Thèse numérisée par la Direction des bibliothèques de l'Université de Montréal.

Hevea brasiliensis est la seule source commerciale du caoutchouc naturel. La maladie CLF (Corynespora Leaf Fall) provoquée par le champignon Corynespora cassiicola est une pathologie foliaire de l’hévéa importante en Asie et en Afrique. Les épidémies sont liées au développement de nouvelles variétés clonales très sensibles. Une méthode conductimétrique de phénotypage de la réponse de l’hévéa aux exsudats de C. cassiicola (filtrats de culture ou toxine purifiée) a été développée et caractérisée. Ce test toxinique, sans contact direct avec le champignon lui-même, peut être mis en œuvre sans risque sur les plantations. Après application de gouttes d’exsudats sur des folioles détachées d’un génotype d’hévéa, deux mesures de conductivité C1 et C2 sont réalisées avant et après autoclavage, et le pourcentage de fuite d’électrolyte dû au traitement, %FE = 100*C1/C2, est calculé. L’influence de différents facteurs tels que le stade foliaire et la durée d’incubation, a été précisée. Une corrélation positive a été trouvée entre les réponses au test toxinique et à l’inoculation par des conidies du champignon. La pertinence de la correction de C1 par C2 a été analysée.Deux études du test toxinique ont été réalisées pour une série de clones d’hévéa, d’une part en France avec des plantes en pots cultivées en serre, et d’autre part en Côte d’Ivoire, dans les conditions normalement envisagées pour ce test, avec des folioles provenant de jardins à bois de greffe. La répétabilité entre ces deux études est imparfaite mais néanmoins satisfaisante. Le critère %FE est apparu plus sensible et plus précis que la mesure de la surface de nécrose induite (SN). Une variabilité importante de la sensibilité des clones et de l’agressivité des exsudats a été observée, avec l’existence d’interactions. L’analyse du déterminisme génétique de la réponse au test toxinique a été menée par détection de QTL sur deux populations de la famille biparentale d’hévéa PB260 x RRIM600, analysées sur deux sites en Côte d’Ivoire. Une carte génétique a été construite au préalable avec 308 marqueurs SSR. Le phénotypage des deux populations a révélé un déterminisme polygénique, sans effet monolocus majeur, avec, pour l’ensemble des exsudats fongiques étudiés, 13 QTL répartis sur 10 groupes de liaison et présentant des pourcentages d’explication R2 variant de 10 à 20 %. Sur les deux sites, deux QTL ont été identifiés conjointement pour le filtrat CCP et pour la toxine Cas1 purifiée à partir du filtrat CCP, démontrant l’importance de Cas1 pour la toxicité de ce filtrat. Des corrélations positives entre certains filtrats ont été observées, avec des profils de réponse similaires et des QTL communsLe test toxinique pourrait remplacer avantageusement la méthode d’inoculation conidienne. Cependant, la valeur prédictive du test pour la sensibilité des clones d’hévéa à la maladie CLF en plantation reste à caractériser. Une meilleure connaissance de l’épidémiologie du pathosystème H. brasiliensis x C. cassiicola contribuerait à l’objectif de sélection, et plus largement à la conception de méthodes de lutte ou d’évitement. / Hevea brasiliensis is the only commercial source of natural rubber. The CLF (Corynespora Leaf Fall) disease caused by the fungus Corynespora cassiicola is an important foliar pathology of the rubber tree in Asia and Africa. The disease progression is related with the development of new highly sensitive clonal varieties. A phenotyping method based on the measurement of electrolyte leakage in response to C. cassiicola exudates (culture filtrates or purified toxin) was developed and characterized. This test, without direct contact with the fungus itself, can be implemented without risk for plantations. After application of exudate samples on detached leaflets, conductivity is measured before (C1) and after (C2) autoclaving, and the percentage of electrolyte leakage %FE=100*C1/C2 is calculated. The influence of different factors such as leaf stage or the duration of incubation has been clarified. A positive correlation was found between the responses to the toxin test and to conidial inoculation. The