Orientadores: Wagner Caradori do Amaral, Ricardo Jose Gabrielli Barreto Campello / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação / Made available in DSpace on 2018-08-06T08:36:39Z (GMT). No. of bitstreams: 1 Medeiros_AndersonViniciusde_M.pdf: 896535 bytes, checksum: 48d0d75d38fcbbd0f47f7c49823806f1 (MD5) Previous issue date: 2006 / Resumo: Esta dissertação apresenta uma metodologia para a geração e otimização de modelos fuzzy Takagi-Sugeno (TS) com Funções de Base Ortonormal (FBO) para sistemas dinâmicos não lineares utilizando um algoritmo genético. Funções de base ortonormal têm sido utilizadas por proporcionarem aos modelos propriedades como ausência de recursão da saída e possibilidade de se alcançar uma razoável capacidade de representação com poucos parâmetros. Modelos fuzzy TS agregam a essas propriedades as características de interpretabilidade e facilidade de representação do conhecimento. Enfim, os algoritmos genéticos se apresentam como um método bem estabelecido na literatura na tarefa de sintonia de parâmetros de modelos fuzzy TS. Diante disso, desenvolveu-se um algoritmo genético para a otimização de duas arquiteturas, o modelo fuzzy TS FBO e sua extensão, o modelo fuzzy TS FBO Generalizado. Foram analisados modelos locais lineares e não lineares nos conseqüentes das regras fuzzy, assim como a diferença entre a estimação local e a global (utilizando o estimador de mínimos quadrados) dos parâmetros desses modelos locais. No algoritmo genético, cada arquitetura contou com uma representação cromossômica específica. Elaborou-se para ambas uma função de fitness baseada no critério de Akaike. Em relação aos operadores de reprodução, no operador de crossover aritmético foi introduzida uma alteração para a manutenção da diversidade da população e no operador de mutação gaussiana adotou-se uma distribuição variável ao longo das gerações e diferenciada para cada gene. Introduziu-se ainda um método de simplificação de soluções através de medidas de similaridade para a primeira arquitetura citada. A metodologia foi avaliada na tarefa de modelagem de dois sistemas dinâmicos não lineares: um processo de polimerização e um levitador magnético / Abstract: This work introduces a methodology for the generation and optimization of Takagi-Sugeno (TS) fuzzy models with Orthonormal Basis Functions (OBF) for nonlinear dynamic systems based on a genetic algorithm. Orthonormal basis functions have been used because they provide models with properties like absence of output feedback and the possibility to reach a reasonable approximation capability with just a few parameters. TS fuzzy models aggregate to these properties the characteristics of interpretability and easiness to knowledge representation in a linguistic manner. Genetic algorithms appear as a well-established method for tuning parameters of TS fuzzy models. In this context, it was developed a genetic algorithm for the optimization of two architectures, the OBF TS fuzzy model and its extension, the Generalized OBF TS fuzzy model. Local linear and nonlinear models in the consequent of the fuzzy rules were analyzed, as well as the difference between local and global estimation (using least squares estimation) of the parameters of these local models. Each architecture had a specific chromosome representation in the genetic algorithm. It was developed a fitness function based on the Akaike information criterion. With respect to the genetic operators, the arithmetic crossover was modified in order to maintain the population diversity and the Gaussian mutation had its distribution varied along the generations and differentiated for each gene. Besides, it was used, in the first architecture presented, a method for simplifying the solutions by using similarity measures. The whole methodology was evaluated in modeling two nonlinear dynamic systems, a polymerization process and a magnetic levitator / Mestrado / Automação / Mestre em Engenharia Elétrica

Current Intel processors implement a variety of power saving features like frequency scaling and idle states. These mechanisms limit the power draw and thereby decrease the thermal dissipation of the processors. However, they also have an impact on the achievable performance. The various mechanisms significantly differ regarding the amount of power savings, the latency of mode changes, and the associated overhead. In this paper, we describe and closely examine the so-called software controlled clock modulation mechanism for different processor generations. We present results that imply that the available documentation is not always correct and describe when this feature can be used to improve energy efficiency. We additionally compare it against the more popular feature of dynamic voltage and frequency scaling and develop a model to decide which feature should be used to optimize inter-process synchronizations on Intel Haswell-EP processors.

