L’objectif principal de ce travail est l’étude de la planification de trajectoires pour des drones de type HALE ou MALE. Les modèles cinématiques de ces drones sont étudiés. Les drones HALE sont modélisés par le système de Dubins. Pour les drones MALE, le modèle est construit en étudiant le repère cinématique du drone. Nous considérons les problèmes de planification de trajectoires point-point et point-pattern. Il s’agit, à partir de la position courante du drone, de rejoindre un point ou une figure prédéfinie dans l’espace. La planification point-point est abordée sous forme d’un problème de contrôle optimal. Deux méthodes sont proposées pour résoudre le problème point-pattern. D’abord nous présentons la synthèse en temps minimal pour le système de Dubins. Ensuite, nous développons une méthode basée sur le principe de LaSalle. La première méthode est utilisée au sein d’un algorithme de planification pour des drones HALE. La deuxième permet de stabiliser les deux types de drones considérés vers un pattern. Nous proposons une extension des algorithmes de planification développés, basée sur une discrétisation del’espace grâce aux graphes de Voronoï et une méthode de planification discrète, pour construire des trajectoiresdans des milieux encombrés. Nous étudions également le problème de couplage drone/capteur. Il s’agit de calculer une trajectoire permettant de satisfaire les objectifs du drone et de son capteur (une caméra). L’algorithme proposé est construit à partir de la résolution d’un problème quadratique sous contraintes.Dans une seconde partie, nous analysons un problème de contrôle optimal inverse. Celui-ci permet d’améliorer les résultats des méthodes de planification en s’inspirant du comportement des pilotes. Après avoir posé le problème, les résultats théoriques sont exposés et le cas particulier du système de Dubins est étudié en pratique. / This thesis is about path planning for HALE or MALE UAVs (Unmanned Aircraft Vehicles), possibly under mission constraints. As such, the study is performed at the kinematic level : HALE UAVs are represented as Dubins systems, and a model for MALE UAVs is constructed by studying their kinematic frame. In the first part, we tackle the path planning problem for a UAV that must join a target (a point or a pattern), starting from any position. The point to point path planning problem is addressed as an optimal control problem. Regarding the point to pattern path planning problem, two different methods are proposed. The former consists in solving the minimum time synthesis for the Dubins system, in order to obtain a basis for a HALE UAVs planning algorithm. The latter method relies on the LaSalle principle ; it permits to stabilize a HALE or MALE UAV to a pattern.In addition, extensions of the previously developed algorithms to cluttered environnement are provided. This extension is achieved thanks to a space discretization using Voronoi diagrams and a discrete planning method. Finally, the mission constraints are dealt with as a coupling problem between the UAV and its sensors. The proposed algorithm is presented in the form of a constrained quadratic problem.In the second part of this thesis, we want to refine the planning algorithm to get a result closer to trajectories of pilots. In order to do that, we solve an inverse optimal control problem where the cost to find is computed from the experience of pilots. Theoretical results are presented and applied to the particular case of the Dubins system.

No description available.

M.Com. (Business Management) / The commercial real estate broking industry is considered one of the most competitive industries globally, with research showing that it has many competitors and the ease of entry is rather easy. This research was conducted in the context of the commercial real estate broking industry in order to examine the extent of competition in the industry and to establish how one of the companies in this industry can sustain the competitive advantage. The study revealed that gaining and sustaining competitive advantage is about strengthening the resources that are not valuable, rare and imitable in the organisation as reflected in the model by Knott (2009: 166). The primary purpose of this study was to explore how JLL SA can obtain and sustain competitive advantage in the competitive commercial real estate broking environment through the Resource Based View. The research methodology applied in this study was a qualitative study, which consisted of 9 semi structured interviews from Jones Lang LaSalle South Africa (JLL SA). The criteria for the population sample was that the respondents had to have adequate experience in the commercial real estate broking industry and had a thorough knowledge of the company under review. The research highlighted a number of challenges regarding the market position of JLL SA and that the company does not have a formal rivalry strategy. From the research, it was found that strategic planning, particularly for competition happens as and when the organisation is challenged by competition. The study also reveals that JLL SA does not communicate a strategic direction for tackling competition. The main recommendations of this study is that JLL SA adopts the model suggested by Knott (2009:166) as a measure to gain and sustain competitive advantage. The model refers to the valuable, rarity and inimitability of resources through, in which JLL SA can selects attributes or resources to evaluate in order to sustain competitive commercial real estate broking industry is considered one of the most competitive industries globally, with research showing that it has many competitors and the ease of entry is rather easy. This research was conducted in the context of the commercial real estate broking industry in order to examine the extent of competition in the industry and to establish how one of the companies in this industry can sustain the competitive advantage. The study revealed that gaining and sustaining competitive advantage is about strengthening the resources that are not valuable, rare and imitable in the organisation as reflected in the model by Knott (2009: 166). The primary purpose of this study was to explore how JLL SA can obtain and sustain competitive advantage in the competitive commercial real estate broking environment through the Resource Based View. The research methodology applied in this study was a qualitative study, which consisted of 9 semi structured interviews from Jones Lang LaSalle South Africa (JLL SA). The criteria for the population sample was that the respondents had to have adequate experience in the commercial real estate broking industry and had a thorough knowledge of the company under review. The research highlighted a number of challenges regarding the market position of JLL SA and that the company does not have a formal rivalry strategy. From the research, it was found that strategic planning, particularly for competition happens as and when the organisation is challenged by competition. The study also reveals that JLL SA does not communicate a strategic direction for tackling competition. The main recommendations of this study is that JLL SA adopts the model suggested by Knott (2009:166) as a measure to gain and sustain competitive advantage. The model refers to the valuable, rarity and inimitability of resources through, in which JLL SA can selects attributes or resources to evaluate in order to sustain competitive advantage.