pertinence of C1 correction by C2 was analyzed.Two toxin test studies have been carried out for a set of rubber clones, in France, with potted plants in greenhouse, and also in Ivory Coast, under the conditions normally considered for this test, with leaflets from budwood gardens. Repeatability between these two experiments was not perfect but nevertheless satisfying. The %FE character was found more sensitive and accurate than surface necrosis (SN) measurement. A significant variability in clone susceptibility and treatment virulence was observed, with clone x treatment interactions. The genetic determinism of sensitivity to C. cassiicola exudates was analyzed using the toxin test, by QTL detection on two populations of the biparental rubber family PB260 x RRIM600, in two field trials in Ivory Coast. A genetic map was previously built, using 308 SSR markers. Phenotyping of the two progeny populations revealed a polygenic determinism, without major monolocus effect. Over all the fungal exudates, 13 QTL distributed over 10 different linkage groups were identified, with percentages of explanation R2 varying from 10 to 20 %. On both sites, two QTL have been identified jointly for both the culture filtrate CCP and the toxin Cas1 purified from CCP. Positive correlations have been observed between some of the filtrates, with similar response profiles and common QTL.The toxin test could usefully replace the conidial inoculation method. However, the value of the test for predicting rubber clones susceptibility to the CLF disease in field conditions remains to be better characterized. Better understanding of the epidemiology of the H. brasiliensis x C. cassiicola pathosystem would contribute to the breeding objective, and more broadly to the design of control or avoidance methods.

L’hévéa (Hevea brasiliensis) est actuellement la seule source commerciale de caoutchouc naturel. Parmi les maladies affectant l’hévéa, la CLF (« Corynespora Leaf Fall ») causée par le champignon Corynespora cassiicola, est devenue en une cinquantaine d’années un fléau pour l’ensemble des pays hévéicoles d’Asie et d’Afrique. Actuellement, la gestion du problème consiste à arracher les clones les plus sensibles et à traiter les arbres avec des fongicides en cas d’épidémie. Cependant, le nombre de clones touchés par la maladie ne cessant d’augmenter, il devient urgent de sélectionner de nouveaux clones à la fois tolérants et aptes à la production. Nos travaux ont permis de caractériser le gène codant la cassiicoline, toxine protéique glycosylée secrétée par C. cassiicola, et d’analyser sa diversité. Une étude comparative portant sur trois isolats de C. cassiicola d’agressivité contrastée a montré la présence du gène Cas1 chez les isolats de forte et moyenne agressivité, alors qu’il n’est pas détecté chez l’isolat de faible agressivité. Les niveaux d’agressivité des isolats sont corrélés aux niveaux de transcrits du gène de cassiicoline. Le rôle de la cassiicoline serait prépondérant dans les phases précoces de l’infection. L’analyse de diversité du gène de cassiicoline à partir d’une collection d’isolats provenant de différents hôtes et d’origines géographiques variées, a révélé l’existence d’au moins six isoformes protéiques (Cas1 à Cas6). La structuration génétique globale des isolats basée sur des marqueurs neutres est similaire à la structuration basée sur le gène de cassiicoline. Les isolats prélevés sur hévéa se regroupent en clades spécialisés, dont un correspondant aux isolats porteurs du gène Cas1, identifiés comme étant les plus agressifs sur hévéa. Cependant, 58 % des isolats testés semblent dépourvus de gène de cassiicoline, bien que certains génèrent des symptômes modérés sur hévéa, ce qui suggère l’existence d’autres effecteurs. Des formes endophytiques de C. cassiicola ont été isolées à partir de feuilles asymptomatiques provenant du Brésil, zone encore indemne de CLF. Les gènes de cassiicoline portés par ces souches (isoformes Cas3 et 4) ne semblent pas exprimés lors de l’interaction avec l’hévéa. Nous avons montré par ailleurs que les champignons endophytes de l’hévéa appartenant aux genres Trichoderma et Xylaria présentent une forte