Utilização de experimentos com relés para identificação de pontos da resposta em freqüência de sistemas multivariáveis quadrados e sintonia de controladores PID descentralizados. / It use relay experiments for square multivariable system frequency response and decentralized PID controller tunning.

Utilização de experimentos com relés para identificação de pontos da resposta em freqüência de sistemas multivariáveis quadrados e sintonia de controladores PID descentralizados. / It use relay experiments for square multivariable system frequency response and decentralized PID controller tunning.

Multi-vehicle systems offer many advantages in engineering applications such as increased efficiency and robustness. However, the disadvantage of multi-vehicle systems is that they require a high level of organization and coordination in order to successfully complete a task. Formation control is a field of engineering that addresses this issue, and provides coordination schemes to successfully implement multi-vehicle systems. Two approaches to group coordination were proposed in this work: geometric and self-organizing formations. A geometric reconfiguring formation was developed using the leader-follower method, and the self-organizing formation was developed using the velocity potential equations from fluid flow theory. Both formation controllers were first tested in simulation in MATLAB, and then implemented on the X80 mobile robot units. Various experiments were conducted to test the formations under difficult obstacle scenarios. The robots successfully navigated through the obstacles as a coordinated as a team using the self-organizing and geometric formation control approaches.

This thesis presents a decentralized electromagnetic formation flying (EMFF) control method using frequency-multiplexed sinusoidal control signals. We demonstrate the EMFF control approach in open-loop and closed-loop control experiments using a single-degree-of-freedom testbed with an electromagnetic actuation system (EAS). The EAS sense the relative position and velocity between satellites and implement a frequency-multiplexed sinusoidal control signal. We use a laser-rangefinder device to capture the relative position and an ARM-based microcontroller to implement the closed-loop control algorithm. We custom-design and build the EAS that implements the formation control in one dimension. The experimental results in this thesis demonstrate the feasibility of the decentralized formation control algorithm between two satellites.

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.

Cette thèse porte sur des techniques de commande à base d’optimisation dans le cadre des systèmes dynamiques Multi-Agents sous contraintes, plus particulièrement liées à l’évitement des collisions. Dans un contexte ensembliste, l’évitement des collisions au sein de la formation se traduit par des conditions de non intersection des régions de sécurité caractéristiques à chaque agent/obstacle. Grace à sa capacité à gérer les contraintes, la commande prédictive a été choisie parmi les méthodes de synthèse fondées sur des techniques d’optimisation. Tout d’abord, une structure de type leader-suiveur est considérée comme une architecture décentralisée élémentaire. La zone de fonctionnement de chaque suiveur est décidée par le leader et puis une loi de commande locale est calculée afin de garantir que les suiveurs restent à l’intérieur de la zone autorisée, permettant d’éviter les collisions. Ensuite, un déploiement des agents fondé sur l’approche de commande prédictive décentralisée, utilisant des partitions dynamiques de Voronoi, est proposé, permettant de ramener chaque agent vers l’intérieur de sa cellule Voronoi. Une des contributions a été de considérer le centre de Chebyshev comme cible à l’intérieur de chaque cellule. D’autres solutions proposent l’utilisation du centre de masse ou du centre obtenu par l’interpolation des sommets. Finalement, des méthodes ensemblistes sont utilisées pour construire un niveau supplémentaire de détection de défauts dans le cadre du système Multi-Agents. Cela permet l’exclusion des agents défectueux ainsi que l’intégration des agents extérieurs certifiés sans défauts dans la formation en utilisant des techniques de commande prédictive centralisée. / This thesis presents optimizationbased control techniques for dynamical Multi-Agent systems (MAS) subject to collision avoidance constraints. From the set-theoretic point of view, collision avoidance objective can be translated into non-overlapping conditions for the safety regions characterizing each agent/obstacle while maintaining the convergence towards a specified formation. Among the successful optimizationbased control methods, Model Predictive Control (MPC) is used for constraints handling. First, a leader-follower structure is considered as a basic decentralized architecture. The followers functioning zone assignment is decided by the leader and then the local linear feedback control is computed such that the follower operates strictly inside its authorized zone, offering anti-collision guarantees. Second, a dynamic Voronoi partition based deployment of the agents using an inner target driver is developed. The main novelty is to consider the Chebyshev center as the inner target for each agent, leading to an optimization-based decentralized predictive control design. In the same topic, other inner targets are considered such as the center of mass or vertex interpolated center. Third, set-theoretic tools are used to design a centralized FDI layer for dynamical MAS, leading to the exclusion of a faulty agent from the MAS formation and the integration of an external healthy/recovered agent in the current formation. The set-based FDI allows detecting and isolating these faulty agents to protect the current formation using centralized predictive control techniques.

