É notável que o Sistema Elétrico de Potência (SEP) está constantemente exposto a adversidades ocasionadas por forças da natureza ou desgastes naturais de seus componentes. A localização com certa exatidão de uma falta no Sistema de Transmissão de Energia Elétrica (STEE), ocasionada por uma destas adversidades, contribui para a confiabilidade e segurança da operação do SEP. Neste contexto, este trabalho apresenta um método computacional de Localização de Faltas (LDF), desenvolvido em ambiente MATLAB®, baseado na Impedância Aparente (IA), utilizando dados de apenas um terminal, sendo o equacionamento desenvolvido em componentes simétricas. Este método tem como dados de entrada as componentes fundamentais de tensão e corrente do terminal local e a Potência de Curto Circuito (SCC) do sistema equivalente conectado ao terminal remoto. Para validar o método proposto foram realizadas diversas simulações para todos os tipos de faltas shunt, contemplando inúmeras condições destas, utilizando para isso, o programa computacional ATP-EMTP, alimentado com dados de parâmetros reais de um sistema de transmissão. Objetivando analisar a eficiência deste método, realizou-se a comparação de seus resultados com os resultados obtidos por um método estado-da-arte. A partir de análises e comparações, os resultados alcançados pertencentes ao método de LDF proposto, apresentam-se promissores para a LDF em sistemas de transmissão. / Electric Power Systems (EPS) are constantly exposed to adversities caused by forces of nature and components lifetime. The precise fault location in the Transmission System (TEPS), caused by these adversities, contributes to the reliability and safety the operation of the EPS. In this context, this work presents a computational method for fault location (FL), developed in MATLAB ® environment, based on the Apparent Impedance (AI) of the transmission line. The proposed method uses only one terminal data and is developed in symmetrical components equations. This method has as input data the fundamental components of voltage and current of the local terminal and the short circuit power (SCP) of the equivalent system connected to the remote terminal. The proposed method is validated through several simulations for all fault types, covering many fault conditions, making use of the software ATP-EMTP, fed with data from the actual parameters of a real life transmission system. Aiming to analyze the efficiency of this method, a comparison was made against the an state-of-the-art method. From analysis and comparison of the results obtained, is observed that the proposed scheme is capable to obtain precise fault location estimations and presents as promising method for the FL in transmission systems.

La cohésion d’un texte repose sur un ensemble de liens syntaxiques et pragmatico-sémantiques entre les unités de discours dont les reprises anaphoriques. Les reprises assurent la continuité référentielle et thématique de phrases successives mais ne suffisent pas à assurer la cohérence des enchaînements. Elles font l’objet d’inférences qui permettent au système cognitif de trouver la relation de cohérence exprimée par l’enchaînement. Dans les cas simples, le système cognitif résout une reprise pronominale en lui attribuant un antécédent contextuellement pertinent, sélectionné parmi les éléments saillants actifs en mémoire. Lorsque la reprise concerne des arguments pluriels, les opérations cognitives sont plus complexes. Rentrent alors en ligne de compte non seulement la saillance de ces référents mais aussi le mode de relation qu’ils entretiennent au sein du prédicat de leur phrase d’accueil. Un prédicat d’accueil collectif tendra à effacer les distinctions entre les référents et facilitera les reprises plurielles au contraire des prédicats distributifs qui faciliteront les reprises de singuliers. Les prédicats réciproques lexicaux, quant à eux, établissent des liens de symétrie entre les entités dont on fait l’hypothèse qu’ils nécessitent des traitements cognitifs spécifiques. Les liens établis entre les arguments de prédicats collectifs, distributifs et symétriques sont étudiés d’un point de vue théorique. L’influence de ces liens sur la résolution de reprises pronominales au pluriel d’arguments dispersés mais aussi au singulier d’un élément d’une pluralité, est examinée expérimentalement (tests psycholinguistiques en production et en compréhension). L’étude théorique permet de mettre en évidence des différences fondamentales entre prédicats réciproques et prédicats symétriques d’une part, et entre prédicats collectifs et prédicats symétriques d’autre part. Les résultats expérimentaux corroborent l’étude théorique et permettent de proposer un pattern argumental pour les verbes lexicalement réciproques. / The cohesion of a text is based on a set of syntactic and pragmatico-semantic links between discourse units, including anaphoric referring terms. These terms ensure referential and thematic continuity in successive sentences, but are not sufficient to ensure the coherence of the chains. They involve inferences that enable the cognitive system to find the relationship of coherence expressed by the chain. In simple cases, the cognitive system resolves a pronominal anaphor by allocating to it a contextually relevant antecedent, selected from the salient elements that are active in memory. When the co-reference concerns plural arguments, the cognitive operations are more complex. In this case, it is necessary to take into account include not only the salience of the referents but also the type of relationship they have within the predicate of their host sentence. A collective hosting predicate will tend to remove distinctions between referents and will facilitate plural co-reference, unlike distributive predicates that facilitate singular coreference. Lexically reciprocal predicates, meanwhile, establish links of symmetry between entities that, are predicted to require specific cognitive processing. The links established between collective, distributive and symmetrical predicates are studied from a theoretical viewpoint. The influence of these links on the resolution of pronominal anaphora whose antecedent forms part of a plurality, or whose antecedents are dispersed in a comitative construction, is examined experimentally (psycholinguistic tests of production and comprehension). The theoretical study points to fundamental differences between reciprocal predicates and symmetrical predicates on the one hand, and between collective predicates and symmetrical predicates on the other. The experimental results support the theoretical study and lead to the proposal of an argumental pattern for symmetrical verbs.