activité mycoparasitaire sur C. cassiicola, in vitro. Ces travaux ouvrent de nouvelles perspectives pour le contrôle de la maladie (diagnostic précoce, sélection de clones tolérants, lutte biologique). / Rubber tree (Hevea brasiliensis) is the only source of commercial natural rubber. The “Corynespora Leaf Fall” (CLF) disease, caused by the fungus Corynespora cassiicola, has become over the last 50 years a very serious problem in the Asian and African rubber producing countries. Currently, the main methods to face the problem are the uprooting of the most susceptible cultivars and chemical treatments in case of severe outbreaks. However, the number of cultivars affected by the disease keeps increasing, and the selection of new cultivars, both tolerant and high yielding, is urgent. In this study, the gene encoding cassiicolin, a toxic glycosylated protein secreted by C. cassiicola, was characterized and its diversity analyzed. The analysis of three C. cassiicola isolates with contrasted levels of aggressiveness revealed the presence of identical cassiicolin genes in the highly and moderately aggressive isolates but none in the isolate of mild aggressiveness. The levels of aggressiveness were correlated to the cassiicolin gene transcript levels. The cassiicolin gene was preferentially expressed in the early phase of the infection. Analysis of the cassiicolin gene diversity among isolates from various host and various geographical origins revealed the existence of at least six protein isoforms (Cas1 to Cas6). The genetic structure of the isolates based on neutral markers was closely related to the genetic structure based on the cassiicolin gene. The isolates sampled on rubber tree were grouped in several specialized clades, including one clade regrouping all the Cas1 isolates, which were the most aggressive on rubber tree. However, 58 % of the isolates seemed to be deprived of cassiicolin gene, although some of them generate moderate symptoms on rubber tree, suggesting the existence of other effectors. Endophytic C. cassiicola isolates were found in asymptomatic rubber tree leaves from Brazil, a region were CLF outbreak was never reported. No expression of the cassiicolin genes carried by these isolates could be detected in interaction with rubber tree. In addition, we found that the Trichoderma and Xylaria species, all fungal endophytes of rubber tree, were antagonists of C. cassiicola, in vitro. This work opens new perspectives for the control of CLF (through early diagnosis, selection of tolerant clones, or biocontrol).

L'hévéa (Hevea brasiliensis) est la seule source de caoutchouc naturel commercialisé à travers le monde. En Afrique et en Asie, la maladie 'Corynespora Leaf Fall' (CLF), causée par le champignon nécrotrophe Corynespora cassiicola, affecte les plantations hévéicoles en provoquant des défoliations massives sur les clones les plus sensibles. L’évaluation précoce de la sensibilité des clones dans les programmes de sélection est un enjeu majeur pour écarter les individus les plus sensibles des programmes de développement et ainsi réduire la pression de la maladie. Une des méthodes d’évaluation envisagée consiste à tester indirectement la sensibilité des clones aux effecteurs fongiques responsables de la virulence (test toxinique). Parmi tous les effecteurs potentiels du champignon identifiés in silico, seule la cassiicoline Cas1 a été purifiée et caractérisée à ce jour. Il s’agit d’une petite glycoprotéine sécrétée qui jouerait un rôle dans les phases précoces de l’infection en induisant la nécrose des tissus. Les souches porteuses du gène Cas1 sont parmi les plus agressives sur les clones d'hévéa testés. Néanmoins, certaines souches de C. cassiicola ne produisant pas de cassiicoline présentent tout de même une agressivité modérée, suggérant l'implication d'autres effecteurs dans l'établissement de la maladie CLF. Les objectifs de cette thèse sont (i) de déterminer si la sensibilité à la cassiicoline Cas1 est un critère de sélection pertinent pour identifier les clones d’hévéa les plus sensibles à la maladie CLF, et (ii) d’identifier chez l’hévéa des facteurs de sensibilité à la cassiicoline Cas1. Nous avons d'abord