abstract: Researchers and practitioners have widely studied road network traffic data in different areas such as urban planning, traffic prediction and spatial-temporal databases. For instance, researchers use such data to evaluate the impact of road network changes. Unfortunately, collecting large-scale high-quality urban traffic data requires tremendous efforts because participating vehicles must install Global Positioning System(GPS) receivers and administrators must continuously monitor these devices. There have been some urban traffic simulators trying to generate such data with different features. However, they suffer from two critical issues (1) Scalability: most of them only offer single-machine solution which is not adequate to produce large-scale data. Some simulators can generate traffic in parallel but do not well balance the load among machines in a cluster. (2) Granularity: many simulators do not consider microscopic traffic situations including traffic lights, lane changing, car following. This paper proposed GeoSparkSim, a scalable traffic simulator which extends Apache Spark to generate large-scale road network traffic datasets with microscopic traffic simulation. The proposed system seamlessly integrates with a Spark-based spatial data management system, GeoSpark, to deliver a holistic approach that allows data scientists to simulate, analyze and visualize large-scale urban traffic data. To implement microscopic traffic models, GeoSparkSim employs a simulation-aware vehicle partitioning method to partition vehicles among different machines such that each machine has a balanced workload. The experimental analysis shows that GeoSparkSim can simulate the movements of 200 thousand cars over an extensive road network (250 thousand road junctions and 300 thousand road segments). / Dissertation/Thesis / Masters Thesis Computer Engineering 2019

This research presents the first use of side-polished Panda fiber in the fabrication of slab-coupled optical sensors (SCOS). It is determined that the Panda SCOS provides a sensor that is comparable to the D-fiber SCOS in cases of electric field sensing. It exhibits greater power transfer and higher bandwidth than a typical D-SCOS. The Panda SCOS is also less costly and easier to splice than the D-fiber alternative. This comes at a cost of slightly decreased sensitivity and a more fragile fabrication process. This research also demonstrates the use of the Panda-SCOS as means of voltage characterization across both the spark gap of an ignition coil circuit along with the spark gap of an automobile. This paper demonstrates the use of a Panda SCOS to measure the voltages and time delays across the spark gaps at different stages of a Marx generator setup .

Outlier detection remains a heated area due to its essential role in a wide range of applications, including intrusion detection, fraud detection in finance, medical diagnosis, etc. Local Outlier Factor (LOF) has been one of the most influential outlier detection techniques over the past decades. LOF has distinctive advantages on skewed datasets with regions of various densities. However, the traditional centralized LOF faces new challenges in the era of big data and no longer satisfies the rigid time constraints required by many modern applications, due to its expensive computation overhead. A few researchers have explored the distributed solution of LOF, but existant methods are limited by their grid-based data partitioning strategy, which falls short when applied to high-dimensional data. In this thesis, we study efficient distributed solutions for LOF. A baseline MapReduce solution for LOF implemented with Apache Spark, named MR-LOF, is introduced. We demonstrate its disadvantages in communication cost and execution time through complexity analysis and experimental evaluation. Then an approximate LOF method is proposed, which relies on locality-sensitive hashing (LSH) for partitioning data and enables fully distributed local computation. We name it MR-LOF-LSH. To further improve the approximate LOF, we introduce a process called cross-partition updating. With cross-partition updating, the actual global k-nearest neighbors (k-NN) of the outlier candidates are found, and the related information of the neighbors is used to update the outlier scores of the candidates. The experimental results show that MR-LOF achieves a speedup of up to 29 times over the centralized LOF. MR-LOF-LSH further reduces the execution time by a factor of up to 9.9 compared to MR-LOF. The results also highlight that MR-LOF-LSH scales well as the cluster size increases. Moreover, with a sufficient candidate size, MR-LOF-LSH is able to detect in most scenarios over 90% of the top outliers with the highest LOF scores computed by the centralized LOF algorithm.

Processing real-world data requires the ability to analyze data in real-time. Data processing engines like Hadoop come short when results are needed on the fly. Apache Spark's streaming library is increasingly becoming a popular choice as it can stream and analyze a significant amount of data. To showcase and assess the ability of Spark various metrics were designed and operated using data collected from the USGODAE data catalog. The latency of streaming in Apache Spark was measured and analyzed against many nodes in the cluster. Scalability was monitored by adding and removing nodes in the middle of a streaming job. Fault tolerance was verified by stopping nodes in the middle of a job and making sure that the job was rescheduled and completed on other node/s. A full stack application was designed that would automate data collection, data processing and visualizing the results. Google Maps API was used to visualize results by color coding the world map with values from various analytics.

Thesis (Ph. D.)--University of Texas at Austin, 2007. / Vita. Includes bibliographical references.

