Resumo: A crescente demanda por alimentos no mundo, com a mesma área agrícola e com os recursos financeiros para a aquisição de fertilizantes, cada vez mais escassos é inevitável à utilização de novas tecnologias para fazer frente a essa nova realidade, pois com o uso dessas pode-se produzir mais alimentos com menos fertilizantes. Na busca por novas tecnologias que fazem o processo produtivo mais eficiente é que surge a chamada "Agricultura de Precisão" utilizada como uma alternativa para aperfeiçoar o processo de produção agrícola. As ações que conduzem à chamada agricultura de precisão exigem diversas etapas para sua implementação. Contudo, uma das mais importantes está relacionada com o tratamento localizado das culturas, para isto, tem-se a necessidade de desenvolver equipamentos capazes de executar tais tarefas. Com esse intuito estudou-se o comportamento de um protótipo que tem como objetivo dosar e aplicar no momento do plantio, três fertilizantes sólidos simultâneos, que são: Nitrogênio, Fósforo e Potássio (N, P e K) em taxas variáveis de acordo com as quantidades desejáveis da cultura em questão. Para a avaliação do protótipo foi construída uma bancada de testes onde foram acoplados os dosadores helicoidais e seus respectivos motores elétricos de corrente contínua e três tanques para armazenamento de fertilizantes. Também foram utilizados equipamentos complementares tais como: uma fonte (12vdc); três drivers de potência (MOSFET); um Desktop; uma placa de aquisição e controle e um software de Instrumentação Virtual. Os ensaios foram realizados com o software de instrumentação e controle virtual (LabVIEW®), que teve como função, controlar a rotação dos eixos dos três motores de maneira independente, baseando-se em dados de um mapa de fertilidade do solo com localização fundamentada em informações geográficas, obtendo a vazão... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: The increasing demand for food in the world with the same agricultural area and the financial resources to purchase fertilizes, more and more scarse it is inevitable to use new technologies to confront this new reality, because with the use of these can produce more food with less fertilizer. In the search for new tecnologies that make the production process more efficient arises the called "precision agriculture" used as an alternative to improve the agricultural production process. The action that lead to the called precision agriculture require several steps for its implementation. However, one of the most important is related to the localized treatment of cultures, for this, there is the need to develop equipment capable of performing such tasks. With this aim we studied the behavior of a prototype which aims to meter and apply at the planning moment three simultaneous solid fertilizers, which are: nitrogen, phosphorus and potassium (N, P and K) with variable rates according with the desired quantities of the culture in question. For the evaluation of the prototype was built a test bench where engaged the helical dosers and their respective direct current electric motors and three tanks for storage of fertilizers. Were also used additional equipment such as: a source (12vdc), threee power drivers (MOSFET), a desktop, an acquisition board and control software and virtual instrumentation. The test were performed with the software virtual instrumentation and control (LabVIEW®), with had as its objective to control the rotation of the axes of the three engines independently, based on data from a map of soil fertility based on location information geographical, obtaining the mass flow of the desired product in the exact location and quatity needed for the plant, according to the size of grid soil sampling. The values of the mass flow rate determined by the time were compared with... (Complete abstract click electronic access below) / Orientador: Edson A. Capello Souza / Coorientador: João Eduardo Guarnetti dos Santos / Banca: José Paulo Molin / Banca: Bento Rodrigues de Pontes Junior / Mestre

