Το θέμα της παρούσας διπλωματικής εργασίας είναι η μελέτη της σύγχρονης μηχανής αξονικής ροής με μόνιμους μαγνήτες, που προορίζεται να λειτουργήσει σε ανεμογεννήτρια, μέσω της δημιουργίας του αντίστοιχου τρισδιάστατου μοντέλου πεπερασμένων στοιχείων. Σκοπός της εργασίας είναι η σχεδίαση και η εξομοίωση του μοντέλου στις τρεις διαστάσεις μέσω του λογισμικού της OPERA 3D. Τα δεδομένα, προκειμένου να επιτευχθεί η σχεδίαση της μηχανής, προέκυψαν έπειτα από ακριβή διαδικασία διαστασιολόγησής της αξιοποιώντας τις θεωρητικές σχέσεις που, σύμφωνα με τη βιβλιογραφία, περιγράφουν τη λειτουργία της. Η δομή της διπλωματικής εργασίας διαρθρώνεται ως εξής. Στο πρώτο κεφάλαιο γίνεται μια σύντομη αναφορά στο θέμα της αιολικής ενέργειας σε σχέση με την αξιοποίησή της τόσο σε εγχώριο όσο και σε παγκόσμιο επίπεδο, καθώς επίσης και μια παράθεση στοιχείων που αφορούν τα χαρακτηριστικά και τον τρόπο λειτουργίας των ανεμογεννητριών. Στο δεύτερο κεφάλαιο παρουσιάζεται η διερεύνηση της βέλτιστης τοπολογίας σύγχρονης μηχανής αξονικής ροής που φέρει μόνιμους μαγνήτες για χρήση σε ανεμογεννήτρια. Η διερεύνηση γίνεται ως προς τη διάταξη δρομέα – στάτη, το τύλιγμα και τους μαγνήτες της μηχανής. Συνεχίζοντας, στο τρίτο κεφάλαιο παρατίθεται η θεωρητική ανάλυση της συγκεκριμένης μηχανής και η βήμα προς βήμα διαδικασία της διαστασιολόγησής της. Σε αυτό το σημείο θα πρέπει να διευκρινιστεί ότι το περιεχόμενο αυτού του κεφαλαίου είναι αποτέλεσμα συνεργασίας με τη φοιτήτρια του τμήματος Ηλεκτρολόγων Μηχανικών και Τεχνολογίας Υπολογιστών του Πανεπιστημίου Πατρών Αγγελίνα Ιωάννου, της οποίας η Διπλωματική Εργασία είναι παρόμοιου θέματος με αυτό της παρούσας Διπλωματικής Εργασίας. Στο τέταρτο κεφάλαιο γίνεται μια σύντομη θεωρητική προσέγγιση της μεθόδου των πεπερασμένων στοιχείων και μια εισαγωγή στον τρόπο λειτουργίας του λογισμικού της OPERA 3D. Στο πέμπτο κεφάλαιο δίνονται οι βασικές πληροφορίες σχετικά με τον τρόπο που σχεδιάστηκε το μοντέλο στην OPERA 3D. Τέλος, στο έκτο κεφάλαιο παρουσιάζονται τα αποτελέσματα που προέκυψαν από τις εξομοιώσεις του μοντέλου στην OPERA 3D οι οποίες περιλαμβάνουν: • Στατική γραμμική ανάλυση της τοπολογίας διπλού δρομέα – μονού στάτη • Στατική μη γραμμική ανάλυση της τοπολογίας διπλού δρομέα – μονού στάτη • Στατική γραμμική ανάλυση της τοπολογίας μονού δρομέα – μονού στάτη • Στρεφόμενη γραμμική ανάλυση της τοπολογίας διπλού δρομέα – μονού στάτη εν κενώ • Στρεφόμενη γραμμική ανάλυση της τοπολογίας διπλού δρομέα – μονού στάτη για διάφορα φορτία / The subject of this thesis is the research of a permanent magnet axial flux synchronous machine, which is intended to operate in a wind turbine, through finite element analysis. The purpose of this thesis is the design and the simulation of the model in three dimensions using the software OPERA 3D. The data we needed, in order to achieve the design of the machine, were results of a detailed process of sizing with the use of all the theoretical relationships that, according to the literature, describe machine’s operation. This thesis is structured as follows. The first chapter is a brief reference to the issue of wind energy in relation to its use both in Greece and abroad, as well as a quote data concerning the characteristics and operation of wind turbines. The second chapter presents the investigation of the optimal topology of a permanent magnet axial flux synchronous machine used in wind turbines. The investigation includes the rotor – stator order, the winding and the magnets of the machine. Continuing, the third chapter presents the theoretical analysis of this machine and the step by step process of sizing. It should be noted that the content of this chapter is the result of cooperation with the student of