Les robots parallèles à câbles sont une variante originale des robots parallèles. L'utilisation de câbles en lieu et place des segments rigides procure à ce type de robots un espace de travail potentiellement très grand car des longueurs importantes de câbles peuvent être déroulées. Toutefois, dans la plupart des études sur les robots à câbles, un modèle de câble sans masse non élastique est utilisé. Si dans le cas de robots de faibles dimensions soumis à de faibles efforts, ce modèle est valide, lorsque l'on considère des applications de très grande dimension pour lesquels la masse des câbles et l'élasticité ne peuvent plus être négligées, ces modèles simples ne sont plus valables. Ces travaux de thèse proposent des nouvelles méthodes d'étude des robots parallèles à câbles de grande dimension. Dans un premier temps, des tests de traction réalisés sur différents câbles permettent de proposer différents modèles élastiques. La modélisation d'un câble par une caténaire élastique est ensuite rappelée, et l'erreur importante obtenue en négligeant la masse des câbles est mise en exergue. La modélisation par caténaire élastique bien que précise, nécessite la résolution d'un système d'équations couplées non-linéaires. Un modèle simplifié de câble pesant est alors présenté. Il permet, sous l'hypothèse de faible déflection du câble, de simplifier la résolution de l'équilibre statique d'un robot à câble. Ce modèle permet également de développer des outils utiles à la détermination de l'ensemble des torseurs d'efforts admissibles à la plate-forme d'un robot parallèle à câbles. La vérification de l'inclusion de l'ensemble des torseurs nécessaires à la réalisation d'une tâche dans l'ensemble des torseurs admissibles est finalement utilisée comme critère d'optimisation pour une méthode de conception de robots à câbles de grandes dimensions. / Cable-driven robot is an original variation of parallel robots. Replacing rigid bodies by cables provides new capabilities to these robots, and particularly large-size workspaces, since long cable lengths can be deployed. In the literature, cables are usually supposed to be inextensible and massless. If this modeling is valid for small robots with moderate payloads, this cable model is not accurate enough to be used for large dimension cable-driven robots. The work presented here focuses on the modeling of such large cable robots. First, from a set of traction tests applied to various cables, elastic models are proposed. Then, the well-know elastic catenary model is recalled, and its effects on the modeling of large dimension cable robots is shown. However, when using this cable model, solving the platform static equilibrium require the resolution of a non-linear coupled equation system. Assuming a low sagging of the cable, some simplifications can be made to this model. The resulting simplified hefty cable model is then presented and the new expression of the static equilibrium is shown to be close to the one obtained with the massless cable model. Thus, it allows us to determine the set of admissible mobile platform wrenches at a given pose. By comparing this set to the set of required wrenches for a specific task a cost function is finally defined and used in a design procedure dedicated to large dimension cable-driven robots.

In the present study, the optimal design of feeding system in steel sand-mold castings is considered. The first part of this research includes fundamental studies on the physics of shrinkage defect formation during the casting process. The results of these studies lead to new findings on the mechanism of shrinkage defect formation, effect of melt quality on the distribution of defects within the castings and the connection between shrinkage and gases defects. The theoretical analysis of thermal criterion functions for the prediction of shrinkage defects in castings and introducing new criterion function with fewer shortcomings can be accounted as the other finding of this part. A new model was introduced in the second part of this research for the purpose of optimal design of feeding system in the shape casting processes. In this model the optimal design problem is formulated as a point-wise constrained topology optimization problem. Unlike alternative methods, the presented method does not require any predesigned feeding system as an initial guess. Using the functional analysis on the infinite-dimensional function spaces, a numerically efficient method was introduced to solve the optimal design problem in this study. By using extensive numerical experiments, capabilities and limitations of the presented method were studied in the last part of this research.

<p>New techniques and approaches for organic synthesis, purification and biological testing are enabling pharmaceutical industries to produce and test increasing numbers of compounds every year. Surprisingly, this has not led to more new drugs reaching the market, prompting two questions – why is there not a better correlation between their efforts and output, and can it be improved? One possible way to make the drug discovery process more efficient is to ensure, at an early stage, that the tested compounds are diverse, representative and of high quality. In addition the biological evaluation systems have to be relevant and reliable. The diversity of the tested compounds could be ensured and the reliability of the biological assays improved by using Design Of Experiments (DOE) more frequently and effectively. However, DOE currently offers insufficient options for these purposes, so there is a need for new, tailor-made DOE strategies. The aim of the work underlying this thesis was to develop and evaluate DOE approaches for diverse compound selection and efficient assay optimisation. This resulted in the publication of two new DOE strategies; D-optimal Onion Design (DOOD) and Rectangular Experimental Designs for Multi-Unit Platforms (RED-MUP), both of which are extensions to established experimental designs.</p><p>D-Optimal Onion Design (DOOD) is an extension to D-optimal design. The set of possible objects that could be selected is divided into layers and D-optimal selection is applied to each layer. DOOD enables model-based, but not model-dependent, selections in discrete spaces to be made, since the selections are not only based on the D-optimality criterion, but are also biased by the experimenter’s prior knowledge and specific needs. Hence, DOOD selections provide controlled diversity.</p><p>Assay development and optimisation can be a major bottleneck restricting the progress of a project. Although DOE is a recognised tool for optimising experimental systems, there has been widespread unwillingness to use it for assay optimisation, mostly because of the difficulties involved in performing experiments according to designs in 96-, 384- and 1536- well formats. The RED-MUP framework combines classical experimental designs orthogonally onto rectangular experimental platforms, which facilitates the execution of DOE on these platforms and hence provides an efficient tool for assay optimisation.</p><p>In combination, these two strategies can help uncovering the crossroads between biology and chemistry in drug discovery as well as lead to higher information content in the data received from biological evaluations, providing essential information for well-grounded decisions as to the future of the project. These two strategies can also help researchers identify the best routes to take at the crossroads linking biological and chemical elements of drug discovery programs.</p>