Cette thèse propose trois extensions de la Régression linéaire par tranches (Sliced Inverse Regression, SIR), notamment Collaborative SIR, Student SIR et Knockoff SIR.Une des faiblesses de la méthode SIR est l’impossibilité de vérifier si la Linearity Design Condition (LDC) est respectée. Il est établi que, si x suit une distribution elliptique, la condition est vraie ; dans le cas d’une composition de distributions elliptiques il n y a aucune garantie que la condition soit vérifiée globalement, pourtant, elle est respectée localement.On va donc proposer une extension sur la base de cette considération. Étant donné une variable explicative x, Collaborative SIR réalise d’abord un clustering. Pour chaque cluster, la méthode SIR est appliquée de manière indépendante.Le résultat de chaque composant contribue à créer la solution finale.Le deuxième papier, Student SIR, dérive de la nécessité de robustifier la méthode SIR.Vu que cette dernière repose sur l’estimation de la covariance et contient une étape APC, alors elle est sensible au bruit.Afin d’étendre la méthode SIR on a utilisé une stratégie fondée sur une formulation inverse du SIR, proposée par R.D. Cook.Finalement, Knockoff SIR est une extension de la méthode SIR pour la sélection des variables et la recherche d’une solution sparse, ayant son fondement dans le papier publié par R.F. Barber et E.J. Candès qui met l’accent sur le false discovery rate dans le cadre de la régression. L’idée sous-jacente à notre papier est de créer des copies de variables d’origine ayant certaines proprietés.On va montrer que la méthode SIR est robuste par rapport aux copies et on va proposer une stratégie pour utiliser les résultats dans la sélection des variables et pour générer des solutions sparse / This thesis proposes three extensions of Sliced Inverse Regression namely: Collaborative SIR, Student SIR and Knockoff SIR.One of the weak points of SIR is the impossibility to check if the Linearity Design Condition (LDC) holds. It is known that if X follows an elliptic distribution thecondition holds true, in case of a mixture of elliptic distributions there are no guaranties that the condition is satisfied globally, but locally holds. Starting from this consideration an extension is proposed. Given the predictor variable X, Collaborative SIR performs initially a clustering. In each cluster, SIR is applied independently. The result from each component collaborates to give the final solution.Our second contribution, Student SIR, comes from the need to robustify SIR. Since SIR is based on the estimation of the covariance, and contains a PCA step, it is indeed sensitive to noise. To extend SIR, an approach based on a inverse formulation of SIR proposed by R.D. Cook has been used.Finally Knockoff SIR is an extension of SIR to perform variable selection and give sparse solution that has its foundations in a recently published paper by R. F. Barber and E. J. Candès that focuses on the false discovery rate in the regression framework. The underlying idea of this paper is to construct copies of the original variables that have some properties. It is shown that SIR is robust to this copies and a strategy is proposed to use this result for variable selection and to generate sparse solutions.

Esta pesquisa analisa o comportamento variável das oclusivas dentais /t/ e /d/ seguidas de [i] na fala de informantes das comunidades linguísticas de Santa Vitória do Palmar e Chuí, de acordo com os parâmetros da Teoria da Variação Linguística de Labov (1972). A amostra é proveniente do corpus pertencente ao projeto “Coleta de corpus nos municípios de Santa Vitória do Palmar e Chuí”, da Universidade Federal do Rio Grande – FURG. Os dados coletados foram submetidos a tratamento estatístico com o uso do programa de análise multivariada Goldvarb, nas versões 2001 e 2005. Este estudo tornou possível a conclusão de que a ocorrência da Africada Alveolar, variante bastante produtiva nos referidos municípios e frequente na fala das mulheres mais velhas e de baixa escolaridade, é uma regra variável, sujeita a condicionamentos linguísticos e extralinguísticos. / This research analyzes the variable behavior of dental stops /t/ e /d/ followed by [i] in the speech of informants of the language communities of Santa Vitória do Palmar and Chuí, according to the parameters of the Theory of Linguistic Variation (Labov, 1972). The sample is from the corpus belonging to the project “Coleta de corpus nos municípios de Santa Vitória do Palmar e Chuí”, from Universidade Federal do Rio Grande – FURG. The data collected were analyzed statistically with the use of the multivariate analysis program Goldvarb, versions 2001 and 2005. This study made possible the conclusion that the occurrence of alveolar affricate, very productive variant in these cities and often in the speech of older women with low education, is a variable rule, subject to linguistic and social conditioning.