the Department of Electrical and Computer Engineering in the University of Patras, Aggelina Ioannou whose thesis subject is similar to that of this thesis. The fourth chapter is a brief theoretical approach to the finite element method and an introduction to the way the software OPERA 3D operates. The fifth chapter provides basic information about how the model was designed to OPERA 3D. Finally, the sixth chapter presents the results obtained from simulations of the model in OPERA 3D which include: • Static linear analysis of the double rotor – single stator topology • Static non linear analysis of the double rotor – single stator topology • Static linear analysis of the single rotor – single stator topology • RΜ linear no load analysis of the double rotor – single stator topology • RΜ linear load analysis of the double rotor – single stator topology

Les aimants NdFeB sont les aimants les plus puissants que l’on peut trouver actuellement sur le marché. Leurs propriétés exceptionnelles proviennent des propriétés intrinsèques de la phase Nd2Fe14B mais également de la microstructure imposée par le mode d’élaboration. Ces aimants sont généralement réalisés par frittage avec phase liquide à partir d’une poudre monocristalline préalablement orientée sous champ magnétique. Cette voie permet d’obtenir une microstructure à grains fin découplés magnétiquement les uns des autres par une fine couche de phase riche en néodyme, dont la distribution est optimisée après un recuit à une température donnée. Pour une utilisation dans des moteurs et générateur électriques, il est courant d’ajouter du dysprosium pour conserver une coercitivité suffisante à la température de fonctionnement. Mais ce gain de coercitivité se fait au détriment de l’induction rémanente et du coût de l’aimant de par la rareté du dysprosium. Il est donc nécessaire de pouvoir s’affranchir de son utilisation par une meilleure compréhension des liens entre le procédé d’élaboration et la microstructure du matériau, afin d’optimiser les propriétés magnétiques finales.Cette thèse s’intéresse tout d’abord au frittage d’aimants NdFeB pour une nuance du commerce et donc fortement alliée. La forte anisotropie de retrait lors de la densification n’est pas clairement expliquée dans la littérature et son interprétation peut apporter un éclairage sur les propriétés magnétiques. Le frittage a ainsi été étudié à l’aide d’essais de dilatométrie dans la direction d’orientation de la poudre ainsi que dans la direction transverse. Le frittage a également été interrompu à différents moments et la microstructure analysée. Des analyses d’images ont permis de comprendre, à l’aide d’un modèle analytique du frittage, qu’une partie de l’anisotropie de retrait s’expliquait par une distribution anisotrope des surfaces de contact entre les grains de poudre, issue de l’étape d’orientation sous champ. Cette analyse a été confirmée par des simulations par éléments discrets.Le deuxième volet de la thèse s’intéresse au rôle des éléments d’alliage les plus couramment rencontrés dans les aimants hors dysprosium. Le rôle de l’aluminium, du cobalt et du cuivre sur les propriétés magnétiques a été examiné en étudiant une vingtaine de compositions simplifiées avec des quantités d’éléments d’alliage voisines des valeurs que l’on retrouve dans les aimants du commerce. Les différents échantillons ont été recuits à des températures déduites d’essais DSC. Les résultats montrent que ces trois éléments d’alliage agissent de manière croisée sur la coercitivité. Au-delà de la mesure de coercitivité, l’évolution de la forme de la courbe de désaimantation en fonction de la composition et de la température de recuit apporte des informations