É notável que o Sistema Elétrico de Potência (SEP) está constantemente exposto a adversidades ocasionadas por forças da natureza ou desgastes naturais de seus componentes. A localização com certa exatidão de uma falta no Sistema de Transmissão de Energia Elétrica (STEE), ocasionada por uma destas adversidades, contribui para a confiabilidade e segurança da operação do SEP. Neste contexto, este trabalho apresenta um método computacional de Localização de Faltas (LDF), desenvolvido em ambiente MATLAB®, baseado na Impedância Aparente (IA), utilizando dados de apenas um terminal, sendo o equacionamento desenvolvido em componentes simétricas. Este método tem como dados de entrada as componentes fundamentais de tensão e corrente do terminal local e a Potência de Curto Circuito (SCC) do sistema equivalente conectado ao terminal remoto. Para validar o método proposto foram realizadas diversas simulações para todos os tipos de faltas shunt, contemplando inúmeras condições destas, utilizando para isso, o programa computacional ATP-EMTP, alimentado com dados de parâmetros reais de um sistema de transmissão. Objetivando analisar a eficiência deste método, realizou-se a comparação de seus resultados com os resultados obtidos por um método estado-da-arte. A partir de análises e comparações, os resultados alcançados pertencentes ao método de LDF proposto, apresentam-se promissores para a LDF em sistemas de transmissão. / Electric Power Systems (EPS) are constantly exposed to adversities caused by forces of nature and components lifetime. The precise fault location in the Transmission System (TEPS), caused by these adversities, contributes to the reliability and safety the operation of the EPS. In this context, this work presents a computational method for fault location (FL), developed in MATLAB ® environment, based on the Apparent Impedance (AI) of the transmission line. The proposed method uses only one terminal data and is developed in symmetrical components equations. This method has as input data the fundamental components of voltage and current of the local terminal and the short circuit power (SCP) of the equivalent system connected to the remote terminal. The proposed method is validated through several simulations for all fault types, covering many fault conditions, making use of the software ATP-EMTP, fed with data from the actual parameters of a real life transmission system. Aiming to analyze the efficiency of this method, a comparison was made against the an state-of-the-art method. From analysis and comparison of the results obtained, is observed that the proposed scheme is capable to obtain precise fault location estimations and presents as promising method for the FL in transmission systems.

