The principle of controlled auto-ignition (CAI) is to mix fuel and air homogeneously before compressing the mixture to the point of auto-ignition. As ignition occurs simultaneously, CAI engines operate with lean mixtures preventing high cylinder pressures. CAI engines produce small amounts of nitrogen oxides (NOx) due to low combustion temperatures while maintaining high compression ratios and engine efficiencies. Due to simultaneous combustion and lean mixtures, CAI engines are restricted between low and mid load operations. Various strategies have been studied to improve the load limit of CAI engines. The scope of the project is to investigate the consequences of varying valve timing, as a method to control the mixture temperature within the combustion chamber and therefore, controlling the mixture auto-ignition point. This study presents computational fluid dynamics (CFD) modelling results of transient flow, inside a 0.45 litre Lotus single cylinder engine. After a validation process, a chemical kinetics model is combined with the CFD code, in order to study in-cylinder temperatures, the mixture distribution during compression and to predict the auto-ignition timing. The first part of the study focuses on validating the calculated in-cylinder velocities. A mesh sensitivity study is performed as well as a comparison of different turbulence models. A method to reduce computational time of the calculations is presented. The effects of engine speed on charge delay and charge amount inside the cylinder, the development of the in-cylinder flow field and the variation of turbulence parameters during the intake and compression stroke, are studied. The second part of the study focuses on the gasoline mixture and the variation of the valve timing, to retain different ratios of residual gases within the cylinder. After validation of the model, a final set of CFD calculations is performed, to investigate the effects of valve timing on flow and the engine parameters. The results are then compared to a fully homogeneous mixture model to study the benefits of varying valve duration. New key findings and contributions to CAI knowledge were found in this investigation. Reducing the intake and exhaust valve durations created a mixture temperature stratification and a fuel concentration distribution, prior to auto-ignition. It resulted in extending the heat release rate duration, improving combustion. However, shorter valve timing durations also showed an increase in heat transfer, pumping work and friction power, with a decrease of cylinder indicated efficiency. Valve timing, as a method to control auto-ignition, should only be used when the load limit of CAI engines, is to be improved.

For the task of near-duplicate document detection, comparison approaches based on bag-of-words used in information retrieval community are not sufficiently accurate. This work presents novel approach when instance-level constraints are given for documents and it is needed to retrieve them, given new query document for near-duplicate detection. The framework incorporates instance-level constraints and clusters documents into groups using novel clustering approach Grouped Latent Dirichlet Allocation (gLDA). Then distance metric is learned for each cluster using large margin nearest neighbor algorithm and finally ranked documents for given new unknown document using learnt distance metrics. The variety of experimental results on various datasets demonstrate that our clustering method (gLDA with side constraints) performs better than other clustering methods and the overall approach outperforms other near-duplicate detection algorithms.

The ability to automatically analyze and segment news articles by their content is a growing research field. This thesis explores the unsupervised machine learning method topic modeling applied on Swedish news articles for generating topics to describe and segment articles. Specifically, the algorithms non-negative matrix factorization (NMF) and the latent Dirichlet allocation (LDA) are implemented and evaluated. Their usefulness in the news media industry is assessed by its ability to serve as a uniform categorization framework for news articles. This thesis fills a research gap by studying the application of topic modeling on Swedish news articles and contributes by showing that this can yield meaningful results. It is shown that Swedish text data requires extensive data preparation for successful topic models and that nouns exclusively and especially common nouns are the most suitable words to use. Furthermore, the results show that both NMF and LDA are valuable as content analysis tools and categorization frameworks, but they have different characteristics, hence optimal for different use cases. Lastly, the conclusion is that topic models have issues since they can generate unreliable topics that could be misleading for news consumers, but that they nonetheless can be powerful methods for analyzing and segmenting articles efficiently on a grand scale by organizations internally. The thesis project is a collaboration with one of Sweden’s largest media groups and its results have led to a topic modeling implementation for large-scale content analysis to gain insight into readers’ interests.

The aim of the thesis is to put a transparent scaled PMMA model of the pressure swirl nozzle into operation, which includes, the selection of working fluids and the preparation of a test set to allow measurements using optical methods (LDA, PDA, PIV, high-speed visualization). The theoretical part describes the basic theory of atomization, optical measurement methods and deals with the problems of optical transition in optically complex systems. It also includes an extensive search for transparent liquids and materials of enlarged models that have been used in experiments, which often aim to match light refractive indices between these materials. In the practical part, attention is paid to the preparation of the test set and tests of chemical effects of several selected liquids on PMMA material are conducted, including a summary of experience with their use, as there was a permanent damage to the material. Several LDA measurements followed, using kerosene, p-cymene, 1-bromonaphthalene and water, evaluated the effect of the refractive index difference on the results. It turned out that no observable influence occurred if the refractive index difference between the nozzle material and the liquid was small. In addition, a visualization of internal flow through a high-speed camera was made. The practical part closes the static pressure measurement in the nozzle chamber, where the pressure ratio was measured on the walls of the chamber as well as on its axis. The measurement confirmed that the pressure on the chamber walls is constant and varies with the distance from the chamber axis.