analysé l’inoculum naturel et montré que les souches porteuses du gène Cas1 représentent un quart de la population de C. cassiicola dans les plantations d’hévéa d’Afrique de l’Ouest, le reste étant majoritairement constitué de souches dépourvues du gène codant la cassiicoline (Type A/Cas0). Nous avons ensuite créé un mutant de délétion du gène Cas1 pour la souche de référence CCP et comparé sa virulence à celle de la souche sauvage. Nous avons ainsi pu montrer que la cassiicoline Cas1 est bien un effecteur de nécrotrophie déterminant pour la virulence de C. cassiicola chez l’hévéa puisque la souche délétée perd toute virulence sur les clones testés. Enfin, nous avons recherché chez l'hévéa des facteurs de sensibilité à la cassiicoline Cas1 à travers deux approches. La technique de "double hybride en levures" nous a permis d'identifier une trentaine de protéines candidates qui pourraient interagir physiquement avec la toxine. Une approche transcriptomique, nous a permis d’identifier les gènes d’hévéa dont l’expression est modifiée suite à l’application de la cassiicoline purifiée, en comparant un clone sensible (PB260) et un clone tolérant (RRIM600). En conclusion, ces travaux ont permis de mieux comprendre les mécanismes impliqués dans l'interaction compatible entre C. cassiicola et l'hévéa et ouvrent la voie de la sélection assistée par effecteur. / The rubber tree (Hevea brasiliensis) is the primary commercial source of natural rubber worldwide. In Asia and Africa, H. brasiliensis is affected by the Corynespora leaf fall (CLF) disease, caused by the broad-spectrum necrotrophic fungus Corynespora cassiicola. During severe attacks, massive fall of young leaves can occur in susceptible cultivars. Early evaluation of the susceptibility of rubber clones in breeding programs is required to avoid developing highly susceptible clones that would amplify the disease. An indirect phenotyping procedure envisaged consists in testing the sensitivity to the fungal toxins (or effectors) rather than the susceptibility to the fungus itself (toxin test). Among all putative effectors identified in silico, only cassiicolin Cas1 has been purified and characterized to date. This small secreted glycoprotein was for long suspected to play a role in the early phase of infection by inducing tissue necrosis. Strains carrying the Cas1 gene are the most aggressive on tested rubber clones. However, strains without cassiicolin gene (called Cas0) still show moderate aggressiveness, suggesting the existence of effectors other than cassiicolin. The objectives of this study are (i) to determine if susceptibility to cassiicolin Cas1 is a relevant selection criterion to eliminate the rubber clones most susceptible to CLF disease, and (ii) identify molecular factors involved in the sensitivity to Cas1, in rubber tree. We have thus analyzed the typology of a large set of C. cassiicola isolates collected from various rubber plantations in West Africa. Our results show that isolates carrying the cassiicolin isoform Cas1 are widely represented, but that the most represented type (A/Cas0) are isolates without cassiicolin gene. Here we show that deletion of the cassiicolin gene in the isolate CCP resulted in a total loss of virulence. This clearly demonstrated that cassiicolin is indeed a necrotrophic effector required for the virulence of isolate CCP in rubber tree. Finally, we have investigated susceptibility factors to cassiicolin Cas1 on rubber tree with two different approaches. We identified about thirty candidate proteins that could physically interact with the toxin, through the two-hybrid assay. A transcriptomic approach allowed us to identify the rubber genes differentially expressed in response to the purified cassiicolin, comparing a susceptible clone (PB260) and a tolerant clone (RRIM600). In conclusion, we think that the necrotrophic effector Cas1 can be an interesting tool for effector-based selection of tolerant clones for African plantations; however, efforts should also be placed on A/Cas0 isolates, in order to identify potential necrotrophic effector(s) responsible for their virulence. This would enlarge the potential of effector-based selection.

