Three different topics of turbulence research that cover modeling, theory and model computation categories are selected and studied in depth. In the first topic, "velocity gradient dynamics in turbulence" (modeling), the Lagrangian linear diffusion model that accounts for the viscous-effect is proposed to make the existing restricted-Euler velocity gradient dynamics model quantitatively useful. Results show good agreement with DNS data. In the second topic, "pressure-strain correlation in homogeneous anisotropic turbulence subject to rapid strain-dominated distortion" (theory), extensive rapid distortion calculation is performed for various anisotropic initial turbulence conditions in strain-dominated mean flows. The behavior of the rapid pressure-strain correlation is investigated and constraining criteria for the rapid pressure-strain correlation models are developed. In the last topic, "unsteady computation of turbulent flow past a square cylinder using partially-averaged Navier-Stokes method" (model computation), the basic philosophy of the PANS method is reviewed and a practical problem of flow past a square cylinder is computed for various levels of physical resolution. It is revealed that the PANS method can capture many important unsteady flow features at an affordable computational effort.

This thesis aims to increase our understanding of mechanisms that influence larval dispersal in marine benthic invertebrates, particularly in the absence of strong oceanographic features (e.g. estuarine plumes, upwelling events, or markedly different water masses). Laboratory experiments identified behavioural mechanisms that regulate the vertical distribution of larvae in response to thermal stratification, and field studies in St. George’s Bay, Nova Scotia (NS), Canada, examined the relationship between larval abundance and physical variables (temperature, salinity, fluorescence, etc) and identified mechanisms that regulate larval distributions in situ. In the laboratory, I demonstrated that thermal stratification affects the vertical distribution of larvae by acting as a barrier to migration, or through temperature-dependent vertical swimming velocities. I also developed a random walk based model which highlighted that the key to successfully simulating larval response to temperature was 1) determining the temperature-dependent distribution of vertical swimming velocities and 2) the temporal autocorrelation in these velocities. In the field, the most striking pattern was that the larval distributions for species with similar swimming abilities were significantly correlated to one another at all scales (0.5 to 40 km). This suggests a common mechanism, related to larval swimming ability, which greatly influences the horizontal larval distribution. I found that the spatial scale of variability in larval distributions (~ 3 km) matches that in both the environmental variables and of coherent structures in current velocities (i.e. the tidal excursion). Results from an aggregation-diffusion model suggest that horizontal larval swimming could not be responsible for the observed level of aggregation in the larval horizontal distributions. I suggest that these horizontal patterns are the result of 1) an aggregative process (i.e. larvae swimming against a vertical current and maintaining their vertical position) and 2) a diffusive process which scales the aggregations to the scale of the coherent structures in current velocity (i.e. tidal excursion). In conclusion, this thesis increases our understanding of larval behaviour and its effects on larval dispersal. The results will be particularly useful to those who are interested in mechanisms regulate population connectivity, particularly those using bio-physical models to model dispersal trajectories.

