Os sistemas de aplicações científicas e comerciais geram, cada vez mais, imensas quantidades de dados os quais dificilmente podem ser analisados sem que sejam usados técnicas e ferramentas adequadas de análise. Além disso, muitas destas aplicações são voltadas para Internet, ou seja, possuem seus dados distribuídos, o que dificulta ainda mais a realização de tarefas como a coleta de dados. A área de Extração de Conhecimento de Base de Dados diz respeito às técnicas e ferramentas usadas para descobrir automaticamente conhecimento embutido nos dados. Num ambiente de rede de computadores, é mais complicado realizar algumas das etapas do processo de KDD, como a coleta e processamento de dados. Dessa forma, pode ser feita a utilização de novas tecnologias na tentativa de auxiliar a execução do processo de descoberta de conhecimento. Os agentes de software são programas de computadores com propriedades, como, autonomia, reatividade e mobilidade, que podem ser utilizados para esta finalidade. Neste sentido, o objetivo deste trabalho é apresentar a proposta de um sistema multi-agente, chamado Minador, para auxiliar na execução e gerenciamento do processo de Extração de Conhecimento de Base de Dados. / Nowadays, commercial and scientific application systems generate huge amounts of data that cannot be easily analyzed without the use of appropriate tools and techniques. A great number of these applications are also based on the Internet which makes it even more difficult to collect data, for instance. The field of Computer Science called Knowledge Discovery in Databases deals with issues of the use and creation of the tools and techniques that allow for the automatic discovery of knowledge from data. Applying these techniques in an Internet environment can be particulary difficult. Thus, new techniques need to be used in order to aid the knowledge discovery process. Software agents are computer programs with properties such as autonomy, reactivity and mobility that can be used in this way. In this context, this work has the main goal of presenting the proposal of a multiagent system, called Minador, aimed at supporting the execution and management of the Knowledge Discovery in Databases process.

Os sistemas de aplicações científicas e comerciais geram, cada vez mais, imensas quantidades de dados os quais dificilmente podem ser analisados sem que sejam usados técnicas e ferramentas adequadas de análise. Além disso, muitas destas aplicações são voltadas para Internet, ou seja, possuem seus dados distribuídos, o que dificulta ainda mais a realização de tarefas como a coleta de dados. A área de Extração de Conhecimento de Base de Dados diz respeito às técnicas e ferramentas usadas para descobrir automaticamente conhecimento embutido nos dados. Num ambiente de rede de computadores, é mais complicado realizar algumas das etapas do processo de KDD, como a coleta e processamento de dados. Dessa forma, pode ser feita a utilização de novas tecnologias na tentativa de auxiliar a execução do processo de descoberta de conhecimento. Os agentes de software são programas de computadores com propriedades, como, autonomia, reatividade e mobilidade, que podem ser utilizados para esta finalidade. Neste sentido, o objetivo deste trabalho é apresentar a proposta de um sistema multi-agente, chamado Minador, para auxiliar na execução e gerenciamento do processo de Extração de Conhecimento de Base de Dados. / Nowadays, commercial and scientific application systems generate huge amounts of data that cannot be easily analyzed without the use of appropriate tools and techniques. A great number of these applications are also based on the Internet which makes it even more difficult to collect data, for instance. The field of Computer Science called Knowledge Discovery in Databases deals with issues of the use and creation of the tools and techniques that allow for the automatic discovery of knowledge from data. Applying these techniques in an Internet environment can be particulary difficult. Thus, new techniques need to be used in order to aid the knowledge discovery process. Software agents are computer programs with properties such as autonomy, reactivity and mobility that can be used in this way. In this context, this work has the main goal of presenting the proposal of a multiagent system, called Minador, aimed at supporting the execution and management of the Knowledge Discovery in Databases process.