Due to increasing skill and awareness of overall functions in programs such as Excel, an increasing number of analysts at real estate firms and consultancies have started developing "desktop" versions of valuation models used for professional appraisal of property value. Due to personal preferences, differences in schools and professional backgrounds, these so called desktop models vary in quality, robustness, accuracy, design and user friendliness. Professional software suites are not suitable either, as they are expensive, hard to learn, hard to adapt to specific needs of the business, outdated design and need of additional IT resources. At a multinational Real Estate consultancy such as Jones Lang LaSalle, requirements on tools used for professional opinions on questions as important as property value, are rigorous. Therefore, decision was made to develop a new DCF model which would be closely monitored by management and have a prismatic approach meaning that the model would satisfy the needs of more than one division at Jones Lang LaSalle. When reviewing existing models and practices at the company, the result became a tailored DCF valuation model that was focused on increasing efficiency of appraisers at Jones Lang LaSalle. Aside from being robust and technically sophisticated, the result also suited the specific needs of Jones Lang LaSalle in terms of features and user interface. Development of the model involved several divisions to ensure that the needs were met for Research & Valuation, Capital Markets, Corporate Solutions and Asset Management at Jones Lang LaSalle.

Stoney Burns (Brent LaSalle Stein) edited and published Dallas Notes, a Dallas, Texas, underground newspaper, from November 1967 through September 1970. This thesis considers whether Burns was the unifying figure in the Dallas counterculture.

Muitos sistemas físicos são modelados por sistemas dinâmicos discretos. Com o advento da tecnologia digital os sistemas discretos tornaram-se ainda mais importantes, sendo assim, o desenvolvimento de ferramentas analíticas para este tipo de sistema é de grande importância. Neste trabalho, estudam-se alguns dos principais resultados relacionados à estabilidade de sistemas dinâmicos discretos, e alguns novos são propostos. É bem conhecido na literatura que a estabilidade de um ponto de equilíbrio pode ser caracterizada pelo Método Direto de Lyapunov, via uma função auxiliar denominada função de Lyapunov. LaSalle, ao estudar a teoria de Lyapunov, estabeleceu uma importante relação entre função de Lyapunov e conjuntos limites de Birkhoff, que deu origem ao Princípio de Invariância de LaSalle. Este, entre outras coisas, permite a análise de estabilidade assintótica. Tanto o Método Direto de Lyapunov quanto o Princípio de Invariância requerem que a variação da função de Lyapunov seja não positiva ao longo das trajetórias do sistema. Em sistemas com comportamentos mais complexos, dificilmente encontra-se uma função com esta propriedade. Neste trabalho, propõe-se uma versão mais geral do Princípio de Invariância para sistemas discretos, a qual não exige que a variação da função de Lyapunov seja sempre não positiva. Com isto, a obtenção de funções deste tipo torna-se mais simples e muitos problemas, que antes não poderiam ser tratados com a teoria convencional, passam a ser tratados através deste novo resultado. Os resultados desenvolvi- dos, neste trabalho, são úteis para encontrar estimativas de atratores de sistemas não-lineares discretos. / Many physical systems are modeled by discrete dynamic systems. With the evolution digital technology, the discrete systems became still more important, so the development of analytic tools for this type of system has high importance nowadays. ln this work, some of the main results in stability of discrete dynamic systems are studied and some new ones are proposed. lt is well known in the literature that the stability of an equilibrium point may be characterized by the Lyapunov\'s Direct Method, with a function known as Lyapunov auxiliary function. LaSalle, when studying the Lyapunov theory, established an important relationship between Lyapunov function and Birkhoff limit sets. Then, he created the Lasalle\'s lnvariance Principle. This, among other features, allows the analysis of asymptotically stability. Both the Lyapunov\'s Direct Method and the lnvariance Principle request the variation of the Lyapunov function to be negative semidefinite along the system trajectory. In systems with more complex behaviors, a function is hardly found with this property. This work developed a more general version of the lnvariance Principle for discrete systems, which does not require the variation of the Lyapunov function to be always negative semidefinite. This new theory enables to find these functions easily and many insoluble problems, which could not be treated with the conventional theory before, become treatable by this new result. The results of this work are useful to find estimates of discrete nonlinear systems atractors.