Ce travail de recherche porte sur la modélisation mathématique des principaux procédés sidérurgiques en suivant une approche systémique. L’objectif est d’élaborer un outil de modélisation de l’ensemble de la filière destiné à l’optimiser du point de vue énergétique et environnemental. Nous avons développé des modèles physico-chimiques du haut fourneau, de la cokerie, de l’agglomération et du convertisseur. Ces modèles ont ensuite été reliés entre eux sous forme d’un diagramme de flux unique en utilisant le logiciel ASPEN Plus. Dans une première partie, nous nous sommes particulièrement intéressés au haut fourneau à recyclage, une variante innovante du haut fourneau dans laquelle les gaz de gueulard sont recyclés et réinjectés aux tuyères après capture du CO2. Nous avons testé une réinjection à un niveau (aux tuyères) et à deux niveaux (tuyères et ventre). Les résultats ont été comparés avec succès à des données expérimentales issues d’un réacteur pilote et montrent que le recyclage permet une baisse de plus de 20 % des émissions de CO2 du haut fourneau. Le recyclage à deux niveaux ne semble pas plus performant que celui à un seul niveau. Dans un deuxième temps, nous avons simulé le fonctionnement d’une usine sidérurgique intégrée dans son ensemble. Différentes configurations ont été testées, pour un haut fourneau classique ou un haut fourneau à recyclage, en considérant un éventuel recyclage du laitier de convertisseur à l’agglomération, et en étudiant l’influence de la teneur en silicium de la fonte sur toute la filière. On montre notamment qu’il est possible de réduire le prix de revient de la tonne d’acier en substituant et recyclant différents sous-produits / This research study deals with mathematical modeling of the main steelmaking processes following a systems approach. The objective was to build a modeling tool of the whole steelmaking route devoted to its energetic and environmental optimization. We developed physical-chemical models for the blast furnace, the coke oven, the sintering plant and the basic oxygen furnace. These models were then linked together in a single flow sheet using the ASPEN Plus software. First, we focused on the top gas recycling blast furnace, a novel variant of the blast furnace in which the top gas is recycled and re-injected into the tuyeres after CO2 removal and capture. We tested both a reinjection at one level (tuyeres only) and at two levels (tuyeres and shaft). The results were successfully compared with experimental data from a pilot reactor and demonstrate that recycling can lower the blast furnace CO2 emissions by more than 20%. Recycling at two levels does not seem more efficient than at a single level. Second, we simulated the operation of an entire integrated steelmaking plant. Different configurations were tested, using a conventional blast furnace or a top gas recycling blast furnace, considering a possible recycling of the converter slag to the sintering plant, and studying the influence of Si content in the hot metal on the entire steelmaking plant operation. We show that it is possible to reduce the cost of producing steel by substituting and recycling various by-products

Este trabalho tem por objetivo desenvolver uma abordagem fuzzy focando na estimação de curto prazo da demanda de potência ativa de um alimentador de sistema de distribuição de energia elétrica. A motivação para este trabalho encontra-se na redução do erro de estimação para que o sistema de distribuição como um todo seja corretamente operado. O destaque da abordagem desenvolvida é a metodologia de seleção de entradas para o sistema de estimação, que o treina fornecendo-lhe informações não redundantes e não desnecessárias sobre o comportamento da série temporal. Os resultados, obtidos com treinamento e teste de um sistema de inferência fuzzy multicamadas, mostram que as estimações realizadas selecionando as entradas do sistema de forma criteriosa apresentam menor erro que quando não há critério de seleção. Conclui-se então que a metodologia foi funcional e eficiente para o caso estudado, o que faz com que este trabalho resulte em válidas contribuições nas áreas de sistemas inteligentes, de sistemas dinâmicos e inclusive na forma metodológica de especificação de modelos de estimação de séries temporais. / This work aims to develop a fuzzy approach focusing on the short-term active power demand forecast in a feeder of an electrical energy distribution system. This work motivation lies on the reduction of the forecast error so that the whole distribution system can be correctly operated. The highlight of the developed approach is the methodology to select the inputs for the estimation system, which trains it giving to it non-redundant and non-unnecessary information about the time series behavior. The results, obtained by training and testing a multilayer fuzzy inference system, show that the estimations made by following a criterion to select the inputs have smaller error than when there is no selection criterion at all. It is therefore concluded that the methodology was functional and efficient for the case under study, what makes this work result in valid contributions for the fields of intelligent systems, dynamic systems and in the methodological way to specify models to estimate time series.

