No description available.

Le placement de caméra en environnement virtuel consiste à positionner et orienter une caméra virtuelle 3D de façon à respecter un ensemble de propriétés visuelles ou cinématographiques définies par l'utilisateur. Réaliser cette tâche est difficile en pratique. En effet, l'utilisateur possède une vision claire du résultat qu'il souhaite obtenir en termes d'agencement des objets à l'image. Toutefois le processus classique de placement de caméra est particulièrement contre-intuitif. L'utilisateur doit effectuer une inversion mentale afin d'inférer la position et l'orientation de la caméra dans l'environnement 3D amenant au résultat souhaité. Des méthodes d'aide au placement de caméra apparaissent donc particulièrement profitables pour les utilisateurs. Dans cette thèse, nous identifions trois axes de recherche relativement peu couverts par la littérature dédiée au placement de caméra et qui nous apparaissent pourtant essentiels. D'une part, les approches existantes n'offrent que peu de flexibilité tant dans la résolution que dans la description d'un problème en termes de propriétés visuelles, en particulier lorsque celui-ci ne possède aucune solution. Nous proposons une méthode de résolution flexible qui calcule l'ensemble des solutions, maximisant la satisfaction des propriétés du problème, que celui-ci soit sur-contraint ou non. D'autre part, les méthodes existantes ne calculent qu'une seule solution, même lorsque le problème possède plusieurs classes de solutions équivalentes en termes de satisfaction de propriétés. Nous introduisons la méthode des volumes sémantiques qui calcule l'ensemble des classes de solutions sémantiquement équivalentes et propose un représentant de chacune d'elles à l'utilisateur. Enfin, le problème de l'occlusion, bien qu'essentiel dans la transmission de l'information, n'est que peu abordé par la communauté. En conséquence, nous présentons une nouvelle méthode de prise en compte de l'occlusion dans des environnements dynamiques temps réel.

Laser in situ scattering and transmissometry (LISST) instruments are used to measure the particle size distributions (PSDs) and volume concentration of individual and groups of phytoplankton in water. The objective of this research was to test the LISST’s ability in detecting monospecific blooms in-situ and the ability to detect aggregation after diatoms were subjected to different temperatures and bacteria concentrations. The PSDs of ten harmful algal bloom (HAB) species were measured with the LISST characterizing the peak location, peak height, peak width, and peak range resulting in a scattering signature for each species. Each species had specific characteristics that would allow for their detection with the LISST, though microscope observations would be needed for complete accuracy. The LISST was able to detect HABs placed in natural seawater collected off the Texas coast. Blooms of four HAB species before they reached full” bloom concentrations were detected making the LISST a possible low cost, effective tool in the early detection and monitoring of HABs. The diatom, Odontella aurita, was used to test how well the LISST could monitor aggregation, an important process in the termination of many phytoplankton blooms. Increasing temperature causes an increase in transparent exopolymer particle (TEP) production in diatoms, which is a critical sticky particle that increases the probability of aggregation. An increase in temperature can also cause an increase in bacteria concentration that can positively effect TEP production and thus aggregation. O. aurita was grown at 20 °C and 28 °C and showed a significant increase in TEP abundance with temperature (p = 0.002), though no relationship between TEP production and bacteria concentration existed. Coomassie stained particles (CSP) are proteinaceous gel-like particles, which are currently understudied. CSP was consistently produced though it did not appear to be dependent upon any single factor. The increase in ocean temperatures has implications for an increase in phytoplankton blooms making the monitoring and understanding of these blooms even more important as they can affect the carbon cycle and potentially the microbial loop.

The runtime for solving constraint satisfaction problems (CSP) and propositional satisfiability problems (SAT) using systematic backtracking search has been shown to exhibit great variability. Randomization and restarts is an effective technique for reducing such variability to achieve better expected performance. Several restart strategies have been proposed and studied in previous work and show differing degrees of empirical effectiveness. <br /><br /> The first topic in this thesis is the extension of analytical results on restart strategies through the introduction of physically based assumptions. In particular, we study the performance of two of the restart strategies on Pareto runtime distributions. We show that the geometric strategy provably removes heavy tail. We also examine several factors that arise during implementation and their effects on existing restart strategies. <br /><br /> The second topic concerns the development of a new hybrid restart strategy in a realistic problem setting. Our work adapts the existing general approach on dynamic strategy but implements more sophisticated machine learning techniques. The resulting hybrid strategy shows superior performance compared to existing static strategies and an improved robustness.

The runtime for solving constraint satisfaction problems (CSP) and propositional satisfiability problems (SAT) using systematic backtracking search has been shown to exhibit great variability. Randomization and restarts is an effective technique for reducing such variability to achieve better expected performance. Several restart strategies have been proposed and studied in previous work and show differing degrees of empirical effectiveness. <br /><br /> The first topic in this thesis is the extension of analytical results on restart strategies through the introduction of physically based assumptions. In particular, we study the performance of two of the restart strategies on Pareto runtime distributions. We show that the geometric strategy provably removes heavy tail. We also examine several factors that arise during implementation and their effects on existing restart strategies. <br /><br /> The second topic concerns the development of a new hybrid restart strategy in a realistic problem setting. Our work adapts the existing general approach on dynamic strategy but implements more sophisticated machine learning techniques. The resulting hybrid strategy shows superior performance compared to existing static strategies and an improved robustness.