Ground penetrating radar (GPR) systems are increasingly being used for the detection and location of buried objects within the upper regions of the earth’s surface. The antenna is the most critical component of such a system. This thesis presents a comprehensive study of resistor-loaded planar dipole antennas for GPR applications using both theory and experiments. The theoretical analysis is performed using the finite difference time domain (FDTD) technique. The analysis starts with the most popular planar dipole, the bow-tie. A parametric study is done to find out how the flare angle, length, and lumped resistors of the antenna should be selected to achieve broadband properties and good target detection with less clutter. The screening of the antenna and the position of transmitting and receiving antennas with respect to each other and ground surface are also studied. A number of other planar geometrical shapes are considered and compared with the bow-tie in order to find what geometrical shape gives the best performance. The FDTD simulations are carried out for both lossless and lossy, dispersive grounds. Also simulations are carried out including surface roughness and natural clutter like rocks and twigs to make the modeling more realistic. Finally, a pair of resistor-loaded bow-tie antennas is constructed and both indoor and outdoor measurements are carried out to validate the simulation results. / <p>QC 20100923</p>

In the present thesis, computational methods for theanalysis and design of photonic bandgap structure areconsidered. Many numerical methods have been used to study suchstructures. Among them, the plane wave expansion method is veryoften used. Using this method, we show that inclusions ofelliptic air holes can be used effectively to obtain a largercomplete band gap for two-dimensional (2D) photonic crystals.An optimal design of a 2D photonic crystal is also consideredin the thesis using a combination of the plane wave expansionmethod and the conjugate gradient method. We find that amaximum complete 2D band gap can be obtained by connectingdielectric rods with veins for a photonic crystal with a squarelattice of air holes in GaAs. For some problems, such as defect modes, the plane waveexpansion method is extremely time-consuming. It seems that thefinite-difference time-domain (FDTD) method is promising, sincethe computational time is proportional to the number of thediscretization points in the computation domain (i.e., it is oforderN). A FDTD scheme in a nonorthogonal coordinate systemis presented in the thesis to calculate the band structure of a2D photonic crystal consisting of askew lattice. The algorithmcan easily be used for any complicated inclusion configuration,which can have both the dielectric and metallic constituents.The FDTD method is also applied to calculate the off-plane bandstructures of 2D photonic crystals in the present thesis. Wealso propose a numerical method for computing defect modes in2D crystals (with dielectric or metallic inclusions). Comparedto the FDTD transmission spectra method, our method reduces thecomputation time and memory significantly, and finds as manydefect modes as possible, including those that are not excitedby an incident plane wave in the FDTD transmission spectramethod. The FDTD method has also been applied to calculateguided modes and surface modes in 2D photonic crystals using acombination of the periodic boundary condition and theperfectly matched layer for the boundary treatment. Anefficient FDTD method, in which only real variables are used,is also proposed for the full-wave analysis of guided modes inphotonic crystal fibers. / QC 20100629

In this paper we investigate (d + 1)-point D-optimal designs for d-th degree polynomial regression with weight function w(x) > 0 on the interval [a, b]. Suppose that w'(x)/w(x) is a rational function and the information of whether the optimal support contains the boundary points a and b is available. Then the problem of constructing (d + 1)-point D-optimal designs can be transformed into a differential equation problem leading us to a certain matrix with k auxiliary unknown constants. We characterize the weight functions corresponding to the cases when k= 0 and k= 1. Then, we can solve (d + 1)-point D-optimal designs directly from differential equation (k = 0) or via eigenvalue problems (k = 1). The numerical results show us an interesting relationship between optimal designs and ordered eigenvalues.

Consider the minimally-supported D-optimal designs for dth degree polynomial regression with bounded and positive weight function on a compact interval. We show that the optimal design converges weakly to the arcsin distribution as d goes to infinity. Comparisons of the optimal design with the arcsin distribution and D-optimal arcsin support design by D-efficiencies are also given. We also show that if the design interval is [−1, 1], then the minimally-supported D-optimal design converges to the D-optimal arcsin support design with the specific weight function 1/¡Ô(£\-x^2), £\>1, as £\¡÷1+.