La conception de systèmes énergétiques innovants et la caractérisation du confort des occupants requièrent de d'être capable d'estimer les détails des écoulements et des transferts de chaleur au sein des zones des bâtiments. Les méthodes de modélisation permettant d'estimer ces détails, telles que les méthodes zonales et CFD, sont difficilement applicables à l'étude d'un bâtiment dans son ensemble et sur de longues périodes de temps. Notre étude consiste à proposer une plate-forme de simulation permettant dans un premier temps de traiter la plupart des zones du bâtiment et de son enveloppe à l'aide de l'approche nodale qui considère chacune des zones comme un volume parfaitement uniforme. Chaque zone est alors caractérisée par un seul noeud de calcul où sont déterminées les variables d'états (température, pression, concentration, etc.). Ensuite, il s'agit d'étudier les détails au sein d'un nombre limité de zones et d'estimer l'impact de ces détails sur le comportement global du bâtiment. Ainsi, nous proposons différentes méthodes de couplage entre d'une part la méthode nodale et d'autre part, les méthodes zonales et CFD. Après avoir présenté les différentes méthodes de modélisation retenues pour la détermination des transferts de masse et de chaleur dans les bâtiments, nous tentons de montrer l'intérêt d'utiliser l'une plutôt qu'une autre en fonction, des spécificités de chacune des zones, et de l'étude à effectuer. Ensuite, nous exposons la plate-forme de simulation développée, permettant de mettre en oeuvre aussi bien les approches nodales et zonales, que les différentes procédures de couplage dépendant du type d'association de modèles. Enfin, différentes applications viennent confirmer les capacités offertes par la plate-forme, pour moduler le niveau de finesse des modèles utilisés pour représenter chacune des zones d'un bâtiment, mais aussi pour proposer de nouvelles orientations de recherche. En effet, la dernière application présente une approche de couplage entre les approches zonale et CFD, dans laquelle la première tire la connaissance de la structure de l'écoulement de la seconde. L'outil développé apporte de nombreuses possibilités d'applications, aussi bien pour la caractérisation du bâtiment lui-même que de son intégration dans son environnement immédiat.

<p>Nano-powders of BaTiO₃, SrTiO₃, Ba_0.6Sr_0.4TiO₃, a mixture of the composition (BaTiO₃)_0.6(SrTiO₃)_0.4 with particle sizes in the range of 60 to 80 nm, and Bi₄Ti₃O₁₂ with an average particle size of 100 nm were consolidated by spark plasma sintering (SPS). The kinetics of reaction, densification and grain growth were studied. An experimental procedure is outlined that allows the determination of a “kinetic window” within which dense nano-sized compacts can be prepared. It is shown that the sintering behaviour of the five powders varies somewhat, but is generally speaking fairly similar. However, the types of grain growth behaviour of these powders are quite different, exemplified by the observation that the kinetic window for the (BaTiO₃)_0.6(SrTiO₃)_0.4 mixture is 125 ^oC, ~75 ^oC for Bi₄Ti₃O₁₂, ~25^oC for BaTiO₃ and SrTiO₃, while it is hard to observe an apparent kinetic window for obtaining nano-sized compacts of Ba_0.6Sr_0.4TiO₃. During the densification of the (BaTiO₃)_0.6(SrTiO₃)_0.4 mixture the reaction 0.6BaTiO₃+0.4SrTiO₃ → Ba_0.6Sr_0.4TiO₃ takes place, and this reaction is suggested to have a self-pinning effect on the grain growth, which in turn explains why this powder has a large kinetic window. Notably, SPS offers a unique opportunity to more preciously investigate and monitor the sintering kinetics of nano-powders, and it allows preparation of ceramics with tailored microstructures.</p><p>The dielectric properties of selected samples of (Ba, Sr)TiO₃ ceramics have been studied. The results are correlated with the microstructural features of these samples, <i>e.g.</i> to the grain sizes present in the compacts. The ceramic with nano-sized microstructure exhibits a diffuse transition in permittivity and reduced dielectric losses in the vicinity of the Curie temperature, whereas the more coarse-grained compacts exhibit normal dielectric properties in the ferroelectric region.</p><p>The morphology evolution, with increasing sintering temperature, of bismuth layer-structured ferroelectric ceramics such as Bi₄Ti₃O₁₂ (BIT) and CaBi₂Nb₂O₉ (CBNO) was investigated. The subsequent isothermal sintering experiments revealed that the nano-sized particles of the BIT precursor powder grew into elongated plate-like grains within a few minutes, via a dynamic ripening mechanism.</p><p>A new processing strategy for obtaining highly textured ceramics is described. It is based on a<i> directional dynamic ripening mechanism</i> <i>induced by superplastic deformation</i>. The new strategy makes it possible to produce a <i>textured</i> microstructure within minutes, and it allows production of textured ferroelectric ceramics with tailored morphology and improved physical properties.</p><p>The ferroelectric, dielectric, and piezoelectric properties of the textured bismuth layer-structured ferroelectric ceramics have been studied, and it was revealed that all textured samples exhibited anisotropic properties and improved performance. The highly textured Bi₄Ti₃O₁₂ ceramic exhibited ferroelectric properties equal to or better than those of corresponding single crystals, and much better than those previously reported for grain-orientated Bi₄Ti₃O₁₂ ceramics. Textured CaBi₂Nb₂O₉ ceramics exhibited a very high Curie temperature, <i>d</i><i>33</i>-values nearly three times larger than those of conventionally sintered materials, and a high thermal depoling temperature indicating that it is a very promising material for high-temperature piezoelectric applications.</p>

