Made available in DSpace on 2016-08-17T14:53:24Z (GMT). No. of bitstreams: 1 Dissertacao Rafael de Souza.pdf: 5581665 bytes, checksum: 4edbe8b1f2b84008b5129a93038f2fee (MD5) Previous issue date: 2013-03-04 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / In this work, we applied the technology of Multi-Agent Systems (MAS) in the capital market, i.e., the stock market, specifically in Bolsa de Mercadorias e Futuros de São Paulo (BM&FBovespa). The research focused mainly on negotiation protocols and learning of investors agents. Within the Stock Exchange competitive field, the development of an agent that could learn to negotiate, could become differential for investors who wish to increase their profits. The decision-making based on historical data is motivation for further research in the same direction, however, we sought a different approach with regard to the representation of the states of q-learning algorithm. The reinforcement learning, in particular q-learning, has been shown to be effective in environments with various historical data and seeking reward decisions with positive results. That way it is possible to apply in the purchase and sale of shares, an algorithm that rewards the profit and punishes the loss. Moreover, to achieve their goals agents need to negotiate according to specific protocols of stock exchange. Therefore, endeavor was also the specifications of the rules of negotiation between agents that allow the purchase and sale of shares. Through the exchange of messages between agents, it is possible to determine how the trading will occur and facilitate communication between them, because it sets a standard of how it will happen. Therefore, in view of the specification of negotiation protocols based on q-learning, this research has been the modeling of intelligent agents and models of learning and negotiation required for decision making entities involved. / Neste trabalho, aplicou-se a tecnologia de Sistemas MultiAgente (SMA) no mercado de capitais, isto é, na Bolsa de Valores, especificamente na Bolsa de Mercadorias e Futuros de São Paulo (BM&FBovespa). A pesquisa concentrou-se principalmente nos protocolos de negociação envolvidos e na aprendizagem dos agentes investidores. Dentro do cenário competitivo da Bolsa de Valores, o desenvolvimento de um agente que aprendesse a negociar poderia se tornar diferencial para os investidores que desejam obter lucros cada vez maiores. A tomada de decisão baseada em dados históricos é motivação para outras pesquisas no mesmo sentido, no entanto, buscou-se uma abordagem diferenciada no que diz respeito à representação dos estados do algoritmo de aprendizagem-q. A aprendizagem por reforço, em especial a aprendizagem-q, tem demonstrado ser eficiente em ambientes com vários dados históricos e que procuram recompensar decisões com resultados positivos. Dessa forma é possível aplicar na compra e venda de ações, um algoritmo que premia o lucro e pune o prejuízo. Além disso, para conseguir alcançar seus objetivos os agentes precisam negociar de acordo com os protocolos específicos da bolsa de valores. Sendo assim, procurou-se também as especificações das regras de negociação entre os agentes que permitirão a compra e venda de títulos da bolsa. Através da troca de mensagens entre os agentes, é possível determinar como a negociação ocorrerá e facilitará comunicação entre os mesmos, pois fica padronizada a forma como isso acontecerá. Logo, tendo em vista as especificações dos protocolos de negociação baseados em aprendizagem-q, tem-se nesta pesquisa a modelagem dos agentes inteligentes e os modelos de aprendizagem e negociação necessários para a tomada de decisão das entidades envolvidas.

Distributed Mobile Robotics (DMR) is a multidisciplinary research area with many open research questions. This is a survey of the state of the art in Distributed Mobile Robotics research. DMR is sometimes referred to as cooperative robotics or multi-robotic systems. DMR is about how multiple robots can cooperate to achieve goals and complete tasks better than single robot systems. It covers architectures, communication, learning, exploration and many other areas presented in this master thesis. / Uppsatsen är en brett spektrum på vilken forskning som pågår rörande distribuerad mobil robotik, dvs hur många robotar kan samverka för att lösa uppgifter. / Adress: Victor Adolfsson Folkparksvägen 12:24 372 38 Ronneby Telefon: 0457-171 42 Mobil: 0709-441389