A swarm is a group of uninformed individuals that exhibit collective behaviors. Without any information about the external world, a swarm has limited ability to achieve complex goals. Prior work on human-swarm interaction methods allow a human to influence these uninformed individuals through either leadership or predation as informed agents that directly interact with humans. These methods of influence have two main limitations: (1) although leaders sustain influence over nominal agents for a long period of time, they tend to cause all collective structures to turn in to flocks (negating the benefit of other swarm formations) and (2) predators tend to cause collective structures to fragment. In this thesis, we present the use of mediators as a novel form for human-swarm influence and use mediators to shape the perimeter of a swarm. The mediator method uses special agents that operate from within the spatial center of a swarm. This approach allows a human operator to coordinate multiple mediators to modulate a rotating torus into various shapes while sustaining influence over the swarm, avoiding fragmentation, and maintaining the swarm's connectivity. The use of mediators allows a human to mold and adapt the torus' behavior and structure to a wide range of spatio-temporal tasks such as military protection and decontamination tasks. Results from an experiment that compares previous forms of human influence with mediator-based control indicate that mediator-based control is more amenable to human influence for certain types of problems.

La généralisation de données géographiques est l'opération de simplification de ces données effectuée lors de la diminution de leur échelle de représentation. Cette thèse s'appuie sur le modèle de généralisation automatique à base d'agents de (Ruas et Duchêne, 2007), qui a été mis en oeuvre avec succès pour la généralisation des objets géographiques discrets (bâtiments, routes, etc.). L'objectif est de permettre une prise en compte d'un nouveau type de thèmes, appelés "thèmes champ", comme le relief ou l'occupation du sol. Ces thèmes ont pour particularité d'être définis en tout point de l'espace. Nous souhaitons permettre une préservation des relations pouvant exister entre les objets géographiques et les champs, comme par exemple le fait qu'un cours d'eau s'écoule sur le relief. Pour répondre à cet objectif, nous proposons le modèle de généralisation GAEL (Généralisation à base d'Agents Elastiques) qui permet d'appliquer des déformations aux champs pour préserver les relations objet-champ. Les champs sont modélisés comme des agents, dits élastiques, qui ont la capacité de se déformer pour amortir les opérations de généralisation appliquées aux objets géographiques (bâtiments, routes, etc.). Ces déformations sont obtenues en s'appuyant sur une décomposition des champs en petits éléments contraints (points, segments, triangles, etc.) et sur une modélisation des points composant les champs sous forme d'agents. Couplé au modèle de (Ruas et Duchêne, 2007), le modèle GAEL permet de disposer d'un modèle de généralisation hybride, capable d'effectuer à la fois des opérations discrètes et continues / Geographic data generalisation is the process of simplification of these data when their representation scale decreases. This thesis is based on the automated generalisation model of (Ruas et Duchêne, 2007), which has been successfully applied to the generalisation of discrete geographic objects (buildings, roads, etc.). Our purpose is to take into account a new kind of themes, the "field themes", such as the relief and the land use cover. These themes have the specificity to be defined at each point of the space. We aim at preserving the relationships between geographic objects and fields, such as for example, the fact that a river should flow down on the relief. To reach this goal, we propose a generalisation model called GAEL (Generalisation based on Agents and Elasticity). This model allows to apply deformations to the fields in order to preserve the object-field relationships. Fiels are modelled as "elastic agents", which are able to deform themselves to propagate the generalisation operations applied on the geographic objects (buildings, roads, etc.). These deformations are performed by using a decomposition of the fields into small constrained elements (points, segments, triangles, etc.) and by modelling the points composing the fields as agents. Because it is merged with the model of (Ruas et Duchêne, 2007), the GAEL model offers a hybrid generalisation model, able to manage both discrete and continuous operations