As técnicas de análise de estabilidade transitória em sistemas elétricos de potência desenvolveram-se significativamente nas últimas duas décadas. Atualmente, o principal desafio dos pesquisadores é a obtenção de técnicas que sejam adequadasa análises em tempo real. Neste sentido, as idéias de Liapunov associadas ao Princípio de Invariância de LaSalle têm sido utilizadas para estimar a bacia de atraçãoo dos sistemas de potência. Embora esta filosofia seja bastante adequada a análises de estabilidade em tempo real, existem alguns obstáculos que impedem a aplicação da mesma à análise de sistemas reais. Dentre estes obstáculos poder-se-ia destacar a impossibilidade de utilização de modelos mais realísticos e a limitação da análise ao primeiro "swing". Em verdade, estes obstáculos estão intimamente relacionados com as limitações do Princípio de Invariância de LaSalle. Para superar estes problemas, propõe-se, neste trabalho, uma extensão deste princípio que é mais geral e portanto mais flexível do que o original. Aproveitando esta maior flexibilidade, duas aplicações em análise de estabilidade transitória são abordadas, ambas com o objetivo de reduzir os obstáculos anteriormente mencionados. Na primeira, propõe-se uma nova função energia para sistemas de potência com perdas nas linhas de transmissão. Mostra-se que esta é uma função de Liapunov no sentido mais geral da extensão do Princípio de Invariância de LaSalle, podendo portanto ser empregada para estudos de estabilidade. Na segunda, uma metodologia de análise de estabilidade multi-"swing" é proposta com base em uma análise de coerência de geradores.

Pour les systèmes de commande sous la forme de dx/dt = f (x, u), dans la littérature, les chercheurs s'intéressaient à la stabilisation de ce système de différentes manières : asymptotique, uniformément asymptotique, partielle, en temps fini, etc. Pour aboutir à ces résultats, les méthodes utilisées font appel aux techniques suivantes : Lyapunov, Lasalle, Barbalat, surface glissante, etc. Dans cette thèse, nous nous sommes intéressés à une autre fonctionnalité de la commande, dite commande répulsive stabilisante. Les résultats ont été généralisés au cas d'un système avec dérive et sans dérive. Comme résultat, l'approche de commande qu'on propose assure la stabilité du système autour d'une position désirée et la répulsion de celui-ci par rapport à un ensemble indésirable, construit dans l'espace de navigation. Toute forme d'application sera concernée par nos résultats théoriques, on peut citer, la navigation terrestre et aérienne dans un environnement peu ou pas connu. De même, la commande qu'on propose préserve la communication inter-agent, une fois planifiées. En terme d'application, on a considéré le modèle d'un véhicule à roues type unicycle, sans tenir compte de l'orientation (cas non holonôme) et dans le cas où l'environnement contient un ou plusieurs obstacle(s). Contrairement aux résultats de la littérature, qui sont basés sur une commande à structure variable pour l'évitement d'un obstacle, la commande répulsive-stabilisante trouvée est une commande continue sur l'espace de navigation. La deuxième partie de cette thèse traite le problème de stabilité d'une formation d'agents (système multi-véhicules) qui évolue dans un environnement hostile tout en préservant la communication entre les agents. Pour réussir la formation, la décentralisation de la commande par rapport aux agents est rendue robuste à travers des graphes de communication. Ces graphes relèvent de la stratégie et objectifs de la formation. Nos résultats de stabilité ont fait l'objet d'une implémentation rigoureuse sur un simulateur réalisé sous Matlab.

