Avec l’épuisement progressif des ressources pétrolières, la production de biocarburant est actuellement en plein développement. Utiliser la biomasse en tant que source d’énergie permettrait également de limiter la concentration en dioxyde de carbone dans l’atmosphère et ainsi l’effet de serre. Cependant son processus de transformation implique une étape d’hydrodésoxygénation qui nécessite d’avoir recours à des catalyseurs de type sulfure de molybdène dopé par du cobalt ou du nickel. Ces catalyseurs sont également indispensables au cours du traitement du pétrole lors du procédé d’hydrodésulfuration qui est primordial pour limiter la pollution de l’air par les gaz de combustion des carburants. Cette thèse propose une méthode de synthèse et de structuration de catalyseurs massiques NiMoS facile à mettre en œuvre et rapide. Deux solvants ont été utilisés : l’eau et le formamide. L’ajout d’un polymère hydrosoluble amphiphile, le PEO113-b-PLLA32, dans l’eau et de liquides ioniques (BMIMNTf2 et BMIMBF4) dans le formamide ont permis la structuration des matériaux et l’obtention de surfaces spécifiques atteignant 55 m2g-1. Ceci représente une surface importante pour ce type de composé, en particulier dans le cadre d’une synthèse par chimie douce. Les premiers tests catalytiques réalisés vis-à-vis de la réaction d’hydrodésoxygénation de l’acide décanoïque montrent une activité catalytique très prometteuse pour ces matériaux. / With the progressive decrease of oil resources, the production of biomass fuel is a fast growing field. This could limit the concentration of carbon dioxide in the atmosphere and thus the global warming. The production of this kind of fuel requires the use of catalysts such as nickel or cobalt promoted molybdenum sulfide. These materials are also crucial during the refining of oil, an essential step to limit atmospheric pollution during the classic fuel combustion. New methods to synthesize and structure NiMoS catalysts have been developed in this work. Two solvants have been studied: formamide and water. The addition of an amphiphilic water-soluble polymer (PEO113-b-PLLA32) in water and of ionic liquids (BMIMNTf2 and BMIMBF4) in formamide led to the structuring of the materials with specific surface areas up to 55 m2g-1. This is a high value for this kind of material, especially since the synthesis is carried out under soft conditions. The first catalytic tests dealt with decanoïc acid hydrodesoxygenation and highlighted a promising activity for these materials

Afin de réduire la teneur en soufre des essences, l’hydrodésulfuration transforme les molécules soufrées en hydrocarbures en présence d’un catalyseur supporté (métaux de transition sulfurés : MoS2) et dopé (Co, Ni). Cette phase active, déposée sur des nanoparticules, présente un nombre plus important de défauts, sites actifs essentiels à la catalyse. Les nanoparticules ont un ratio S/V élevé, une grande réactivité de surface avec une juste utilisation des quantités de métaux. L’objectif de ce projet de thèse est de synthétiser des catalyseurs nanométriques de type coeur@coquille possédant une meilleure activité catalytique qu’un simple mélange mécanique de deux métaux sulfurés. Le coeur est composé de Fe3O4 ou de nanodiamants et la coquille de MoS2, NiMoS, CoMoS ou NiCoMoS, supporté sur du TiO2 ou de la γ-Al2O3. Une réaction modèle (HDS du thiophène) a été utilisée afin d’évaluer l’activité catalytique et d’optimiser la structure du catalyseur. L’étude portera sur les paramètres de synthèse et l’effet de la taille du coeur, la synthèse utilisée, les interactions entre le coeur et la coquille, le support, la (co)– promotion (Ni/Co) et l’activation par la température. / In hydrodesulfurization of fossil fuels, the sulfur levels are reduced by sulfur extraction from hydrocarbons by using supported catalysts (MoS2), doped (Co, Ni). Ultra-deep hydrodesulfurization will be achieved by improving new catalysts. Nanoparticles are a promising candidate with their high S/V ratio and permit to use the precise amount of metallic sulphide. The aim of this thesis is the synthesis of core@shell nanometric catalyst with improved activities. Core composed of Fe3O4 or nanodiamonds will be surrounded by a shell formed of MoS2, NiMoS, CoMoS or NiCoMoS, supported on TiO2, γ-Al2O3. Model reaction (thiophene) has allowed to compare conversion rates between each catalyst. Additionally, characterizations have provided a better understanding of the HDS catalyst structure and performances. Some factors have been investigated such as the size of the core, theinteractions between the core and the shell, the type of synthesis, the support chosen, the synergetic effect with doping ions and also the activation of the catalyst at low temperature.