Les jeux sérieux (serious-games en anglais) sont de plus en plus utilisés dans le domaine de la formation, le but étant alors d'apprendre une ou plusieurs compétences liée(s) à un métier ou à des qualités plus générales, à un ou plusieurs apprenants (joueurs), ceci avec une dimension ludique. Mesurer qualitativement ou quantitativement le succès ou non de cet apprentissage, ce qui relève au final d'un processus d'évaluation du ou des l'apprenant(s), constitue ces dernières années un enjeu majeur dans les jeux sérieux s'ajoutant aux problématiques plus classiques de conception et de scénarisation.Dans un contexte de jeux sérieux multi-acteurs, avec apprentissage de tâches collaboratives réalisées par différents acteurs humains ou virtuels partageant un but commun, cette recherche s'intéresse au processus d'évaluation des apprenants, conduisant à une évaluation tant individuelle que collective. Nous proposons tout d'abord un cadre méthodologique pour l'évaluation des apprenants pour de tels jeux. Ce cadre méthodologique est basé sur le concept « d'espace d'évaluation » associé à un point de vue correspondant à un objectif d'évaluation particulier, Nous proposons ensuite une architecture logicielle spécifique à base d'agents, d'une part permettant à des apprenants humains et des joueurs virtuels d'interagir (simulation de comportements), et d'autre part supportant une évaluation tant individuelle que collective des apprenants. Cette architecture logicielle est implémentée et intégrée au jeu sérieux SIMFOR, et sa mise en oeuvre est illustrée sur un scénario réaliste relatif à la formation à la gestion de crise. / Serious Games (SG) are increasingly used in the field of training, SG aims to teach one or more skills(s) related to a business or to more general qualities, to one or more learners (players), this with a playful dimension. Usually, SG aims a specific training objective related to a specific context training which can be linked to a trade (eg fire ), a body of knowledge (school or university), or the acquisition of social skills (conflict management, cooperation ...). To Measure the success or failure of this learning is ultimately an assessment process of the learner(s) and in recent years became a major issue in SG in addition to issues classic design and screenwriting.In the context of multi-player SG, with a collaborative tasks learning performed by different learners (or virtual actor) sharing a common goal, our research focuses on the assessment of learning processes, leading to both individual and collective assessment. We propose first a methodological framework for learners assessment. This methodological framework is based on the "Evaluation space" concept associated with a specific view corresponding to a particular assessment objective, leading to the calculation of indicators, based on specific knowledge models representation and assessment. In order to operationalize this methodological assessment framework, we propose a agents based architecture, on the one hand allowing human learners and virtual players (simulation of behavior) interaction, and other hand supporting individual and collective learners' assessment. This architecture is implemented and integrated to the SIMFOR SG, and illustrated in a crisis management scenario exercise.

Les besoins, toujours plus grands, en énergie et la pollution de la planète, due à l'utilisation d'énergies polluantes non-renouvelables, obligent à concevoir de nouveaux modèles énergétiques durables et fiables. Ces nouveaux modèles se fondent, aujourd'hui, sur une intégration massive des énergies renouvelables dans le réseau électrique. Le problème des énergies renouvelables est leur caractère intermittent, dépendant des conditions météorologiques, la plupart du temps. L'arrivée des nouvelles technologies de l'information et de la communication permet l’intégration d’une couche informationnelle au réseau énergétique lui permettant d'être intelligent et d'entrevoir la possibilité d'une gestion distribuée des énergies renouvelables. Ces énergies étant principalement décentralisées, contrairement aux imposantes centrales nucléaires, au charbon, au gaz et au fioul, sont produites directement chez le consommateur. Le consommateur devient alors un prosumer capable de répondre à ses besoins énergétiques, voire même d'agir comme un producteur s'il produit plus d'énergie qu'il n'en consomme. Mais l'arrivée d'une pléthore de petits acteurs capables d'acheter et de vendre de l'énergie, en temps réel, dans un marché comprenant les puissants acteurs traditionnels du marché peut être une source de volatilité pour les prix de l'énergie. Des variations importantes des prix peuvent conduire à des situations néfastes en déstabilisant le réseau. Pour faire face à ce problème, nous avons développé un premier mécanisme de négociation automatique, sur trois échelles de temps, qui impose des contraintes sur la demande ainsi que sur les prix pour garantir leur stabilité. Ce mécanisme s'appuie sur des entités représentatives (producteurs, prosumers et agrégateurs) pour gérer l'offre et la demande sans toutefois prendre en compte l'impact sur le réseau des contrats négociés entre ces entités. Le second mécanisme, fondé sur la technologie Blockchain, permet des négociations bilatérales décentralisées et prend en compte les impacts physiques de chaque échange d'énergie entre prosumers, garantissant ainsi l'intégrité du réseau. Le mécanisme ainsi proposé se soustrait à une gestion de la stabilité du réseau par des tiers / Increasing needs in energy, and earth pollution, coming from the use of polluting and non-renewable energy, make it indispensable the design of new energetic models, sustainable and reliable. Today, these new models are based on a massive integration of renewable generators in the grid. The renewable integration issues come from their stochastic features, depending on the weather conditions, most of the time. The coming of new information and communication technologies allows the integration of an information layer to the energy grid allowing it to be smarter and allow to have a glimpse of the possibility of a decentralised management of renewable energy. These energy being mainly decentralised, unlike imposing nuclear, coal and gas power plants, are produced directly at the customer's location. Consumers become, then, a prosumer able to answer to its own energetic needs,and maybe to act as a producer if it produces more energy than it consumes. But, the coming of a plethora of small actors, able to buy and sell energy, in real time, in a market containing the more powerful actors, traditional in the market, can be a source of volatility for energy prices. Considerable variations of the price can lead to detrimental situations by disturbing the grid. To face this problem, we developed a first mechanism for automatic negotiations, on three time scales, which decrees constraints on demand and on prices in order to guarantee their stability. This mechanism rely on representative entities (producers, prosumers and aggregators) to manage demand and supply without taking into account the impact on the grid of the negotiated contracts between these entities. The second mechanism, based on blockchain technology, allows bilateral decentralised negotiations and take into account the physical impacts on the grid of each energy exchange between prosumers, guaranteing then, the grid integrity. The proposed mechanism exempt itself from a management of the grid stability from a centralised third part

Ces dernières années ont consacré l’émergence du parallélisme dans la plupart des branches de l’informatique.Au niveau matériel, tout d’abord, du fait de la stagnation des fréquences de fonctionnement des unités decalcul. Au niveau logiciel, ensuite, avec la popularisation de nombreuses plates-formes d’exécution parallèle.Une forme de parallélisme est également présente dans les systèmes multi-agents, qui facilitent la description desystèmes complexes comme ensemble d’entités en interaction. Si l’adéquation entre ce parallélisme d’exécutionlogiciel et conceptuel semble naturelle, la parallélisation reste une démarche difficile, du fait des nombreusesadaptations devant être effectuées et des dépendances présentes explicitement dans de très nombreux systèmesmulti-agents.Dans cette thèse, nous proposons une solution pour faciliter l’implémentation de ces modèles sur une plateformed’exécution parallèle telle que le GPU. Notre bibliothèque MCMAS vient répondre à cette problématiqueau moyen de deux interfaces de programmation, une couche de bas niveau MCM permettant l’accès direct àOpenCL et un ensemble de plugins utilisables sans connaissances GPU. Nous étudions ensuite l’utilisation decette bibliothèque sur trois systèmes multi-agents existants : le modèle proie-prédateur, le modèle MIOR etle modèle Collemboles. Pour montrer l’intérêt de cette approche, nous présentons une étude de performancede chacun de ces modèles et une analyse des facteurs contribuant à une exécution efficace sur GPU. Nousdressons enfin un bilan du travail et des réflexions présentées dans notre mémoire, avant d’évoquer quelquespistes d’amélioration possibles de notre solution. / These last years have seen the emergence of parallelism in many fields of computer science. This is explainedby the stagnation of the frequency of execution units at the hardware level and by the increasing usage ofparallel platforms at the software level. A form of parallelism is present in multi-agent systems, that facilitatethe description of complex systems as a collection of interacting entities. If the similarity between this softwareand this logical parallelism seems obvious, the parallelization process remains difficult in this case because ofthe numerous dependencies encountered in many multi-agent systems.In this thesis, we propose a common solution to facilitate the adaptation of these models on a parallel platformsuch as GPUs. Our library, MCMAS, provides access to two programming interface to facilitate this adaptation:a low-level layer providing direct access to OpenCL, MCM, and a high-level set of plugins not requiring anyGPU-related knowledge.We study the usage of this library on three existing multi-agent models : predator-prey,MIOR and Collembola. To prove the interest of the approach we present a performance study for each modeland an analysis of the various factors contributing to an efficient execution on GPUs. We finally conclude on aoverview of the work and results presented in the report and suggest future directions to enhance our solution.