O presente trabalho tem por objetivo a implementação de algoritmos de controle do tipo PID com sintonia automática de seus parâmetros para controle de processos térmicos de primeira ordem mais tempo morto. Existe um grande número de métodos para a sintonia automática, os quais requerem informações variadas sobre o processo. A sintonia automática dos parâmetros do controlador escolhida para implementação é baseada no método da realimentação a relê que utiliza o ganho estático e o período e amplitude críticos do processo e pode ser facilmente adotada na indústria. Este método foi implementado em um controlador industrial microprocessado que utiliza um microcontrolador da família 8051. A correta identificação dos parâmetros críticos do processo presume a geração de oscilações simétricas durante a realimentação a relê. O algoritmo de sintonia automática desenvolvido inclui correção em tempo real da amplitude do relê para evitar identificação falsa de um ciclo da oscilação e da assimetria na saída do processo durante a realimentação a relê. A implementação do algoritmo é simples e direta em termos de programação no ambiente de microcontroladores mesmo em linguagem de baixo nível (assembly). São apresentados resultados de simulação e os resultados experimentais realizados em um aparelho industrial comercial. / The objective of this work is the implementation of a PID control algorithm with auto-tuning for a thermal first order process plus dead-time. There is a number of methods for the auto-tuning of PID controllers which require different information on the process. The auto-tuning algorithm is implemented in a industrial controller based on the 8051 microcontroller family. The auto-tuning method considered for implementation is based on the relay feedback procedure which require the process steady state gain and the ultimate point on the Nyquist plot and can be easily adopted by industry. The method is also suited for pre-tuning adaptive controllers. The auto-tuning method used assume symmetrical oscillations during the relay feedback experiment to render the correct ultimate point on the Nyquist plot The proposed auto-tuning algorithm includes on-line correction for asymmetry oscillation and relay amplitude to avoid false capture of the limit cycle. The symmetrical oscillations are ensured by a DC bias signal added to the relay output. The algorithm is simply and straight to be written in a microcontroller programming language. Experimental results accomplished in a industrial PID auto-tuner are presented.

É notável que o Sistema Elétrico de Potência (SEP) está constantemente exposto a adversidades ocasionadas por forças da natureza ou desgastes naturais de seus componentes. A localização com certa exatidão de uma falta no Sistema de Transmissão de Energia Elétrica (STEE), ocasionada por uma destas adversidades, contribui para a confiabilidade e segurança da operação do SEP. Neste contexto, este trabalho apresenta um método computacional de Localização de Faltas (LDF), desenvolvido em ambiente MATLAB®, baseado na Impedância Aparente (IA), utilizando dados de apenas um terminal, sendo o equacionamento desenvolvido em componentes simétricas. Este método tem como dados de entrada as componentes fundamentais de tensão e corrente do terminal local e a Potência de Curto Circuito (SCC) do sistema equivalente conectado ao terminal remoto. Para validar o método proposto foram realizadas diversas simulações para todos os tipos de faltas shunt, contemplando inúmeras condições destas, utilizando para isso, o programa computacional ATP-EMTP, alimentado com dados de parâmetros reais de um sistema de transmissão. Objetivando analisar a eficiência deste método, realizou-se a comparação de seus resultados com os resultados obtidos por um método estado-da-arte. A partir de análises e comparações, os resultados alcançados pertencentes ao método de LDF proposto, apresentam-se promissores para a LDF em sistemas de transmissão. / Electric Power Systems (EPS) are constantly exposed to adversities caused by forces of nature and components lifetime. The precise fault location in the Transmission System (TEPS), caused by these adversities, contributes to the reliability and safety the operation of the EPS. In this context, this work presents a computational method for fault location (FL), developed in MATLAB ® environment, based on the Apparent Impedance (AI) of the transmission line. The proposed method uses only one terminal data and is developed in symmetrical components equations. This method has as input data the fundamental components of voltage and current of the local terminal and the short circuit power (SCP) of the equivalent system connected to the remote terminal. The proposed method is validated through several simulations for all fault types, covering many fault conditions, making use of the software ATP-EMTP, fed with data from the actual parameters of a real life transmission system. Aiming to analyze the efficiency of this method, a comparison was made against the an state-of-the-art method. From analysis and comparison of the results obtained, is observed that the proposed scheme is capable to obtain precise fault location estimations and presents as promising method for the FL in transmission systems.

Magister Scientiae (Physiotherapy) - MSc(Physio) / Peripheral neuropathy (PN) is a common neurological complication occurring in the asymptomatic and symptomatic stages of human immune deficiency virus (HIV) infection. The pain and other symptoms caused by PN can impair functional ability and limit physical activity that could affect quality of life (QoL). Additionally, studies done on quality of life of people living with HIV/AIDS have shown that, HIV-related neurological syndromes, including PN, significantly reduce QoL. The aim of this study was to determine the prevalence of peripheral neuropathy amongst and the quality of life of adults living with HIV/AIDS attending the out-patient clinic at Rutongo Hospital in Rulindo District in Rwanda. / South Africa

