction motor driving a reducer, a very flexible torsion bar, a step-up gearbox and a variable inertia, was realised and equipped with transducers allowing dynamic measures. A system of measure by optical encoders was specially developed to follow up the evolution of the difference between the input and the output of the reducer and the step-up gearbox. From this measures, the author analysed the dynamic phenomena observed during start-up : taking up play, teeth contact loss, natural frequencies, etc. . . These results will also be used to validate the simulation. A pro gram was developed to simulate the dynamic torsional behaviour of the transmission by taking the interaction between the electrical system and the mechanical system into account This pro gram is based on the model of the generalized electric machine, applied to the induction motors and coup led to a dynamic torsional model of the transmission. The calculation of the eigenvalue and the study of the modal sensitivity coincides with the experimental observations. The confrontation of the calculated responses with those measured shows up the necessity to take the backlash and the casing deformations into account The aim of the dynamic study is to know the influence of the transient state on the fatigue life of the gears. This effect is determined by braking op the load cycles according to the passage of the teeth and by applying the cumulative damage laws
Mouyon, Philippe. « Pseudolinéarisation et pseudocompensation des systèmes non linéaires ». Toulouse, ENSAE, 1986. http://www.theses.fr/1986ESAE0015.
Texte intégralAntonucci, Laura. « Spectroscopie résolue en temps de la femtoseconde à la milliseconde ». Palaiseau, Ecole polytechnique, 2014. https://tel.archives-ouvertes.fr/tel-01094630v2/document.
Texte intégralThe aim of the work was the development of a time-resolved pump-probe measurement system allowing temporal scanning from femtosecond to millisecond time scale. An original method for asynchronous optical sampling, named AD-ASOPS, was developed, patented, and is now undergoing the industrialization phase. The method can be applied to couples of lasers oscillating at arbitrary repetition rates, which is a novelty and opens up many experimental possibilities. Two specific set-ups were realized in the case of laser oscillators (MHz AD-ASOPS) or in the case of amplified laser systems (kHz AD-ASOPS). A feasibility study confirmed many application possibilities. Different tests were performed for a complete characterization and demonstrated a sub-picosecond time resolution of about 400 fs for temporal sampling windows only limited by the period of the pump laser. The implementation flexibility related to AD-ASOPS allowed a pump-probe experiment on biological samples (reaction centre of Rhodobacter Sphaeroides), by using a conventional Ti:Sapphire oscillator at 74. 5 MHz as probe and a CPO oscillator at 5. 1 MHz as pump. The multi-scale dynamics for returning to the fundamental state, from picoseconds to about fifty nanoseconds, could be recorded into only a few minutes of acquisition time with one test bench. The results are in agreement with the literature. Finally, a demonstration experiment of the kHz AD-ASOPS version was performed with two amplified laser systems, by comparison with Fourier transform spectral interferometry. A sub-picosecond time resolution was confirmed over 1 millisecond temporal window
Taouali, Wahiba. « Modélisation de populations neuronales pour l'intégration visuo-motrice : dynamiques et décisions ». Thesis, Université de Lorraine, 2012. http://www.theses.fr/2012LORR0105/document.
Texte intégralWithin the context of enaction and a global approach to perception, we focused on the characteristics of neural computation necessary to understand the relationship between structures in the brain and their functions. We first considered computational problems related to the discretization of differential equations that govern the studied systems and the synchronous and asynchronous evaluation schemes. Then, we investigated a basic functional level : the transformation of spatial sensory representations into temporal motor actions within the visual-motor system. We focused on the visual flow from the retina to the superior colliculus to propose a minimalist model of automatic encoding of saccades to visual targets. This model, based on simple local rules (CNFT and logarithmic projection) in a homogeneous population and using a sequential processing, reproduces and explains several results of biological experiments. It is then considered as a robust and efficient basic model. Finally, we investigated a more general functional level by proposing a computational model of the basal ganglia motor loop. This model integrates sensory, motor and motivational flows to perform a global decision based on local assessments. We implemented an adaptive process for action selection and context encoding through an innovative mechanism that allows to form the basic circuit for other cortico-basal loops. This mechanism allows to create internal representations according to the enactive approach that opposes the computer metaphor of the brain. Both models have interesting dynamics to study from whether a biological point of view or a computational numerical one
Bosilca, George. « Contribution à l'étude des environnements d'exécution parallèle fortement dynamiques à passage de messages ». Paris 11, 2004. http://www.theses.fr/2004PA112316.