Dans ce travail, nous proposons un système auto-organisé composé d'agents-robots contrôlés par une architecture de subsomption et des règles locales probabilistes de prises et de dépôts. Ces agents-robots sont capables, grâce au développement de leurs capacités cognitives de se créer une carte cognitive, d'apprendre plusieurs lieux buts et de planifier le retour vers ces buts. Bien que formellement l'algorithme ne permette pas à chaque agent de "mélanger ni de fusionner ou d'optimiser" plusieurs objectifs, nous montrerons que le système global est capable de réaliser une optimisation multi-objectifs. Particulièrement, la fusion de l'apprentissage local avec l'accumulation de décisions individuelles permet de faire émerger (i) des structures dans l'environnement et (ii) des dynamiques tel que le comportement de spécialisation ou les comportements que nous pouvons considérer comme étant "égoïstes" ou "altruistes". Nous montrons qu'un mécanisme d'imitation simple contribue à l'amélioration des performance de notre SMAC et de notre SMRC, à savoir l'optimisation de la durée pour découvrir des différentes ressources, le temps moyen de planification, le niveau global de satisfaction des agents et enfin le temps moyen de convergence vers une solution stable. Particulièrement, l'ajout d'une capacité d'imitation améliore la construction des cartes cognitives pour chaque agent et stimule le partage implicite des informations dans un environnement a priori inconnu. En effet, les découvertes individuelles peuvent avoir un effet au plan social et donc inclure l'apprentissage de nouveaux comportements au niveau individuel. Pour finir, nous comparons les propriétés émergentes de notre SMAC à un modèle mathématique basé sur la programmation linéaire (PL). Cette évaluation montre les bonnes performances de notre SMAC qui permet d'avoir des solutions proches des solutions de la PL pour un coût de calcul réduit. Dans une dernière série d'expériences, nous étudions notre système d'agrégation dans un environnement réel. Nous mettons en place un SMRC, composé par des robots qui sont capables d'effectuer les opérations de prise, de dépôt et de maintien. Nous montrons via les premiers tests d'agrégation que les résultats sont prometteurs. / In this work, we propose a self-organized system composed by agents-robots, controlled by a subsumption architecture with probabilistic local rules of deposits and taking. These agents-robots are able, thanks to the development of their cognitive abilities to create a cognitive map, to learn various goals' locations and to plan the return to these goals. Although formally the algorithm does not allow each agent to « mix or merge or optimize » several objectives, we show that the overall system is able to perform a multi-objective optimization. Specifically, the fusion of local learning with the accumulation of the individual decisions allows to emerge (i) structures in the environment and (ii) several dynamics such as specialization behavior or those that we can consider as « selfish » or « altruistic ». We show that the imitation strategy contributes to the improvement of the performance of our SMAC and our SMRC, namely the optimization of time to explore the various resources, the average planning time, the overall satisfaction level of agents and finally the the average time of convergence to a stable solution. Specifically, the addition of an imitation ability improves the construction of cognitive maps for each agent and stimulates the implicit sharing of informations in an unknown environment. Indeed, individual discoveries can affect the social level and therefore include learning new behaviors at the individual level. Finally, we compare the emergent properties of our SMAC with a mathematical model based on linear programming (LP). This evaluation shows the good performance of our SMAC which allows to obtain solutions close to the solution of the PL for a low cost of computation. In a final series of experiments, we study our aggregation system in a real environment. We set up a SMRC, composed by robots that are able to perform taking operations, deposits operations and refueling operations. We show through the first tests of aggregation that the results are promising.