L'utilisation de biohuiles comme carburants provenant de la transformation de différents types de biomasse (lignocellulosique, huile végétale, algue,...) est une alternative intéressante aux produits pétroliers développée depuis quelques années par l'Union Européenne. Cependant, le taux d'oxygène de ces biohuiles doit être fortement réduit avant leur introduction dans le pool carburant.Afin de modéliser ces huiles, différents composés oxygénés modèles ont été utilisés (acide décanoïque, décanal, phénol) seuls ou en mélange. Leur désoxygénation a été étudiée à 340 °C et sous 4 MPa, à la fois sur catalyseurs sulfures supportés sur alumine (Mo, CoMo, NiMo), et sur catalyseurs massiques de type NiMoS.Dans ces conditions, la désoxygénation de l'acide décanoïque conduit à la formation d'hydrocarbures en C10 et en C9, ces derniers étant majoritaires sur catalyseurs promus (CoMo et NiMo). La désoxygénation du phénol conduit à la formation d'aromatiques et de naphtènes, ces derniers étant majoritaires quel que soit le catalyseur utilisé. Un fort effet inhibiteur de l'acide sur la désoxygénation du phénol a été mis en évidence, attribué à un phénomène de compétition à l'adsorption en faveur de l'acide.Afin de s'affranchir de l'effet de support, des catalyseurs massiques constitués exclusivement de phases NiMoS ont été synthétisés en présence de différents agents (liquide ionique, formamide, eau). Leurs activités et sélectivités ont été mesurées en désoxygénation de l'acide décanoïque et du phénol. Sur ces deux composés, le solide formé dans l'eau a montré les meilleures activités. / Using bio-oil as engine fuels coming from conversion of biomass (lignocellulosic, vegetable oils, microalgae) is an interesting alternative to petroleum-based products depeloped over the last few years by the European Union. However, oxygen concentrations observed in bio-oil need to be sharply reduced before their introduction in the fuel pool.In order to model bio-oils, the effect of various oxygenates compounds was studied (decanoic acid, decanal and phenol) alone or in a mixture. Their deoxygenation was performed at 340 °C under 4 MPa over sulfided supported catalysts (Mo/Al2O3, CoMo/Al2O3, NiMo/Al2O3) and on bulk catalysts (NiMoS).Under these operating conditions, the deoxygenation of decanoic acid involved production of C10 and C9 hydrocarbons, the latter being the major products obtained on promoted catalysts (CoMo and NiMo). Phenol deoxygenation led to the production of aromatic and naphthenes, the latter being the major products whatever the catalyst used. A strong inhibiting effect of acid on the transformation of phenol was highlighted, assigned to a competitive adsorption in favour of the acid.In order to break out support effect, bulk catalysts formed exclusively by NiMoS phase were synthesized in presence of differents agents (ionic liquid, formamide, water). Activity and selectivity of these catalysts were measured in deoxygenation of two oxygenated model compounds (decanoic acid and phenol). For both oxygenated model compounds, the catalyst synthesized in water presented the highest activity.