A ansiedade e a depressão são as condições psicopatológicas mais comuns na atualidade. As pesquisas científicas em neurociência, psicologia, psiquiatria e medicina geral procuram encontrar diferentes formas para o tratamento e prevenção. No entanto, a maioria das pesquisas não leva em conta as diferenças individuais ou história do individuo. Quase toda a investigação farmacológica é feita em populações heterogêneas de animais, conduzindo assim à \"média\" dos dados. Recentemente, no Laboratório de Neurociência Comportamental na Pontifícia Universidade Católica do Rio de Janeiro (PUC-RJ), duas novas linhagens de ratos Wistar foram selecionadas fenotipicamente da resposta emocional de congelamento em um teste de medo condicionado . Uma das linhagens, denominada Carioca High ConditionedFreezing (CHF), apresenta uma resposta significativamente aumentada de congelamento no teste, enquanto que a outra (Carioca Low Conditioned-Freezing, CLF) mostra uma resposta de congelamento baixa. Em todos os experimentos utilizamos a linhagem controle (Random, RND) Wistar resultado de cruzamento aleatório entre ratos Wistar. Aqui apresentamos dados de diversos experimentos que avaliaram as diferenças em condições basais e vulnerabilidade ao estresse entre as duas linhagens e a linhagem RND. A concentração plasmática de corticosterona dos ratos foi comparada entre as duas linhagens e o grupo RND, primeiramente no nível basal e posteriormente após a exposição ao mesmo teste, onde foram expostos ao medo condicionado, como estímulo aversivo. Também foi avaliada a proliferação celular e sobrevivência. A imuno-histoquímica para c-fos foi conduzida com o objetivo de se determinar as diferenças de atividade neural entre as duas linhagens e o grupo RND. Um teste farmacológico com ketanserina, antagonista 5-HT2A, foi realizado e os animais foram avaliados quanto à ansiedade e locomoção, no labirinto em cruz elevado. Todos os nossos resultados mostraram diferenças estatisticamente significativas entre as linhagens selecionadas. Estes resultados sugerem que as diferentes condições basais podem ter impacto sobre a reação de comportamento exibido por cada animal. Assim, o uso de grupos selecionados é melhor para se modelar transtornos de ansiedade e procurar quais são caraterísticas que protegem aos animais que não apresentam os sintomas modelados. A partir dos nossos resultados foi evidente que os perfis comportamentais e neuroquímicos dos três grupos CHF, CLF e RND foram qualitativamente e quantitativamente diferentes. Na linha de base após ser fenotipados foi observado que existem diferenças na proliferação celular. Após a reexposição ao contexto observaram-se diferenças na concentração de corticosterona e na ativação neuronal. O estresse crônico levou a respostas diferenciais no número de neurônios imaturos, e o teste farmacológico levou a respostas diferencias tanto a nível local como sistêmico. Os dados obtidos podem ajudar a um melhor entendimento às bases neurobiológicas de aspectos associados às psicopatologias como ansiedade e depressão que estão relacionadas com a vulnerabilidade ao estresse. / Anxiety and depression are the most common psychopathological conditions today. Scientific research in neuroscience, psychology, psychiatry and general medicine seeks to find different ways for treatment and prevention. However, most research does not take into account individual differences and history. Almost all pharmacological research is done in heterogeneous populations of animals, thereby leading to the \"average\" data. Recently, in the Laboratory of Behavioral Neuroscience at the Pontifical Catholic University of Rio de Janeiro (PUC - RJ), two new strains of rats were selected phenotypically by their emotional response (freezing) in a test of cond itioned fear. One of the lines, so called \"Carioca High Conditioned - Freezing\" (CHF) has an increased freezing in the test , whereas the other (\"Carioca Low -Conditioned Freezing \" CLF) shows low freezing response . In all experiments we used the Wistar control strain (RND) derived from random bred between Wistar rats. Here we present data from several experiments evaluating the differences in basal conditions and vulnerability to stress between the two strains and RND strain. The plasma corticosterone of rats was compared between the two lines and the RND group, first at baseline and then after exposure to the same test where the y were exposed to conditioned fear as aversive stimulus. We also analyzed the cell proliferation and survival in the hippocampus after chronic stress. Immunohistochemistry for c -fos was conducted in order to determine the differences in neural activity between the two strains and the RND group. A pharmacological test with ketanserin, 5 -HT2A antagonist, was performed and the animals were assessed for anxiety and locomotion in the elevated plus maze. All our results showed statistically significant differences between the selected strains. These results suggest that different baseline conditions may have an impact on the behavior exhibited by each animal. Thus the use of selected groups is better approach in order to model anxiety disorders and to determine which features prevent animal from presenting symptoms. It is evident that the neurochemical and behavioral profiles of the three groups CHF, CLF and RND were different qualitatively and quantitatively. At baseline after been phenotyped it has been observed that there are differences in cell proliferation and corticosterone concentration. After re-exposure to the context differences were observed in the concentration of corticosterone and neuronal activation. Chronic stress has led to differential responses in the number of immature neurons. And the pharmacological test led to differential responses at both local and systemic. The data obtained are useful for a better understanding of the neurobiological aspects associated with psychopathologies such as anxiety and depression that are associated with vulnerability to stress.