Ce travail de thèse s’inscrit dans le projet logistique de l’Assistance Publique Hôpitaux de Marseille. En effet l’AP-HM a ouvert une plateforme logistique en avril 2013 afin de centraliser les activités de production de repas, de stérilisation, de stockage de produits hôteliers et de blanchiment de linge. Ces produits finis sont ensuite transportés dans des contenants, grâce à une équipe de transport, vers les quatre centres hospitaliers de Marseille. Le but de cette étude de recherche est de proposer une méthode et un outil permettant d’aider l’équipe de régulation des transports pour la gestion des ressources de transport. Cette étude prend en compte la variabilité des temps de trans- port ainsi que les aléas pouvant intervenir dans le cycle de vie d’une tournée de transport.Pour cela nous commençons par réaliser un modèle de connaissance du système logistique existant grâce à la méthodologie ASCI. Nous proposons ensuite une méthode et un outil permettant la génération de planning de tournées quotidien. Cette méthode est une solution ad-hoc qui intègre la résolution d’un problème de charge- ment, la planification de véhicules et d’équipages, ainsi qu’une représentation et une mo- délisation statistique de la variabilité des temps de transports en milieu urbain. En effet, le taux de congestion quotidien peut faire varier un temps de transport du simple au double. Enfin pour la gestion des aléas, nous proposons une méthode de réparation de planning que nous modélisons grâce aux systèmes multi agents. Ce dernier point de ce travail de thèse permet, en fonction des scénarios de défaillance, de proposer la solution la plus adaptée aux équipes de transport. / This thesis work is part of the logistics project of the Assistance Publique-Hôpitaux de Marseille. Indeed the AP-HM opened a logistics platform in April 2013 in order to central- ize production activities of meals, sterilization, storage product and bleaching of linen. These products are then transported in containers, thanks to a team of transport, to the four hospitals in Marseille. After consumption of the products, by healthcare units, used containers must be re- ported to the logistics platform that they are disinfected and reinstated in the production loop. The purpose of this research study is to propose a method and a tool to help the team of regulation of transport for the management of transport resources. This study takes into account the variability of the transport time and the hazards that could inter- vene in the life cycle of a tour of transport.For this we make a knowledge model of logistics system using the ASCI methodology. This model of knowledge is then validated with a simulation model. We offer then a method and a tool allowing the generation of daily tour schedule. This method is an ad - hoc solu- tion that integrates solving a problem loading, planning for vehicle and crew, as well as representation and statistical modelling of variability in the time of transport in urban areas. Indeed, the daily congestion rate can vary a transport time of one to two. Finally, for the management of the ups and downs, we propose a method of repair of planning that we model with multi agent systems. This last point of this thesis according to failure sce- narios, makes it possible to propose the best solution to the transport staff.

Construction projects in developing countries may suffer from time overruns, which are associated with cost overruns. This research project investigated both time and cost overruns on high-rise building projects in Dhaka, Bangladesh. Surprisingly, preliminary data analysis showed negligible cost overruns in comparison to time overruns. So, further analysis o f cost overruns was not considered in this thesis. This research project also investigated how the causes o f time-overruns can be mitigated. 72 time-overrun and 22 cost-overrun variables were identified through a literature review. These variables were taken as parameters and a personal interview survey was conducted with developers, consultants, contractors and project managers using semistructured questionnaire. A similar second survey was conducted using 22 measures, which can mitigate time- overruns. Data analysis involved the relative importance index to rank the variables, factors analysis to reduce variables to factors with minimum loss of data, stepwise regression to find links among factors in successive stages of construction process and multiple regression to explain delays in terms of factors. The main causes o f time-overruns were ‘cash flow’, ‘planning and scheduling deficiency’ and ‘design changes’. A scree graph identified 31 important variables that caused delays but factor analysis reduced these to 14 factors. Stepwise regression found no strong links among the factors to identify them as reasons for delay in successive stages of the construction project. A multiple regression model explained about 85% of the variance of the delays using eight factors. The main individual measures mitigating time-overruns were ‘improvement of cash flow’, ‘improvement o f communication and coordination among project participants’ and ‘development o f robust planning and scheduling instruments’. Factor analysis produced ten representative factors. Stepwise regression could not find strong links among factors mitigating time-overruns in successive stages of the construction project.