Made available in DSpace on 2016-12-23T13:50:03Z (GMT). No. of bitstreams: 1 Michelle Carvalho de Souza.pdf: 1066706 bytes, checksum: 3c78ec62d0a1683cf0df485633effc39 (MD5) Previous issue date: 2013-08-21 / O yacon é um tubérculo de origem andina, e seu consumo tem se disseminado pelos demais países devido as suas características de alimento funcional. Os benefícios advindos do yacon são devido à presença dos FOS (frutooligossacarídeos), que são carboidratos não digeríveis pelo organismo humano, que atuam como fibra e são fermentados pelas bactérias intestinais, produzindo compostos benéficos à saúde. Devido ao alto teor de água, média de 90%, o yacon é suscetível a uma rápida degradação e a uma vida útil de aproximadamente sete dias. A secagem é uma alternativa para aumentar o período de consumo do yacon, além disso, a farinha de yacon pode ser utilizada como ingrediente em diversos produtos industrializados, entretanto, não há uma padronização desse processo para a produção da farinha de yacon em larga escala, o que dificulta a inserção desse produto na agroindústria. Para o estudo do processo de secagem do yacon e obtenção de farinhas que preservassem as características mais próximas as dos produtos in natura, foram utilizadas cinco temperaturas de secagem (40, 50, 60, 70 e 80º C). O modelo matemático que melhor se ajustou ao experimento foi o modelo de Page, e a partir disso, pode-se elaborar um modelo matemático generalizado que representasse a secagem do yacon. Estimou-se a difusividade efetiva do yacon nas diferentes temperaturas testadas através do modelo de difusão, além disso, a influência da temperatura de secagem sobre a difusividade efetiva foi avaliada ajustando-se o modelo de Arrhenius. A energia de ativação para a difusão líquida no processo de secagem do yacon foi de 22,088 kJ/mol. O yacon atingiu sua umidade de equilíbrio a 60º C próximo a 400 minutos, o que também foi observado nas temperaturas mais altas. O teor de fibra bruta das farinhas não variou com o aumento da temperatura de secagem, entretanto, o maior valor de açúcar redutor foi obtido a 80º C, enquanto que a 50º C encontrou-se o maior valor de açúcar não redutor. Isso pode ser devido a hidrólise dos FOS, consequência do efeito térmico. A farinha elaborada a 60º C possui baixa umidade, além de preservar suas características intrínsecas mais próximas as do produto in natura / The yacon tuber is an Andean origin, and its consumption has been disseminated by other countries due to its characteristics of functional food. The benefits of yacon are due to the presence of FOS (fructo-oligosaccharides), which are non-digestible carbohydrates by the human body, which act as fiber and are fermented by intestinal bacteria, producing compounds beneficial to health. Due to the high water content, 90% average, the yacon is susceptible to rapid degradation and a lifespan of about seven days. The drying is an alternative to increasing the consumption period of yacon, moreover, yacon flour can be used as an ingredient in various manufactured products. To study the drying of yacon, and getting flour which preserved the characteristics as close as possible to the fresh products, this study tested five drying temperatures (40, 50, 60, 70 and 80 º C). The mathematical model that best fit the experiment was the Page model, and from this, one can develop a generalized mathematical model to represent the drying of yacon. We estimated the effective diffusivity of yacon tested at different temperatures by diffusion model, moreover, the influence of the drying temperature on the effective diffusivity was measured by adjusting the Arrhenius model. The activation energy for liquid diffusion in the drying process of yacon was 22.088 kJ / mol. At 60 ° C the yacon reached its equilibrium moisture content close to 400 minutes, which was also observed at higher temperatures. The crude fiber content of the flour did not vary with increasing drying temperature, however, the largest amount of reducing sugar was obtained at 80 ° C, while at 50 ° C was found the highest amount of non-reducing sugar. This may be due to hydrolysis of FOS result of the thermal effect. Thus, the flour produced at 60 º C has low humidity, while preserving its merits closer the product fresh

The purpose of this thesis is to compare option pricing models. We have investigated the constant volatility models Black-Scholes-Merton (BSM) and Merton’s Jump Diffusion (MJD) as well as the stochastic volatility models Heston and Bates. The data used were option prices from Microsoft, Advanced Micro Devices Inc, Walt Disney Company, and the S&P 500 index. The data was then divided into training and testing sets, where the training data was used for parameter calibration for each model, and the testing data was used for testing the model prices against prices observed on the market. Calibration of the parameters for each model were carried out using the nonlinear least-squares method. By using the calibrated parameters the price was calculated using the method of Carr and Madan. Generally it was found that the stochastic volatility models, Heston and Bates, replicated the market option prices better than both the constant volatility models, MJD and BSM for most data sets. The mean average relative percentage error for Heston and Bates was found to be 2.26% and 2.17%, respectively. Merton and BSM had a mean average relative percentage error of 6.90% and 5.45%, respectively. We therefore suggest that a stochastic volatility model is to be preferred over a constant volatility model for pricing options. / Syftet med denna tes är att jämföra prissättningsmodeller för optioner. Vi har undersökt de konstanta volatilitetsmodellerna Black-Scholes-Merton (BSM) och Merton’s Jump Diffusion (MJD) samt de stokastiska volatilitetsmodellerna Heston och Bates. Datat vi använt är optionspriser från Microsoft, Advanced Micro Devices Inc, Walt Disney Company och S&P 500 indexet. Datat delades upp i en träningsmängd och en test- mängd. Träningsdatat användes för parameterkalibrering med hänsyn till varje modell. Testdatat användes för att jämföra modellpriser med priser som observerats på mark- naden. Parameterkalibreringen för varje modell utfördes genom att använda den icke- linjära minsta-kvadratmetoden. Med hjälp av de kalibrerade parametrarna kunde priset räknas ut genom att använda Carr och Madan-metoden. Vi kunde se att de stokastiska volatilitetsmodellerna, Heston och Bates, replikerade marknadens optionspriser bättre än båda de konstanta volatilitetsmodellerna, MJD och BSM för de flesta dataseten. Medelvärdet av det relativa medelvärdesfelet i procent för Heston och Bates beräknades till 2.26% respektive 2.17%. För Merton och BSM beräknades medelvärdet av det relativa medelvärdesfelet i procent till 6.90% respektive 5.45%. Vi anser därför att en stokastisk volatilitetsmodell är att föredra framför en konstant volatilitetsmodell för att prissätta optioner.