Cette étude a pour objet de montrer qu'il est possible de générer automatiquement des modèles zonaux pour l'étude du comportement thermique et aéraulique des bâtiments. Les modèles zonaux sont basés sur le partitionnement des pièces en un petit nombre de sous-volumes. Cette approche est intermédiaire entre celle des modèles à un noeud (qui considèrent que la température est homogène dans chaque pièce, et pour cette raison ne permettent pas de prédire le confort thermique dans une pièce) et celle des codes CFD (qui sont très coûteux en temps de calcul). Pour atteindre notre objectif, nous avons reformulé le modèle zonal. Ceci a consisté à regrouper les équations de description du comportement du bâtiment dans des sous-systèmes d'équations. Ce regroupement est calqué sur le découpage spatial des pièces. Ainsi, les équations de bilan et d'état appliquées à un sous-volumes forment les modules de la famille des « cellules » et celles de transfert entre deux sous-volumes forment les modules de la famille des « interfaces ». Ces familles sont constituées de plusieurs modèles correspondant aux différents types d'écoulement qui se développent dans les bâtiments. Ceux-ci ont été traduits en objets SPARK, lesquels forment la bibliothèque de modèles. Construire une simulation consiste à choisir les modèles appropriés pour décrire les pièces et à les connecter. Cette dernière étape a été automatisée, si bien qu'il ne reste plus à l'utilisateur qu'à donner le partitionnement et à choisir les modèles qu'il désire implémenter. Le système d'équations résultant est résolu par le solveur de SPARK. Des résultats de simulations pour différentes configurations d'écoulement dans des pièces sont présentés et comparés à des données expérimentales. Nous donnons également des exemples d'application de la méthode zonale à l'étude d'un groupe de deux pièces, d'un bâtiment et d'une pièce de géométrie complexe.

Thesis (Ph.D.)--University of Missouri-Columbia, 2006. / The entire dissertation/thesis text is included in the research.pdf file; the official abstract appears in the short.pdf file (which also appears in the research.pdf); a non-technical general description, or public abstract, appears in the public.pdf file. Title from title screen of research.pdf file viewed on (February 27, 2007) Vita. Includes bibliographical references.

Spark Plasma Sintering (SPS) is a promising rapid consolidation technique that allows a better understanding and manipulating of sintering kinetics and therefore makes it possible to obtain Si3N4-based ceramics with tailored microstructures, consisting of grains with either equiaxed or elongated morphology. The presence of an extra liquid phase is necessary for forming tough interlocking microstructures in Yb/Y-stabilised α-sialon by HP. The liquid is introduced by a new method, namely by increasing the O/N ratio in the general formula RExSi12-(3x+n)Al3x+nOnN16-n while keeping the cation ratios of RE, Si and Al constant. Monophasic α-sialon ceramics with tailored microstructures, consisting of either fine equiaxed or elongated grains, have been obtained by using SPS, whether or not such an extra liquid phase is involved. The three processes, namely densification, phase transformation and grain growth, which usually occur simultaneously during conventional HP consolidation of Si3N4-based ceramics, have been precisely followed and separately investigated in the SPS process. The enhanced densification is attributed to the non-equilibrium nature of the liquid phase formed during heating. The dominating mechanism during densification is the enhanced grain boundary sliding accompanied by diffusion- and/or reaction-controlled processes. The rapid grain growth is ascribed to a dynamic ripening mechanism based on the formation of a liquid phase that is grossly out of equilibrium, which in turn generates an extra chemical driving force for mass transfer. Monophasic α-sialon ceramics with interlocking microstructures exhibit improved damage tolerance. Y/Yb- stabilised monophasic α-sialon ceramics containing approximately 3 vol% liquid with refined interlocking microstructures have excellent thermal-shock resistance, comparable to the best β-sialon ceramics with 20 vol% additional liquid phase prepared by HP. The obtained sialon ceramics with fine-grained microstructure show formidably improved superplasticity in the presence of an electric field. The compressive strain rate reaches the order of 10-2 s-1 at temperatures above 1500oC, that is, two orders of magnitude higher than that has been realised so far by any other conventional approaches. The high deformation rate recorded in this work opens up possibilities for making ceramic components with complex shapes through super-plastic forming.