Orientador: Cláudio Kitano / Abstract: Interferometers are extremely sensitive measurement devices, which use the principle of interference between two or more sources of light to generate a pattern of constructive and destructive interference. This pattern contains information about the physical phenomenon under study, and their light intensity can be used to calculate the optical path difference traveled by the two beams. The optical path difference and light intensity relationship is given by a cosine type function. Large disturbances can change the interferometer operation point, reaching nonlinear regions of the interferometric curve and even inducing ambiguities due to the periodicity of the input/output relationship. The present work concerns with the modeling, development and application of a control strategy based on sliding mode control, in a two-beam quadrature interferometer. It was used the high gain approach, which consists in to fully compensate the phase shifts induced on the sensor arm with the control system, in such a way that the voltage control signal becomes proportional to the phase disturbances. Therefore, the demodulation process does not require phase unwrapping algorithms. This implemented system showed capability to improve dynamic range and bandwidth when compared with other control systems in literature that were based on different high gain approach topologies. Also a new method of interferometric phase demodulation is proposed allying this control strategy to a virtual emulated inte... (Complete abstract click electronic access below) / Resumo: Interferômetros são dispositivos de medição extremamente sensíveis, os quais utilizam o princípio de interferência entre duas ou mais fontes de luz para gerar um padrão de interferência construtiva e destrutiva. Este padrão contém informação sobre os fenômenos físicos sob estudo, e sua intensidade luminosa pode ser usada para calcular a diferença de caminho óptico acumulada pelos dois feixes de luz. A diferença de caminho óptico e a intensidade de luz são relacionadas por uma função cossenoidal. Grandes distúrbios podem alterar o ponto de operação do interferômetro, alcançando regiões não lineares da curva característica do interferômetro e até mesmo induzindo ambiguidades, devido à periodicidade da relação entrada/saída. Este trabalho preocupou-se com o modelamento, desenvolvimento e aplicação de uma estratégia de controle baseada em controle com modos deslizantes, em um interferômetro de dois feixes em quadratura. Foi utilizada a abordagem de alto ganho, a qual consiste em utilizar o sistema de controle para compensar completamente os deslocamentos de fase induzidos no braço sensor, de tal forma que o sinal de controle se relaciona com os deslocamentos de fase por uma equação de reta. Portanto o processo de demodulação não necessita de algoritmos de desdobramento de fase. O sistema implementado mostrou capacidade de melhorar a faixa dinâmica e largura de banda quando comparado com outros sistemas de controle na literatura, também baseados na abordagem de alto ganho. Destaca... (Resumo completo, clicar acesso eletrônico abaixo) / Doutor

A number of missions to comets and asteroids have been undertaken by major space organizations driving a need to accurately characterize their gravitational fields. This is complicated however by their irregular shapes. To accurately and safely navigate spacecraft in these environments, a simple point-mass gravity model is insufficient and instead higher-fidelity models are required. Several such models exist for this purpose but all posess drawbacks. Moreover, there are some applications for which the currently available models are not particular well suited. In this dissertation, numerical quadrature and curvilinear meshing techniques are applied to the small body gravity problem. The goal of this work is to to create a gravitational model suitable for integating large numbers of low altitude trajectories and rapidly characterizing the near-surface potential field. In total three new models are developed. The first applies two-dimensional quadrature formulas to calculate the gravitational field of an arbitrary triangular surface mesh. The second extends this result to curvilinear surface meshes that more accurately approximate the surface topology. The third applies three-dimensional quadrature to curvilinear tetrahedral meshes to generate accurate distributions of point-masses. The accuracy of the new models is fully characterized and simple relations are presented for predicting the error of integrated trajectories. The efficiency of the models is then compared to other high-fidelity models currently in use. The new models perform well between the body's circumsphere and a thin layer that surrounds the surface.