In modern manufacturing, industrial robots are essential components that allow saving cost, increase quality and productivity for instance. To achieve such goals, high accuracy and speed are simultaneously required. The design of control laws compliant with such requirements demands high-fidelity mathematical models of those robots. For this purpose, dynamic models are built from experimental data. The main objective of this thesis is thus to provide robotic engineers with automatic tools for identifying dynamic models of industrial robot arms. To achieve this aim, a comparative analysis of the existing methods dealing with robot identification is made. That allows discerning the advantages and the limitations of each method. From those observations, contributions are presented on three axes. First, the study focuses on the estimation of the joint velocities and accelerations from the measured position, which is required for the model construction. The usual method is based on a home-made prefiltering process that needs a reliable knowledge of the system’s bandwidths, whereas the system is still unknown. To overcome this dilemma, we propose a method able to estimate the joint derivatives automatically, without any setting from the user. The second axis is dedicated to the identification of the controller. For the vast majority of the method its knowledge is indeed required. Unfortunately, for copyright reasons, that is not always available to the user. To deal with this issue, two methods are suggested. Their basic philosophy is to identify the control law in a first step before identifying the dynamic model of the robot in a second one. The first method consists in identifying the control law in a parametric way, whereas the second one relies on a non-parametric identification. Finally, the third axis deals with the home-made setting of the decimate filter. The identification of the noise filter is introduced similarly to methods developed in the system identification community. This allows estimating automatically the dynamic parameters with low covariance and it brings some information about the noise circulation through the closed-loop system. All the proposed methodologies are validated on an industrial robot with 6 degrees of freedom. Perspectives are outlined for future developments on robotic systems identification and other complex problems.

Correlated ordinal data typically arises from multiple measurements on a collection of subjects. Motivated by an application in credit risk, where multiple credit rating agencies assess the creditworthiness of a firm on an ordinal scale, we consider multivariate ordinal regression models with a latent variable specification and correlated error terms. Two different link functions are employed, by assuming a multivariate normal and a multivariate logistic distribution for the latent variables underlying the ordinal outcomes. Composite likelihood methods, more specifically the pairwise and tripletwise likelihood approach, are applied for estimating the model parameters. Using simulated data sets with varying number of subjects, we investigate the performance of the pairwise likelihood estimates and find them to be robust for both link functions and reasonable sample size. The empirical application consists of an analysis of corporate credit ratings from the big three credit rating agencies (Standard & Poor's, Moody's and Fitch). Firm-level and stock price data for publicly traded US firms as well as an unbalanced panel of issuer credit ratings are collected and analyzed to illustrate the proposed framework.

This thesis presents a novel compliant mechanism with adjustable force output. The force comes from the bending of a rectangular cross section beam within the mechanism. By rotating this beam with a stepper motor, the force output of the mechanism changes. A model was made to simulate this mechanism, and a prototype was made based off of this data. A test apparatus was constructed around this mechanism, and a series of tests were performed. These tests adjusted parameters such as beam rotation speed and weight in order to characterize the system. Adjustments were made based on this information and the mechanism was refined. The results suggest the following. The speed has a negligible effect on the behavior of the system, while the weight, length of top link r3, and position of bottom stop have a significant effect. Also, there is a large, consistent amount of hysteresis in the system. This is likely caused by the beam storing torsion or friction from the slider.