In the beginning we outline some important properties of blood and de- scribe it from the biological point of view. In the next section we show how we derived our model based on the mixture theory. For the final model we suggest a mathematical method based on the finite element method and subject it to tests for flow in a simple domain. In the middle part we prove the existence of solution for a model with simplified constitutive relation for the stress tensor, which still includes an anisotropic model for the platelet diffusion. In the last section we show numerical results. We start with sim- ple testing computations in simple domains, followed by computations in a two-dimensional simulation of an aneurysm, and narrowed blood vessel re- spectively. In the end we also show some illustrative computations in three dimensions. 1

Understanding how innovations are accepted in a dynamic and complex market environment is a crucial factor for competitive advantage. To understand the relevant factors for this diffusion and to predict success, empirically grounded agent-based models have become increasingly popular in recent years. Despite the popularity of these innovation diffusion models, no common framework that integrates their diversity exists. This article presents a flexible, modular and extensible common description and implementation framework that allows to depict the large variety of model components found in existing models. The framework aims to provide a theoretically grounded description and implementation framework for empirically grounded agent-based models of innovation diffusion. It identifies 30 component requirements to conceptualize an integrated formal framework description. Based on this formal description, a java-based implementation allowing for flexible configuration of existing and future models of innovation diffusion is developed. As a variable decision support tool in decision-making processes on the adoption of innovations the framework is valuable for the investigation of a range of research questions on innovation diffusion, business model evaluation and infrastructure transformation.

No description available.

Les méthodes d’apprentissage profond connaissent une croissance fulgurante. Une explication de ce phénomène est l’essor de la puissance de calcul combiné à l’accessibilité de données en grande quantité. Néanmoins, plusieurs applications de la vie réelle présentent des difficultés: la disponibilité et la qualité des données peuvent être faibles, l’étiquetage des données peut être ardu, etc. Dans ce mémoire, nous examinons deux contextes : celui des données limitées et celui du modèle économique CATS. Pour pallier les difficultés rencontrées dans ces contextes, nous utilisons des modèles d’apprentissage profond pour images avec structure additionnelle. Dans un premier temps, nous examinons les réseaux de scattering et étudions leur version paramétrée sur des petits jeux de données. Dans un second temps, nous adaptons les modèles de diffusion afin de proposer une alternative aux modèles à base d’agents qui sont complexes à construire et à optimiser. Nous vérifions empiriquement la faisabilité de cette démarche en modélisant le marché de l’emploi du modèle CATS. Nous constatons tout d’abord que les réseaux de scattering paramétrés sont performants sur des jeux de données de classification pour des petits échantillons de données. Nous démontrons que les réseaux de scattering paramétrés performent mieux que ceux non paramétrés, c’est-à-dire les réseaux de scattering traditionnels. En effet, nous constatons que des banques de filtres adaptés aux jeux de données permettent d’améliorer l’apprentissage. En outre, nous observons que les filtres appris se différencient selon les jeux de données. Nous vérifions également la propriété de robustesse aux petites déformations lisses expérimentalement. Ensuite, nous confirmons que les modèles de diffusion peuvent être adaptés pour modéliser le marché de l’emploi du modèle CATS dans une approche d’apprentissage profond. Nous vérifions ce fait pour deux architectures de réseau de neurones différentes. De plus, nous constatons que les performances sont maintenues pour différents scénarios impliquant l’apprentissage avec une ou plusieurs séries temporelles issues de CATS, lesquelles peuvent être tirées à partir d’hyperparamètres standards ou de perturbations de ceux-ci. / Deep learning methods are booming. An explanation of this phenomenon is the rise of computing power combined with the accessibility of large data quantity. Nevertheless, several real-life applications present difficulties: the availability and quality of data can be low, data labeling can be tricky, etc. In this thesis, we examine two contexts: that of limited data and that of the CATS economic model. To overcome the difficulties encountered in these contexts, we use deep learning models for images with additional structure. First, we examine scattering networks and study their parameterized version on small datasets. In a second step, we adapt diffusion models in order to propose an alternative to agent-based models which are complex to build and to optimize. We empirically verify the feasibility of this approach by modeling the labor market of the CATS model. We first observe that the parameterized scattering networks perform well on classification datasets for small samples of data. We demonstrate that parameterized scattering networks perform better than those not parametrized, i.e. traditional scattering networks. Indeed, we find that filterbanks adapted to the datasets make it possible to improve learning. Moreover, we observe that the learned filters differ according to the datasets. We also verify the property of robustness to small smooth deformations experimentally.. Then, we confirm that diffusion models can be adapted to model the labor market of the CATS model in a deep learning approach. We verify this fact for two different neural network architectures. Moreover, we find that performance is maintained for different scenarios involving training with one or more time series from CATS, which can be derived from standard hyperparameters or perturbations thereof.