Texte intégralGlobal Computing platforms, large scale clusters and future TeraGRID systems gather thousands of nodes for computing parallel scientific applications. These parallel architectures currently uses different approaches depending on the hardware memory model of the architecture, the abstraction level of the programming environment and the nature of the application. On the first part, I introduce an original client-server execution model based on RPCs called out-of-order parallel virtual machine (OVM). OVM aims to provide three main features: portability through a unique memory model, load-balancing using a plug-in support and high performance provided by several optimizations. The main optimizations are: non-blocking RPCs, data-flow management, persistent and non-persistent data, dynamic scheduling and asynchronous global operations. I present OVM general architecture and demonstrate high performance for several types of parallel applications: regular, with load balancing needs and with real-time constraints. At this scale, node failures or disconnections are frequent events. This Volatility reduces the MTBF of the whole system in the range of hour or minutes. On the second part, I present MPICH-V, an automatic fault tolerant MPI environment based on uncoordinated checkpoint/rollback and distributed message logging. MPICH-V architecture relies on Channel Memories, Checkpoint servers and theoretically proven protocols to execute SPMD and Master-Worker MPI applications on volatile nodes. Focusing on the channel memory, I present and analyse the performances for non-trivial parallel applications. Experimental results demonstrate good scalability and high tolerance to faults
Jacquot, Benoît. « Conception, étude et réalisation des algorithmes de commande des systèmes de traction asynchrone pour les TGV de nouvelle génération ». Toulouse, INPT, 1995. http://www.theses.fr/1995INPT115H.
Texte intégralFeddaoui-Papin, Aïda. « Observateurs non linéaires pour les systèmes à mesures asynchrones : application robotique mobile ». Electronic Thesis or Diss., Toulon, 2020. http://www.theses.fr/2020TOUL0008.
Texte intégralThe aim of observability studies and observer design is to reconstruct the state of a dynamic system using the measurements available. In particular, the Kalman filter algorithm is considered. This widely-studied and used observer exists in several versions : for linear or nonlinear systems, for discrete, continuous or even continuous-discrete time, in the stochastic or deterministic framework. However, Most of the time, these observers are used with the assumption that the measurements provided by the sensors are synchronous. Most of the time, this assumption can be far from the physical reality, in particular when dealing with robotic systems. In this memoir, an observer tailored for nonlinear continuous-discrete asynchronous systems is presented. These systems are made of a continuous state equation and a multirate sampled output equation. Based on the existing high-gain Extended Kalman Filter for continuous nonlinear systems and continuous-discrete nonlinear systems with synchronous outputs, we develop an ad-hoc formalism and design an observer with a deterministic point of view. Its convergence is proven analytically and illustrated by an application on a mobile robotic system
Taouali, Wahiba. « Modélisation de populations neuronales pour l'intégration visuo-motrice : dynamiques et décisions ». Electronic Thesis or Diss., Université de Lorraine, 2012. http://www.theses.fr/2012LORR0105.