Grâce à l'optimisation de l'effet TWIP (Twinning Induced Plasticity) et à la maîtrise de la microstructure (basse énergie de défaut d'empilement, austénite stable à grains fins), les aciers austénitiques Fe-Mn-C combinent une excellente ductilité à une résistance mécanique élevée. L'étude s'intéresse aux paramètres mécaniques critiques menant à la rupture ductile en mode slant (biseau dans l'épaisseur) et aux aspects physiques de l'endommagement dans le but de déterminer un critère de rupture quantitatif. Le comportement plastique de l'acier TWIP présente un caractère anisotrope et cinématique avec une déformation hétérogène en traction uniaxiale due à un mécanisme de pseudo-vieillissement dynamique. Après avoir déterminé la loi de comportement, la courbe limite de formage a pu être établie par comparaison entre résultats expérimentaux et prédictions du modèle mécanique, dans une large gamme de triaxialité des contraintes. Quel que soit le mode de déformation considéré, la rupture s'effectue brutalement en mode slant sans striction localisée. Peu d'endommagement, y compris en trois dimensions, a été observé autour des zones de rupture. Un critère de rupture phénoménologique basé sur la contrainte équivalente et sur l'angle de Lode, redéfinis tous deux dans l'équivalent de Barlat, permet de prédire de façon correcte la formabilité de l'acier TWIP. L'influence de la pression hydrostatique et du mécanisme de pseudo vieillissement dynamique sur la courbe limite de formage est discutée.

<p>The subject of this thesis is the design of a control law for a magnetic levitation system, which in this case is the system 33-210. The method used is backstepping technique and specifically adaptive observer backstepping due to parameter uncertainties and lack of access to all the states of the system. The second state of the system, the speed of the steel ball, was estimated by a reduced order observer. The model used gave us the opportunity to estimate a parameter which in the literature is denoted virtual control coefficient. Backstepping method gives us a rather straight forward way to design the controlling unit for a system with these properties. Stabilization of the closed-loop system is achieved by incorporating a Lypapunov function, which were chose a quadratic one in this thesis. If thederivative of this function is rendered negative definite by the control law, then we achieve stability. The results of the design were evaluated in simulations and real-time measurements by testing the tracking performance of the system. The simulation results were very promising and the validations in real-time were satisfying. Note that this has been done in previous studies; the new aspect here is the limitation of the voltage input. The real-time results showed that the parameter estimation converges only locally.</p>

The subject of this thesis is the design of a control law for a magnetic levitation system, which in this case is the system 33-210. The method used is backstepping technique and specifically adaptive observer backstepping due to parameter uncertainties and lack of access to all the states of the system. The second state of the system, the speed of the steel ball, was estimated by a reduced order observer. The model used gave us the opportunity to estimate a parameter which in the literature is denoted virtual control coefficient. Backstepping method gives us a rather straight forward way to design the controlling unit for a system with these properties. Stabilization of the closed-loop system is achieved by incorporating a Lypapunov function, which were chose a quadratic one in this thesis. If thederivative of this function is rendered negative definite by the control law, then we achieve stability. The results of the design were evaluated in simulations and real-time measurements by testing the tracking performance of the system. The simulation results were very promising and the validations in real-time were satisfying. Note that this has been done in previous studies; the new aspect here is the limitation of the voltage input. The real-time results showed that the parameter estimation converges only locally.

