[pt] Neste trabalho, um - software - foi desenvolvido para prover predições do sinal em ambientes exteriores e interiores. O programa inclui dois diferentes modelos de propagação. Para situações urbanas com antenas baixas (Microcélulas) e para situações urbanas onde o receptor pode ser posicionado numa área - indoor - e o transmissor numa área urbana ( - outdoor -). O procedimento automatizado combina técnicas de várias diferentes disciplinas, incluindo computação gráfica e teoria eletromagnética assintótica, para reduzir o tempo de computação sem comprometer a precisão. Usando técnicas de computação gráfica, o tempo necessário para o lançamento de raios através da geometria urbana é grandemente reduzido. Os dados teóricos foram testados e validados por comparação com dados experimentais. Para modelar os diferentes mecanismos de propagação, a aproximação eletromagnética usada foi a OG/UTD. / [en] In this work, a software tool has been developed that provides accurate, sitespecific radio signal predictions for outdoor to indoor/outdoor locations. The program includes two different propagation models. For urban calculations with low antennas (Microcells) and for urban situations where receiver can be located at indoor area and transmitter antennas can be located at Urban area (outdoor). The software combines techniques from several different disciplines, including computer graphics and asymptotic electromagnetic theory, to reduce the calculation time without compromising accuracy. Using computer graphics techniques, the time necessary for shooting the rays through the urban geometry is greatly reduced. The calculation engine used has been tested and validated by comparison with measured data. In order to model the diferent propagation mechanisms, the eletromagnetic approach considered has been the GO/UTD combination.

This doctoral thesis concerns the designs, creation, and a pilot survey of real remotely controlled laboratories that can be used for the introduction to modern physics within the integrated e-learning strategy. With grant support four new remote laboratory experiments could be created, featuring some innovations like the possibility to modify experimental setup, and the design allowing to make a mistake and to obtain nonsensical experimental data in order to test students' critical thinking. The new remote experiment topics are: laws of radioactivity, polarisation state of light, photoelectric effect, and the origin of spectral lines. The emphasis is put on easy experimental data acquision (measurement, record, and download) for further scientific statistical or advanced graphical data proces- sing. The pilot survey aims to select adequate research designs, tools, and data sources and it describes the results and experience from the observation at several Czech secondary schools. The most interesting findings were revealed by action log file analysis and interviews with teachers and students. Finally, real usability of remote laboratory experiments in lessons and physics education is discussed. 1

The knowledge of phase equilibrium is essential for the planning and realisation of separation processes in chemical engineering. In this work an equipment for measurement of precise isothermal vapour–liquid equilibria (VLE) using the dynamic method was developed. The pool of experimental data for cyclohexylamine was extended significantly. Isothermal VLE were measured in 3 binary and 4 ternary systems, liquid-liquid equilibria (LLE) were measured in 4 ternary systems and in one quarternary system, in each case for two temperatures, whereas in 2 ternary systems and in the quarternary system even liquid-liquid-liquid equilibria (LLLE) occur. Furthermore, activity coefficients at infinite dilution in 4 binary systems and excess molar volumes in 7 binary systems have been estimated. Binary VLE and LLE data of the components water, octane, cyclohexylamine and aniline of this work and data from literature were fitted with the activity coefficient models NRTL and UNIQUAC, as well as with the equations of state Elliott-Suresh-Donohue (ESD) and Perturbed-Chain-Statistical Associating Fluid Theory (PC-SAFT) which contain both a term to consider explicit hydrogen bonds. In addition, the predictive capabilities of the equations of state (EoS) were investigated. With parameters obtained by simultaneous fitting of VLE and, if available, LLE data similar results with the models NRTL and UNIQUAC could be obtained. Each time the deviations for the vapour pressure were lower than 3 % and lower than 2 % in vapour phase composition. The deviations, in three out of the six systems for vapour pressure and vapour phase composition, were larger with the ESD-EoS than with the activity coefficient models. NRTL, UNIQUAC and ESD delivered similar results with the simultaneously fitted parameters for the LLE, whereas the deviations were lower than 5 %. Comparable results were delivered by the ESD-EoS and PC-SAFT for the fitting and the prediction in the investigated binary systems. Also a fitting for NRTL, UNIQUAC and ESD was carried out, but only to one data set. The intention was to show the effect of parameterization on prediction in ternary systems. Predictions were made for VLE and LLE in ternary systems of the above mentioned components, solely with interaction parameters fitted to binary data. For the models NRTL, UNIQUAC und ESD predictions of simultaneously and separately fitted parameters are