[pt] O progresso tecnológico da computação móvel associada à Internet promove a concretização de uma visão futurista em que os usuários acessam dados e serviços a qualquer momento e em qualquer lugar. Entretanto, esta visão expõe questões importantes no desenvolvimento de aplicações, pois se deve dar maior atenção para a comunicação entre dispositivos móveis e serviços web, considerando restrições de processamento, armazenamento de dados, diversidade de padrões e disponibilidade dos serviços. Esta dissertação explora aspectos de engenharia de software como computação autonômica, frameworks e sistemas multiagentes no desenvolvimento de aplicações para dispositivos móveis dando ênfase ao gerenciamento do uso de serviços web. Para consolidar os conceitos abordados e prover um guia que facilite o desenvolvimento de aplicações autoadaptáveis, propõe-se um framework para auxiliar o desenvolvimento de aplicações para computação móvel. Para ilustrar a utilização do framework são apresentadas duas aplicações. A primeira aplicação é um sistema que apresenta o risco de deslizamento de massa em uma área a partir do cálculo de susceptibilidade baseado em informações sobre clima, relevo e solo. A segunda aplicação faz parte de um sistema de agência de viagens que define um pacote de viagens que melhor atende às necessidades do usuário. / [en] Technological progress in mobile computing and the Internet promotes an achievement of the futuristic vision in which users have access to data and services anytime and anywhere. However, this vision realization brings important issues to application development, because more attention is required to the communication between mobile devices and web services, especially regarding processing and data storage constraints, diversity standards and availability of services. This dissertation explores aspects of software engineering as autonomic computing, multi-agent systems and frameworks in developing applications for mobile devices, emphasizing autonomic management of web services use. Aiming the consolidation of the investigated concepts and provide a guide to facilitate the development of self-adapting applications, a framework is proposed to help in applications development for mobile computing. The development of two applications illustrates the framework appliance. The first application is a system that presents the risk of mass sliding in an area through a susceptibility calculation based on information of climate, topography and soil. The second application is part of a travel agency system that defines a travel package that best fit user needs.

Custom software development and maintenance is one of the key expenses associated with developing automated systems for mass customization. This paper presents a method for reducing the risk associated with this expense by developing a flexible environment for determining and executing dynamic workflow paths. Strategies for developing an autonomous agent-based framework and for identifying and creating web services for specific process tasks are presented. The proposed methods are outlined in two different case studies to illustrate the approach for both a generic process with complex workflow paths and a more specific sequential engineering process.

This thesis is focused on design of framework for creation an articial opponents in strategy games. We will analyze different types of strategy games and artificial intelligence systems used in these types of games. Next we will describe problems, which can occur  in these systems and why agent-based systems makes better artificial opponents. Next we will use knowledge from this research to design and implement framework, which will act as support for creating an artificial intelligence in strategy games.

This thesis is focused on design of framework for creation an articial opponents in strategy games. We will analyze different types of strategy games and artificial intelligence systems used in these types of games. Next we will describe problems, which can occur  in these systems and why agent-based systems makes better artificial opponents. Next we will use knowledge from this research to design and implement framework, which will act as support for creating an artificial intelligence in strategy games.

This diploma thesis is focusing on description of processing social network analysis, design and implementation of a model that simulates a particular social network and its analysis. Social networks are modern and very used in this time. They are very good point for exploring. This project deal with static analysis social network, where social network is constructed by graph. We nd out di erent properties of single component and than we establish signi cance of them. Relationships between components are important too for us, because they have a big influence on propagation information in network. Structural properties figure out existence of di fferent communities. We simulate social network with multi-agent systems, they are desirable for represent changes in network. Multi-agent systems have implemented a simulation model that represents a particular social network. His behaviour was analyzed and examinated by chosen methods.