Les cyanobactéries sont des micro-organismes qui préoccupent les autorités de santé publique dans le monde entier, en raison de la toxicité des cyanotoxines qu'elles produisent. Certaines cyanotoxines dont les microcystines (MC) sont des hépatotoxines inhibitrices de protéines phosphatases à sérine/thréonine. Aujourd'hui, plus de 200 variants de MCs ont été identifiés. Il s'agit d'heptapeptides monocycliques synthétisés par voie non-ribosomale dont la MC-LR (cyclo- (D-Ala-L-Leu-D-érythro-β-méthylAsp-L-Arg-ADDA-D-Glu-N-méthyl-hydro-Ala) est le variant le plus étudié en raison de sa fréquence et de sa forte toxicité. L’objectif de cette étude est le développement d'une méthode d'immunoanalyse rapide, sensible et fiable pour détecter les MCs. Le projet vise donc à développer un outil alternatif de détection de la MC-LR, qui serait mieux adapté aux analyses sur le terrain que les méthodes analytiques, biologiques ou les méthodes d'inhibition d'activité enzymatique actuellement disponibles. L'originalité de ce projet réside dans l'utilisation de deux approches différentes pour sélectionner de nouveaux anticorps spécifiques de la MC-LR. La première repose sur l'immunisation d'animaux de laboratoire, la technologie d'hybridation cellulaire et la sélection d'hybridomes sécréteurs d'anticorps monoclonaux. Si la méthodologie mise en œuvre a effectivement permis d'obtenir des immun-sérums spécifiques, la sélection des hybridomes d'intérêt reste à optimiser. La seconde stratégie mise en œuvre est basée sur la technologie du phage display pour sélectionner des fragments d'anticorps spécifiques de MC-LR à partir d'une banque de taille d’environ 109 phages, exprimant en surface des anticorps sous un format scFv (Shahsavarian et al., 2014). Plusieurs méthodes de criblage ont été développées et trois scFv ont été sélectionnés et étudiés, parallèlement à un quatrième scFv identifié dans une étude précédente (McElhiney et al., 2002), tous spécifiques à la MC-LR. Ces scFv ont été produits sous forme libre, soluble et leur spécificité à la MC-LR a été évaluée par ELISA et résonance plasmonique de surface. Les résultats obtenus montrent que les scFv sélectionnés sont tous capables de reconnaître la MC-LR. Néanmoins, ces résultats sont peu reproductibles et remettent en question le protocole de renaturation utilisé. Un travail de fond sur l’optimisation du protocole de renaturation s’avèrerait nécessaire pour les scFv ici sélectionnés, afin d’identifier les paramètres précis aboutissant à la perte ou au gain de leur fonctionnalité. / Cyanobacteria are ubiquitous microorganisms that present a worldwide concern to public health authorities because of the toxicity of the cyanotoxins they produce. Some cyanotoxins are hepatotoxins such as microcystins (MCs). At least 200 variants of MCs have been identified till today. In our study, we focus on MC-LR, a monocyclic heptapeptide (cyclo-(D-Ala-L-Leu-D-erythro-β-methylAsp-L-Arg-ADDA-D-Glu-N-methyldehydro-Ala), since it is the most frequently detected and one of the most toxic. In our study, we are interested in developing a fast, sensitive and reliable method to detect MCs. The project aims to develop an alternative pollution detection method that would be better suited to field measurements than the physicochemical methods currently available. The originality of this project lies in the use of two different approaches to select a panel of antibodies suitable for the development of immunodetection tests. The first one is based on the hybridoma technology for the production of monoclonal antibodies. The second one is based on phage display technique to select antibody fragments that are specific to MC-LR from a library of approximately 109 phages, expressing on the surface scFv fragments (Shahsavarian et al., 2014). Two monoclonal antibodies were selected using the first approach, and their specificity was evaluated using ELISA technique. Along with three scFvs selected from phage display approach. An additional scFv was added to this list: 3A8, selected from a previous study (McElhiney et al., 2002) and also specific to MC-LR. The scFvs were cloned into an expression vector in order to get each clone in its scFv soluble form. Then, their specificity to MC-LR was evaluated using ELISA technique and Surface plasmon resonance. The results show a potential specificity to MC-LR. Nevertheless, these results are not very reproducible and call into question the refolding protocol used. A thorough work on this protocol optimization would be necessary, in order to find the key parameters that control the loss or gain of their functionality