Texte intégralWithin the context of enaction and a global approach to perception, we focused on the characteristics of neural computation necessary to understand the relationship between structures in the brain and their functions. We first considered computational problems related to the discretization of differential equations that govern the studied systems and the synchronous and asynchronous evaluation schemes. Then, we investigated a basic functional level : the transformation of spatial sensory representations into temporal motor actions within the visual-motor system. We focused on the visual flow from the retina to the superior colliculus to propose a minimalist model of automatic encoding of saccades to visual targets. This model, based on simple local rules (CNFT and logarithmic projection) in a homogeneous population and using a sequential processing, reproduces and explains several results of biological experiments. It is then considered as a robust and efficient basic model. Finally, we investigated a more general functional level by proposing a computational model of the basal ganglia motor loop. This model integrates sensory, motor and motivational flows to perform a global decision based on local assessments. We implemented an adaptive process for action selection and context encoding through an innovative mechanism that allows to form the basic circuit for other cortico-basal loops. This mechanism allows to create internal representations according to the enactive approach that opposes the computer metaphor of the brain. Both models have interesting dynamics to study from whether a biological point of view or a computational numerical one
Taouali, Wahiba, et Wahiba Taouali. « Modélisation de populations neuronales pour l'intégration visuo-motrice : Dynamiques et décisions ». Phd thesis, Université de LORRAINE, 2012. http://tel.archives-ouvertes.fr/tel-00749924.
Texte intégralBalloul, Iyad. « Commande robuste des systemes non linéaires ». Grenoble INPG, 2000. http://www.theses.fr/2000INPG0001.
Texte intégralFavier, Arnaud. « Eventual Leader Elections in Dynamic Networks ». Electronic Thesis or Diss., Sorbonne université, 2022. http://www.theses.fr/2022SORUS059.
Texte intégralLeader election is important for many fault-tolerant services in asynchronous distributed systems. By coordinating actions of a set of processes, it allows solving agreement problems like the consensus, a fundamental problem of distributed computing. Several consensus algorithms, such as Paxos, rely on an eventual leader election service, also known as the Omega failure detector. Omega returns the identity of a process in the system, ensuring that eventually the identity of the same correct process is always returned. Many leadership algorithms were proposed in the literature to implement Omega. Among those that consider dynamic systems, most of them do not choose the leader according to a topological criterion. However, the position of the leader in the network directly impacts the performance of algorithms using the leader election service, since the leader must often interact with other processes, for example, to collect information from a majority of processes in consensus algorithms. This thesis studies the eventual leader election problem in dynamic evolving networks and performance related issues. Two eventual leader election algorithms are proposed for Mobile Ad Hoc Networks. They maintain and exploit the knowledge of the network topology to eventually elect one leader per connected component with the best closeness centrality. Evaluations were conducted on simulators with different mobility models and performance results show that these algorithms present better performance than other algorithms of the literature, including fewer messages, shortest paths to the leader, and better stability
Rouquier, Jean-Baptiste. « Robustesse et émergence dans les systèmes complexes : le modèle des automates cellulaires ». Phd thesis, Ecole normale supérieure de lyon - ENS LYON, 2008. http://tel.archives-ouvertes.fr/tel-00354042.
Texte intégralUne première partie étudie expérimentalement l'apparition de la percolation dirigée dans les automates cellulaires, notamment dans le cadre du "damage spreading". Le dernier chapitre de cette partie prouve une équivalence entre une classe d'automates cellulaires probabilistes et les automates cellulaires asynchrones.
La seconde partie étudie dans un premier chapitre l'interaction des deux perturbations évoquées: asynchronisme et topologie. Alors que le modèle habituel utilise une grille Zd, nous étudions une grille où certains liens sont temporairement coupés. Puis un second chapitre démontre des propriétés théoriques sur la règles minorité lorsque la topologie est un arbre.
Nous avons dans cette thèse mené à la fois des études expérimentales et des études théoriques. Une préoccupation transversale est la simulation formelle entre modèles. L'enjeu de ces travaux est, à terme, de savoir comment obtenir des systèmes ayant un comportement global prédéfini, ou bien comment rendre robuste à certaines perturbations un système complexe donné.
Buhrig, Aurélien. « Optimisation de la consommation des noeuds de réseaux de capteurs sans fil ». Phd thesis, Grenoble INPG, 2008. http://www.theses.fr/2008INPG0027.