As part of the natural evolution and continued optimisation of their designs, current and future magnetic recording heads, used and proposed in technologies such as perpendicular recording, shingled magnetic recording and two-dimensional magnetic recording, often exhibit asymmetry in their structure. They consist of two semi-infinite poles separated by a gap (where the recording field is produced), with an inner gap faces inclined at an angle. Modelling of the fields from asymmetrical structures is complex, and no explicit solutions are currently available (only implicit conformal mapping solutions are available for rational inclination angles). Moreover, there is limited understanding on the correlation between the gap corner angle and the magnitude, distribution and wavelength response of these head structures. This research was therefore set out to investigate approximate analytical and semi-analytical methods for modelling the magnetic potentials and fields of two-dimensional symmetrical and asymmetrical magnetic recording heads, and deliver a quantitative understanding of the behaviour of the potentials and fields as functions of gap corner angles. The accuracy of the derived expressions (written in terms of the normalised root-mean-square deviation) was assessed by comparison to exact available solutions for limited cases, and to finite-element calculations on Comsol Multiphysics. Two analytical methods were derived to approximately model the fields from two-dimensional heads with tilted gap corners in the presence and absence of a soft magnetic underlayer (SUL): in the first method, the potential near a single, two-dimensional corner held at a constant potential is derived exactly through solution of Laplace's equation for the scalar potential in polar coordinates. Then through appropriate choice of enclosing boundary conditions, the potentials and fields of two corners at equal and opposite potentials and displaced from each other by a distance equal to the gap length were superposed to map the potential and field for asymmetrical and symmetrical heads. For asymmetrical heads, the superposition approximation provided good agreement to finite-element calculations for the limited range of exterior corner angles 0 from 0 (right-angled corner) to 45, due to the mismatch of surface charge densities on both poles for this geometry. For symmetrical head structures, the superposition approximation was found to yield remarkable agreement to exact solutions for all gap corner orientations from 0 (right-angled head) to 90 ("thin" gap head). In the second method derived in this research for modelling asymmetrical heads involved using a rational function approximation with free parameters to model the surface potential of asymmetrical heads. The free parameters and their functional dependence on corner angle were determined through fitting to finite-element calculations, enabling the derivation of analytical expressions for the magnetic fields that are in good agreement with exact solutions for all corner angels (0 to 90). To complement the two approximate methods for modelling the fields from asymmetrical and symmetrical heads, a new general approach based on the sine integral transform was derived to model the reaction of soft underlayers on the surface potential or field of any two-dimensional head structure, for sufficiently close head-to-underlayer separations. This method produces an infinite series of correction terms whose coefficients are functions of the head-to-underlayer separation and gap corner angle, that are added to the surface potential or field in the absence of an underlayer. This new approach demonstrated good agreement with finite-element calculations for sufficiently close head-to-underlayer separations, and with the classical Green's functions solutions for increasing separations. Using the derived analytical method and explicit expressions in this work, an understanding of the nature of the magnetic fields and their spectra as functions of the gap corner angles is gained. This understanding and analytical theory will benefit the modelling, design and optimisation of high performance magnetic recording heads.

The aim of this thesis, which is based on one-year participant observation in Petřín (a public park in the centre of Prague), is to approach key processes for creating and maintenance of this space in symmetrical-anthropological way. The park is not consider to be mere static background for human interaction, but, reversely, hybrid dynamical process created through a wide actor-web of both material and immaterial matter. The park is not just a public, thus human, space but a space shared by a great variety of actors. Among this hybrid actors human is the one who holds the power over park management, who can include, or reversely, exclude chosen actors from participation and maintenance of this space. The author of this thesis tries to answer the question: What is the basis for this kind of decisions? Who shall be given the "right to Petřín"? From which power position? The author argues that these decisions are politicised in all cases because there is no possibility to describe the complicated reality of Petřín in exhaustive scientific way and make ever-lasting claims about it. For this reason, all those decisions originate in historically particular context. As a consequence of some actors exclusion there is a threat of dissolution of both cultural and biological diversity of this space. Key words:...