Patient’s mobility throughout his lifetime leaves a trial of information scattered in laboratories, clinical institutes, primary care units, and other hospitals. Hence, the medical history of a patient is valuable when subjected to special healthcare units or undergoes home-care/personal-care in elderly stage cases. Despite the rhetoric about patient-centred care, few attempts were made to measure and improve in this arena. In this thesis, we will describe and implement a high-level view of a Patient Centric information management, deploying at a preliminary stage, the use of Agent Technologies and Grid Computing. Thus, developing and proposing an infrastructure that allows us to monitor and survey the patient, from the doctor’s point of view, and investigate a Persona, from the patients’ side, that functions and collaborates among different medical information structures. The Persona will attempt to interconnect all the major agents (human and software), and realize a distributed grid info-structure that directly affect the patient, therefore, revealing an adequate and cost-effective solution for most critical information needs. The results comprehended in the literature survey, consolidating Healthcare Information Management with emerged intelligent Multi-Agent System Technologies (MAS) and Grid Computing; intends to provide a solid basis for further advancements and assessments in this field, by bridging and proposing a framework between the home-care sector and the flexible agent architecture throughout the healthcare domain.

Le changement de croyances vise à trouver des moyens adéquats pour faire évoluer les croyances d'un agent lorsqu'il est confronté à de nouvelles informations. Dans la plupart des travaux sur la révision de croyances, l'ensemble de croyances d'un agent est composé de croyances au sujet de l'environnement (le monde) et est représenté par un ensemble de formules de la logique classique. Dans de nombreuses applications, un agent n'est pas seul dans l'environnement, mais le partage avec d'autres agents, qui ont aussi des croyances. Ainsi les croyances sur les croyances des autres agents constituent un élément d'information important pour l'agent, afin d'être en mesure de prendre les meilleures décisions et d'effectuer les meilleures actions. L'utilisation de croyances sur les croyances des autres agents est par exemple cruciale dans la théorie des jeux. Dans cette thèse, nous étudions dans un premier temps les opérateurs de contraction propositionnelle correspondant aux opérateurs de révision de Katsuno et Mendelzon. Nous étudions ensuite une connexion entre les logiques épistémiques et la théorie du changement de croyances, proche de l'approche AGM. Nous nous sommes intéressés à l'utilisation des opérateurs qui modifient les croyances des agents dans les modèles KD45n standard. Cette tâche est plus compliquée que dans le cadre AGM standard, car, dans un contexte multi-agents, les nouvelles informations peuvent prendre différentes formes. Par exemple, chaque nouvelle information peut être observée/transmise/disponible à tous les agents ou seulement à certains d’entre eux. / Belief change is about finding appropriate ways to evolve an agent's beliefs when confronted with new pieces of information. In most works on belief revision, the set of beliefs of an agent is composed of beliefs about the environment (the world) and is represented by a set of formulas of classical logic. In many applications, an agent is not alone in the environment, but sharing with other agents, which also have beliefs. Thus beliefs about the beliefs of other agents are an important piece of information for the agent in order to be able to make the best decisions and perform the best actions. The use of beliefs about the beliefs of other agents is, for exampel, crucial in game theory. In this thesis, we first study the operators of propositional contraction corresponding to the revision operators proposed by Katsuno and Mendelzon. Then, we study a connection between epistemic logics and belief change theory, close to the AGM approach. We are interested in the use of operators that modify agent beliefs in standard KD45n models. This task is more complicated than in the standard AGM framework because, in a multi-agent context, new information can take different forms. For example, each new information can be observed/transmitted/available to all agents or only some of them.