L'objectif de cette thèse de doctorat est (i) d'étudier et de développer des méthodes d’analyse et de commande de systèmes de contrôle en réseau linéaires et non linéaires et (ii) de montrer le potentiel de ces approches dans des applications complexes pertinentes. À cet égard, la théorie des systèmes à plusieurs agents, la théorie des graphes algébriques et le consensus sont des outils méthodologiques les plus intéressants. Une attention particulière est accordée à la caractérisation des relations entre, d'une part, la topologie du graphe de communication qui sous-tend l'évolution du système à plusieurs agents considéré et, d'autre part, les propriétés spectrales de la matrice Laplacienne associée au graphe lui-même. Le contrôle d'un groupe d'agents autonomes est étudié sous différents angles. Le principal objectif de contrôle est de s’assurer que les agents travaillent ensemble de manière coopérative, où la coopération représente la relation étroite entre tous les agents de l'équipe, le partage de l'information jouant un rôle important. En particulier, beaucoup de problèmes de consensus/accord/ synchronisation /rendez-vous sont étudiés afin de guider un groupe d’agents vers un état commun. Le consensus est étudié dans un contexte à temps discret parce que la dynamique du système est en général continue alors que les mesures et les entrées de contrôle sont des données échantillonnées. En outre, la théorie des jeux est utilisée pour faire face aux problèmes de coordination distribués à plusieurs agents, avec une application aux réseaux connus sous le nom de Software Defined Networks. À cet égard, on peut montrer que, sous des protocoles correctement conçus, les joueurs convergent vers un équilibre unique de Wardrop. On concentre l’attention sur le contrôle distribué, car cette approche présente des avantages évidents par rapport à la centralisation, comme l'évolutivité et la robustesse. Pourtant, le contrôle distribué a également ses propres inconvénients : avant tout, un inconvénient est que chaque agent ne peut pas prédire efficacement le comportement global du groupe en se basant uniquement sur des informations locales. Une certaine attention est également accordée à la nécessité de sécuriser les réseaux électriques contre le danger des attaques cyber-physiques grâce au développement de technologies d'intelligence distribuée. À cet égard, sur la base de topologies de réseaux d'énergie réalistes, nous présentons brièvement la conception d'un schéma de protection contre les attaques dynamiques à un point et à points multiples en boucle fermée. Nous formulons et résolvons un problème d'optimisation non convexe soumis à une contrainte de stabilité de Lyapunov pour la représentation à plusieurs agents autonome d'un réseau électrique obtenue après la linéarisation et l'application des lois d’attaque et de contrôle de fréquence. Finalement, nous présentons des résultats obtenus sur : le pilotage exact de la dynamique non linéaire finie à données échantillonnées avec des retards sur les entrées, au sujet de la stabilisation à données échantillonnées et de la poursuite de l'orbite quasi-halo autour du point de libration translunaire L₂, et au sujet des algorithmes heuristiques basés sur des méthodes d'apprentissage par renforcement à plusieurs agents capables d'effectuer un contrôle adaptatif optimal de qualité de service / qualité de l’expérience dans des scénarios sans modèle. / The objective of this PhD thesis is (i) to investigate and develop methods for the analysis and design of linear and nonlinear networked control systems and (ii) to show the potential of such approaches in relevant complex applications. In this respect, multi-agent systems theory, algebraic graph theory and consensus are the most interesting methodological tools, and specific attention is paid to the characterization of the relationships between, on the one hand, the topology of the communication graph that underlies the evolution of the considered multiagent system and, on the other hand, the spectral properties of the Laplacian matrix associated with the graph itself. The control of a group of autonomous agents is investigated from different perspectives. The main control objective is to make sure that the agents work together in a cooperative fashion, where cooperation accounts for the close relationship among all agents in the team, with information sharing playing an important role. In particular, various problems regarding consensus/agreement/synchronization/rendezvous are investigated with the specific aim of driving a group of agents to some common state. Consensus is investigated in a discrete-time setting due to the fact that the system dynamics is normally continuous while the measurements and control inputs might only be made in a sampled-data setting. Moreover, game theory is relied upon in order to cope with distributed multi-agent coordination problems, with application to Software Defined Networks. In this respect, it can be shown that, under properly designed protocols, the players converge to a unique Wardrop equilibrium. We focus on distributed control, since this approach shows obvious benefits over centralization, such as scalability and robustness. Yet, it also has its own drawbacks: among all, one drawback is that each agent cannot effectively predict the overall group behaviour based on only local information. Some attention is also devoted to the need for securing power grids against the danger of cyber-physical attacks through the development of distributed intelligence technologies accompanied by appropriate security enforcements. In this respect, based on realistic power network topologies, we briefly present the design of a protection scheme against closed-loop single-point and multi-point dynamic load altering attacks. This is done by formulating and solving a non-convex optimization problem subject to a Lyapunov stability constraint for the autonomous multiagent representation of a power system obtained after linearization and application of the attack and frequency control laws. Eventually, we show some other results achieved in terms of the exact steeering of finite sampled nonlinear dynamics with input delays, of sampled-data stabilization and quasi-halo orbit following around the L₂ translunar libration point, and of heuristic algorithms based on multi-agent reinforcement learning methods capable of performing optimal adaptive Quality of Service/Quality of Experience control in model-free scenarios.