importantes sur l’action des éléments d’alliage sur la microstructure. / NdFeB hard magnets are the most powerful magnets commercially available. Their outstanding properties originate from Nd2Fe14B intrinsic properties and from the microstructure imposed by the manufacturing process. These magnets are generally obtained by liquid phase sintering of an oriented monocrystalline powder which enables a microstructure made of grain magnetically decoupled by a thin neodymium-rich layer which is optimally distributed in the magnet after a low temperature annealing. For them to be used in electrical engines and generators, dysprosium is usually added so that the coercivity is high enough at the working temperature. But dysprosium is rare and expensive and lowers the remnant induction. It is therefore important to get free of its usage by a better understanding of the links between the manufacturing process and the microstructure, so that the final magnetic properties can be optimized.First, this thesis deals with NdFeB sintering on an alloyed commercial grade. The high shrinkage anisotropy during densification is not clearly explained and its interpretation could bring information on the magnetic properties. Dilatometric studies have been performed along orientation direction as well as along the transverse direction. Sintering has been interrupted at different times and the microstructure was observed. Image analysis has enabled to understand, thanks to an analytical model, that a part of the anisotropy could be explained by an anisotropic contact orientation distribution, originating from the magnetic orientation step. Discrete element modelling has confirmed this approach.The second part of the thesis deals with the role of the most commonly used alloying elements on the magnetic properties: aluminum, cobalt and copper. Around twenty different model grades were examined with composition close to the ones of commercial magnets. After having been sintered, the samples were annealed at temperatures deduced from DSC measurements. Results show that the three elements have cross effects on coercivity. Beyond coercivity, demagnetizing curve shape is sensitive to the composition and annealing temperature and gives important information on the role of the alloying elements on the microstructure.

A teoria de duplo eixo é geralmente utilizada para modelar e caracterizar máquinas síncronas de polos salientes. O conhecimento das indutâncias de eixo direto e eixo em quadratura são fundamentais para o desenvolvimento destes modelos e para o projeto de controladores destas máquinas. No entanto, métodos tradicionais para determinação destas indutâncias através de ensaios tornam-se impraticáveis em máquinas síncronas de ímãs permanentes pela impossibilidade de controlar a corrente de campo. Recentemente, a IEEE publicou o Std. 1812, chamado Trial-Use Guide for Testing Permanent Magnet Machines, que pode servir de guia para obtenção da reatância de eixo direto Xd. Esta norma não sugere formas de determinar a reatância de eixo em quadratura Xq, mas existem diferentes métodos para calcular este parâmetro a partir de dados coletados em testes, geralmente envolvendo medição de ângulo de carga. O objetivo deste trabalho é propor e analisar dois métodos alternativos para determinar as reatâncias do modelo de duplo eixo através de ensaios aplicados a um gerador síncrono de ímãs permanentes (GSIP) de 1,5 kW. O primeiro é baseado em predição de ângulo de carga e o segundo é baseado em medição de ângulo de carga. O primeiro combina ensaios a vazio e de curto circuito com análise do diagrama fasorial para descrever o ângulo de carga como função de grandezas elétricas facilmente mensuráveis. O