presented. It is shown that with parameters simultaneously fitted to several data sets significantly better results could be obtained compared to the parameters separately fitted to a single data set. Additionally, for the equations of state ESD and PC-SAFT predictions for the LL(L)E in ternary systems are compared, but here only with separately fitted parameters. For three out of the four investigated ternary systems a too large miscibility gap is calculated with the models NRTL, UNIQUAC and ESD. In the system water+octane+aniline good results could be obtained for the prediction of the LLLE. In summary the equations of state deliver similar results. In the systems water+octane+CHA and octane+CHA+aniline also too large two phase regions were delivered. Better predictions could be obtained in the systems water+octane+aniline and water+CHA+aniline. The forecasts of the VLE in the ternary systems are good with the simultaneously fitted parameters. The deviations for the vapour phase compositions are as for the vapour pressures under 6 %. Larger deviations occur for the system water+octane+aniline only. As evaluation result for the thermodynamic models can be mentioned that the activity coefficient models NRTL and UNIQUAC deliver somewhat better results for the fitting of the binary data than the equations of state ESD and PC-SAFT however, with a larger number of adjustable parameters. The prediction of the VLE is satisfactorily in the ternary systems and with similar quality of all considered models. Larger deviations occur for the prediction of the LL(L)E in the ternary systems. The results of the ESD-EoS were, with one exception, each time better than those of the activity coefficient models. There is no significant difference between the prediction of the ternary systems for the ESD-EoS and the PC-SAFT.

Testing and analysis of measurement data is a major part of the workday for development engineers within the section Fluid and Emission Treatment to assure the quality of products developed by Scania CV. The ineluctable variation and therefore uncertainty of results can be quantified and discussed with the use of statistical methods. The overall objective of this study is to examine how the competence of development engineers can be improved within the area of statistics. This has been done by relating the concept of competence with the use of statistical methods to analyse experimental data and by investigating which factors that influence the competence of the engineers. Observations of the daily operations, interviews and a focus group with development engineers and a workshop with management personnel, has been used as research methods in this study. The results show that the desired competence, from the perspective of the individual, consists of knowledge, motivation and support. It is important that none of these parts are neglected when the competence of engineers is to be strengthened. The demand of statistical analysis by management and decision makers is a factor that has been shown to have great influence on the competence of development engineers. The competence of development engineers is also influenced by which tools and other types of support the organisation offers. The conclusion is that competence development is more comprehensive than a one-time-effort, as for example a short course, since the competence needs to be maintained and develop as the business evolves. A continuous support from the organisation would probably contribute to a more active development of competence. Knowledge sharing between colleagues on internal digital platforms has great potential and could be a component of a continuous support for the development engineers. / Provtagningar och analys av mätdata är en stor del av utvecklingsingenjörernas arbetsvardag inom sektionen Fluid and Emission Treatment för att säkerställa kvalitén på de produkter som Scania CV AB utvecklar. Den ofrånkomliga variationen och därmed osäkerheten i resultat kan kvantifieras och diskuteras med statistiska metoder. Syftet med denna studie är att undersöka hur utvecklingsingenjörernas kompetens inom statistikämnet kan förstärkas. Detta har gjorts genom att relatera kompetensbegreppet till användningen av statistiska metoder för att analysera experimentella data samt utreda vilka faktorer som påverkar ingenjörens kompetens. Observationer av verksamheten, intervjuer och en fokusgrupp med utvecklingsingenjörer samt en workshop med personer i ledande befattning har använts som forskningsmetod i studien. Resultatet visar att den önskade kompetensen, ur individens perspektiv, består av tre delar, kunskap, motivation och stöd. Det är viktigt att ingen av dessa delar bortses ifrån när ingenjörernas kompetens ska utvecklas. Efterfrågan av statistiska analyser från chefer och beslutsfattare är en faktor som har visat sig ha stor påverkan på utvecklingsingenjörernas kompetens. Vilka hjälpmedel och övriga stöd som erbjuds från organisationen har även stor inverkan på kompetensen. Slutsatsen är att kompetensutvecklingen är mer omfattande än en punktinsats, som exempelvis en kort kurs, eftersom kompetensen måste underhållas samt utvecklas vartefter verksamheten utvecklas. Ett kontinuerligt stöd från organisationen bidrar sannolikt till en mer aktiv kompetensutveckling. Kunskapsdelning mellan kollegor via interna digitala plattformar har stor potential och skulle kunna utgöra en del av ett kontinuerligt stöd för utvecklingsingenjörerna.