As técnicas de análise de estabilidade transitória em sistemas elétricos de potência desenvolveram-se significativamente nas últimas duas décadas. Atualmente, o principal desafio dos pesquisadores é a obtenção de técnicas que sejam adequadasa análises em tempo real. Neste sentido, as idéias de Liapunov associadas ao Princípio de Invariância de LaSalle têm sido utilizadas para estimar a bacia de atraçãoo dos sistemas de potência. Embora esta filosofia seja bastante adequada a análises de estabilidade em tempo real, existem alguns obstáculos que impedem a aplicação da mesma à análise de sistemas reais. Dentre estes obstáculos poder-se-ia destacar a impossibilidade de utilização de modelos mais realísticos e a limitação da análise ao primeiro "swing". Em verdade, estes obstáculos estão intimamente relacionados com as limitações do Princípio de Invariância de LaSalle. Para superar estes problemas, propõe-se, neste trabalho, uma extensão deste princípio que é mais geral e portanto mais flexível do que o original. Aproveitando esta maior flexibilidade, duas aplicações em análise de estabilidade transitória são abordadas, ambas com o objetivo de reduzir os obstáculos anteriormente mencionados. Na primeira, propõe-se uma nova função energia para sistemas de potência com perdas nas linhas de transmissão. Mostra-se que esta é uma função de Liapunov no sentido mais geral da extensão do Princípio de Invariância de LaSalle, podendo portanto ser empregada para estudos de estabilidade. Na segunda, uma metodologia de análise de estabilidade multi-"swing" é proposta com base em uma análise de coerência de geradores.

Muitos sistemas físicos são modelados por sistemas dinâmicos discretos. Com o advento da tecnologia digital os sistemas discretos tornaram-se ainda mais importantes, sendo assim, o desenvolvimento de ferramentas analíticas para este tipo de sistema é de grande importância. Neste trabalho, estudam-se alguns dos principais resultados relacionados à estabilidade de sistemas dinâmicos discretos, e alguns novos são propostos. É bem conhecido na literatura que a estabilidade de um ponto de equilíbrio pode ser caracterizada pelo Método Direto de Lyapunov, via uma função auxiliar denominada função de Lyapunov. LaSalle, ao estudar a teoria de Lyapunov, estabeleceu uma importante relação entre função de Lyapunov e conjuntos limites de Birkhoff, que deu origem ao Princípio de Invariância de LaSalle. Este, entre outras coisas, permite a análise de estabilidade assintótica. Tanto o Método Direto de Lyapunov quanto o Princípio de Invariância requerem que a variação da função de Lyapunov seja não positiva ao longo das trajetórias do sistema. Em sistemas com comportamentos mais complexos, dificilmente encontra-se uma função com esta propriedade. Neste trabalho, propõe-se uma versão mais geral do Princípio de Invariância para sistemas discretos, a qual não exige que a variação da função de Lyapunov seja sempre não positiva. Com isto, a obtenção de funções deste tipo torna-se mais simples e muitos problemas, que antes não poderiam ser tratados com a teoria convencional, passam a ser tratados através deste novo resultado. Os resultados desenvolvi- dos, neste trabalho, são úteis para encontrar estimativas de atratores de sistemas não-lineares discretos. / Many physical systems are modeled by discrete dynamic systems. With the evolution digital technology, the discrete systems became still more important, so the development of analytic tools for this type of system has high importance nowadays. ln this work, some of the main results in stability of discrete dynamic systems are studied and some new ones are proposed. lt is well known in the literature that the stability of an equilibrium point may be characterized by the Lyapunov\'s Direct Method, with a function known as Lyapunov auxiliary function. LaSalle, when studying the Lyapunov theory, established an important relationship between Lyapunov function and Birkhoff limit sets. Then, he created the Lasalle\'s lnvariance Principle. This, among other features, allows the analysis of asymptotically stability. Both the Lyapunov\'s Direct Method and the lnvariance Principle request the variation of the Lyapunov function to be negative semidefinite along the system trajectory. In systems with more complex behaviors, a function is hardly found with this property. This work developed a more general version of the lnvariance Principle for discrete systems, which does not require the variation of the Lyapunov function to be always negative semidefinite. This new theory enables to find these functions easily and many insoluble problems, which could not be treated with the conventional theory before, become treatable by this new result. The results of this work are useful to find estimates of discrete nonlinear systems atractors.