Universiẗat, Diss., 2002--Stuttgart. / Erscheinungsjahr an der Haupttitelstelle: 2002.

Thesis (Ph. D.)--University of Canterbury, 2006. / Typescript (photocopy). Includes bibliographical references (p. 343-364). Also available via the World Wide Web.

Nürnberg, Universiẗat, Diss., 2004--Erlangen.

Thesis (Honors)--College of William and Mary, 2009. / Includes bibliographical references (p. 54-55). Also available via the World Wide Web.

Zugl.: Zürich, Techn. Hochsch., Diss.

Examensarbetet innefattade ekonomiska beräkningar gällande införskaffande av sol- samt vindkraftsproduktion för en ö i Luleå skärgård som fungerar vid ö-drift gentemot kostnaden för att flyga ut ett reservaggregat i tre delar med helikopter. En förenklad kostnadskalkyl utfördes gällande investering samt underhåll för den förnyelsebara elproduktionen och dess nuvärde sett till 25-års livslängd för vindkraft såväl som solkraft. Detta vägdes mot utgifterna för utflygning av ett reservaggregat med helikopter som är den reservkraftslösning som finns att tillgå i dagsläget. Beräkningar av kortslutningseffekt med den förnyelsebara produktionen och hur det påverkar felbortkopplingar samt spänningsnivåer togs också med. Den ekonomiska investeringen kom också att vägas mot kundnyttan i form av ökad trygghetskänsla med en färdig reservkraftslösning samt fortsatt utveckling för företaget inom förnyelsebar energi. Uppdraget begränsades till att innefatta två olika scenarion med förnyelsebar energi, ett med 100 % solkraft och ett med cirka 10 kW vindkraft och 35 kW solkraft. Solcellssystemet kom att vara fristående på markställning vriden direkt mot söder med lämplig vinkling på solcellspanelerna. För scenarierna användes färdiga produktionslösningar samt färdiga batterilösningar för att behålla visst överskott av produktion till senare behov samt ett nät med nog låg spänningsvariation samt hög frekvensstabilitet under kortare tid. Kortare tid i detta fall gällde mellan 3 timmar upp till 3 veckor. Projektet tog effektbehovsdata samt möjliga meteorologiska data från en ö i Luleå Skärgård som ligger i skärgårdsnätet. Där fanns det 38 kunder där majoriteten hade kraftbehov under årets varmare månader, med några få kunder som var där nästan året runt. Ön kommer hädanefter refereras till som driftplatsen i rapporten. Den meteorologiska data som ej fanns att tillgå på driftplatsen togs vid närmaste väderstation, i detta fall Luleå stad. Sett till instrålningsdata samt avlästa vindhastigheter visade det sig att en produktionsanläggning med endast solcellspaneler och minst 12 h batteribank inte var realistisk som reservkraftlösning vid möjlig ö-drift vintertid. En större batteribank på 3 dygns försörjning skulle inte vara ekonomiskt försvarbar jämfört med investeringskostnaden till produktionsanläggningen, och det skulle trots det inte avhjälpa de produktionsproblem som var aktuella under vinterhalvåret. En kombination av solkraft och vindkraft med 3h batteriförsörjning kompletterade varandra väl till effektbehoven som återfanns på driftplatsen, klarade av de kraven som ställdes gällande variation i spänning och möjliga överströmmar samt skulle möjliggöra elförsörjning året runt.

This thesis explores approaches to two problems in the frame-rate computation of a priori unknown 3D scene structure and camera pose using a single camera, or monocular simultaneous localisation and mapping. The thesis reflects two trends in vision in general and structure from motion in particular: (i) the move from directly recovered and towards learnt geometry; and (ii) the sparsification of otherwise dense direct methods. The first contributions mitigate scale drift. Beyond the inevitable accumulation of random error, monocular SLAM accumulates error via the depth/speed scaling ambiguity. Three solutions are investigated. The first detects objects of known class and size using fixed descriptors, and incorporates their measurements in the 3D map. Experiments using databases with ground truth show that metric accuracy can be restored over kilometre distances; and similar gains are made using a hand-held camera. Our second method avoids explicit feature choice, instead employing a deep convolutional neural network to yield depth priors. Relative depths are learnt well, but absolute depths less so, and recourse to database-wide scaling is investigated. The third approach uses a novel trained network to infer speed from imagery. The second part of the thesis develops sparsified direct methods for monocular SLAM. The first contribution is a novel camera tracker operating directly using affine image warping, but on patches around sparse corners. Camera pose is recovered with an accuracy at least equal to the state of the art, while requiring only half the computational time. The second introduces a least squares adjustment to sparsified direct map refinement, again using patches from sparse corners. The accuracy of its 3D structure estimation is compared with that from the widely used method of depth filtering. It is found empirically that the new method's accuracy is often higher than that of its filtering counterpart, but that the method is more troubled by occlusion.