Un sistema de gestión de sequías apropiado requiere de la anticipación de los posibles efectos que un episodio de este tipo tenga sobre el sistema de recursos hídricos. Esta tarea sin embargo resulta más complicada de lo que parece. En primer lugar, debido al alto grado de incertidumbre existente en la predicción de variables hidrológicas futuras. Y en segundo, debido al riesgo de sobrerreacción en la activación de medidas de mitigación generando falsa sensación de escasez, o sequía artificial. A este respecto, los planes especiales de sequía proveen de herramientas para la gestión eficiente de situaciones con escasez de recursos y la preparación de cara a futuros eventos. De todos modos, las diferentes estrategias de operación seguidas en cada sistema de recursos hídricos hacen que las herramientas que en algunos casos resultaron altamente útiles no lo sean tanto cuando se aplican en sistemas distintos. Debido a la falta de tiempo y/o al exceso de confianza en los trabajos realizados por terceros, con excelentes resultados en sus respectivos casos, a veces se cae en el error de implementar metodologías no del todo apropiadas en sistemas con requisitos completamente distintos. El desarrollo y utilización de metodologías generalizadas aplicables a diferentes sistemas y capaces de proporcionar resultados adaptados a cada caso es, por tanto, muy deseable. Este es el caso de las herramientas de modelación de sistemas de recursos hídricos generalizadas. Estas permiten homogeneizar los procesos mientras siguen siendo los suficientemente adaptables para proporcionar resultados apropiados para cada caso de estudio. Esta tesis presenta una serie de herramientas destinadas a avanzar en el análisis y comprensión de los sistemas de recursos hídricos, haciendo énfasis en la prevención de sequías y la gestión de riesgos. Las herramientas desarrolladas incluyen: un modelo de optimización generalizado para esquemas de recursos hídricos, con capacidad para la representación detallada de cualquier sistema de recursos hídricos, y una metodología de análisis de riesgo basada en la optimización de Monte Carlo con múltiples series sintéticas. Con estas herramientas es posible incluir tanto la componente superficial como la subterránea del sistema estudiado dentro del proceso de optimización. La optimización está basada en la resolución iterativa de redes de flujo. Se probó la consistencia y eficiencia de diferentes algoritmos de resolución para encontrar un balance entre la velocidad de cálculo, el número de iteraciones, y la consistencia de los resultados, aportando recomendaciones para el uso de cada algoritmo dadas las diferencias entre los mismos. Las herramientas desarrolladas se aplican en dos casos de estudio reales en la evaluación y posibilidad de complementación de los sistemas de monitorización y alerta temprana de sequías existentes en los mismos. En el primer caso, se propone un enfoque alternativo para la monitorización de la sequía en el sistema de operación anual del río Órbigo (España), complementándolo con la utilización de la metodología de análisis de riesgo. En el segundo caso, las herramientas se emplean en un sistema con una estrategia de operación completamente distinta. Se estudia como el análisis de riesgo de la gestión óptima puede ayudar a la activación anticipada de los escenarios de sequía en los sistemas de los ríos Júcar y Turia, cuya operación es hiperanual. En esta ocasión, el sistema de indicadores existente goza de una gran confianza por parte de los usuarios. La metodología de análisis de riesgo es, sin embargo, capaz de anticipar los eventos de sequía con mayor alarma, aspecto que es deseable si se quiere evitar que los episodios en desarrollo vayan a más. En ambos casos se muestra como la evaluación anticipada de las posibles situaciones futuras del sistema permiten una definición confiable de los escenarios de sequía con suficiente antelación para la activación efectiva de medidas de prevención y/o mitigación en caso de ser necesarias. La utilización de indicadores provenientes de modelos frente a indicadores basados en datos observados es complementaria y ambos deberían utilizarse de forma conjunta para mejorar la gestión preventiva de los sistemas de recursos hídricos. El empleo de modelos de optimización en situaciones de incertidumbre hidrológica es muy apropiado gracias a la no necesidad de definir reglas de gestión para obtener los mejores resultados del sistema, y teniendo en cuenta que las reglas de operación habituales pueden no ser completamente adecuadas en estas ocasiones. / Haro Monteagudo, D. (2014). Methodology for the optimal management design of water resources system under hydrologic uncertainty [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/45996 / TESIS