Made available in DSpace on 2014-12-17T15:14:48Z (GMT). No. of bitstreams: 1 RooseweltFC.pdf: 5021440 bytes, checksum: 406e0f21dd64983ce0f8f2ea84fc6d62 (MD5) Previous issue date: 2007-12-17 / The complex behavior of a wide variety of phenomena that are of interest to physicists, chemists, and engineers has been quantitatively characterized by using the ideas of fractal and multifractal distributions, which correspond in a unique way to the geometrical shape and dynamical properties of the systems under study. In this thesis we present the Space of Fractals and the methods of Hausdorff-Besicovitch, box-counting and Scaling to calculate the fractal dimension of a set. In this Thesis we investigate also percolation phenomena in multifractal objects that are built in a simple way. The central object of our analysis is a multifractal object that we call Qmf . In these objects the multifractality comes directly from the geometric tiling. We identify some differences between percolation in the proposed multifractals and in a regular lattice. There are basically two sources of these differences. The first is related to the coordination number, c, which changes along the multifractal. The second comes from the way the weight of each cell in the multifractal affects the percolation cluster. We use many samples of finite size lattices and draw the histogram of percolating lattices against site occupation probability p. Depending on a parameter, ρ, characterizing the multifractal and the lattice size, L, the histogram can have two peaks. We observe that the probability of occupation at the percolation threshold, pc, for the multifractal is lower than that for the square lattice. We compute the fractal dimension of the percolating cluster and the critical exponent β. Despite the topological differences, we find that the percolation in a multifractal support is in the same universality class as standard percolation. The area and the number of neighbors of the blocks of Qmf show a non-trivial behavior. A general view of the object Qmf shows an anisotropy. The value of pc is a function of ρ which is related to its anisotropy. We investigate the relation between pc and the average number of neighbors of the blocks as well as the anisotropy of Qmf. In this Thesis we study likewise the distribution of shortest paths in percolation systems at the percolation threshold in two dimensions (2D). We study paths from one given point to multiple other points / O comportamento complexo de uma ampla variedade de fen?menos que s?o de interesse de matem?ticos, f?sicos, qu?micos e engenheiros ? caracterizado quantitativamente por meio de id?ias de distribui??es de fractais e multifractais, que correspondem de modo ?nico ? forma geom?trica e a propriedades din?micas dos sistemas em estudo. Nesta tese apresentamos o Espa?o dos Fractais e os m?todos de Hausdorff-Besicovitch, de Contagem de Caixas e de Escala, para calcular a Dimens?o Fractal de um Conjunto. Estudamos tamb?m fen?menos de percola??o em objetos multifractais constru?dos de maneira simples. O objeto central de nossas an?lises ? um objeto multifractal que chamamos de Qmf . Nestes objetos a multifractalidade surge diretamente da sua forma geom?trica. Identificamos algumas diferen?as entre percola??o nos multifractais que propusemos e percola??o em uma rede quadrada. Existem basicamente duas fontes destas diferen?as. A primeira est? relacionada com o n?mero de coordena??o, c, que muda ao longo do multifractal. A segunda vem da maneira como o peso de cada c?lula no multifractal afeta o aglomerado percolante. Usamos muitas amostras de redes de tamanho finito e fizemos o histograma de redes percolantes versus a probabilidade de ocupa??o p. Dependendo de um par?metro, ρ, que caracteriza o multifractal e o tamanho da rede, L, o histograma pode ter dois picos. Observamos que a probabilidade de ocupa??o no limiar de percola??o, pc, para o multifractal, em suporte d = 2, ? menor do que para a rede quadrada. Calculamos a dimens?o fractal do aglomerado percolante e o expoente cr?tico β. A despeito das diferen?as topol?gicas, encontramos que a percola??o em um suporte multifractal est? na mesma classe de universalidade da percola??o padr?o. A ?rea e o n?mero de vizinhos dos blocos de Qmf apresentam um comportamento n?o-trivial. Uma vis?o geral do objeto Qmf mostra uma anisotropia. O valor de pc ? uma fun??o de ρ que est? relacionada com esta anisotropia. Analisamos a rela??o entre pc e o n?mero m?dio de vizinhos dos blocos, assim como, a anisotropia de Qmf. Nesta tese estudamos tamb?m a distribui??o de caminhos m?nimos em sistemas percolativos no limiar de percola??o em duas dimens?es (2D). Estudamos caminhos que come?am em um determinado ponto e terminam em v?rios outros pontos. Na terminologia da ind?stria do petr?leo, ao ponto inicial dado associamos um po?o de inje??o (injetor) e aos outros pontos associamos po?os de produ??o (produtores). No caso padr?o apresentado anteriormente de um po?o de inje??o e um po?o de produ??o, separados por uma dist?ncia euclidiana r, a distribui??o de caminhos m?nimos l, P(l|r), apresenta um comportamento de lei-de-pot?ncia com expoente gl = 2, 14 em 2D. Analisamos a situa??o de um injetor e uma matriz A de produtores. Configura??es sim?tricas de produtores levam a uma distribui??o, P(l|A), com um ?nico pico, que ? a probabilidade que o caminho m?nimo entre o injetor e a matriz de produtores seja l, enquanto que as configura??es assim?tricas levam a v?rios picos na distribui??o P(l|A). Analisamos situa??es em que o injetor est? fora e situa??es em que o injetor est? no interior do conjunto de po?os produtores. O pico em P(l|A) nas configura??es assim?tricas decai mais r?pido do que no caso padr?o. Para os caminhos muito longos todas as configura??es estudadas exibiram um comportamento de lei-de-pot?ncia com o expoente g ≃ gl.