In clustered survival data, unobservable cluster effects may exert powerful influences on the outcomes and thus induce correlation among subjects within the same cluster. The ordinary partial likelihood approach does not account for this dependence. Frailty models, as an extension to Cox regression, incorporate multiplicative random effects, called frailties, into the hazard model and have become a very popular way to account for the dependence within clusters. We particularly study the two-level nested lognormal frailty model and propose an estimation approach based on the complete data likelihood with frailty terms integrated out. We adopt B-splines to model the baseline hazards and adaptive Gauss-Hermite quadrature to approximate the integrals efficiently. Furthermore, in finding the maximum likelihood estimators, instead of the Newton-Raphson iterative algorithm, Gauss-Seidel and BFGS methods are used to improve the stability and efficiency of the estimation procedure. We also study competing risks models with missing cause of failure in the context of Cox proportional hazards models. For competing risks data, there exists more than one cause of failure and each observed failure is exclusively linked to one cause. Conceptually, the causes are interpreted as competing risks before the failure is observed. Competing risks models are constructed based on the proportional hazards model specified for each cause of failure respectively, which can be estimated using partial likelihood approach. However, the ordinary partial likelihood is not applicable when the cause of failure could be missing for some reason. We propose a weighted partial likelihood approach based on complete-case data, where weights are computed as the inverse of selection probability and the selection probability is estimated by a logistic regression model. The asymptotic properties of the regression coefficient estimators are investigated by applying counting process and martingale theory. We further develop a double robust approach based on the full data to improve the efficiency as well as the robustness.

On adapte des méthodes numériques efficaces pour équations de Fredholm à la résolution d'équations singulières. On développe en particulier les variantes itératives de Brakhage et d' Atkinson de la méthode de Nyström. Les exemples numériques traités confirment la nette supériorité de la méthode itérative de Brakhage

<p>Image segmentation is the problem of partitioning an image into meaningful parts, often consisting of an object and background. As an important part of many imaging applications, e.g. face recognition, tracking of moving cars and people etc, it is of general interest to design robust and fast segmentation algorithms. However, it is well accepted that there is no general method for solving all segmentation problems. Instead, the algorithms have to be highly adapted to the application in order to achieve good performance. In this thesis, we will study segmentation methods for blood vessels in medical images. The need for accurate segmentation tools in medical applications is driven by the increased capacity of the imaging devices. Common modalities such as CT and MRI generate images which simply cannot be examined manually, due to high resolutions and a large number of image slices. Furthermore, it is very difficult to visualize complex structures in three-dimensional image volumes without cutting away large portions of, perhaps important, data. Tools, such as segmentation, can aid the medical staff in browsing through such large images by highlighting objects of particular importance. In addition, segmentation in particular can output models of organs, tumors, and other structures for further analysis, quantification or simulation.</p><p>We have divided the segmentation of blood vessels into two parts. First, we model the vessels as a collection of lines and edges (linear structures) and use filtering techniques to detect such structures in an image. Second, the output from this filtering is used as input for segmentation tools. Our contributions mainly lie in the design of a multi-scale filtering and integration scheme for de- tecting vessels of varying widths and the modification of optimization schemes for finding better segmentations than traditional methods do. We validate our ideas on synthetical images mimicking typical blood vessel structures, and show proof-of-concept results on real medical images.</p>

Coherent detection employing multilevel modulation formats has become one of the most promising technologies for next generation high speed transmission systems due to the high power and spectral efficiencies. Using the powerful digital signal processing (DSP), coherent optical receivers allow the significant equalization of chromatic dispersion (CD), polarization mode dispersion (PMD), phase noise (PN) and nonlinear effects in the electrical domain. Recently, the realizations of these DSP algorithms for mitigating the channel distortions in the coherent transmission systems are the most attractive investigations. The CD equalization can be performed by the digital filters developed in the time and the frequency domain, which can suppress the fiber dispersion effectively. The PMD compensation is usually performed in the time domain with the adaptive least mean square (LMS) and constant modulus algorithms (CMA) equalization. Feed-forward and feed-back carrier phase estimation (CPE) algorithms are employed to mitigate the phase noise (PN) from the transmitter (TX) and the local oscillator (LO) lasers. The fiber nonlinearities are compensated by using the digital backward propagation methods based on solving the nonlinear Schrödinger (NLS) equation and the Manakov equation. In this dissertation, we present a comparative analysis of three digital filters for chromatic dispersion compensation, a comparative evaluation of different carrier phase estimation methods considering digital equalization enhanced phase noise (EEPN) and a brief discussion for PMD adaptive equalization. To implement these investigations, a 112-Gbit/s non-return-to-zero polarization division multiplexed quadrature phase shift keying (NRZ-PDM-QPSK) coherent transmission system with post-compensation of dispersion is realized in the VPI simulation platform. In the coherent transmission system, these CD equalizers have been compared by evaluating their applicability for different fiber lengths, their usability for dispersion perturbations and their computational complexity. The carrier phase estimation using the one-tap normalized LMS (NLMS) filter, the differential detection, the block-average (BA) algorithm and the Viterbi-Viterbi (VV) algorithm is evaluated, and the analytical predictions are compared to the numerical simulations. Meanwhile, the phase noise mitigation using the radio frequency (RF) pilot tone is also investigated in a 56-Gbit/s NRZ single polarization QPSK (NRZ-SP-QPSK) coherent transmission system with post-compensation of chromatic dispersion. Besides, a 56-Gbit/s NRZ-SP-QPSK coherent transmission system with CD pre-distortion is also implemented to analyze the influence of equalization enhanced phase noise in more detail. / QC 20120528