This master’s thesis deals with measurement of spray characteristics of Effervescent atomizers intended for burning waste and heavy fuels. Atomizers were tested on cold test bench by means of Phase Doppler Anemometry. Spray characteristics were evaluated for many different regimes of pressure, temperature and Gas to liquid ratio. The aim of this measurement is to compare flow regimes and their influence on the quality of spray characteristics. The theoretical part describes basic fundamental principles of liquid atomization, effervescent atomization and principles of laser diagnostic methods. The practical part is engaged with improvements in test bench and setting up and optimization of the laser measuring system. Results contain visualization of spray, velocity profiles and drop size distribution in various operating flow regimes of the jet.

Das Ziel der Dissertation ist die Charakterisierung und Optimierung der Prozessbedingungen beim Hochgeschwindigkeitsdrahtflammspritzen. Dazu werden diagnostische Methoden wie das LDA-Verfahren, die Bewertung von Schichten und numerische Betrachtungen herangezogen. Verschiedene Spritzzusatzwerkstoffe wie Massiv- und Fülldrähte sowie hoch- und niedrigschmelzende Materialien werden verarbeitet. Zur Bewertung der Gebrauchseigenschaften erfolgen Korrosions- und Verschleißtests. Aus den Untersuchungsergebnissen resultieren Hinweise und Konzepte zur Verbesserung der Brennertechnik. Die Entwicklung eines neuen HVCW-Systems wird vorgestellt, welches Spritzpartikelgeschwindigkeiten im Überschallbereich ermöglicht, die deutlich über denen klassischer Systeme liegen. In einem gesonderten Abschnitt der Dissertation wird ein neuartiges Verfahren zur Herstellung hydrolysefähigen Materials vorgestellt. Der hergestellte Al-Sn-Werkstoff zersetzt sich in Kontakt mit Wasser unter Abgabe großer Mengen Wasserstoff in kürzester Zeit vollständig.

This final thesis deals primarily with the comparison of two calibration approaches. It determines the degree of mutual agreement and tries to answer the reasons of their deviations. This agreement was measured by the LDA method. It was found that the calibration approach based on the reference framework of values of the previous bachelor thesis differs systematically from the calibration using calibration equations obtained from certified laboratories by -2 %. The causes of this deviation were investigated using LDA and PIV. Although a significant number of measurements were performed, all tested hypotheses were refuted and therefore it was not possible to determine their cause.

In the traditional agenda-setting theory, the agenda-setters were the news media and the public has a minimal role in the process of agenda-setting, which makes the public a passive receiver located at the bottom in the top-down agenda-setting dynamics. This study claims that with the development of Information communication technologies, primarily social media, the networked public may be able to set their own agendas through connective actions, outside the influence of the news media agenda. There is little empirical research focused on development and dynamics of public agenda-setting through social media platforms. Understanding the development and dynamics of public agenda-setting may be key to accounting for and overcoming conflicting findings in previous reverse agenda-setting research. This study examined the public agenda-setting dynamics through a case of gun violence prevention activism Twitter network, the #MarchForOurLives Twitter network. This study determined that the agenda setters of the #MarchForOurLives Twitter network are the key Never Again MSD student leaders and the March For Our Lives. The weekly reflected important events and issues and the identified topics were highly co-related with the themes examined in the tweets created by the agenda setters. The amplifiers comprised the vast majority of the tweets. The advocates and the supporters consisted of 0.44% and 4.43% respectively. The tweets made by the agenda setters accounted for 0.03%. The young activists and the like-minded and participatory public could continuously make changes taking advantage of technologies, and they could be the hope in the current and future society.

This work is describing basic methods of face recognition. The methods PCA, LDA, ICA, trace tranfsorm, elastic bunch graph map, genetic algorithm and neural network are described. In practical part, the PCA, PCA + RBF neural network and genetic algorithms are implemented. The RBF neural network is used in the way of clasificator and genetic algorithm is used for RBF NN training in one case and for selecting eigenvectors from PCA method in the other case. This method, PCA + GA, called EPCA, outperform other methods tested in this work on the ORL testing database.

Measurement science has seen fast growth of data in both volume and complexity in recent years, new algorithms and methodologies have been developed to aid the decision<br>making in measurement sciences, and this process is automated for the liberation of labor. In light of the adversarial approaches shown in digital image processing, Chapter 2 demonstrate how the same attack is possible with spectroscopic data. Chapter 3 takes the question presented in Chapter 2 and optimized the classifier through an iterative approach. The optimized LDA was cross-validated and compared with other standard chemometrics methods, the application was extended to bi-distribution mineral Raman data. Chapter 4 focused on a novel Artificial Neural Network structure design with diffusion measurements; the architecture was tested both with simulated dataset and experimental dataset. Chapter 5 presents the construction of a novel infrared hyperspectral microscope for complex chemical compound classification, with detailed discussion in the segmentation of the images and choice of a classifier to choose.<br>