Une meilleure performance économique du séchage industriel consiste à la fois en l'amélioration de la qualité des produits et de l’optimisation énergétique. Plusieurs technologies peuvent être appliquées pour intensifier le processus de séchage. Ce travail s’intéresse à l'effet des vibrations, qui peuvent accélérer le séchage via un échauffement par dissipation visqueuse, une amélioration de la migration de l’eau liquide et de la vapeur à l'intérieur du milieu poreux et une augmentation des coefficients de transferts externes de chaleur de masse. Toutefois, malgré les études numériques et expérimentales antérieures, l'évaluation de l'importance relative de ces mécanismes d'intensification doit encore être améliorée.Ce travail porte sur l'échelle du pore, simulée par un tube capillaire, partiellement rempli d'eau soumis à des vibrations sinusoïdales. Un dispositif expérimental complet a été conçu, développé et utilisé. Grâce à une procédure de traitement d'images pertinente, le comportement de l'eau à l'intérieur du tube (déplacement et le rayon du ménisque) ont été étudiés pour différentes configurations de vibrations (fréquence, amplitude, taille du tube). Les expériences avec des géométries symétriques et asymétriques sont destinées à imiter le transfert d'humidité pendant le séchage d'un milieu poreux. Quelle que soit la configuration étudiée, l'intensification du séchage a été clairement confirmée.Pour la configuration symétrique, une nouvelle cartographie 2D basée sur des nombres adimensionnels est définie pour résumer les effets capillaires et visqueux, et ainsi expliquer le comportement de l'eau en fonction des conditions vibratoires (rayon du tube, fréquence et amplitude). Pour la configuration asymétrique, les résultats expérimentaux sont discutés et comparés aux simulations obtenues avec deux modèles numériques (un modèle 0-D et un modèle 1-D). Le potentiel de la modélisation a été mis en évidence, mais il est encore limité par les hypothèses physiques aux interfaces. D'autres développements sont proposés afin d'augmenter la plage de validité et d'approfondir l'étude de ce système physique. / A better economic performance of industrial drying consists both in product quality improvement and energy efficiency enhancement. Several technologies can be applied to intensify the drying process. This work is focused on the effect of vibration that can assist the drying through heating-up due to viscous dissipation, enhancement of liquid and vapor migration inside the porous medium and the increase of the external heat and mass transfer. However, despite previous numerical and experimental studies, the assessment of the relative importance of these intensification mechanisms has still to be improved.This work is focused on the pore scale, simulated by a capillary tube, partially filled with water subjected to sinusoidal vibrations. A full experimental device was conceived, developed and used. Thanks to a relevant image processing chain, the behavior of water inside the tube (displacement and meniscus radius), was investigated for various vibration configurations (frequency, amplitude, tube size). Experiments with both symmetrical and asymmetrical geometries are intended to mimic the moisture transfer during drying of a porous medium. Whatever the configuration, the drying intensification was clearly confirmed.For the symmetrical configuration, a new 2D dimensionless number map is set to summarize the capillary and viscous effects and thus to explain the water behavior depending on vibration conditions (tube radius, frequency and amplitude). For the asymmetrical configuration, the experimental results are discussed and compared to the simulations obtained with two computational models (a 0-D model and 1-D model). The potential of modeling was evidenced but still limited by the surrounding physical assumptions. Further developments are proposed to increase the validity range and further investigate this physical system

<p> InxGa1_xAs (0.00</= x </= 0.42) metamorphic pseudosubstrate layers (MSLs) were studied as a means to change the lattice constant of the substrates and to modify the growth conditions of InAs quantum dots (QDs) by varying the strain. The MSLs showed symmetrical mosaicity about the 110 axes but the spread was different in the two orthogonal [110] and [110] directions. The anisotropy in the mosaic spread in two < 11 O > directions was correlated to asymmetry of kinks and multilevel-terrace growth front during the growth of InxGa1_xAs buffer layers. X-ray and electron diffraction along with the least squares criterion can interchangeably be employed to determine the lattice constant of the MS Ls. It is possible to grow a defect free MSL with employing a compositional undershoot relative to the terminating buffer layer. Asymmetric tilt was found in an In0.42Ga0.58As MSL grown on a singular (001) GaAs substrate with an initial layer of a low temperature ( < 300 °C) grown InGaP prior to the growth of step-graded InxGa1-xAs (x = 0.02 to 0.42) buffer layers. The tilt around [110] axis was correlated with the imbalance of the tilt component of the Burgers vector (BV) of the 60° α-dislocations. Climb and jog formation of β-dislocations in the presence of P-interstitials were considered as a plausible mechanism for multiplication of the like-sign BV α-dislocations. These results show that an asymmetric tilt boundary can be induced in mismatched heterointerfaces grown on singular substrates. The evolution of InAs QDs on InxGa1-xAs(0.0</= x </= 0.3) MSLs on GaAs substrates was studied. The results indicate that the ratio of the height (h) over lateral diameter (d) of the QDs decreases with decreasing strain (i.e. the morphology of the coherent islands evolve toward a uniform film morphology (h/d=0) with deceasing strain). This evolution is analogous to the current </p> / Thesis / Doctor of Philosophy (PhD)