As grandes cidades brasileiras, como São Paulo, Rio de Janeiro e Porto Alegre, sofreram um crescimento acelerado ao longo das últimas décadas. Em conseqüência, mostrou-se necessário controlar o seu planejamento, garantindo à população condições favoráveis de desenvolvimento habitacional. Para Porto Alegre/RS, o instrumento básico de controle urbanístico é o Plano Diretor de Desenvolvimento Urbano e Ambiental (PDDUA). As definições apresentadas no PDDUA identificam, a partir do início da sua vigência, uma nova tipologia arquitetônica para as várias regiões da cidade. Neste contexto, o estudo de tipologias torna-se importante, pois, além de indicar os critérios de projeto adotados em determinada época e local, pode servir de parâmetro para os estudos de viabilidade, quando da necessidade de tomada de decisão para a implantação de uma nova edificação. Em função disto, o presente trabalho tem como objetivo identificar as características geométricas típicas dos projetos arquitetônicos de edificações residenciais multifamiliares construídos em Porto Alegre/RS a partir da vigência do PDUUA. Para cumprir com o objetivo proposto, foi realizado um levantamento documental junto à Secretaria Municipal de Obras e Viação da Prefeitura Municipal de Porto Alegre/RS (SMOV/PMPA) onde, a partir de uma população de projetos elaborados seguindo as diretrizes do PDDUA, foram realizadas análises com uma amostra de projetos estatisticamente representativa. Como na amostra houve grande concentração de projetos localizados nas Macrozonas 1 e 3, as edificações destas regiões da cidade foram detalhadamente estudadas. Os aspectos analisados procuraram englobar os principais elementos arquitetônicos dos projetos constituintes da amostra. Dentre os resultados alcançados, cabe salientar que a área total projetada, a presença de apartamento de zelador, a área útil total das unidades privativas e o número de dependências apresentam resultados diferentes segundo a localização. Além disso, os apartamentos de 3 dormitórios possuem, de maneira geral, valores de área superiores aos apartamentos de 2 dormitórios. Confirmam-se, assim, decisões de projeto particulares segundo a região. / Brazilian big cities, like São Paulo, Rio de Janeiro and Porto Alegre, have suffered accelerated growth throughout the last decades. Due to this growth, it has been necessary to control city planning, in order to guarantee favorable conditions of housing development for the population. In Porto Alegre/RS, the basic instrument of urban building control is the Management Plan of Urban and Environmental Development (PDDUA). The definitions presented in the PDDUA identify, from the beginning of its validity, a new architectural standard for the various regions of the city. In this context, the study of the standards has become important, for, besides indicating the project criteria to be adopted in determined time and place, they can serve as parameter for viable studies, when making planning decisions for the implementation of a new construction. For this reason, the present work has, as objective, to identify the typically geometric characteristics found in the architectural projects of multifamily residential constructions built in Porto Alegre /RS from the validity date of the PDUUA. To fulfill the considered objective, a documentary survey was carried out along with the City Department of Work and Means of Transportation of the Municipal City Hall of Porto Alegre /RS (SMOV/PMPA) where a statistically valid sample was selected from a population of elaborated projects which follow the norms of the PDDUA. As in the sample there was great concentration of projects located in Macrozones 1 and 3, the projects of these areas of the city were studied in detail. The analyzed aspects tried to include the main architectural elements of the constituent projects of the sample. Amongst the results reached, the projected total area, the presence of keeper's apartment, the useful total area of the apartments and the number of dependences present different results according to the location. Besides, apartments with 3 bedrooms possess, in a general way, higher area values than apartments with 2 bedrooms. Thus, it confirms different decisions according to the region.