<div>Self-organization in a multi-robot system is a spontaneous process where some form of overall order arises from local interactions between robots in an initially disordered system. Cooperative coordination strategies for self-organization promote teamwork to complete a task while increasing the total utility of the system. In this dissertation, we apply prosocial behavioral concepts such as altruism and cooperation in multi-robot systems and investigate their effects on overall system performance on given tasks. We stress the significance of this research in long-term applications involving minimal to no human supervision, where self-sustainability of the multi-robot group is of utmost importance for the success of the mission at hand and system re-usability in the future.</div><div><br></div><div>For part of the research, we take bio-inspiration of cooperation from the huddling behavior of Emperor Penguins in the Antarctic which allows them to share body heat and survive one of the harshest environments on Earth as a group. A cyclic energy sharing concept is proposed for a convoying structured multi-robot group inspired from penguin movement dynamics in a huddle with carefully placed induction coils to facilitate directional energy sharing with neighbors and a position shuffling algorithm, allowing long-term survival of the convoy as a group in the field. Simulation results validate that the cyclic process allows individuals an equal opportunity to be at the center of the group identified as the most energy conserving position, and as a result robot groups were able to travel over 4 times the distance during convoying with the proposed method without any robot failing as opposed to without the shuffling and energy sharing process. </div><div><br></div><div>An artificial potential based Adaptive Inter-agent Spacing (AIS) control law is also proposed for efficient energy distribution in an unstructured multi-robot group aimed at long-term survivability goals in the field. By design, as an altruistic behavior higher energy bearing robots are dispersed throughout the group based on their individual energy levels to counter skewed initial distributions for faster group energy equilibrium attainment. Inspired by multi-huddle merging and splitting behavior of Emperor Penguins, a clustering and sequential merging based systematic energy equilibrium attainment method is also proposed as a supplement to the AIS controller. The proposed system ensures that high energy bearing agents are not over crowded by low energy bearing agents. The AIS controller proposed for the unstructured energy sharing and distribution process yielded 55%, 42%, 23% and 33% performance improvements in equilibrium attainment convergence time for skewed, bi-modal, normal and random initial agent resource level distributions respectively on a 2D plane using the proposed energy distribution method over the control method of no adaptive spacing. Scalability analysis for both energy sharing concepts confirmed their application with consistently improved performances different sized groups of robots. Applicability of the AIS controller as a generalized resource distribution method under certain constraints is also discussed to establish its significance in various multi-robot applications.</div><div><br></div><div>A concept of group based survival from damaging directional external stimuli is also adapted from the Emperor Penguin huddling phenomenon where individuals on the damaging stimuli side continuously relocate to the leeward side of the group following the group boundary using Gaussian Processes Machine Learning based global health-loss rate minima estimations in a distributed manner. The method relies on cooperation from all robots where individuals take turns being sheltered by the group from the damaging external stimuli. The distributed global health loss rate minima estimation allowed the development of two settling conditions. The global health loss rate minima settling method yielded 12.6%, 5.3%, 16.7% and 14.2% improvement in average robot health over the control case of no relocation, while an optimized health loss rate minima settling method further improved on the global health loss rate settling method by 3.9%, 1.9%, 1.7% and 0.6% for robot group sizes 26, 35, 70 and 107 respectively.</div><div><br></div><div>As a direct application case study of collaboration in multi-robot systems, a distributed shape formation strategy is proposed where robots act as beacons to help neighbors settle in a prescribed formation by local signaling. The process is completely distributed in nature and does not require any external control due to the cooperation between robots. Beacon robots looking for a robot to settle as a neighbor and continue the shape formation process, generates a surface gradient throughout the formed shape that allow robots to determine the direction of the structure forming frontier along the dynamically changing structure surface and eventually reach the closest beacon. Simulation experiments validate complex shape formation in 2D and 3D using the proposed method. The importance of group collaboration is emphasized in this case study without which the shape formation process would not be possible, without a centralized control scheme directing individual agents to specific positions in the structure. </div><div> </div><div>As the final application case study, a collaborative multi-agent transportation strategy is proposed for unknown objects with irregular shape and uneven weight distribution. Although, the proposed system is robust to single robot object transportation, the proposed methodology of transport is focused on robots regulating their effort while pushing objects from an identified pushing location hoping other robots support the object moment on the other end of the center of mass to prevent unintended rotation and create an efficient path of the object to the goal. The design of the object transportation strategy takes cooperation cues from human behaviors when coordinating pushing of heavy objects from two ends. Collaboration is achieved when pushing agents can regulate their effort with one another to maintain an efficient path for the object towards the set goal. Numerous experiments of pushing simple shapes such as disks and rectangular boxes and complex arbitrary shapes with increasing number of robots validate the significance and effectiveness of the proposed method. Detailed robustness studies of changing weight of objects during transportation portrayed the importance of cooperation in multi-agent systems in countering unintended drift effects of the object and maintain a steady efficient path to the goal. </div><div><br></div><div>Each case study is presented independent of one another with the Penguin huddling based self-organizations in response to internal and external stimuli focused on fundamental self-organization methods, and the structure formation and object transportation strategies focused on cooperation in specific applications. All case studies are validated by relevant simulation and experiments to establish the effectiveness of altruistic and cooperative behaviors in multi-robot systems.</div>