Made available in DSpace on 2014-12-17T15:14:50Z (GMT). No. of bitstreams: 1 RooseweltFS.pdf: 5124149 bytes, checksum: e3abde7691299157913b5304193cd2d5 (MD5) Previous issue date: 2007-12-17 / The complex behavior of a wide variety of phenomena that are of interest to physicists, chemists, and engineers has been quantitatively characterized by using the ideas of fractal and multifractal distributions, which correspond in a unique way to the geometrical shape and dynamical properties of the systems under study. In this thesis we present the Space of Fractals and the methods of Hausdorff-Besicovitch, box-counting and Scaling to calculate the fractal dimension of a set. In this Thesis we investigate also percolation phenomena in multifractal objects that are built in a simple way. The central object of our analysis is a multifractal object that we call Qmf . In these objects the multifractality comes directly from the geometric tiling. We identify some differences between percolation in the proposed multifractals and in a regular lattice. There are basically two sources of these differences. The first is related to the coordination number, c, which changes along the multifractal. The second comes from the way the weight of each cell in the multifractal affects the percolation cluster. We use many samples of finite size lattices and draw the histogram of percolating lattices against site occupation probability p. Depending on a parameter, ρ, characterizing the multifractal and the lattice size, L, the histogram can have two peaks. We observe that the probability of occupation at the percolation threshold, pc, for the multifractal is lower than that for the square lattice. We compute the fractal dimension of the percolating cluster and the critical exponent β. Despite the topological differences, we find that the percolation in a multifractal support is in the same universality class as standard percolation. The area and the number of neighbors of the blocks of Qmf show a non-trivial behavior. A general view of the object Qmf shows an anisotropy. The value of pc is a function of ρ which is related to its anisotropy. We investigate the relation between pc and the average number of neighbors of the blocks as well as the anisotropy of Qmf. In this Thesis we study likewise the distribution of shortest paths in percolation systems at the percolation threshold in two dimensions (2D). We study paths from one given point to multiple other points. In oil recovery terminology, the given single point can be mapped to an injection well (injector) and the multiple other points to production wells (producers). In the previously standard case of one injection well and one production well separated by Euclidean distance r, the distribution of shortest paths l, P(l|r), shows a power-law behavior with exponent gl = 2.14 in 2D. Here we analyze the situation of one injector and an array A of producers. Symmetric arrays of producers lead to one peak in the distribution P(l|A), the probability that the shortest path between the injector and any of the producers is l, while the asymmetric configurations lead to several peaks in the distribution. We analyze configurations in which the injector is outside and inside the set of producers. The peak in P(l|A) for the symmetric arrays decays faster than for the standard case. For very long paths all the studied arrays exhibit a power-law behavior with exponent g ∼= gl. / O comportamento complexo de uma ampla variedade de fen?menos que s?o de interesse de matem?ticos, f?sicos, qu?micos e engenheiros ? caracterizado quantitativamente por meio de id?ias de distribui??es de fractais e multifractais, que correspondem de modo ?nico ? forma geom?trica e a propriedades din?micas dos sistemas em estudo. Nesta tese apresentamos o Espa?o dos Fractais e os m?todos de Hausdorff-Besicovitch, de Contagem de Caixas e de Escala, para calcular a Dimens?o Fractal de um Conjunto. Estudamos tamb?m fen?menos de percola??o em objetos multifractais constru?dos de maneira simples. O objeto central de nossas an?lises ? um objeto multifractal que chamamos de Qmf . Nestes objetos a multifractalidade surge diretamente da sua forma geom?trica. Identificamos algumas diferen?as entre percola??o nos multifractais que propusemos e percola??o em uma rede quadrada. Existem basicamente duas fontes destas diferen?as. A primeira est? relacionada com o n?mero de coordena??o, c, que muda ao longo do multifractal. A segunda vem da maneira como o peso de cada c?lula no multifractal afeta o aglomerado percolante. Usamos muitas amostras de redes de tamanho finito e fizemos o histograma de redes percolantes versus a probabilidade