Forecasting of highway traffic is a common practice for real traffic information system, and is of vital importance to traffic management and control on highways. As a typical time-series forecasting task, we want to propose a deep learning model to map the historical sensory traffic values (e.g., speed, flow) to future traffic forecasts. Prevailing traffic forecasting methods focus on the graph representation of the urban road. However, compared to the dense connectivity of urban road networks, highway traffic flows normally run on road segments of serial topology. This indicates that the highway traffic flows do not have the same type of spatial interaction, therefore motivating us to resort to a new forecasting paradigm. While traffic patterns can be intuitively represented by spatial-temporal (ST) images, this study transforms the traffic forecasting task into the conditional image generation task. Our approach explores the inherent properties of ST-images from the perspectives of physical meaning and traffic dynamics. An innovative deep learning based architecture is designed to process the ST-image, and a diffusion model is trained to obtain traffic forecasts by generating future ST-image based on the historical STimages. We demonstrate the effectiveness of the architecture in processing ST-image through ablation studies and the effectiveness of the model through comparison with popular baseline models, i.e., LSTM and T-GCN. / Prognos av vägtrafik är en vanlig praxis för riktiga trafikinformationssystem och är av vital betydelse för trafikhantering och kontroll på motorvägar. Som en typisk tidsserieförutsägelseuppgift vill vi föreslå en djupinlärningsmodell för att kartlägga historiska sensoriska trafikvärden (t.ex. hastighet, flöde) till framtida trafikprognoser. Rådande trafikprognosmetoder fokuserar på grafrepresentationen av stadsvägar. Jämfört med den täta anslutningen av stadsvägnät, löper motorvägstrafik normalt på vägsegment med seriell topologi. Detta indikerar att motorvägstrafikflöden inte har samma typ av rumslig interaktion, vilket motiverar oss att använda en ny prognosparadigm. Medan trafikmönster intuitivt kan representeras av spatial-temporala (ST) bilder, omvandlar denna studie trafikprognosuppgiften till en uppgift för betingad bildgenerering. Vår metod utforskar de inneboende egenskaperna hos ST-bilder från perspektiven fysisk betydelse och trafikdynamik. En innovativ djupinlärningsbaserad arkitektur är utformad för att behandla STbilden, och en diffusionsmodell tränas för att erhålla trafikprognoser genom att generera framtida ST-bilder baserat på historiska ST-bilder. Vi demonstrerar effektiviteten hos arkitekturen genom avbränningsstudier och modellens effektivitet genom jämförelse med populära baslinjemodeller, dvs. LSTM och T-GCN.

No description available.