Texte intégralWireless sensor networks pose many design challenges. They must collect information coming from the environment, treat gathered data, receive and forward it with a lifetime that must reach tens of years depending on the applications without any external intervention. It is thus necessary to optimize the energy consumption at every design level. This work proposes to reduce the energy consumption of the digital part of a network node thanks to the use of asynchronous logic and synchronous communications between software and hardware. In addition, it is important to dynamically manage consumption by scaling the supply voltages. This aspect is treated on the whole system, at the material level, thanks to the specification of a dedicated coprocessor, at the algorithmic level and real-time software, and at the communication interfaces between the software and the hardware
Buhrig, Aurélien. « Optimisation de la consommation des noeuds de réseaux de capteurs sans fil ». Phd thesis, Grenoble INPG, 2008. http://tel.archives-ouvertes.fr/tel-00319073.
Texte intégralAshglaf, Mohmed Omran. « Development of Hybridization concept for horizontal axis wind / tidal systems using functional similarities and advanced real-time emulation methods ». Thesis, Normandie, 2019. http://www.theses.fr/2019NORMLH07/document.
Texte intégralThe ability of conventional wind and tidal generation systems to provide the grid with reliable and stable power at all times is a new challenge due to weather fluctuations, which have a significant and direct impact on energy production. This is why the hybridization of wind and tidal power generation systems has been studied to improve the integration of wind and tidal power into the electricity grid.This study led us to develop contributions related to two main axes:The first axis is focused on a new concept of hybridization of two different energy sources in terms of physical properties, wind and horizontal axis turbines, based on an electromechanical coupling of these two systems. The two resources are wind energy and marine energy. The concept is developed using the functional similarities of turbines and similarities in energy conversion of their energy chains. To apply this concept first, the parameters of the double fed asynchronous generator installed in the GREAH emulator are identified. Then, the power conversion chain is modeled mathematically and simulated in a MATLAB / SIMULINK environment. We have developed two control strategies.A fixed speed strategy called "Direct Speed Control", and a variable speed strategy based on the search for maximum power, called "Indirect Speed Control". Finally, this concept has been implemented practically on the real-time emulator of the laboratory. The results obtained were analyzed and discussed following this work.The second axis is devoted to a concept called "accelerated time" simulation or "virtual time". Subsequently, this concept was implemented on the multi-physics emulator available at the GREAH laboratory. This concept (accelerated time) is based on reducing wind profile samples in order to decrease simulation time and facilitate real-time control.The main results are obtained first in MATLAB / SIMULINK, then verified on the emulator in real time.The main objective of this thesis is to study the concept of offshore wind / tidal turbine hybridization based on the flexibility of a multi-function emulator that allows various emulation architectures: wind turbines, tidal turbines, and hybrid wind - tidal turbines systems. We analyze its impact on the output power of the system; the obtained results are correlated with wind and tidal speed profiles, in which statistical properties impacting global power chains could be complementary and in particular in function of the given sites. Main contributions and perspectives- Development of the concept of electromechanical coupling.When two renewable energy sources are "integrated", the rapid fluctuation of the power generated is stabilized, but under certain conditions such as the presence of storage units or an automatic clutch system.- The accelerated time conceptThis method is used to reduce the size of the recorded wind or sea current data, to speed up the simulation time of the power generation units with reasonable results that are close to actual situations.- Study and develop the concept of electric shaft regime: If the electromechanical coupling is difficult to achieve from the mechanical point of view and the single shaft decouples are too frequent so high mechanical stress, one can study the electric shaft regime with two DFIG induction machines.There is a regime in which the ratios between the speeds of the different machines are rigorously constant. The system can operate in synchronous mode with specific structures and configurations
Canat, Sylvain. « Contribution à la modélisation dynamique d’ordre non entier de la machine asynchrone à cage ». Phd thesis, 2005. http://oatao.univ-toulouse.fr/7302/1/canat.pdf.
Texte intégralGouy-Pailler, Cedric. « Vers une modélisation dynamique de l'activité cérébrale pour la conception d'interfaces cerveau-machines asynchrones ». Phd thesis, 2009. http://tel.archives-ouvertes.fr/tel-00440836.
Texte intégral