de ocupa??o p. Dependendo de um par?metro, ρ, que caracteriza o multifractal e o tamanho da rede, L, o histograma pode ter dois picos. Observamos que a probabilidade de ocupa??o no limiar de percola??o, pc, para o multifractal, em suporte d = 2, ? menor do que para a rede quadrada. Calculamos a dimens?o fractal do aglomerado percolante e o expoente cr?tico β. A despeito das diferen?as topol?gicas, encontramos que a percola??o em um suporte multifractal est? na mesma classe de universalidade da percola??o padr?o. A ?rea e o n?mero de vizinhos dos blocos de Qmf apresentam um comportamento n?o-trivial. Uma vis?o geral do objeto Qmf mostra uma anisotropia. O valor de pc ? uma fun??o de ρ que est? relacionada com esta anisotropia. Analisamos a rela??o entre pc e o n?mero m?dio de vizinhos dos blocos, assim como, a anisotropia de Qmf . Nesta tese estudamos tamb?m a distribui??o de caminhos m?nimos em sistemas percolativos no limiar de percola??o em duas dimens?es (2D). Estudamos caminhos que come?am em um determinado ponto e terminam em v?rios outros pontos. Na terminologia da ind?stria do petr?leo, ao ponto inicial dado associamos um po?o de inje??o (injetor) e aos outros pontos associamos po?os de produ??o (produtores). No caso padr?o apresentado anteriormente de um po?o de inje??o e um po?o de produ??o, separados por uma dist?ncia euclidiana r, a distribui??o de caminhos m?nimos l, P(l|r), apresenta um comportamento de lei-de-pot?ncia com expoente gl = 2, 14 em 2D. Analisamos a situa??o de um injetor e uma matriz A de produtores. Configura??es sim?tricas de produtores levam a uma distribui??o, P(l|A), com um ?nico pico, que ? a probabilidade que o caminho m?nimo entre o injetor e a matriz de produtores seja l, enquanto que as configura??es assim?tricas levam a v?rios picos na distribui??o P(l|A). Analisamos situa??es em que o injetor est? fora e situa??es em que o injetor est? no interior do conjunto de po?os produtores. O pico em P(l|A) nas configura??es assim?tricas decai mais r?pido do que no caso padr?o. Para os caminhos muito longos todas as configura??es estudadas exibiram um comportamento de lei-de-pot?ncia com o expoente g ≃ gl

Submitted by Automa??o e Estat?stica (sst@bczm.ufrn.br) on 2017-04-03T19:16:36Z No. of bitstreams: 1 JeanyAraujoMendesDaSilva_DISSERT.pdf: 8639285 bytes, checksum: 04582afa34b412438a38094bc21a9541 (MD5) / Approved for entry into archive by Arlan Eloi Leite Silva (eloihistoriador@yahoo.com.br) on 2017-04-04T22:15:25Z (GMT) No. of bitstreams: 1 JeanyAraujoMendesDaSilva_DISSERT.pdf: 8639285 bytes, checksum: 04582afa34b412438a38094bc21a9541 (MD5) / Made available in DSpace on 2017-04-04T22:15:25Z (GMT). No. of bitstreams: 1 JeanyAraujoMendesDaSilva_DISSERT.pdf: 8639285 bytes, checksum: 04582afa34b412438a38094bc21a9541 (MD5) Previous issue date: 2016-02-29 / Esta disserta??o insere-se no ?mbito da Perspectiva Textual-Interativa (JUBRAN; KOCH, 2006), que constitui uma vertente da Lingu?stica Textual. A pesquisa tem como prop?sito investigar a estrutura??o dos Segmentos T?picos M?nimos (SegTs m?nimos) em aulas para o ensino m?dio, bem como descrever esse processo. A unidade de an?lise tomada para a consecu??o desses objetivos ? o t?pico discursivo, uma vez que essa categoria permite operar no recorte de segmentos t?picos como tamb?m dos segmentos t?picos m?nimos. Do ponto de vista metodol?gico, o trabalho adotou o m?todo emp?rico-indutivo de an?lise, a partir de uma abordagem qualitativa. Partindo da proposta de investigar como ocorre o processo de estrutura??o interna de SegTs m?nimos (PENHAVEL, 2010) em aulas expositivas (SILVA, 2008), coletamos dados correspondentes a aulas de Biologia, de Geografia, de Hist?ria e de Portugu?s, em uma escola de Ensino M?dio da rede p?blica do estado de Pernambuco. Nessa dire??o, para realizar uma an?lise mais detalhada dos segmentos t?picos m?nimos, selecionamos uma das aulas para an?lise, por considerarmos que ? representativa dentro do conjunto das aulas expositivas. Finalmente, verificamos nas an?lises que o processo de estrutura??o dos SegTs m?nimos, particularmente da aula expositiva, tem como princ?pio geral o encadeamento sequencial de quatro unidades t?picas: Esquematiza??o inicial, Problema(quest?o), Explica??o e Ratifica??o-avalia??o. Al?m dessa, h? outras formas de estrutura??o que se realizam com tr?s, duas ou uma unidade t?pica, cujas bases permanecem no princ?pio geral de estrutura??o dos SegTs m?nimos. Assim, identificamos entre as formas mais recorrentes, SegTs organizados com base no encadeamento sequencial Quest?o-Explica??o-Avalia??o, Esquematiza??o inicial-Explica??o-Avalia??o; ou apenas Esquematiza??o