This thesis presents advanced torque control methods for torque ripple reduction and performance improvement in switched reluctance motor (SRM) drives. A new offline torque sharing function (TSF) method is proposed for torque ripple reduction in SRMs. The proposed TSF achieves lower current tracking error by establishing a new current reference generation strategy. The phase current reference is first derived from the torque command using offline calculations and also from the phase current response that is obtained from the dynamic model of the SRM. Then, an optimization problem is formulated to shape the current reference for the objective of minimizing the torque ripple and copper losses, while maintaining the required average output torque at the given operating speed. The dynamic simulation of the SRM model is also utilized in the optimization problem. A new online TSF method is proposed for torque ripple reduction in SRMs. The proposed TSF takes the current dynamics and induced electromotive force into account by establishing a new online current profile generation technique. First, a primary phase current reference derived from the torque reference is applied to the SRM. Then, the decaying phase current after the turn-off angle is sampled, and it is used to update the current reference. A new online optimization strategy is performed to shape the current reference during the operation of the machine. Owing to the proposed current profile generation technique, the optimization process is decoupled to independently minimize the torque ripple by optimizing the turn-on angle and minimizing copper losses by optimizing the turn-off angle. Compared to the conventional TSFs and existing optimization-based TSFs, the proposed two TSFs achieve accurate torque control, improved torque-speed capability, reduced torque ripple, and better current tracking performance. All the proposed TSF methods are validated by both simulations and experiments on a 3-phase, 12/8 SRM. / Dissertation / Doctor of Philosophy (PhD)

Ce travail est une contribution à l’économie de la famille visant à analyser l’importance des interactions familiales sur le comportement des agents économiques. En effet, ces interactions peuvent influencer le pouvoir de négociation des ménages, leurs consommations des biens, ou encore leur offre de travail. Nous nous proposons donc au cours de ce document une évaluation de l’importance de ces interactions dans la prise de décision des ménages. Le modèle de base qui sera utilisé dans les différents résultats est celui de Browing et Chiappori (1998) et généralisé par Chiappori et Ekeland (2006). Ce travail comporte éventuellement trois chapitres dont l’objectif est unique ; il s’agit d’analyser l’effet des interactions familiales sur la prise de décision des ménages : La première partie de la thèse a pour objectif de modéliser l’effet de l'aversion au risque des conjoints sur leur choix marital. Les fondements théoriques de ce travail se trouvent dans une généralisation de l'article de Arrondel et Calvo (2011). Dans la deuxième partie de la thèse, on estime l'effet du statut marital, mariage vs concubinage, sur les dépenses des enfants et ce en utilisant les données du Family Expenditure Survey, FES, pour la période 1995-2006. Dans la troisième partie de la thèse, nous illustrons la relation matrimoine consommation par une application empirique plus structurel en utilisant toujours des données Britanniques mais pour une plus longue période (1988-2007). Plus particulièrement, nous nous intéressons à l'estimation du coût individuel des enfants. L’approche adoptée dans cette partie est étroitement liée aux développements récents dans la littérature des modèles collectifs( Browning, Chiappori et Lewbel, 2008 et Lewbel et Pendakur, 2008). / This work is a contribution to the economy of the family to analyze the importance of the family interactions on the behavior of the economic agents. So we propose during this document an evaluation of the importance of these interactions in the decision-making of the households. The basic model which will be used in the various results is the one of Browing and Chiappori ( 1998 ) and generalized by Chiappori and Ekeland (2006 ). This work contains possibly three chapters the objective of which is unique; it is a question of analyzing the effect of the family interactions on the decision-making of the households: The first part of the thesis has for objective to model the effect of parental's risk aversion on their marital choice. The theoretical foundations of this work are in a generalization of the article of Arrondel and Calvo(2011). In the second part of the thesis, we estimate the effect of the marital status, marriage vs cohabitation, on children's well-being by using the data of the Family Expenditure Survey, FES, for period 1995-2006. In the third part of the thesis, we illustrate the relation “matrimoine consumption” by a more structural empirical application using the same British data but for longer period ( 1988-2007 ). More particularly, we are interested to estimate individual children’s cost. The proposed model is inspired by recent collective model of the behavior of the household developed by Browning, Chiappori et Lewbel (2008) and Lewbel et Pendakur (2008).