The Coffin-Manson equations of Sn/Ag/Cu and Sn/Pb solder joints are presented in this thesis. The experimental results of CSP thermal cycle fatigue test and ball shear test are used to formulate Coffin-Manson equations. The maximum amplitude of equivalent plastic shear strain corresponding to these two experiments are employed. The MARC finite element package is used to calculate the plastic shear strain. Different published fatigue experiment results have been used to show the accuracy and the feasibility of these proposed equations. The 3-D finite element models of the BGA type¡¦s CSP and VCSEL assembly are employed to simulate the thermal cycling fatigue. Results indicate that the fatigue lifes of solder predicted by using the proposed equations have good agreement with those measured from experimental tests.

In recent year, a lot of scholar try to generate test sets for combinatorial strategy automatically. But these algorithms based on combinatorial strategy don¡¦t consider conflicts of input parameter model. A conflict exists when the result of combining two or more values of different parameter dose not make sense. Thus, invalid sub-combinations may be included in test cases in the test suite, and these are useless to us. Besides, these algorithms all directly generate all test cases once, in other words, it is unable to utilize test cases generated at present to feedback and revise the algorithm, so it is easy to generate useless combinations. So, this paper proposes new test generation algorithm for combinatorial testing based on constraint satisfaction problem(CSP) to solve problem which invalid sub-combinations may be included in test cases, and we can add constraints flexibly during generating test cases to avoid generate useless or repeated combinations. The experimental result indicate that our algorithm perform well, with respect to the amount of time required for test generation, otherwise, we can generate conflict-free test cases directly.

Abstract The thesis is aimed to analyze the flip chip in chip scale package (CSP) by finite element method incorporated with software ANSYS due to thermally cyclic loading. The coefficient of thermal expansion (CTE) of underfill and different mechanical properties of four kinds underfill-A, B, C, D and with/without metal cap are considered as parameters. The effects of above-mentioned parameters on package¡¦s displacement, strain and stress fields are studied. The results show that the maximum equivalent strain and stress take place at the interface between chip and underfill far away from the center of the whole package and on the top of the most outside solder bump in the solder joint. The larger the CTE of underfill is, the larger the maximum equivalent strain and stress are. Package with metal cap can reduce the displacement to almost half or more of that without cap, but increase the values of maximum equivalent strain and stress. No matter with metal cap or not, the underfill D is the best choice. Hence, the underfill material properties possess lower CTE and larger Young¡¦s modulus than those of solder bump.

In recent year the research field of the combinatorial testing, which can roughlybe divided into two kinds including pair-wise coverage and multi-wise coverage. a lot of scholar try to use various strategies to generate test data automatically. In order to weight the generated test set, the generated test data must satisfy certain criterion. But these combinatorial strategy neglected the flexibility of using on the practice. Considering software testing from the practice, which often be restricted by the cost. For this reason, how to obtaint the greatest testing benefits under the limited cost must be considered on the parctice. But in the extant combinatorial strategy, there is no flexible use. In other words, we must testing test set totally. Therefore, there is very great restriction to exist on using the test data generated by the extant combinatorial strategy on the practice. So, this paper proposes a flexible combinatorial strategy based on CSP, which allow users to do the most valid testing under the limited cost, which also allow users join the constraints that needs at any time during the testing process, revise the test data that we produced dynamically. The experimental result indicate that our method perform well, it can avoid including the test data whether some users think the interests less or unnecessarier, in order to achive the greatest testing beneifts.Further, we can achive the goal of reducing the quantity of testing data.

Notwithstanding several years of robust growth, solar energy still only accounts for<1% of total electrical generation in the US. Before solar energy can substantially replace fossil fuels subsidy-free at utility scale, further cost reductions and efficiency improvements are needed in complete generating systems. Flat panel silicon PV modules are by far the most dominant solar technology today, but have little room for improvement in efficiency and are limited by balance of system costs. Concentrated PV (CPV) is an alternate approach with long-term potential for much higher efficiency in sunny climates. In CPV modules, large area optics collect and concentrate direct sunlight onto small multi-junction cells with>40% conversion efficiency. Concentrated Solar Power (CSP) uses mirrors to concentrate sunlight onto thermally absorbing receivers, which generate electricity with convention thermal cycles. In this dissertation, four new optical approaches to CPV and CSP with potential for lower cost are analyzed. Common to each approach is the use of large square glass reflectors, which have very low areal cost (~$35/m^2) and field-proven reliability in the CSP industry. Chapter 2 describes a freeform toroidal lens array used to intercept the low concentration line focus of a parabolic trough to produce multiple high concentration foci (>800X) for multi-junction cells. In Chapter 3, three embodiments of dish mirrors and freeform lenslet arrays are explored, including an off-axis system. In each case, a dish mirror illuminates a freeform lenslet array, which divides sunlight equally to a sparse matrix of multi-junction cells. The off-axis optical system achieves +/-0.45° acceptance angle and averages 1215X geometric concentration over 400 multi-junction cells. Chapter 4 proposes a new architecture for CSP central receivers that achieves extremely high collection efficiency (>70%) with unconventional heliostat field tracking. In Chapter 5, the design and preliminary testing of a spectrum-splitting hybrid PV/thermal generator is discussed. This system has the advantage of 'drop-in' capability in existing CSP trough plants and allows for thermal storage, an important mitigation to the intermittency of the solar resource.