Consider the D-optimal designs for the dth-degree polynomial regression model with a bounded and positive weight function on a compact interval. As the degree of the model goes to infinity, we show that the D-optimal design converges weakly to the arcsin distribution. If the weight function is equal to 1, we derive the formulae of the values of the D-criterion for five classes of designs including (i) uniform density design; (ii) arcsin density design; (iii) J_{1/2,1/2} density design; (iv) arcsin support design and (v) uniform support design. The comparison of D-efficiencies among these designs are investigated; besides, the asymptotic expansions and limits of their D-efficiencies are also given. It shows that the D-efficiency of the arcsin support design is the highest among the first four designs.

New techniques and approaches for organic synthesis, purification and biological testing are enabling pharmaceutical industries to produce and test increasing numbers of compounds every year. Surprisingly, this has not led to more new drugs reaching the market, prompting two questions – why is there not a better correlation between their efforts and output, and can it be improved? One possible way to make the drug discovery process more efficient is to ensure, at an early stage, that the tested compounds are diverse, representative and of high quality. In addition the biological evaluation systems have to be relevant and reliable. The diversity of the tested compounds could be ensured and the reliability of the biological assays improved by using Design Of Experiments (DOE) more frequently and effectively. However, DOE currently offers insufficient options for these purposes, so there is a need for new, tailor-made DOE strategies. The aim of the work underlying this thesis was to develop and evaluate DOE approaches for diverse compound selection and efficient assay optimisation. This resulted in the publication of two new DOE strategies; D-optimal Onion Design (DOOD) and Rectangular Experimental Designs for Multi-Unit Platforms (RED-MUP), both of which are extensions to established experimental designs. D-Optimal Onion Design (DOOD) is an extension to D-optimal design. The set of possible objects that could be selected is divided into layers and D-optimal selection is applied to each layer. DOOD enables model-based, but not model-dependent, selections in discrete spaces to be made, since the selections are not only based on the D-optimality criterion, but are also biased by the experimenter’s prior knowledge and specific needs. Hence, DOOD selections provide controlled diversity. Assay development and optimisation can be a major bottleneck restricting the progress of a project. Although DOE is a recognised tool for optimising experimental systems, there has been widespread unwillingness to use it for assay optimisation, mostly because of the difficulties involved in performing experiments according to designs in 96-, 384- and 1536- well formats. The RED-MUP framework combines classical experimental designs orthogonally onto rectangular experimental platforms, which facilitates the execution of DOE on these platforms and hence provides an efficient tool for assay optimisation. In combination, these two strategies can help uncovering the crossroads between biology and chemistry in drug discovery as well as lead to higher information content in the data received from biological evaluations, providing essential information for well-grounded decisions as to the future of the project. These two strategies can also help researchers identify the best routes to take at the crossroads linking biological and chemical elements of drug discovery programs.

Dans le but de réduire la consommation d’énergie et d’augmenter la part des énergies renouvelables, la conception optimale des futurs bâtiments (bâtiments intelligents) apparaît comme un facteur important. Cette thèse vise donc à développer des modèles, des méthodes innovantes d’aide à la conception pour ces bâtiments. Notre nouvelle approche de conception est une optimisation globale et simultanée de l’enveloppe, des systèmes énergétiques et de leurs stratégies de gestion dès la phase d’esquisse, qui prend en compte plusieurs critères de coût (investissement et exploitation) et de confort (thermique, visuel et aéraulique). Le problème d’optimisation multi-objectif est donc un problème de couplage fort de grande taille avec de nombreuses variables et contraintes, qui induisent des difficultés lors de sa résolution. Après avoir fait des analyses sur des cas tests, une méthode d’optimisation d’ordre 1 est choisie, en association à des modèles analytiques dérivés formellement de manière automatique. Notre méthodologie est appliquée à la conception de maisons individuelles, et plus particulièrement des maisons à énergie positive. Les résultats obtenus par cette approche globale apportent des informations importantes aux concepteurs pour l’aider à faire des choix en phase amont du processus de conception. / In order to reduce the energy consumption and to increase the use of renewable energy, the optimal design of future buildings (smart-buildings) appears as an important factor.This thesis aims to develop models, innovative methods aiding decision-making during the design of buildings. Our approach of design is a global and simultaneous optimization of envelope, energy systems and their management strategies from the sketch phase, which takes into account multi-criterions of costs (investment et exploitation) and comforts (thermal, visual, aeraulic). The multi-objective optimization problem is so a strong coupling problem of large scale with a lot of variables and constraints, which leads to difficulties to solve.After the tests, an optimization method of order 1 is chosen in combination with analytical models formally derived automatically. Our methodology is applied to the design of individual houses, especially positive energy houses. The results of this global approach provide important information to designers to help make choices from the preliminary phase of the design process.