Current Intel processors implement a variety of power saving features like frequency scaling and idle states. These mechanisms limit the power draw and thereby decrease the thermal dissipation of the processors. However, they also have an impact on the achievable performance. The various mechanisms significantly differ regarding the amount of power savings, the latency of mode changes, and the associated overhead. In this paper, we describe and closely examine the so-called software controlled clock modulation mechanism for different processor generations. We present results that imply that the available documentation is not always correct and describe when this feature can be used to improve energy efficiency. We additionally compare it against the more popular feature of dynamic voltage and frequency scaling and develop a model to decide which feature should be used to optimize inter-process synchronizations on Intel Haswell-EP processors.

This dissertation develops systems engineering and machine learning models for aviation maintenance support. With the constant increase in demand for air travel, aviation organizations compete to maintain airworthy aircraft to ensure the safety of passengers. Given the importance of aircraft safety, the aviation sector constantly needs technologies to enhance the maintenance experience, ensure system safety, and limit aircraft downtime. Based on the current literature, the aviation maintenance sector still relies on outdated technologies to maintain aircraft maintenance documentation, including paper-based technical orders. Aviation maintenance documentation contains a mixture of structured and unstructured technical text, mainly inputted by operators, making them prone to error, misunderstanding communication, and inconsistency. This dissertation intends to develop decision support models based on systems engineering and artificial intelligence models that can automate the maintenance documentation system, extract useful information from maintenance work orders, and predict the aircraft's top degrader signals based on textual data. The first chapter of this dissertation introduces the significant setbacks of the aviation industry and provides a working ground for the following chapters. The dissertation's second chapter develops a system engineering framework using model-based systems engineering (MBSE) methodology to model the aviation maintenance process using the systems engineering language (SysML). The outcome of this framework is the design of an automated maintenance system model that can be used to automate maintenance documentation, making it less prone to error. The third chapter of the dissertation uses textual data in maintenance work orders to develop a hybrid approach that uses natural language processing (NLP) and transformer models to predict the readiness of a legacy aircraft. The model was tested using a real-life case study of the EA-6B military aircraft. The fourth chapter of this dissertation develops an ensemble transformer model based on three different transformer models. The ensemble model leverages the benefits of three different transformer architectures and is used to classify events based on an aviation log-based dataset. This dissertation's final and fifth chapter summarizes key findings, proposes future work directions, and provides the dissertation's limitations.

Com a crescente complexidade das aplicações e sistemas computacionais, atualmente tem se tornado importante o uso de formalismos de várias naturezas na representação e modelagem de problemas complexos, como os sistemas reativos e concorrentes. Este trabalho apresenta uma contribuição na Tecnologia Adaptativa e uma nova técnica no desenvolvimento de uma aplicação para execução de alguns tipos de jogos, (General Game Playing), cuja característica está associada à capacidade de o sistema tomar conhecimento das regras do jogo apenas em tempo de execução. Com esse trabalho, amplia-se a classe de problemas que podem ser estudados e analisados sob a perspectiva da Tecnologia Adaptativa, através dos Dispositivos Adaptativos Cooperantes. A aplicação desenvolvida como exemplo neste trabalho introduz uma nova ótica no desenvolvimento de aplicações para jogos gerais (GGP) e abre novos horizontes para a aplicação da Tecnologia Adaptativa, como a utilização das regras para extração de informação e inferência. / The complexity of computer applications has grown so much that several formalisms of different kinds became important nowadays. Many systems (e.g. reactive and concurrent ones) employ such formalisms to represent and model actual complex problems. This work contributes to the field of Adaptive Technology, and proposes a new approach for developing general game playing system, whose feature is the capability to play a game by acknowledging the game rules only at run time. This work expands the set of problems that can be studied and analyzed under the Adaptive Technology perspective, by means of cooperating adaptive devices. The developed application used a new approach for general game playing development bringing and widens the application field of Adaptive Technology with subjects related to information extraction and inference based in the devices rules.