segundo propõe um método de medição de ângulo de carga com auxílio de um encoder incremental. Resultados comprovam que é possível determinar indutâncias do modelo de duplo eixo com as equações desenvolvidas desde que o conteúdo harmônico não seja elevado. / A teoria de duplo eixo é geralmente utilizada para modelar e caracterizar máquinas síncronas de polos salientes. O conhecimento das indutâncias de eixo direto e eixo em quadratura são fundamentais para o desenvolvimento destes modelos e para o projeto de controladores destas máquinas. No entanto, métodos tradicionais para determinação destas indutâncias através de ensaios tornam-se impraticáveis em máquinas síncronas de ímãs permanentes pela impossibilidade de controlar a corrente de campo. Recentemente, a IEEE publicou o Std. 1812, chamado Trial-Use Guide for Testing Permanent Magnet Machines, que pode servir de guia para obtenção da reatância de eixo direto Xd. Esta norma não sugere formas de determinar a reatância de eixo em quadratura Xq, mas existem diferentes métodos para calcular este parâmetro a partir de dados coletados em testes, geralmente envolvendo medição de ângulo de carga. O objetivo deste trabalho é propor e analisar dois métodos alternativos para determinar as reatâncias do modelo de duplo eixo através de ensaios aplicados a um gerador síncrono de ímãs permanentes (GSIP) de 1,5 kW. O primeiro é baseado em predição de ângulo de carga e o segundo é baseado em medição de ângulo de carga. O primeiro combina ensaios a vazio e de curto circuito com análise do diagrama fasorial para descrever o ângulo de carga como função de grandezas elétricas facilmente mensuráveis. O segundo propõe um método de medição de ângulo de carga com auxílio de um encoder incremental. Resultados comprovam que é possível determinar indutâncias do modelo de duplo eixo com as equações desenvolvidas desde que o conteúdo harmônico não seja elevado.

学位授与大学：京都大学 ; 取得学位: 博士(工学) ; 学位授与年月日: 2007-09-25 ; 学位の種類: 新制・課程博士 ; 学位記番号: 工博第2864号 ; 請求記号: 新制/工/1421 ; 整理番号: 25549 / Kyoto University (京都大学) / 0048 / 新制・課程博士 / 博士(工学) / 甲第13393号 / 工博第2864号 / 新制||工||1421(附属図書館) / 25549 / UT51-2007-Q794 / 京都大学大学院工学研究科電気工学専攻 / (主査)教授 引原 隆士, 教授 小林 哲生, 准教授 中村 武恒 / 学位規則第4条第1項該当

Cette thèse décrit dans un premier temps la synthèse entièrement régiosélective de phtalocyanines tétra-(endo-alcoxy)-fonctionnalisées puis la formation de leurs complexes avec des ions métalliques. Dans un second temps sont étudiées leurs propriétés magnétiques, l'induction de mésophases colonnaires dans des gammes de températures convenables et avec des tailles modérées de substituants, ainsi que leur utilisation dans des diodes organiques. L'approche synthétique à suivre est la cyclo-tétramérisation de 3-(2-alkylalcoxy)-phtalonitriles suivie de la coordination d'un ion métallique. La symétrisation des chaînes aliphatiques dans le précurseur 3-alcoxy-phtalonitrile, en allant de 2 butyloctyl à 2-pentylheptyl, maintient la régiosélectivité et le mésomorphisme, tandis que les courtes chaînes 2-butylhexyl mènent à la formation d'un mélange de phtalocyanines régioisomères et à une plus grande tentance à la crystallisation. La combinaison de températures de clarification raisonnables avec un empilement colonnaire à température ambiante et avec une proportion assez importante de centres conjurés au sein de la masse moléculaire rend les deux premières séries de matériaux, c'est-à-dire MPc(OCH2CHBuHex)4, and MPc(OCH2CHPent2)4, potentiellement utiles comme transporteurs de charges uniformément orientables dans des dispositifs électroniques organiques. Nous démontrons que ces matériaux phtalocyanines tétra-α-alcoxy-substitués, représentés par H2Pc(OCH2CHBuHex)4, NiPc(OCH2CHBuHex)4 