Esta pesquisa apresenta a implementação de um sistema de proteção digital de distância adaptativo dispondo de um sistema multi-agentes (SMA). Tal SMA trabalhará de forma distribuída, possuindo entidades especialistas (Agentes) inseridas em locais específicos e agindo de forma eficiente na solução de problemas locais. Espera-se pela aplicação deste SMA, atribuir aos relés de distância características desejáveis e difíceis de serem alcançadas pelas técnicas convencionais. Dentre as deficiências observadas nos sistemas de proteção convencionais, tem-se a não sensibilidade à dinâmica dos sistemas elétricos de potência em função das alterações nas condições operacionais. Logo, tem-se então a necessidade de um sistema de proteção mais flexível, seletivo e eficaz para assegurar a confiabilidade e a continuidade no suprimento da energia elétrica. Para desenvolver tal proposição, foram realizadas diversas e distintas simulações computacionais dispondo do software ATP (Alternative Transients Program) para a geração das prováveis curvas de operação adaptativa. Estas, quando inseridas em um banco de dados, serão representativas das esperadas alterações nas condições operacionais de um sistema elétrico em especifico. Para a implementação e validação desta abordagem, diversos testes foram realizados, tendo como finalidade a comparação desta abordagem dispondo de SMA com a filosofia tradicional de proteção de distância. Os resultados obtidos são promissores e demonstram o alto grau de precisão e confiabilidade do método proposto. / This research presents the development of an adaptive digital distance protection system by a multi-agent system (MAS). This MAS will work in a distributed way, with expert entities (agents) inserted in specific places and acting efficiently in the solution of local problems. With the application of the MAS, the distance relays will have desirable characteristics, which are difficult to be reached by conventional techniques. Among the deficiencies observed in conventional protection systems, there is the insensitivity to the dynamics of the electrical power system due to changes in operating conditions. So, there is the necessity of a protection system which is more flexible, selective and effcient to assure reliability and continuity in the electricity supply. To improve these characteristics, several computational simulations were done using alternative transients program (ATP) to generate the desirable adaptive operation curves. These curves, when inserted into a database, will represent the expected changes in the operating conditions of an electrical system. For the implementation and validation of this approach, several tests were done with the purpose of comparison of this approach (using MAS) to the traditional philosophy of distance protection. The results obtained are promising and demonstrate a highly satisfactory degree of accuracy and reliability of the proposed method.

Esta pesquisa apresenta a implementação de um sistema de proteção digital de distância adaptativo dispondo de um sistema multi-agentes (SMA). Tal SMA trabalhará de forma distribuída, possuindo entidades especialistas (Agentes) inseridas em locais específicos e agindo de forma eficiente na solução de problemas locais. Espera-se pela aplicação deste SMA, atribuir aos relés de distância características desejáveis e difíceis de serem alcançadas pelas técnicas convencionais. Dentre as deficiências observadas nos sistemas de proteção convencionais, tem-se a não sensibilidade à dinâmica dos sistemas elétricos de potência em função das alterações nas condições operacionais. Logo, tem-se então a necessidade de um sistema de proteção mais flexível, seletivo e eficaz para assegurar a confiabilidade e a continuidade no suprimento da energia elétrica. Para desenvolver tal proposição, foram realizadas diversas e distintas simulações computacionais dispondo do software ATP (Alternative Transients Program) para a geração das prováveis curvas de operação adaptativa. Estas, quando inseridas em um banco de dados, serão representativas das esperadas alterações nas condições operacionais de um sistema elétrico em especifico. Para a implementação e validação desta abordagem, diversos testes foram realizados, tendo como finalidade a comparação desta abordagem dispondo de SMA com a filosofia tradicional de proteção de distância. Os resultados obtidos são promissores e demonstram o alto grau de precisão e confiabilidade do método proposto. / This research presents the development of an adaptive digital distance protection system by a multi-agent system (MAS). This MAS will work in a distributed way, with expert entities (agents) inserted in specific places and acting efficiently in the solution of local problems. With the application of the MAS, the distance relays will have desirable characteristics, which are difficult to be reached by conventional techniques. Among the deficiencies observed in conventional protection systems, there is the insensitivity to the dynamics of the electrical power system due to changes in operating conditions. So, there is the necessity of a protection system which is more flexible, selective and effcient to assure reliability and continuity in the electricity supply. To improve these characteristics, several computational simulations were done using alternative transients program (ATP) to generate the desirable adaptive operation curves. These curves, when inserted into a database, will represent the expected changes in the operating conditions of an electrical system. For the implementation and validation of this approach, several tests were done with the purpose of comparison of this approach (using MAS) to the traditional philosophy of distance protection. The results obtained are promising and demonstrate a highly satisfactory degree of accuracy and reliability of the proposed method.