Cette thèse s’inscrit dans une thématique de recherche prioritaire développée par le laboratoire GREAH et portant sur l’optimisation de l'efficacité énergétique des systèmes de gestion et de production de l’énergie électrique. Dans ce cadre, les performances de convertisseurs d’énergie (machines électriques pour la conversion de l'énergie électromécanique) ont un impact déterminant sur l'efficacité énergétique de la conversion et sur la qualité de gestion de l'énergie électrique en termes de rendement, de maximisation de la puissance massique, de réduction des émissions, de réduction des coûts, ce qui nécessite, par conséquent, un effort particulier de conception et de dimensionnement. L’objectif principal de cette thèse vise à développer un modèle léger permettant d’exploiter assez rapidement l’espace des solutions potentiellement optimales dans la première phase de la conception des machines synchrones à aimants permanents. Les travaux présentés dans ce mémoire de thèse portent sur la modélisation électromagnétique pour le pré-dimensionnement et la conception des machines synchrones à aimants permanents intégrées dans l’éolien de fort puissance. Dans ce manuscrit, deux approches de modélisation des machines électriques ont été proposées. La modélisation par réseau de réluctances présentée consiste à découper le domaine d’étude en un certain nombre d’éléments volumiques dont chacun est décomposé en tubes de flux. La modélisation analytique hybride proposée est basée sur un couplage fort entre un réseau de réluctances généré à partir d’un maillage du domaine d’étude et une solution formelle des équations de Maxwell dans les régions de faible perméabilité (entrefer magnétique). Les résultats obtenus à partir de ces modèles sont validés par les résultats correspondants issus de la méthode des éléments finis. Pour montrer le gain obtenu en temps de calcul, les temps d’exécutions des codes de calcul sont comparés aux temps mis par le logiciel Flux. / This subject of the thesis is part of a priority research theme developed by the GREAH laboratory on the optimization of the energy efficiency of electrical energy management and production systems. In this context, the performance of energy converters (electrical machines for the conversion of electromechanical energy) have a decisive impact on the energy efficiency of the conversion and on the quality of electrical energy management in terms of efficiency, reduction of emissions, reduction of costs, which therefore requires a special effort to design. The main objective of this thesis is to develop a light model allowing to exploit fairly quickly the space of potentially optimal solutions in the first phase of machine design. The work presented in this thesis focuses on electromagnetic modeling for the pre-design of permanent magnets synchronous machines integrated into high power wind turbines. In this manuscript, two modeling approaches of electric machines have been proposed. The presented reluctance network modeling is generated from the mesh of studied domain as the finite element method. Reluctance elements are used for the mesh. The proposed hybrid analytical modeling is based on a strong coupling between a reluctances network generated from a mesh of the study domain and analytical models based on the formal solution of Maxwell’s equations in regions of low permeability (magnetic gap). This approach can help solve the problem of air-gap modeling in MEC method, and the consideration of the local magnetic saturation in modeling approaches involving analytical technique. The results obtained from these models are validated by the corresponding results from the finite element method and very good has been observed. To indicate the gain obtained in time, the execution times of the programs are compared to the times set by the Flux software.

Os resíduos urbanos descartados de forma irregular constituem um problema sério, principalmente nas grandes cidades, causando entupimento de bueiros e drenagem com conseqüentes inundações, sujeira e transmissão de doenças tais como leptospirose e dengue além de ser um estorvo para o trânsito e acarretarem gastos para a Prefeitura. Por outro lado, reciclar este material é uma fonte de receitas e um gerador de empregos. A Geovisualização Analítica pode ser de grande auxílio para a análise desse problema complexo e para a tomada de decisão. Com essa motivação, a presente dissertação procura fornecer uma visão geral sobre o estado da arte quanto a conceitos e pesquisas em Geovisualização (GVis) e Processamento Analítico (OLAP e SOLAP). Apresenta também processos, metodologias e tecnologias que foram utilizadas no desenvolvimento de um protótipo de um sistema analítico de informações aplicado à área de gestão da coleta seletiva de resíduos sólidos recicláveis em ambiente urbano. Este protótipo, efetivamente implantado, combinou navegação e consulta para recuperar informações através de seleções espaciais e alfanuméricas. Através de exemplos foi mostrado que este tipo de solução auxilia gestores nos processos de exploração, análise e tomada de decisão. / Domestic waste illegally disposed constitute a serious problem, especially in large cities, causing clogging of drains and drainage with subsequent flooding, dirt and transmission of diseases such as leptospirosis and dengue as well as being a hindrance to traffic and would entail city hall costs. Moreover, recycle this material is a source of revenue and a generator of jobs. The Analytical Geovisualization can be of great help in the analysis of this complex problem and the decision-making. With this motivation, this paper seeks to provide an overview of the state of the art as the concepts and research in Geovisualization (GVis) and Analytical Processing (OLAP and SOLAP). It also presents processes, methodologies and technologies that were used in developing a prototype of an analytical system applied to the area of selective collection information management of recyclable solid waste in urban environments. This prototype, effectively deployed, combined navigation and query to retrieve information through spatial and alphanumeric selections. Through examples it was shown that this type of solution helps managers in operating procedures, analysis and decision making.