et CuPc(OCH2CHBuHex)4, mènent à des performances originales des dispositifs quand ils sont utilisés comme couche active organique dans des structures simples de diode ITO/PEDOT :PSS/PC/Al. Un redressement prononcé du courant est obtenu dans des diodes malgré l’alignement planaire prépondérant des colonnes dans les couches. Le ligand Pc tétra-α-alcoxy-substitué très soluble donne des molécules-aimants mononucléaires originales par coordination avec MnIII et DyIII. Remarquablement, l’isomère de symétrie C4h du complexe sandwich octa-alcoxy se forme sélectivement grâce aux subsistants encombrants sur les deux cycles Pc. / This thesis describes the fully regioselective synthesis of symmetric all-endo tetra-alkoxy-functionalized phthalocyanines and their metal ion complexes accompanied by induction of columnar mesophases in convenient temperature ranges at moderate substituent sizes, as well as their use in organic diodes and the study of their magnetic properties. The synthetic approach to follow is lithium-induced macrocyclization of 3-(2-alkylalkoxy)-phthalonitriles prior to transition meatl ion insertion. Symmetrization of the aliphatic chains in the 3-alkoxy-phthalonitrile precursor from 2-butylocytyl to 2-pentylheptyl maintains both the regioisomeric mixture during the cyclo-tetramerization and to a somewhat greater tendency to crystallization. The combination of attainable clearing temperatures with room temperature columnar stacking and with a relatively high content of conjugated core within the molecular mass makes the first two series of materials, i.e. MPc(OCH2CHBuHex)4, and MPc(OCH2CHPent2)4, potentially useful as uniformly orientable charge transposters in organic electronic devices. We establish that these tetra-α-alkoxy substituted phthalocyanine materials, as exemplified with H2Pc(OCH2CHBuHex)4, NiPc(OCH2CHBuHex)4 and CuPc(OCH2CHBuHex)4, lead to original device performances when applied as an active organic layer in simple ITO/PEDOT:PSS/PC/Al diode structures. A pronounced current rectification of the diodes is obtained despite the preponderantly planar alignment of the columns in the films. The highly soluble tetra-α-alkoxy-substituted Pc ligand, when combined with MnIII and DyIII, gives rise to original mononuclear single molecule magnets. Remarkably, the C4h-symmetric isomer of the octa-alkoxy double decker complex is formed selectively due of presence of the bulky substituents on both Pc rings.

A teoria de duplo eixo é geralmente utilizada para modelar e caracterizar máquinas síncronas de polos salientes. O conhecimento das indutâncias de eixo direto e eixo em quadratura são fundamentais para o desenvolvimento destes modelos e para o projeto de controladores destas máquinas. No entanto, métodos tradicionais para determinação destas indutâncias através de ensaios tornam-se impraticáveis em máquinas síncronas de ímãs permanentes pela impossibilidade de controlar a corrente de campo. Recentemente, a IEEE publicou o Std. 1812, chamado Trial-Use Guide for Testing Permanent Magnet Machines, que pode servir de guia para obtenção da reatância de eixo direto Xd. Esta norma não sugere formas de determinar a reatância de eixo em quadratura Xq, mas existem diferentes métodos para calcular este parâmetro a partir de dados coletados em testes, geralmente envolvendo medição de ângulo de carga. O objetivo deste trabalho é propor e analisar dois métodos alternativos para determinar as reatâncias do modelo de duplo eixo através de ensaios aplicados a um gerador síncrono de ímãs permanentes (GSIP) de 1,5 kW. O primeiro é baseado em predição de ângulo de carga e o segundo é baseado em medição de ângulo de carga. O primeiro combina ensaios