Dans cette thèse nous développons des structures de supervision permettant de définir des consignes optimales pour des actionneurs thermiques, ainsi que des stratégies de commande appropriées pour le pilotage d’une pompe à chaleur (PAC). Pour répondre à ces objectifs, plusieurs étapes ont été réalisées :- Modélisation orientée commande d’une PAC réversible, des thermistances, et de l’environnement permettant de les lier à l’intérieur de l’habitacle. Des modèles physiques ont été définis et intégrés dans une plateforme du type Model-in-the-Loop pour permettre a posteriori la validation des stratégies de commande et d’optimisation. - Commande d’une PAC. La linéarisation du modèle de PAC autour de certains points de fonctionnement a permis le développement de la commande de l’actionneur principal. La structure de commande proposée permet de prendre en compte, en boucle fermée, des contraintes d’état et d’entrée du système. Les performances de cette structure ont été analysées en considérant successivement des régulateurs principaux de type PI et Hinf. Enfin, des algorithmes réalisant le pilotage d’un actionneur secondaire du système ont été également proposés. - Optimisation des actionneurs thermiques. L’utilisation combinée de thermistances et de la PAC présente des avantages en termes de réduction de la consommation énergétique et/ou du maintien de la puissance thermique demandée dans des conditions aux limites de fonctionnement. Le problème d’optimisation a été résolu en deux temps : des solutions hors-ligne ont été obtenues par résolution d’un problème mixte en nombre entier avec modèle prédictif, puis utilisées pour déduire des stratégies embarquables sur le véhicule. / Through this thesis we develop some strategies to define the optimal set points for thermal actuators, as well as an adequate control strategy for a heat pump. To this extent, different steps were carried out:- Control-oriented modeling of a reversible heat pump, thermistors, and the environment connecting these elements to the interior of the car’s cabin. Simplified but non-linear physical models were defined and used to build a Model-in-the-Loop platform, which would later be considered for the validation of the control and optimization strategies. - Control of a heat pump. The linearization of the heat pump model around some operating points was used to develop the control of the electric compressor, being the main actuator. The proposed control structure takes into account, in closed loop, input and state constraints. The performance of the structure was analyzed by using main controllers of PI and Hinf type. Lastly, some control algorithms were also proposed to control a second actuator of the system. - Thermal actuators optimization. Despite the high efficiency of the heat pump, the use of thermistors can be advantageous both for reducing the energetic consumption and/or to ensure the thermal power requests in extreme conditions. The optimization problem was carried out in two steps: offline solutions were firstly obtained solving a mixed-integer problem with predictive model, then used to derive some strategies that could be embedded in the vehicle.

Com a crescente complexidade das aplicações e sistemas computacionais, atualmente tem se tornado importante o uso de formalismos de várias naturezas na representação e modelagem de problemas complexos, como os sistemas reativos e concorrentes. Este trabalho apresenta uma contribuição na Tecnologia Adaptativa e uma nova técnica no desenvolvimento de uma aplicação para execução de alguns tipos de jogos, (General Game Playing), cuja característica está associada à capacidade de o sistema tomar conhecimento das regras do jogo apenas em tempo de execução. Com esse trabalho, amplia-se a classe de problemas que podem ser estudados e analisados sob a perspectiva da Tecnologia Adaptativa, através dos Dispositivos Adaptativos Cooperantes. A aplicação desenvolvida como exemplo neste trabalho introduz uma nova ótica no desenvolvimento de aplicações para jogos gerais (GGP) e abre novos horizontes para a aplicação da Tecnologia Adaptativa, como a utilização das regras para extração de informação e inferência. / The complexity of computer applications has grown so much that several formalisms of different kinds became important nowadays. Many systems (e.g. reactive and concurrent ones) employ such formalisms to represent and model actual complex problems. This work contributes to the field of Adaptive Technology, and proposes a new approach for developing general game playing system, whose feature is the capability to play a game by acknowledging the game rules only at run time. This work expands the set of problems that can be studied and analyzed under the Adaptive Technology perspective, by means of cooperating adaptive devices. The developed application used a new approach for general game playing development bringing and widens the application field of Adaptive Technology with subjects related to information extraction and inference based in the devices rules.