Space activities have been increasing dramatically in the past decades. As a result, the number of space debris has also increased significantly. Therefore, it is necessary to clean up and remove them to prevent a collision between space debris and spacecraft. In this thesis, we focus on load transportation using tethers, which connect multiple robots and loads together with lightweight cables. We propose a generalized framework to model and calculate the interaction force for the tethered multi-robot system. Then, we develop centralized and decentralized non-linear Model Predictive Control (MPC) controllers to complete a transportation task. Two simulators, a numerical and physical simulator, are presented and used to evaluate the performance of the controllers. The numerical simulator is used to verify the proposed model and evaluate the controllers for the ideal case. The physical simulator is then used to validate the performance of both centralized and decentralized controllers in real-time settings. Finally, we demonstrate how the proposed controllers perform in two and three-dimensional experiments. / Rymdaktiviteter har ökat dramatiskt under de senaste årtiondena. Som en följd av detta har mängden rymdskräp också ökat avsevärt. Därför är det nödvändigt att rensa upp och avlägsna detta skräp för att förhindra kollisioner mellan rymdskräp och rymdfarkoster. I denna rapport fokuserar vi på transporter av rymdobjekt som är sammanbundna via en lätt kabel. Vi föreslår en allmän metod för att modellera och beräkna interaktionskraften för det förenade multirobotsystemet. Sedan utvecklar vi centraliserad och decentraliserad icke-linjär modell-prediktiv reglering, MPC (eng. Model Predictive Control), för att uppnå transportuppgiften. Två simulatorer, en numerisk och en fysisk simulator, presenteras och används för att utvärdera styrsystemets prestanda. Den numeriska simuleringen används för att verifiera den föreslagna modellen och utforma styrsystemet för det idealiska fallet. Den fysiska simuleringen används sedan för att validera prestandan för både det centraliserade och decentraliserade styrsystem i realtid. Slutligen demonstrerar vi hur de föreslagna styrsystemen utför sig i tre- respektive två-dimensionella experiment.

No description available.

Esta tese de doutorado visa o estudo e o desenvolvimento de topologias e técnicas de controle para a integração de fontes alternativas tais como, solar e eólica acopladas a um barramento comum em corrente continua (CC) e conectá-las à rede de distribuição. O sistema contará também com elementos armazenadores como bancos de baterias e ultracapacitores, formando assim uma estrutura híbrida de armazenamento. Algoritmos de gerenciamento de energia serão implementados para que o perfil de injeção de potência na rede seja suave, eliminando as oscilações que são criadas, naturalmente, por fontes dependentes de fatores climáticos. Como consequência, os sistemas formados por fontes alternativas podem se tornar confiáveis e previsíveis, melhorando a capacidade de planejamento em um cenário cujos sistemas apresentem uma participação elevada na matriz energética. Duas metodologias de gerenciamento de energia são executadas neste trabalho, na primeira o ultracapacitor é gerenciado de modo a permitir a transferência de potência constante para a rede de distribuição em intervalos da ordem de minutos. A segunda estratégia se baseia no uso de banco de baterias combinado com ultracapacitores, formando uma estrutura híbrida de armazenamento. Nessa estrutura de gerenciamento, os armazenadores se comunicam entre si de forma a realizar um compartilhamento e filtragem de energia, fazendo com que transitórios de potência não sejam transmitidos para a rede de distribuição. Nesta estratégia, as baterias são responsáveis pelo fornecimento/absorção da potência média enquanto os ultracapacitores se encarregam dos transitórios. No segundo instante outras duas metodologias de divisão de carga são propostas para microrredes híbridas, contudo são baseadas em estratégias descentralizadas, ou seja, os armazenadores não se comunicam entre si para realizar o compartilhamento. Resultados experimentais e simulações irão comprovar a efetividade das metodologias de gerenciamento propostas. / This Ph.D. dissertation aims the study and development of topologies and control techniques to integrate various alternative sources such as solar and wind, coupled to a direct current (DC) common bus and connect them to the distribution grid. Storage devices such as battery banks and ultracapacitors will form a hybrid storage structure that is responsible for the power supplying in periods in which the sources are unable (times of the day in which the light incidence is low or when the wind amount is scarce). Power management algorithms will be implemented so the alternative sources and storage devices exchange energy, in order to make smoother the power injection profile in the grid, eliminating the fluctuations that are created naturally by alternative sources. With a smooth power profile, energy management systems based on alternative sources may become more reliable and predictable, improving planning capacity in a scenario in which the renewable energy sources have a high penetration in the energy matrix. To obtain such a result, two power management methodologies are executed; the first one is based on ultracapacitors and aims to deliver constant power to the distribution network, even when the power production is zero. However, this technique allows constant power just for a few minutes, once the ultracapacitor capacity is limited. The second strategy is based on the bank of batteries combined with ultracapacitors, forming the hybrid storage system. In this management structure, the storage devices communicate with each other in order to perform a power sharing, resulting in a filtrated power profile delivered to the distribution network. In this strategy, the batteries are responsible to providing average power while ultracapacitors are in-charge of the transient power, sparing the batteries from supplying power peaks. In a second moment, two other load sharing methodologies are proposed for hybrid systems, but are based on decentralized techniques, i.e. storage devices do not communicate with each other to make the power sharing. Experimental and simulated results will prove the effectiveness of the control strategies and management methodologies.