Background: Multiple Objective Optimization problems(MOOPs) are common and evident in every field. Container port terminals are one of the fields in which MOOP occurs. In this research, we have taken a case in logistics management and modelled Multi-agent systems to solve the MOOP using Non-dominated Sorting Genetic Algorithm-II (NSGA-II). Objectives: The purpose of this study is to build AI-based models for solving a Multiple Objective Optimization Problem occurred in port terminals. At first, we develop a port agent with an objective function of maximizing throughput and a customer agent with an objective function of maximizing business profit. Then, we solve the problem using the single-objective optimization model and multi-objective optimization model. We then compare the results of both models to assess their performance. Methods: A literature review is conducted to choose the best algorithm among the existing algorithms, which were used previously in solving other Multiple Objective Optimization problems. An experiment is conducted to know how well the models performed to solve the problem so that all the participants are benefited simultaneously. Results: The results show that all three participants that are port, customer one and customer two have gained profits by solving the problem in multi-objective optimization model. Whereas in a single-objective optimization model, a single participant has achieved earnings at a time, leaving the rest of the participants either in loss or with minimal profits. Conclusion: We can conclude that multi-objective optimization model has performed better than the single-objective optimization model because of the impartial results among the participants.

Decentralized decision-making is an active research topic in artificial intelligence. In a distributed system, a number of individually acting agents coexist. If they strive to accomplish a common goal, the establishment of coordinated cooperation between the agents is of utmost importance. With this in mind, our focus is on multi-agent reinforcement learning (RL) methods which allow for automatically acquiring cooperative policies based solely on a specification of the desired joint behavior of the whole system.The decentralization of the control and observation of the system among independent agents, however, has a significant impact on problem complexity. Therefore, we address the intricacy of learning and acting in multi-agent systems by two complementary approaches.First, we identify a subclass of general decentralized decision-making problems that features regularities in the way the agents interact with one another. We show that the complexity of optimally solving a problem instance from this class is provably lower than solving a general one.Although a lower complexity class may be entered by sticking to certain subclasses of general multi-agent problems, the computational complexitymay be still so high that optimally solving it is infeasible. Hence, our second goal is to develop techniques capable of quickly obtaining approximate solutions in the vicinity of the optimum. To this end, we will develop and utilize various model-free reinforcement learning approaches.Many real-world applications are well-suited to be formulated in terms of spatially or functionally distributed entities. Job-shop scheduling represents one such application. We are going to interpret job-shop scheduling problems as distributed sequential decision-making problems, to employ the multi-agent RL algorithms we propose for solving such problems, and to evaluate the performance of our learning approaches in the scope of various established scheduling benchmark problems.

京都大学 / 0048 / 新制・課程博士 / 博士(情報学) / 甲第18408号 / 情博第523号 / 新制||情||92(附属図書館) / 31266 / 京都大学大学院情報学研究科社会情報学専攻 / (主査)教授 多々納 裕一, 教授 石田 亨, 准教授 畑山 満則 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DFAM