With advanced technology and information explosion, the data of interest often have complex structures, with the large size and dimensions in the form of continuous or discrete features. There is an emerging need for data reduction, efficient modeling, and model inference. For example, data can contain millions of observations with thousands of features. Traditional methods, such as linear regression or LASSO regression, cannot effectively deal with such a large dataset directly. This dissertation aims to develop several techniques to effectively analyze large datasets with complex structures in the observational, experimental and time series data. In Chapter 2, I focus on the data reduction for model estimation of sparse regression. The commonly-used subdata selection method often considers sampling or feature screening. Un- der the case of data with both large number of observation and predictors, we proposed a filtering approach for model estimation (FAME) to reduce both the size of data points and features. The proposed algorithm can be easily extended for data with discrete response or discrete predictors. Through simulations and case studies, the proposed method provides a good performance for parameter estimation with efficient computation. In Chapter 3, I focus on modeling the experimental data with quantitative-sequence (QS) factor. Here the QS factor concerns both quantities and sequence orders of several compo- nents in the experiment. Existing methods usually can only focus on the sequence orders or quantities of the multiple components. To fill this gap, we propose a QS transformation to transform the QS factor to a generalized permutation matrix, and consequently develop a simple Gaussian process approach to model the experimental data with QS factors. In Chapter 4, I focus on forecasting multivariate time series data by leveraging the au- toregression and clustering. Existing time series forecasting method treat each series data independently and ignore their inherent correlation. To fill this gap, I proposed a clustering based on autoregression and control the sparsity of the transition matrix estimation by adap- tive lasso and clustering coefficient. The clustering-based cross prediction can outperforms the conventional time series forecasting methods. Moreover, the the clustering result can also enhance the forecasting accuracy of other forecasting methods. The proposed method can be applied on practical data, such as stock forecasting, topic trend detection. / Doctor of Philosophy / This dissertation focuses on three projects that are related to data reduction and statistical modeling of data with complex structures. In chapter 2, we propose a filtering approach of data for parameter estimation of sparse regression. Given data with thousands of ob- servations and predictors or even more, large storage and computation spaces is need to handle these data. It is challenging to computational power and takes long time in terms of computational cost. So we come up with an algorithm (FAME) that can reduce both the number of observations and predictors. After data reduction, this subdata selected by FAME keeps most information of the original dataset in terms of parameter estimation. Compare with existing methods, the dimension of the subdata generated by the proposed algorithm is smaller while the computational time does not increase. In chapter 3, we use quantitative-sequence (QS) factor to describe experimental data. One simple example of experimental data is milk tea. Adding 1 cup of milk first or adding 2 cup of tea first will influence the flavor. And this case can be extended to cases when there are thousands of ingredients need to be input into the experiment. Then the order and amount of ingredients will generate different experimental results. We use QS factor to describe this kind of order and amount. Then by transforming the QS factor to a matrix containing continuous value and set this matrix as input, we model the experimental results with a simple Gaussian process. In chapter 4, we propose an autoregression-based clustering and forecasting method of multi- variate time series data. Existing research works often treat each time series independently. Our approach incorporates the inherent correlation of data and cluster related series into one group. The forecasting is built based on each cluster and data within one cluster can cross predict each other. One application of this method is on topic trending detection. With thousands of topics, it is unfeasible to apply one model for forecasting all time series. Considering the similarity of trends among related topics, the proposed method can cluster topics based on their similarity, and then perform forecasting in autoregression model based on historical data within each cluster.