In this thesis, which consists of five papers (A,B,C,D,E), we are interested in questions related to quadrature domains. Among the problems studied are the possibility of changing the type of measure in a quadrature identity (from complex to real and from real signed to positive), properties of partial balayage, which in a sense can be used to generate quadrature domains, and mother bodies which are closely related to inversion of partial balayage. These three questions are discussed in papers A,D respectively B. The first of these questions (when trying to go from real signed to positive measures) leads to the study of approximation in the cone of positive harmonic functions. These questions are closely related to properties of the harmonic measure on the Martin boundary, and this relationship leads to the study of harmonic measures on ideal boundaries in paper E. Some other approaches to the same problem also lead to some extent to the study of properties of classical balayage in paper C. / QC 20101007

This thesis presents the Differential Quadrature Method (DQM) for solving time-dependent or heat conduction problem. DQM discretizes the space derivatives giving a system of ordinary differential equations with respect to time and the fourth order Runge Kutta Method (RKM) is employed for solving this system. Stabilities of the ordinary differential equations system and RKM are considered and step sizes are arranged accordingly. The procedure is applied to several time dependent diffusion problems and the solutions are presented in terms of graphics comparing with the exact solutions. This method exhibits high accuracy and efficiency comparing to the other numerical methods.

Because of IC technology evolution and the increase of market demand, the communication industry grows vigorously in recent years. The voltage-controlled oscillator plays a key role in the RF transceiver and provides oscillation signals needed for upconversin and downconvertion. Usually, we separate the signals into I/Q channels for modulation and demodulation in upconversin and downconvertion. Because the quality of the local oscillator influences the performance of communication system, designing a voltage-controlled oscillator that can provide two identical signals in accurate quadrature is necessary. In this thesis, a new quadrature voltage-controlled oscillator is presented. We use two identical relaxation oscillators with adjustable Schmitt triggers to construct the cross-coupled architecture. This oscillator has accurate ( <1¢X) and stable quadrature outputs which are independent of operating frequency and process variations. This oscillator circuit is fabricated in TSMC 0.35£gm CMOS Mixed-Signal process provided by National Chip Implementation Center (CIC). Our design is verified by simulation and measurement results.

This thesis researchs the design of quadrature modulator consists of 120MHz quadrature modulator that is fabricated using hybrid elements and print circuit board (PCB) technology for digital signal generator and quadrature modulator monolithic-microwave integrated-circuit (MMIC) that is fabricated using GaAs heterojunction bipolar transistor (HBT) technology for Personal Communication Service (PCS) applications. The 120MHz quadrature modulator incorporates power divider/combiner, phase shifter and doubly balanced mixer; the design architecture, principle and measurement results of division are presented in this thesis. A quadrature modulator is implemented by combining every division and measures specifications accurately, comparing with that of Agilent ESG-D series digital signal generator with the same carrier frequency and digital modulation. The quadrature modulator MMIC for PCS applications incorporates phase shifter, Gilbert cell mixer, differential to single-ended converter and RF amplifier at output; the design architecture, principle and simulation results of division are presented in this thesis. A quadrature modulator is integrated by combining every division and simulates parameters strictly.For troublesome specification measurement of quadrature modulator, this thesis also presents measurement method and instrument setup detailedly.