Submitted by Marcos Samuel (msamjunior@gmail.com) on 2017-02-06T15:18:32Z No. of bitstreams: 1 Dissertação_Natalia_versaofinal_CD.pdf: 2740629 bytes, checksum: 2ad0d4cbbf049e9e14424691d159d518 (MD5) / Approved for entry into archive by Vanessa Reis (vanessa.jamile@ufba.br) on 2017-02-06T16:10:18Z (GMT) No. of bitstreams: 1 Dissertação_Natalia_versaofinal_CD.pdf: 2740629 bytes, checksum: 2ad0d4cbbf049e9e14424691d159d518 (MD5) / Made available in DSpace on 2017-02-06T16:10:18Z (GMT). No. of bitstreams: 1 Dissertação_Natalia_versaofinal_CD.pdf: 2740629 bytes, checksum: 2ad0d4cbbf049e9e14424691d159d518 (MD5) / A Unidade de Bombeio Mecânico (UBM) é um dos equipamentos mais utilizados na produção de petróleo. Ele sofre com efeitos de desbalanceamento devido às características das cargas aplicadas e aos desgastes mecânicos causados ao longo do tempo por conta do esforço repetitivo que exerce. O desbalanceamento dos equipamentos é um dos principais efeitos usualmente estudados pela academia e indústria em função do grande impacto que provocam na durabilidade dos equipamentos. Este trabalho propõe um sistema capaz de balancear dinamicamente uma UBM através de algoritmos embarcados no controlador lógico programável do inversor de frequência, utilizado no controle de UBMs automatizadas. Para a detecção do desbalanceamento, duas formas indiretas de avaliar esse efeito foram abordados: o Fator de Carga Cíclica (CLF – Ciclic Load Factor) e análise de vibrações, através da Transformada Discreta de Fourier, ambos utilizando a variável torque. O CLF foi o critério escolhido para validação experimental do sistema. Em uma abordagem convencional, o CLF é calculado a cada ciclo de bombeio, mas tal estratégia não permite a atuação para o controle instantâneo da velocidade, que deve ser baseado na minimização do CLF. Para tanto, este trabalho propõe a aplicação de conceito baseado em janela deslizante, permitindo o cálculo de CLF a cada nova amostra. Esta abordagem permitiu também superar as limitações encontradas nos sistemas de inversão de frequência necessários à implementação da solução proposta neste trabalho. A atuação no controle de velocidade do motor usado no sistema de bombeio mecânico naturalmente pode interferir a capacidade do sistema de bombeio. Uma avaliação dessa capacidade foi implementada de forma a demonstrar a aplicabilidade do método de controle de desbalanceamento. Os resultados alcançados mostram que o sistema desenvolvido é capaz de minimizar o CLF através de um controle de velocidade, essencialmente, sem perda na capacidade do sistema de bombeio. Os testes foram realizados em bancada e validados na planta disponibilizadas pelo Laboratório de Elevação Artificial da Universidade Federal da Bahia

A ansiedade e a depressão são as condições psicopatológicas mais comuns na atualidade. As pesquisas científicas em neurociência, psicologia, psiquiatria e medicina geral procuram encontrar diferentes formas para o tratamento e prevenção. No entanto, a maioria das pesquisas não leva em conta as diferenças individuais ou história do individuo. Quase toda a investigação farmacológica é feita em populações heterogêneas de animais, conduzindo assim à \"média\" dos dados. Recentemente, no Laboratório de Neurociência Comportamental na Pontifícia Universidade Católica do Rio de Janeiro (PUC-RJ), duas novas linhagens de ratos Wistar foram selecionadas fenotipicamente da resposta emocional de congelamento em um teste de medo condicionado . Uma das linhagens, denominada Carioca High ConditionedFreezing (CHF), apresenta uma resposta significativamente aumentada de congelamento no teste, enquanto que a outra (Carioca Low Conditioned-Freezing, CLF) mostra uma resposta de congelamento baixa. Em todos os experimentos utilizamos a linhagem controle (Random, RND) Wistar resultado de cruzamento aleatório entre ratos Wistar. Aqui apresentamos dados de diversos experimentos que avaliaram as diferenças em condições basais e vulnerabilidade ao estresse entre as duas linhagens e a linhagem RND. A concentração plasmática de corticosterona dos ratos foi comparada entre as duas linhagens e o grupo RND, primeiramente no nível basal e posteriormente após a exposição ao mesmo teste, onde foram expostos ao medo condicionado, como estímulo aversivo. Também foi avaliada a proliferação celular e sobrevivência. A imuno-histoquímica para c-fos foi conduzida