In car racing simulation games, finishing first is the main goal. To reach that goal, it is required to go around a racing track, competing against other cars aiming for the same goal. Implementing a bot for doing so may have its difficulties, although using a technique called multi-agent systems combined with artificial potential field, let- ting the agents take care of subtasks like keeping the car on the track, minimize how much the car turns in a curvature and basics in navigation around the track, has showed that artificial potential fields very well fit the problem of driving a car in simulated car racing in a competitive way. / Mobiltelefon: 0707422666

This thesis discusses how Multi-agent Systems (MAS) should be designed in the context of diabetic health care. Three fields are touched: computer science, socio-psychology and systems science. Agent Technology is the core technology in the research. Theories from socio-psychology and systems science are applied to facilitate the discussion about computer agents. As the integration of socio-psychology and systems science, Activity Systems Theory is introduced to give a synthesized description of MAS. Laws and models are introduced with benefits on both individual agent and agent communities. Cybernetics from systems science and knowledge engineering from computer science are introduced to approach the design and implementation of the individual agent architecture. A computer agent is considered intelligent if it is capable of reactivity, proactivity and social activity. Reactivity and proactivity can be realized through a cybernetic approach. Social activity is much more complex, since it considers MAS coordination. In this thesis, I discuss it from the perspectives of socio-psychology. The hierarchy and motivation thinking from Activity Systems Theory is introduced to the MAS coordination. To behave intelligent, computer agents should work with knowledge. Knowledge is considered as a run-time property of a group of agents (MAS). During the MAS coordination, agents generate new information through exchanging the information they have. A knowledge component is needed in agent’s architecture for the knowledge related tasks. In my research, I adopt CommonKADS methodology for the design and implementation of agent’s knowledge component. The contribution of this research is twofold: first, MAS coordination is described with perspectives from socio-psychology. According to Activity Systems Theory, MAS is hierarchically organized and driven by the motivation. This thesis introduces a motivation-driven mechanism for the MAS coordination. Second, the research project Integrated Mobile Information Systems for health care (IMIS) indicates that the diabetic health care can be improved by introducing agent-based services to the care-providers and care-receivers. IMIS agents are designed with capabilities of information sharing, organization coordination and task delegation. To perform these tasks, the IMIS agents interact with each other based on the coordination mechanism that is discussed above.

L'analyse et la prévision du comportement des réseaux de transport sont aujourd'hui des éléments cruciaux pour la mise en place de politiques de gestion territoriale. La simulation informatique du trafic routier est un outil puissant permettant de tester des stratégies de gestion avant de les déployer dans un contexte opérationnel. La simulation du trafic à l'échelle d'un ville requiert cependant une puissance de calcul très importante, dépassant les capacité d'un seul ordinateur.Dans cette thèse, nous étudions des méthodes permettant d'effectuer des simulations de trafic multi-agent à large échelle. Nous proposons des solutions permettant de distribuer l'exécution de telles simulations sur un grand nombre de coe urs de calcul. L'une d'elle distribue directement les agents sur les coeurs disponibles, tandis que la seconde découpe l'environnement sur lequel les agents évoluent. Les méthodes de partitionnement de graphes sont étudiées à cet effet, et nous proposons une procédure de partitionnement spécialement adaptée à la simulation de trafic multi-agent. Un algorithme d'équilibrage de charge dynamique est également développé, afin d'optimiser les performances de la distribution de la simulation microscopique.Les solutions proposées ont été éprouvées sur un réseau réel représentant la zone de Paris-Saclay.Ces solutions sont génériques et peuvent être appliquées sur la plupart des simulateurs existants.Les résultats montrent que la distribution des agents améliore grandement les performances de la simulation macroscopique, tandis que le découpage de l'environnement est plus adapté à la simulation microscopique. Notre algorithme d'équilibrage de charge améliore en outre significativement l'efficacité de la distribution de l'environnement / Nowadays, analysis and prediction of transport network behavior are crucial elements for the implementation of territorial management policies. Computer simulation of road traffic is a powerful tool for testing management strategies before deploying them in an operational context. Simulation of city-wide traffic requires significant computing power exceeding the capacity of a single computer.This thesis studies the methods to perform large-scale multi-agent traffic simulations. We propose solutions allowing the distribution of such simulations on a large amount of computing cores.One of them distributes the agents directly on the available cores, while the second splits the environment on which the agents evolve. Graph partitioning methods are studied for this purpose, and we propose a partitioning procedure specially adapted to the multi-agent traffic simulation. A dynamic load balancing algorithm is also developed to optimize the performance of the microscopic simulation distribution.The proposed solutions have been tested on a real network representing the Paris-Saclay area.These solutions are generic and can be applied to most existing simulators.The results show that the distribution of the agents greatly improves the performance of the macroscopic simulation, whereas the environment distribution is more suited to microscopic simulation. Our load balancing algorithm also significantly improves the efficiency of the environment based distribution