a vazio e de curto circuito com análise do diagrama fasorial para descrever o ângulo de carga como função de grandezas elétricas facilmente mensuráveis. O segundo propõe um método de medição de ângulo de carga com auxílio de um encoder incremental. Resultados comprovam que é possível determinar indutâncias do modelo de duplo eixo com as equações desenvolvidas desde que o conteúdo harmônico não seja elevado. / A teoria de duplo eixo é geralmente utilizada para modelar e caracterizar máquinas síncronas de polos salientes. O conhecimento das indutâncias de eixo direto e eixo em quadratura são fundamentais para o desenvolvimento destes modelos e para o projeto de controladores destas máquinas. No entanto, métodos tradicionais para determinação destas indutâncias através de ensaios tornam-se impraticáveis em máquinas síncronas de ímãs permanentes pela impossibilidade de controlar a corrente de campo. Recentemente, a IEEE publicou o Std. 1812, chamado Trial-Use Guide for Testing Permanent Magnet Machines, que pode servir de guia para obtenção da reatância de eixo direto Xd. Esta norma não sugere formas de determinar a reatância de eixo em quadratura Xq, mas existem diferentes métodos para calcular este parâmetro a partir de dados coletados em testes, geralmente envolvendo medição de ângulo de carga. O objetivo deste trabalho é propor e analisar dois métodos alternativos para determinar as reatâncias do modelo de duplo eixo através de ensaios aplicados a um gerador síncrono de ímãs permanentes (GSIP) de 1,5 kW. O primeiro é baseado em predição de ângulo de carga e o segundo é baseado em medição de ângulo de carga. O primeiro combina ensaios a vazio e de curto circuito com análise do diagrama fasorial para descrever o ângulo de carga como função de grandezas elétricas facilmente mensuráveis. O segundo propõe um método de medição de ângulo de carga com auxílio de um encoder incremental. Resultados comprovam que é possível determinar indutâncias do modelo de duplo eixo com as equações desenvolvidas desde que o conteúdo harmônico não seja elevado.

Este trabalho tem por objetivo estudar e avaliar uma técnica de determinação de posição inicial de rotor em máquinas síncronas de ímã permanente de superfície. A técnica de determinação de posição inicial é baseada na aplicação de pulsos ao motor. É proposto também um algoritmo de ajuste automático desses pulsos de forma a criar um mecanismo genérico de determinação de posição inicial de ângulo de rotor para o tipo de máquina em estudo. Simulações são realizadas de forma a garantir funcionamento correto do algoritmo que foi implementado em código C com matemática de ponto fixo. A implementação foi realizada utilizando-se um microcontrolador comercial de baixo custo que não possui hardware para aritmética de ponto flutuante. Um descritivo detalhado da implementação, do sistema de testes e resultados são apresentados ao longo do texto. / This work aims the study and evaluation of a rotor initial angle estimation technique in surface mounted permanent magnet synchronous machine. The technique of the rotor initial angle estimation is based on pulse injection to the motor. It is also proposed in this work an automatic pulse adjustment algorithm in order to create a generic mechanism to estimate rotor initial angle for the machine type under study. Simulations are performed in order to ensure the proper functionality of the algorithm which was implemented in C code with fixed point mathematics. The implementation was performed using a comercial low cost microcontroller which does not have floating point arithmetic capability. A detailed description of the implementation, test system and results are presented in this text.