com o objetivo de se determinar as diferenças de atividade neural entre as duas linhagens e o grupo RND. Um teste farmacológico com ketanserina, antagonista 5-HT2A, foi realizado e os animais foram avaliados quanto à ansiedade e locomoção, no labirinto em cruz elevado. Todos os nossos resultados mostraram diferenças estatisticamente significativas entre as linhagens selecionadas. Estes resultados sugerem que as diferentes condições basais podem ter impacto sobre a reação de comportamento exibido por cada animal. Assim, o uso de grupos selecionados é melhor para se modelar transtornos de ansiedade e procurar quais são caraterísticas que protegem aos animais que não apresentam os sintomas modelados. A partir dos nossos resultados foi evidente que os perfis comportamentais e neuroquímicos dos três grupos CHF, CLF e RND foram qualitativamente e quantitativamente diferentes. Na linha de base após ser fenotipados foi observado que existem diferenças na proliferação celular. Após a reexposição ao contexto observaram-se diferenças na concentração de corticosterona e na ativação neuronal. O estresse crônico levou a respostas diferenciais no número de neurônios imaturos, e o teste farmacológico levou a respostas diferencias tanto a nível local como sistêmico. Os dados obtidos podem ajudar a um melhor entendimento às bases neurobiológicas de aspectos associados às psicopatologias como ansiedade e depressão que estão relacionadas com a vulnerabilidade ao estresse. / Anxiety and depression are the most common psychopathological conditions today. Scientific research in neuroscience, psychology, psychiatry and general medicine seeks to find different ways for treatment and prevention. However, most research does not take into account individual differences and history. Almost all pharmacological research is done in heterogeneous populations of animals, thereby leading to the \"average\" data. Recently, in the Laboratory of Behavioral Neuroscience at the Pontifical Catholic University of Rio de Janeiro (PUC - RJ), two new strains of rats were selected phenotypically by their emotional response (freezing) in a test of cond itioned fear. One of the lines, so called \"Carioca High Conditioned - Freezing\" (CHF) has an increased freezing in the test , whereas the other (\"Carioca Low -Conditioned Freezing \" CLF) shows low freezing response . In all experiments we used the Wistar control strain (RND) derived from random bred between Wistar rats. Here we present data from several experiments evaluating the differences in basal conditions and vulnerability to stress between the two strains and RND strain. The plasma corticosterone of rats was compared between the two lines and the RND group, first at baseline and then after exposure to the same test where the y were exposed to conditioned fear as aversive stimulus. We also analyzed the cell proliferation and survival in the hippocampus after chronic stress. Immunohistochemistry for c -fos was conducted in order to determine the differences in neural activity between the two strains and the RND group. A pharmacological test with ketanserin, 5 -HT2A antagonist, was performed and the animals were assessed for anxiety and locomotion in the elevated plus maze. All our results showed statistically significant differences between the selected strains. These results suggest that different baseline conditions may have an impact on the behavior exhibited by each animal. Thus the use of selected groups is better approach in order to model anxiety disorders and to determine which features prevent animal from presenting symptoms. It is evident that the neurochemical and behavioral profiles of the three groups CHF, CLF and RND were different qualitatively and quantitatively. At baseline after been phenotyped it has been observed that there are differences in cell proliferation and corticosterone concentration. After re-exposure to the context differences were observed in the concentration of corticosterone and neuronal activation. Chronic stress has led to differential responses in the number of immature neurons. And the pharmacological test led to differential responses at both local and systemic. The data obtained are useful for a better understanding of the neurobiological aspects associated with psychopathologies such as anxiety and depression that are associated with vulnerability to stress.