The wheel-rail contact mechanics and dynamics that are of great importance to the railroad industry are evaluated by applying statistical methods to the large volume of data that is collected on the VT-FRA state-of-the-art roller rig. The intent is to use the statistical principles to highlight the relative importance of various factors that exist in practice to longitudinal and lateral tractions and to develop parametric models that can be used for predicting traction in conditions beyond those tested on the rig. The experiment-based models are intended to be an alternative to the classical traction-creepage models that have been available for decades. Various experiments are conducted in different settings on the VT-FRA Roller Rig at the Center for Vehicle Systems and Safety at Virginia Tech to study the relationship between the traction forces and the wheel-rail contact variables. The experimental data is used to entertain parametric and non-parametric statistical models that efficiently capture this relationship. The study starts with single regression models and investigates the main effects of wheel load, creepage, and the angle of attack on the longitudinal and lateral traction forces. The assumptions of the classical linear regression model are carefully assessed and, in the case of non-linearities, different transformations are applied to the explanatory variables to find the closest functional form that captures the relationship between the response and the explanatory variables. The analysis is then extended to multiple models in which interaction among the explanatory variables is evaluated using model selection approaches. The developed models are then compared with their non-parametric counterparts, such as support vector regression, in terms of "goodness of fit," out-of-sample performance, and the distribution of predictions. / Master of Science / The interaction between the wheel and rail plays an important role in the dynamic behavior of railway vehicles. The wheel-rail contact has been extensively studied through analytical models, and measuring the contact forces is among the most important outcomes of such models. However, these models typically fall short when it comes to addressing the practical problems at hand. With the development of a high-precision test rig—called the VT-FRA Roller Rig, at the Center for Vehicle Systems and Safety (CVeSS)—there is an increased opportunity to tackle the same problems from an entirely different perspective, i.e. through statistical modeling of experimental data. Various experiments are conducted in different settings that represent railroad operating conditions on the VT-FRA Roller Rig, in order to study the relationship between wheel-rail traction and the variables affecting such forces. The experimental data is used to develop parametric and non-parametric statistical models that efficiently capture this relationship. The study starts with single regression models and investigates the main effects of wheel load, creepage, and the angle of attack on the longitudinal and lateral traction forces. The analysis is then extended to multiple models, and the existence of interactions among the explanatory variables is examined using model selection approaches. The developed models are then compared with their non-parametric counterparts, such as support vector regression, in terms of "goodness of fit," out-of-sample performance, and the distribution of the predictions. The study develops regression models that are able to accurately explain the relationship between traction forces, wheel load, creepage, and the angle of attack.

Thesis (DLitt (African Languages))--University of Stellenbosch, 2008. / ENGLISH ABSTRACT: This dissertation is an experimental investigation of vowel duration. It focuses on phonetic and phonological aspects of vowel duration in Civili, and African language spoken in Gabon and some of its neighboring countries. It attempts to bring new insights into the phenomenon of vowel lengthening, and to assess the implictions of these insights for standardizing the orthography of this language. ...

Earthquake engineering is increasingly focusing on large international collaborations to address complex problems. Recent computing advances have greatly contributed to the way scientific collaborations are conducted, where web-based solutions are an emerging trend to manage and present results to the scientific community and the general public. However, collaborations in earthquake engineering lack a common interoperability framework, resulting in tedious and complex processes to integrate results, which cannot be efficiently used by third-party institutions. The work described in this thesis applies novel computing techniques to enable the interoperability of earthquake engineering resources, by integrating data, distributed simulation services and laboratory facilities. This integration focuses on distributed approaches rather than centralised solutions, and has been materialised in a platform called Celestina, that supports the integration of hazard mitigation resources. The prototype of Celestina has been implemented and validated within the context of two of the current largest earthquake engineering networks, the SERIES network in Europe and the NEES network in the USA. It has been divided into three sub-systems to address different problems: (i) Celestina Data, to develop best methods to define, store, integrate and share earthquake engineering experimental data. Celestina Data uses a novel approach based on Semantic Web technologies, and it has accomplished the first data integration between earthquake engineering institutions from the United States and Europe by means of a formalised infrastructure. (ii) Celestina Tools, to research applications that can be implemented on top of the data integration, in order to provide a practical benefit for the end user. (iii) Celestina Simulations, to create the most efficient methods to integrate distributed testing software and to support the planning, definition and execution of the experimental workflow from a high-level perspective. Celestina Simulations has been implemented and validated by conducting distributed simulations between the Universities of Oxford and Kassel. Such validation has demonstrated the feasibility to conduct both flexible, general-purpose and high performance simulations under the framework. Celestina has enabled global analysis of data requirements for the whole community, the definition of global policies for data authorship, curation and preservation, more efficient use of efforts and funding, more accurate decision support systems and more efficient sharing and evaluation of data results in scientific articles.