Orientador: Kleber Roberto Pirota / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Física Gleb Wataghin / Made available in DSpace on 2018-08-23T16:30:39Z (GMT). No. of bitstreams: 1 Almeida_PriscilaToderode_M.pdf: 4163268 bytes, checksum: 28aa8c398a21212739c3588c9ce78593 (MD5) Previous issue date: 2013 / Resumo: O principal objetivo desse trabalho consiste em uma nova abordagem para o tunelamento da magnetização do magneto molecular Mn12, embasado em uma ampliação do modelo de Preisach. Introduziremos novos operadores que levam em conta a possibilidade de efeitos quânticos. Implementamos esse novo modelo num programa de simulação que é capaz de simular curvas de histerese e curvas de relaxação magnética sem recorrer a resolução de hamiltonianas de spin. Além disso, este programa utiliza simulação estocástica, apresentando os resultados em poucos minutos. Os resultados obtidos concordam com os experimentos realizados de histerese e relaxação magnética. Apesar de ser um modelo de simulação simples, reproduz adequadamente a fenomenologia, pois introduz os dois ingredientes essenciais de um sistema com inversão da magnetização por efeito túnel termicamente ativado: a ativação térmica, descrita pela ocupação de níveis segundo a distribuição de Boltzmann e a possibilidade do efeito túnel descrita pelo modelo de Landau¿Zenner. A consistência física do modelo é estudada através da variação de parâmetros do modelo de forma sistemática / Abstract: The main objective of this work consists of a new approach concerning the tunneling of the magnetization of the molecular magnet Mn12, based on an extension of the Preisach model. We will introduce new operators that take into account the possibility of quantum effects. Thus, we have implemented this new model in a simulation software that is capable of simulating hysteresis curves and magnetic relaxation curves without utilizing resolution of spin Hamiltonians. Also, this program uses stochastic simulation, presenting the results in only a few minutes. The results obtained agree with the hysteresis and relaxation experiments. Despite being a simple simulation model, it adequately reproduces the phenomenology, because it introduces two key ingredients of a system with inversion of magnetization by thermally activated tunnel effect: the thermal activation, described by the occupation of levels according to the Boltzmann distribution and the possibility of tunnel effect described by the Landau-Zenner model. The physical consistency of the model is studied by systematically varying the model¿s parameters / Mestrado / Física / Mestra em Física

During these decades, Permanent Magnet Synchronous Motor (PMSM) has become a vital part of military, industry and civil applications due to the advantages of high power density, high efficiency, high reliability and simple structure, small volume and light weight. Sometimes, multiple PMSMs are used to carry out cooperative functions. For example, the bogie of a locomotive, the flight control surface of an airplane. These PMSMs usually operates at the same speed. To reduce the volume and weight, an idea of sharing the static power conversion devices, which is called Mono-Inverter Multi-PMSM system (MIMPMSM), is raised. Although many researchers have given different controller solutions for the MIMPMSM system, most of them are not clear in the aspects of system stability and efficiency issues. This has become the biggest obstacle to the practical use of MIMPMSM. Oriented with these problems, starting with a MIMPMSM system with 2 motors, in the first step, we have tested some control strategies by an experiment to verify the feasibility and performance of them. In final, based on the experiment data, we have figured that the overconstraint problem exists in some control strategies. Then, an analysis and controller design based on steady-state model of a Mono-Inverter Dual-PMSM (MIDPMSM) system is carried out.By studying the solution existence problem of the steady-state model, we give out the design guideline to the controller structure. Combining the open-loop stability and steady-state solution, the region of controllability and stability is obtained. Lagrange Multiplier is used develop theexpression of efficiency-optimal steady-staterelated to torque and speed. The experiment has shown that the efficiency of the new controller has improved significantly. Meanwhile, we have explored the influence of parameter variation in system stability and efficiency-optimization. The variation will influence the stability region. But its influence can be eliminated by using Master- Slave strategy. On the other hand, in the aspect of efficiency optimization, the simulation results have shown that parameter mismatch, especially the permeant flux, can cause high efficiency loss. In the last step, this controller is also adapted to a MIMPMSM system with more than two motors. The simulation results demonstrate the effectiveness.