Avec l’avènement des images satellitaires à très haute résolution, les approches pixelliques ne donnant plus entière satisfaction ont été remplacées par les approches objets. Cependant, ces approches restent tributaires de la première étape qui permet le passage du pixel vers l’objet, à savoir l’étape de construction. L’architecture séquentielle de ces approches fait que les erreurs de l’étape de construction se répercutent sur l’étape d’identification. Il devient donc primordial de passer de cette architecture séquentielle vers une architecture itérative permettant la collaboration entre les étapes de construction et d’identification. Dans le cadre de cette thèse, nous nous sommes concentrés sur l’étude de l’étape de construction(i.e., la segmentation) comme base de départ pour les approches proposées. Nous avons proposé deux approches objets basées sur les techniques de segmentation les plus propices à la collaboration, à savoir les techniques régions et les techniques collaboratives région/contour. La première approche proposée se base sur une croissance sémantique hiérarchique. Elle permet de combiner les algorithmes de croissance de régions et les approches d’analyse d’images orientées objets. La croissance étant spécifique à la classe du germe de départ, nous avons proposé deux adaptations de l’approche sur les objets les plus rencontrés dans le contexte urbain, à savoir, les routes et les bâtiments. La deuxième approche utilise un algorithme évolutionnaire local permettant un paramétrage local des différents agents régions et contours évoluant au sein d’un système multi-agents. / Object based image analysis is a rising research area in remote sensing. However, existing approaches heavily rely on the object construction process, mainly due to the lack of interaction between the two steps, i.e., Construction and identification.In this thesis, we focused on the study of the construction phase (i.e., segmentation) as a basis for the proposed approaches. The first proposed approach is based on a hierarchical semantic growth. This approach allows merging region-growing algorithms and Object Based Image Analysis approaches. Due to the dependency of the semantic growth on the seed class, we propose two adaptations of the approach on the most used class in the urban context, i.e., roadsand buildings. The second approach benefits of both multi-agent systems and genetic algorithms characteristics. It overcomes the threshold’s dependency of the proposed cooperative multi-agent system between an edge approach and a region approach. The genetic algorithm is used to automatically find building extraction parameters for each agent based on expert knowledge. The proposed approaches have been validated on a very high-resolution image of the urban area of Strasbourg.

Conselho Nacional de Desenvolvimento Científico e Tecnológico / The tasks allocation process is a key activity in project management. Being able to conciliate the people profile, as well their professional aspirations is a real challenge for managers. Software projects differ from the rest because of its intangibility and complexity, characteristics that contribute to the failure of these projects. In addition, such projects depend on skilled and highly specialized manpower and the knowledge to motivate and develop them is an extremely important factor for achieving success in this area. Thus, this document sought to present a model of task allocation in software projects that considers, in addition to organizational goals (cost, time, etc.), the professional aspirations of the participating members. The proposed strategy for research is through the simulation of multi-agent systems, systems generally used for behavioral simulation entities. Through this model was intended to provide the manager a tool to aid in decision making in the process of task allocation in software development companies, allowing them to view different scenarios and strategies, looking for producing a better performance in project implementation and greater people satisfaction. / O processo de atribuição de tarefas é uma atividade chave na gerência de projetos. Ser capaz de conciliar o perfil das pessoas, bem como seus anseios profissionais nesse processo é um desafio para gestores. Projetos de software diferem dos demais devido a sua intangibilidade e complexidade, características que contribuem ao fracasso desses projetos. Além disso, tais projetos dependem de mão-de-obra altamente especializada e capacitada e saber motivá-las e desenvolvê-las é um fator de extrema importância para alcançar êxito nessa área. Dessa forma, o presente trabalho buscou apresentar um modelo de alocação de tarefas em projetos de software que considera, além dos objetivos organizacionais (custos, prazos, entre outros), os anseios profissionais dos membros participantes. A estratégia proposta para a investigação é através da simulação de sistemas multiagentes, sistemas geralmente utilizados para simulação de entidades comportamentais. Com esse modelo pretendeu-se prover ao gestor uma ferramenta de auxílio na tomada de decisão no processo de alocação de tarefas em empresas de desenvolvimento de software, permitindo a visualização de diversos cenários e estratégias, gerando dessa forma melhor desempenho na execução de projetos e maior satisfação dos profissionais.