inicial-Explica??o, Explica??o-Avalia??o. H? ainda SegTs estruturados t?o-somente com a unidade de Explica??o que ? a ?nica unidade t?pica de estrutura??o que ocorre em todos os SegTs. / This paper is part of the framework of the Textual-Interactive Grammar or Textual-Interactive Perspective (Jubran, KOCH, 2006), which is a branch of Textual Linguistic. The research aims to investigate the structure of the Minimum Topic segments in classes for high school, and to describe this process. The unit of analysis taken to accomplish these objectives is the discursive topic, because that category will operate in cropping topic segments as well as Minimum Topic segments. From a methodological point of view, the study adopted the empirical-inductive method of analysis from a qualitative approach. From the proposal to investigate how is the internal structuring of Minimum Topic SegTs (PENHAVEL, 2010) in expositive classes (SILVA, 2008), we collected corresponding biology, Geography, History and Portuguese classes, in a public school from Pernambuco. In this direction, to carry out a more detailed analysis of the topic segments, we selected one of the classes for analysis, because we believe that is representative in a set of classes. Finally, we noted in the analysis that the process of internal structuring of minimum SegTs, particularly the expositive classes, has as a general principle of structuring the sequential chaining four topical units: Outlining initial, Problem (question), Explanation and Ratification-assessment. In addition to this, there are other ways of structuring carried out with three, two or one topical unit but they based on the general principle of structuring of minimum SegTs. Thus, we identified among the most recurrent forms, SegTs are organized with the sequential chaining Question, Explanation, Ratification-assessment; Outlining initial, Explanation, Ratification-assessment; or just Outlining initial, Explanation, Ratification-assessment; Explanation, Ratification-assessment. There are still SegTs structured only with the Explanation unit. Furthermore, it is the only topic unit that occurs in all SegTs.

Made available in DSpace on 2014-12-17T14:56:05Z (GMT). No. of bitstreams: 1 PatriciaNG.pdf: 767778 bytes, checksum: 287c497d20e307221386eeee17df3f9d (MD5) Previous issue date: 2005-02-03 / This work presents a modelling and identification method for a wheeled mobile robot, including the actuator dynamics. Instead of the classic modelling approach, where the robot position coordinates (x,y) are utilized as state variables (resulting in a non linear model), the proposed discrete model is based on the travelled distance increment Delta_l. Thus, the resulting model is linear and time invariant and it can be identified through classical methods such as Recursive Least Mean Squares. This approach has a problem: Delta_l can not be directly measured. In this paper, this problem is solved using an estimate of Delta_l based on a second order polynomial approximation. Experimental data were colected and the proposed method was used to identify the model of a real robot / Esta disserta??o apresenta uma metodologia de modelagem e identifica??o para um rob? m?vel dotado de rodas. O modelo din?mico discreto proposto leva em considera??o a din?mica dos atuadores. Ao contr?rio da abordagem cl?ssica, onde as coordenadas de posi??o do rob? (x, y) s?o usadas como vari?veis de estado (o que resulta em um modelo n?o linear), o modelo discreto proposto utiliza o incremento na dist?ncia percorrida pelo rob? Delta_l. Com isto, o modelo resultante ? linear e invariante no tempo, podendo ser identificado utilizando qualquer m?todo cl?ssico, como por exemplo os m?nimos quadrados recursivos. Um problema desta abordagem ? que a vari?vel Delta_l n?o ? diretamente mensur?vel. Nesta proposta, este problema ? contornado usando uma estimativa de Delta_l calculada assumindo que o caminho percorrido durante um per?odo de amostragem pode ser aproximado por uma curva de segundo grau. O m?todo proposto ? validado atrav?s de resultados de simula??o computacional e experi?ncias pr?ticas