abstract: The past two decades have been monumental in the advancement of microchips designed for a diverse range of medical applications and bio-analysis. Owing to the remarkable progress in micro-fabrication technology, complex chemical and electro-mechanical features can now be integrated into chip-scale devices for use in biosensing and physiological measurements. Some of these devices have made enormous contributions in the study of complex biochemical processes occurring at the molecular and cellular levels while others overcame the challenges of replicating various functions of human organs as implant systems. This thesis presents test data and analysis of two such systems. First, an ISFET based pH sensor is characterized for its performance in a continuous pH monitoring application. Many of the basic properties of ISFETs including I-V characteristics, pH sensitivity and more importantly, its long term drift behavior have been investigated. A new theory based on frequent switching of electric field across the gate oxide to decrease the rate of current drift has been successfully implemented with the help of an automated data acquisition and switching system. The system was further tested for a range of duty cycles in order to accurately determine the minimum length of time required to fully reset the drift. Second, a microfluidic based vestibular implant system was tested for its underlying characteristics as a light sensor. A computer controlled tilt platform was then implemented to further test its sensitivity to inclinations and thus it‟s more important role as a tilt sensor. The sensor operates through means of optoelectronics and relies on the signals generated from photodiode arrays as a result of light being incident on them. ISFET results show a significant drop in the overall drift and good linear characteristics. The drift was seen to reset at less than an hour. The photodiodes show ideal I-V comparison between photoconductive and photovoltaic modes of operation with maximum responsivity at 400nm and a shunt resistance of 394 MΩ. Additionally, post-processing of the tilt sensor to incorporate the sensing fluids is outlined. Based on several test and fabrication results, a possible method of sealing the open cavity of the chip using a UV curable epoxy has been discussed. / Dissertation/Thesis / M.S. Electrical Engineering 2011

Submitted by Alessandra Portugal (alessandradf@ioc.fiocruz.br) on 2013-09-24T15:06:55Z No. of bitstreams: 1 Dissertação de Mestrado Paola Cristina Resende Silva.pdf: 1538372 bytes, checksum: 8c68a244d3e53f424ddb574f985bcad8 (MD5) / Made available in DSpace on 2013-09-24T15:06:55Z (GMT). No. of bitstreams: 1 Dissertação de Mestrado Paola Cristina Resende Silva.pdf: 1538372 bytes, checksum: 8c68a244d3e53f424ddb574f985bcad8 (MD5) Previous issue date: 2010 / Fundação Oswaldo Cruz. Instituto Oswaldo Cruz. Rio de Janeiro, RJ, Brasil / Globalmente, as infecções causadas pelos vírus Influenza constituem um importante desafio para a Saúde Pública. Os vírus Influenza B pertencem à família Orthomyxoviridae, seu genoma viral é constituído por um RNA de fita simples e polaridade negativa. O processo de drift antigênico favorece o contínuo aparecimento de novas variantes virais, o que demanda a reformulação anual da vacina. No início década de 80, foi observada a divergência do vírus Influenza B em duas linhagens antigênica e filogeneticamente distintas: B/Victoria/2/87-like (Vic87) e B/Yamagata/16/88-like (Yam88), que tem co-circulado em diferentes países na última década. O objetivo deste estudo consiste na identificação e caracterização molecular das linhagens de Influenza B circulantes em diferentes regiões brasileiras durante as epidemias de 2004 a 2008, com base no sequenciamento dos genes Hemaglutinina (HA) e Neuraminidase (NA). Ainda, padronizamos a metodologia de Eletroforese em Gel com Gradientes Desnaturantes (DGGE), visando à rápida tipagem dos vírus B. Diferentes substituições nos genes da HA e NA foram encontradas. Evidenciamos a co-circulação de ambas as linhagens no período estudado, contudo, não observamos a ocorrência de rearranjo gênico e nem a emergência de cepas resistentes aos inibidores de neuraminidase, com base nos genes investigados. No período 2006-2008, observamos a adequada concordância entre as cepas circulantes e as cepas vacinais preconizadas para uso no Hemisfério Sul. Entretanto, o mesmo não foi verdadeiro para o período 2004-2005. Finalmente, o protocolo de DGGE desenvolvido pode ser eficientemente utilizado para fins de rápida tipagem das linhagens de Influenza B. Os achados deste estudo contribuem para a melhor compreensão sobre a variabilidade dos vírus Influenza B e os mecanismos envolvidos na sua evolução molecular, bem como o padrão de circulação das linhagens virais no Brasil e sua correspondência com as vacinas para Influenza, anualmente administradas no Hemisfério Sul. Este conjunto de informações são de grande relevância para a contínua adequação e implementação das políticas e estratégias voltadas ao controle e prevenção de infecções por Influenza na nossa população. / Worldwide, Influenza infections are a major Public Health issue. Influenza B virus is classified into the Orthomyxoviridae family, the viral genome consists of a single strand RNA and negative polarity. Because of antigenic drift, novel viral variants are continuously rising, what demands the annual review of vaccine formulation. In the early 80´s, was observed the divergence of Influenza B into two distinct antigenic and phylogenetic lineages – B/Victoria/2/87-like (Vic87) and B/Yamagata/16/88-like (Yam88), was observed. In some countries, these strains have been co-circulating in the last 10 years. The aim of this study was to investigate the circulation patterns of Vic87 and Yam88 among different Brazilian regions during the 2004-2008 epidemics, based on haemagglutinin (HA) and neuraminidase (NA) sequencing. Moreover, a Denaturing Gradient Gel Electrophoresis (DGGE) protocol was standardized for rapid typing of Flu B typing into Yam88-like and Vic87-like strains. Different Aminoacid substitutions in the HA and NA were encountered. Along the studied period, our findings showed that both viral lineages have been co-circulating in Brazil. Moreover, no evidence of reassortant nor NA inhibitor-resistant viruses was found. From 2006 to 2008, an adequate match between vaccine and circulating strains was met. However, it was not true for 2004-2005 years. Finally, our DGGE protocol can be successfully used as a rapid Flu B strain typing test. Our findings contribute for a better figure of Flu B genetic variability and its molecular evolution mechanisms, in addition to the circulation patterns of Flu B lineages in Brazil, and their respective association with the vaccine strains used in the Southern Hemisphere. Altogether, these are pivotal information to continuously tailor and implement Public Health policies on behalf of the control and prevention of Influenza infections.