Le partage entre sources indépendantes hétérogènes et distribuées peut être résolu par la construction d’un système d’information coopératif (SIC). Un SIC est un ensemble de composants plus ou moins autonomes, souvent préexistants qui travaillent de manière synergique en échangeant information, expertise et en coordonnant leurs activités. Cela implique notamment la prise en compte de l’interopérabilité liée aux différences de description des données et de représentation de la sémantique. La gestion des données est alors assurée sans recours à un schéma global complet pour respecter l’autonomie des bases locales.Dans ce contexte, nous avons choisi de traiter le thème de la sécurité d’accès dans le but de garantir la confidentialité et l’intégrité des données de la coopération. La sécurité ajoute de nouveaux problèmes d’hétérogénéité et de résolution de conflits à ceux déjà existants en terme de coopération de données. Nous utilisons deux modèles canoniques proposés par l’équipe MODEME, permettant de représenter de manière unifiée les schémas locaux de données et de sécurité. Nous construisons un système pour la gestion de l’interopération des données et des politiques de sécurité, ainsi que la résolution sécurisée de requêtes globales. Nous avons choisi d’implémenter notre système en suivant le paradigme multi-agents avec une approche par médiation et intégration de schémas. Deux protocoles sont définis au regard des deux fonctionnalités du système : - Un protocole de gestion des connaissances permettant de traiter le problème d’interopérabilité entre les différents modes de représentation des données et des modèles de sécurité, et de générer les appariements entre ces différents modèles. - Un protocole de résolution de requêtes à partir des connaissances globales construites a priori, dont l’objectif est de présenter des résultats sémantiquement cohérents et sécurisés. Les agents sont décrits dans leur buts, leurs interactions, leurs connaissances en définissant leur rôle pour chaque protocole, avec notamment le rôle clé de médiateur de sécurité. Un scénario d’expérimentation permet d’illustrer sur un cas concret la génération des connaissances à partir des schémas locaux ainsi que le développement complexe du protocole pour le contrôle d’accès. / Sharing heterogeneous and distributed independent data sources can be solved by building a Cooperative Information System (CIS). A CIS is a set of components, exchanging information, expertise and coordinating their activities. This must consider interoperability related to differences of data description and semantic. Data management is provided without using a comprehensive global scheme to respect the autonomy of local databases.In this context, we chose to treat access security in order to ensure confidentiality and data integrity in a cooperation. This adds new security issues regarding heterogeneity and conflict resolution, on top of those in terms of data cooperation. We use two canonical models proposed by the MODEME team, to represent the local schemas and security policies in a unified way.We build a system to manage the interoperation of data and security policies, and the resolution of secure global queries. We have chosen to implement our system using the multi-agent paradigm, with an schema integration and mediation approach. Two protocols have been defined related to the features of the system: - A knowledge management protocol to address the problem of interoperability between different modes of data representation and security models, and to generate the matches between these different models. - A query resolution protocol using the global knowledge, which aims to present the results semantically consistent and secure. Agents are described in their goals, their interactions, their knowledge by defining their role for each protocol, including the key role of security mediator. An experimental scenario illustrates the knowledge generation from local schemas and the development of protocol for access control.