L’étude de la chimie des actinides est essentielle dans le cadre de la technologie nucléaire pour le développement de nouveaux combustibles, pour l’étude du retraitement des déchets nucléaires et la migration des actinides dans l’environnement mais aussi pour la compréhension fondamentale des interactions actinide/ligand et la formation de liaisons multiples. Les propriétés magnétiques des molécules polymétalliques d’actinides sont particulièrement intéressantes pour explorer la communication magnétique entre différents centres métalliques. De plus, ces molécules ont été identifiées comme particulièrement prometteuses pour la conception de molécules aimants. L’uranium a une grande réactivité redox notamment due à ses multiples degrés d’oxydation accessibles et forme aisément des assemblages polynucléaires. Néanmoins, très peu de synthèses contrôlées de complexes polymétalliques d’uranium et de neptunium ont été décrites dans la littérature. La première approche de ce travail repose sur la synthèse de clusters oxo/hydroxo d’uranium à partir de l’hydrolyse contrôlée d’uranium tétravalent en présence d’un ligand organique rencontré dans l’environnement. Cette étude a mené à une famille de clusters aux topologies originales, dont la taille varie en fonction des conditions réactionnelles employées. Cependant les clusters obtenus ne mènjavascript:nouvelleZone('abstract');_ajtAbstract('abstract');ent pas à des propriétés de molécules aimants. Dans le but de favoriser une plus grande interaction entre les métaux par le ligand pontant, l’interaction cation-cation a été utilisée pour la synthèse rationnelle d’assemblages d’uranyle(V). Par le passé, peu de complexes d’uranyle(V) ont été isolés à cause de son instabilité vis-à-vis de la dismutation ; cependant, l’optimisation du ligand organique et des conditions de synthèse ont finalement permis de stabiliser l’uranyle(V). Nous avons utilisé des complexes stables d’uranyle(V) comme brique de base pour former des molécules hétéronucléaires avec des métaux 3d et 4f. Un réglage fin des conditions de réactions a mené à une conception rationnelle d’assemblages discrets ou polymériques. L’étude des propriétés magnétiques de ces assemblages d’uranium a mis en valeur des propriétés de molécules ou chaînes aimants avec de hautes valeurs d’énergie de relaxation. L’uranyle(V) a également été utilisé comme modèle structural du neptunium qui est plus radioactif permettant d’isoler un complexe isostructural homométallique de neptunyle(V) grâce à des conditions réactionnelles similaires. Finalement, des ligands nitrures favorisant la formation de liaison multiples uranium-ligand, ont été utilisés pour construire de nouveaux complexes binucléaires d’ujavascript:nouvelleZone('abstract');_ajtAbstract('abstract');ranium supportés par des ligands silanols. De nouvelles molécules, sans précédent, contenant des nitrures comme ligand pontant associés à de l’uranium au degré d’oxydation +III ont été isolées et caractérisées. / The study of actinide chemistry is not only essential for the development of nuclear fuel, nuclear fuel reprocessing or environmental clean up, but also for the understanding of fundamental actinide/ligand interactions and multiple bounding. The magnetic properties of polynuclear actinide molecules are of significant interest to investigate the magnetic communication between the metallic centres. Furthermore, they are highly promising for the design of molecular magnets. Uranium undergoes redox reactions due to a wide range of available oxidation states and easily forms polynuclear assemblies. However, only a few controlled synthetic routes towards these polynuclear uranium assemblies are described in the literature. In this context, the first part of this work was dedicated to the synthesis of oxo/hydroxo uranium clusters from the controlled hydrolysis of tetravalent uranium in the presence of an environmentally relevant ligand. This led to the synthesis of clusters with novel topologies, for which size could be varied as a function of the reaction conditions employed. However, the obtained clusters do not behave as SMM. In order to gain a stronger interaction between metallic centres, the cation-cation interaction was used to rationally design polynuclear uranyl(V) complexes. The isolation of uranyl(V) complexes had been limited in the past by its disproportionation, however, a fine tuning of the organic ligand and reaction conditions finally allowed to stabilise uranyl(V). We used stable uranyl(V) units as building block to form heteronuclear complexes with 3d and 4f metals with polymeric or discrete structures. The study of the magnetic properties of the uranium polynuclear assemblies was carried out and revealed single molecule or chain magnet behaviours with high energy barriers. The uranyl(V) unit was also used as a structural model for the more radioactive neptunium element, allowing the isolation of an isostructural trinuclear neptunyl(V) assembly in similar reaction conditions. Finally, the use of a nitride ligand as a bridging unit, allowing the formation of uranium-ligand multiple bonds, was explored to build novel di-uranium complexes supported by siloxy ligands. Nitride molecules containing unprecedented uranium in the +III oxidation state were isolated and characterised.