Made available in DSpace on 2014-12-17T14:55:56Z (GMT). No. of bitstreams: 1 EllonPM_DISSERT.pdf: 1231242 bytes, checksum: 49456bef5c0d0bfdc5bf49d689568b60 (MD5) Previous issue date: 2012-01-27 / Conselho Nacional de Desenvolvimento Cient?fico e Tecnol?gico / Several mobile robots show non-linear behavior, mainly due friction phenomena between the mechanical parts of the robot or between the robot and the ground. Linear models are efficient in some cases, but it is necessary take the robot non-linearity in consideration when precise displacement and positioning are desired. In this work a parametric model identification procedure for a mobile robot with differential drive that considers the dead-zone in the robot actuators is proposed. The method consists in dividing the system into Hammerstein systems and then uses the key-term separation principle to present the input-output relations which shows the parameters from both linear and non-linear blocks. The parameters are then simultaneously estimated through a recursive least squares algorithm. The results shows that is possible to identify the dead-zone thresholds together with the linear parameters / V?rios rob?s m?veis apresentam comportamentos n?o-lineares, principalmente ocasionados por fen?menos de atrito entre as partes mec?nicas do rob? ou entre o rob? e o solo. Modelagens puramente lineares apresentam-se eficientes em alguns casos, mas ? preciso levar em considera??o as n?o-linearidades do rob? quando se deseja movimentos ou posicionamentos precisos. Este trabalho prop?e um procedimento de identifica??o param?trica do modelo de um rob? m?vel com acionamento diferencial, no qual s?o consideradas as n?o-linearidades do tipo zona-morta presentes nos atuadores do rob?. A proposta baseia-se no modelo de Hammerstein para dividir o sistema em blocos lineares e n?o-lineares. O princ?pio da separa??o do termo chave ? utilizado para demonstrar a rela??o entre as entradas e sa?das do sistema com os par?metros tanto da parcela linear quanto da n?o-linear. Os par?metros de ambas as parcelas s?o identificados simultaneamente, atrav?s de um algoritmo de m?nimos quadrados recursivo. Os resultados mostram que ? poss?vel identificar o valor os limites da zona-morta assim como os par?metros da parcela linear do modelo do sistema

Made available in DSpace on 2014-12-17T14:56:13Z (GMT). No. of bitstreams: 1 CarlosAAP_DISSERT.pdf: 1150903 bytes, checksum: a90e625336bbabe7e96da74cb85ee7aa (MD5) Previous issue date: 2013-01-31 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior / The pattern classification is one of the machine learning subareas that has the most outstanding. Among the various approaches to solve pattern classification problems, the Support Vector Machines (SVM) receive great emphasis, due to its ease of use and good generalization performance. The Least Squares formulation of SVM (LS-SVM) finds the solution by solving a set of linear equations instead of quadratic programming implemented in SVM. The LS-SVMs provide some free parameters that have to be correctly chosen to achieve satisfactory results in a given task. Despite the LS-SVMs having high performance, lots of tools have been developed to improve them, mainly the development of new classifying methods and the employment of ensembles, in other words, a combination of several classifiers. In this work, our proposal is to use an ensemble and a Genetic Algorithm (GA), search algorithm based on the evolution of species, to enhance the LSSVM classification. In the construction of this ensemble, we use a random selection of attributes of the original problem, which it splits the original problem into smaller ones where each classifier will act. So, we apply a genetic algorithm to find effective values of the LS-SVM parameters and also to find a weight vector, measuring the importance of each machine in the final classification. Finally, the final classification is obtained by a linear combination of the decision values of the LS-SVMs with the weight vector. We used several classification problems, taken as benchmarks to evaluate the performance of the algorithm and compared the results with other classifiers / A classifica??o de padr?es ? uma das sub?reas do aprendizado de m?quina que possui maior destaque. Entre as v?rias t?cnicas para resolver problemas de classifica??o de padr?es, as M?quinas de Vetor de Suporte (do ingl?s, Support Vector Machines ou SVM) recebem grande ?nfase, devido a sua facilidade de uso e boa capacidade de generaliza??o. A formula??o por M?nimos Quadrados da SVM (do ingl?s, Least Squares Support Vector Machines ou LS-SVM) encontra um hiperplano de separa??o ?tima atrav?s da solu??o de um sistema de equa??es lineares, evitando assim o uso da programa??o quadr?tica implementada na SVM. As LS-SVMs fornecem alguns par?metros livres que precisam ser corretamente selecionados para alcan?ar resultados satisfat?rios em uma determinada tarefa. Apesar das LS-SVMs possuir elevado desempenho, v?rias ferramentas tem sido desenvolvidas para aprimor?