Cette thèse est le résultat de recherche effectuée à Longwy au sein du département CID « Contrôle Identification et Diagnostic» du Centre de Recherche en Automatique de Nancy (CRAN). Elle concerne, d’une part, la synthèse des observateurs dynamiques (d’ordre plein et d’ordre réduit) et la commande basée observateur d’une classe de systèmes linéaires incertains, d’autre part, l’application de ces résultats aux systèmes de grande dimension. Dans une première partie, une nouvelle forme d’observateurs dynamiques H-infini est conçue pour les systèmes linéaires en présence d’entrées inconnues et de perturbations, pour les systèmes continus et discrets. L’observateur proposé généralise ceux existants tels que les observateurs proportionnels et proportionnels-intégrales. La conception d’observateur est fondée sur la résolution des inégalités matricielles linéaires (LMI). Ensuite, ces observateurs ont été utilisés dans la synthèse de contrôleurs basés observateur pour les systèmes incertains en présence de perturbations. Cette synthèse est basée sur le paramétrage des solutions des contraintes algébriques obtenues à partir des erreurs d’estimation. La solution est obtenue à partir de la résolution des inégalités matricielles bilinéaires en utilisant un algorithme à 2 étapes.Dans la dernière partie, les résultats obtenus ont été étendus aux systèmes de grande dimension. Dans ce cadre, les systèmes considérés sont décomposés en plusieurs sous-systèmes interconnectés de faible dimension, où les interconnections sont supposées non linéaires et satisfaire des contraintes quadratiques. Une commande décentralisée basée observateur dynamique est proposée pour les systèmes interconnectés incertains en présence de perturbations / The present thesis is the result of research conducted in Longwy, within the department Control, Identification, Diagnosis (CID) of Research Center for Automatic Control of Nancy (CRAN). This thesis investigates the problem of dynamic observer (full- and reduced-order) and observer-based control design and their applications to large-scale systems. Firstly, a new form of H-infinity dynamic observer is designed for linear systems in the presence of unknown inputs and disturbances. The proposed observer generalizes the existing results on proportional observer and proportional integral observer. The observer design is based on the solution of linear matrix inequalities (LMI). Both continuous-time and discrete-time systems are considered. Thereafter, by inserting the proposed observer into a closed-loop, an observer-based control is presented for uncertain systems in the presence of disturbances. Based on the parameterization of algebraic constraints obtained from the analysis of the estimation error, the control design is derived from the solution of bilinear matrix inequality, by using a two-steps algorithm. Finally, the obtained results have been extended to large-scale systems. A decentralized observer-based control is proposed for large-scale uncertain systems in the presence of disturbances. These systems are composed of several interconnected subsystems of low dimensions, where the interconnections are assumed to be nonlinear and satisfy quadratic constraints