[pt] Os conceitos do mundo real muitas vezes não são estáveis: eles mudam com o tempo. Assim como os conceitos, a distribuição de dados também pode se alterar. Este problema de mudança de conceitos ou distribuição de dados é conhecido como concept drift e é um desafio para um modelo na tarefa de aprender a partir de dados. Este trabalho apresenta um novo modelo neuroevolutivo com inspiração quântica, baseado em um comitê de redes neurais do tipo Multi-Layer Perceptron (MLP), para a aprendizagem em ambientes não estacionários, denominado NEVE (Neuro-EVolutionary Ensemble). Também apresenta um novo mecanismo de detecção de concept drift, denominado DetectA (Detect Abrupt) com a capacidade de detectar mudanças tanto de forma proativa quanto de forma reativa. O algoritmo evolutivo com inspiração quântica binário-real AEIQ-BR é utilizado no NEVE para gerar automaticamente novos classificadores para o comitê, determinando a topologia mais adequada para a nova rede, selecionando as variáveis de entrada mais apropriadas e determinando todos os pesos da rede neural MLP. O algoritmo AEIQ-R determina os pesos de votação de cada rede neural membro do comitê, sendo possível utilizar votação por combinação linear, votação majoritária ponderada e simples. São implementadas quatro diferentes abordagens do NEVE, que se diferem uma da outra pela forma de detectar e tratar os drifts ocorridos. O trabalho também apresenta resultados de experimentos realizados com o método DetectA e com o modelo NEVE em bases de dados reais e artificiais. Os resultados mostram que o detector se mostrou robusto e eficiente para bases de dados de alta dimensionalidade, blocos de tamanho intermediário, bases de dados com qualquer proporção de drift e com qualquer balanceamento de classes e que, em geral, os melhores resultados obtidos foram usando algum tipo de detecção. Comparando a acurácia do NEVE com outros modelos consolidados da literatura, verifica-se que o NEVE teve acurácia superior na maioria dos casos. Isto reforça que a abordagem por comitê neuroevolutivo é uma escolha robusta para situações em que as bases de dados estão sujeitas a mudanças repentinas de comportamento. / [en] Real world concepts are often not stable: they change with time. Just as the concepts, data distribution may change as well. This problem of change in concepts or distribution of data is known as concept drift and is a challenge for a model in the task of learning from data. This work presents a new neuroevolutive model with quantum inspiration called NEVE (Neuro- EVolutionary Ensemble), based on an ensemble of Multi-Layer Perceptron (MLP) neural networks for learning in non-stationary environments. It also presents a new concept drift detection mechanism, called DetectA (DETECT Abrupt) with the ability to detect changes both proactively as reactively. The evolutionary algorithm with binary-real quantum inspiration AEIQ-BR is used in NEVE to automatically generate new classifiers for the ensemble, determining the most appropriate topology for the new network and by selecting the most appropriate input variables and determining all the weights of the neural network. The AEIQ-R algorithm determines the voting weight of each neural network ensemble member, and you can use voting by linear combination and voting by weighted or simple majority. Four different approaches of NEVE are implemented and they differ from one another by the way of detecting and treating occurring drifts. The work also presents results of experiments conducted with the DetectA method and with the NEVE model in real and artificial databases. The results show that the detector has proved efficient and suitable for data bases with high-dimensionality, intermediate sized blocks, any proportion of drifts and with any class balancing. Comparing the accuracy of NEVE with other consolidated models in the literature, it appears that NEVE had higher accuracy in most cases. This reinforces that the neuroevolution ensemble approach is a robust choice to situations in which the databases are subject to sudden changes in behavior.