[pt] Uma preocupação constante referente às operações de reinício de escoamento em poços produtores de óleo é a previsibilidade da mínima pressão necessária para iniciar o fluxo após uma parada de produção. Tal pressão se refere àquela que promove tensões de cisalhamento junto à parede do duto que superam o valor da tensão limite de escoamento do referido fluido. A necessidade de desenvolvimento de diferentes tipos de campos de petróleo ao redor do mundo tem promovido discussões sobre importantes aspectos de garantia de escoamento dos fluidos, especialmente em cenários de produção em águas ultra-profundas. Nesses casos, quando o fluido apresenta valores elevados para a temperatura mínima de aparecimento de cristais (TIAC) ou ainda para a tensão limite de escoamento, o procedimento de repartida de poço após uma parada de produção pode representar um problema. Em geral, os modelos tixotrópicos utilizados na avaliação de reinício do escoamento são definidos por uma equação constitutiva baseada no modelo de Bingham, na qual a tensão de cisalhamento depende do grau de estruturação do fluido no interior do duto, representado por um parâmetro adimensional positivo (lambda). A evolução deste parâmetro no tempo é governada por uma equação constitutiva que considera um termo de construção e um termo de quebra do gel. Este trabalho tem como principal objetivo avaliar o comportamento do reinício da produção através da variação destes parâmetros/termos segundo modelos tixotrópicos específicos (Houska e SMT) e a representatividade dos resultados de simulação quando comparados aos dados experimentais. / [en] The main concern regarding to the restart operations for production wells in oil and gas systems is the forecast of minimum pressure needed to overcome the gel strength, i.e. the pressure which generates a wall stress higher than the yield stress of the gelled oil. The petroleum accumulations around the world have raised big issues regarding flow assurance aspects, especially in some ultra-deep water scenarios. In these cases, when the oil presents unusual high values for the Wax Appearance Temperature (WAT) and even for the yield stress, the flow restart procedure can be an issue, after shutdown times of the oil production. Usually, the thixotropic models applied for start-up flow are defined by a Bingham-like stress equation whose yield stress depends on a structure parameter (lambda), following Houska assumptions. This parameter is a non-negative scalar number that represents the structuring level of the material inside the pipeline. The structure parameter is governed by an evolution equation that considers a build-up term and a breakdown term. The main objective of this work is to evaluate the relationship between the parameters of the constitutive equations for Houska and SMT models (dimensionless coefficients of the evolution equations for the structure parameter) and the representativeness of the simulation results obtained by these models, regarding times for the production stabilization after a restart, delay times and yield stresses, when compared to experimental data.

No presente trabalho quatro ligas Al-Si hipereutéticas, três das quais foram produzidas por conformação por \"spray\" e a outra por fundição, foram caracterizadas quanto a textura, dureza, microestrutura e resistência à corrosão em meio de condensado sintético automotivo (CSA). Duas das ligas conformadas por \"spray\" foram retiradas de camisas de cilindro e a outra de um pré-formado obtido em laboratório. A conformação por \"spray\" envolve a atomização de uma liga e a deposição de gotículas em um substrato, antes mesmo que todas se encontrem no estado sólido. Este processo permite a obtenção de materiais que se caracterizam por uma microestrutura livre de macrossegregações e bastante refinada, implicando melhor trabalhabilidade a quente. A caracterização da microestrutura das quatro ligas hipereutéticas revelou a presença de porosidades na liga obtida em laboratório, e microestrutura com distribuição homogênea de precipitados primários nas três ligas conformadas por \"spray\". A microestrutura de uma das ligas apresentou-se muito diferenciada, com a presença de eutético, sugerindo que esta foi fabricada por fundição. Nas camisas de cilindro foram feitas medidas de rugosidade, e em todas as ligas foram realizados ensaios de microdureza. A liga conformada por \"spray\" e obtida em laboratório foi laminada a quente e a frio. Foram realizados também estudos de textura, para tentar estabelecer uma correlação entre todas as ligas quanto ao processo de fabricação. A avaliação da textura indicou que a presença de fases de silício primário, finamente distribuídas impedem o aparecimento de texturas típicas de deformação de ligas de alumínio, mesmo após severas deformações, como as necessárias para a transformação