-la, principalmente o desenvolvimento de novos m?todos de classifica??o e a utiliza??o de comit?s de m?quinas, ou seja, a combina??o de v?rios classificadores. Neste trabalho, n?s propomos tanto o uso de um comit? de m?quinas quanto o uso de um Algoritmo Gen?tico (AG), algoritmo de busca baseada na evolu??o das esp?cies, para aprimorar o poder de classifica??o da LS-SVM. Na constru??o desse comit?, utilizamos uma sele??o aleat?ria de atributos do problema original, que divide o problema original em outros menores onde cada classificador do comit? vai atuar. Ent?o, aplicamos o AG para encontrar valores efetivos para os par?metros de cada LS-SVM e tamb?m encontrando um vetor de pesos, medindo a import?ncia de cada m?quina na classifica??o final. Por fim, a classifica??o final ? dada por uma combina??o linear das respostas de cada m?quina ponderadas pelos pesos. Foram utilizados v?rios problemas de classifica??o, tidos como benchmarks, para avaliar o desempenho do algoritmo e comparamos os resultados obtidos com outros classificadores

Made available in DSpace on 2014-12-17T14:56:16Z (GMT). No. of bitstreams: 1 DiegoAF_DISSERT.pdf: 1045785 bytes, checksum: 38dd893df9531c8e71f483fc22aa3802 (MD5) Previous issue date: 2012-12-18 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior / This work proposes a new technique for phasor estimation applied in microprocessor numerical relays for distance protection of transmission lines, based on the recursive least squares method and called least squares modified random walking. The phasor estimation methods have compromised their performance, mainly due to the DC exponential decaying component present in fault currents. In order to reduce the influence of the DC component, a Morphological Filter (FM) was added to the method of least squares and previously applied to the process of phasor estimation. The presented method is implemented in MATLABr and its performance is compared to one-cycle Fourier technique and conventional phasor estimation, which was also based on least squares algorithm. The methods based on least squares technique used for comparison with the proposed method were: forgetting factor recursive, covariance resetting and random walking. The techniques performance analysis were carried out by means of signals synthetic and signals provided of simulations on the Alternative Transient Program (ATP). When compared to other phasor estimation methods, the proposed method showed satisfactory results, when it comes to the estimation speed, the steady state oscillation and the overshoot. Then, the presented method performance was analyzed by means of variations in the fault parameters (resistance, distance, angle of incidence and type of fault). Through this study, the results did not showed significant variations in method performance. Besides, the apparent impedance trajectory and estimated distance of the fault were analysed, and the presented method showed better results in comparison to one-cycle Fourier algorithm / Prop?e-se neste trabalho uma nova t?cnica de estima??o fasorial, a ser utilizada em rel?s num?ricos microprocessados digitais para prote??o de dist?ncia de linhas de transmiss?o, baseada no m?todo dos m?nimos quadrados recursivo, denominada m?nimos quadrados em caminhada aleat?ria modificada. Os m?todos de estima??o fasorial t?m seu desempenho comprometido devido, principalmente, ? componente DC de decaimento exponencial presente nas correntes de falta. Para reduzir a influ?ncia da componente DC, agregou-se ao m?todo de m?nimos quadrados um Filtro Morfol?gico (FM) aplicado previamente ao processo de estima??o fasorial. O m?todo apresentado foi implementado em ambiente MATLABr e o seu desempenho comparado aos m?todos convencionais de estima??o fasorial, tamb?m baseados em m?nimos quadrados, e ao algoritmo de Fourier de um ciclo. Os m?todos baseados na t?cnica de m?nimos quadrados utilizados para compara??o com o m?todo proposto foram: recursivo ponderado, com reinicializa??o da covari?ncia e caminhada aleat?ria. As an?lises de desempenho das t?cnicas foram realizadas por meio de sinais sint?ticos e sinais oriundos de simula??es no Alternative Transient Program (ATP). Em compara??o aos demais m?todos de estima??o fasorial, o m?todo proposto apresentou resultados satisfat?rios no que se refere ? velocidade de resposta, oscila??o em regime permanente e percentual de overshoot. Em seguida, o m?todo proposto teve seu desempenho analisado frente ?s varia??es nos par?metros de falta (resist?ncia, dist?ncia, ?ngulo de incid?ncia e tipo de falta). Nesse estudo, constatou-se que o m?todo n?o sofreu varia??es significativas em seus resultados. Al?m disso, analisou-se a trajet?ria da imped?ncia aparente e dist?ncia estimada da falta, ao qual o m?todo proposto apresentou melhores resultados em compara??o ao algoritmo de Fourier de um ciclo