de pré-formados em tubos que originam as camisas de cilindro. As medidas de rugosidade indicaram características próprias do acabamento superficial usado para a produção das camisas, por brunimento ou por ataque químico. Os ensaios de microdureza apresentaram variações de acordo com as rotas de fabricação sendo que a liga eutética apresentou os maiores valores de microdureza em comparação às ligas conformadas por \"spray\". Todas as ligas foram avaliadas quanto a resistência à corrosão por ensaios de espectroscopia de impedância eletroquímica em dois meios, CSA com pH 3,3 e CSA com pH 11. As quatro ligas hipereutéticas estudadas apresentaram mecanismo de corrosão similar em condensado sintético automotivo (CSA) pH 3,3. Em todas ocorreu o ataque intenso da matriz de alumínio e as partículas de silício primário atuaram como regiões catódicas. A liga 2 apresentou maior resistência à corrosão entre todas as ligas ensaiadas, tanto em CSA pH 3,3 como em pH 11. Neste último meio, uma camada de produtos de corrosão formou-se sobre todas as ligas, e os resultados indicaram mecanismos diferentes para o ataque da matriz de Al e para o crescimento da camada depositada na superfície das ligas estudadas. O ataque da matriz da liga 2 neste meio foi aparentemente mais lento do que sobre as demais ligas, com a formação de uma camada mais compacta de produtos de corrosão, estabelecendo um controle por difusão dos processos interfaciais para maiores períodos de ensaio. A camada formada sobre as demais ligas neste meio apresentou-se mais defeituosa, e nestas ocorreu também menor controle da velocidade das reações interfaciais por processos difusionais. / In the present study four hypereutectic Al-Si alloys, three produced by spray forming and one by casting, were characterized for microhardness, roughness, microstructure, texture and corrosion resistance in a synthetic automotive condensed solution (SACS). Two of the spray formed alloys tested were obtained from cylinder liners and the other was laboratory made. Spray forming involves alloy atomization and droplets deposition on a substrate, previous to the solidification of all of the droplets. This process favours the production of materials with a fine microstructure free of macrosegregation that is related to improved hot workability. The microstructure characterization of the four alloys revealed the presence of porosities in the laboratory made alloy. All the three alloys produced by spray forming showed a homogeneous distribution of primary precipitates. The microstructure of one of the alloys showed eutectic microstructure, indicating that this alloy was fabricated by casting. In the cylinder liners, the surface roughness was measured and the microhardness of all the alloys was also evaluated. Furthermore, the laboratory made alloy was hot and cold rolled. Texture determinations were carried out to investigate the correlation between the alloy type and their fabrication process. The texture investigation indicated that the fine distribution of primary silicon phase in the alloy hindered the development of texture typical of aluminium alloys deformation, even after severe mechanical work, such as those used in the conversion of pre-formed in cylinder liners. The surface roughness results indicated typical characteristics of the surface finishing used, honing or chemical etching. The microhardness results were dependent on the fabrication process used, with higher microhardness associated to the eutectic alloy comparatively to the spray formed ones. All hypereutectic alloys were tested for corrosion resistance using electrochemical impedance spectroscopy in two electrolytes, SACS with pH 3.3 and SACS with pH 11. The four tested alloys showed similar corrosion mechanisms in the acid electrolyte (SACS pH 3.3). The intense attack of the aluminium matrix occurred in ali tested alloys and the primary silicon phase was unattacked and appeared as emerging from the surface after corrosion immersion test. The primary silicon particles acted as cathodic sites. The alloy 2 showed the highest corrosion resistance among the tested alloys in both electrolytes, SACS pH 3.3 and SACS pH 11. In this last medium, a layer of corrosion products formed on all the alloys, and the results indicated different mechanisms for the aluminium matrix corrosive attack and growth of the deposited layer on the alloys surface. The kinetics of aluminium matrix attack was apparently slower in the alloy 2 than for the other alloys, resulting in the formation of a more compact layer of corrosion products, leading to diffusion controlled interfacial processes for longer test periods. The layer of corrosion products on the other three alloys (1, 3, and 4) had more defects and for these alloys diffusional controlled interfacial processes were not as significant as for alloy 2.