<p>In recent years, two estimators have been proposed to correct the bias exhibited by the leastsquares (LS) estimator of the lagged dependent variable (LDV) coefficient in dynamic regression models when the sample is finite. They have been termed as &lsquo / mean-unbiased&rsquo / and &lsquo / medianunbiased&rsquo / estimators. Relative to other similar procedures in the literature, the two locationbased estimators have the advantage that they offer an exact and uniform methodology for LS estimation of the LDV coefficient in a first order autoregressive model with or without exogenous regressors i.e. ARX(1).</p> <p><br /> However, no attempt has been made to accurately establish and/or compare the statistical properties among these estimators, or relative to those of the LS estimator when the LDV coefficient is restricted to realistic values. Neither has there been an attempt to&nbsp / compare their performance in terms of their mean squared error (MSE) when various forms of the exogenous regressors are considered. Furthermore, only implicit confidence intervals have been given for the &lsquo / medianunbiased&rsquo / estimator. Explicit confidence bounds that are directly usable for inference are not available for either estimator. In this study a new estimator of the LDV coefficient is proposed / the &lsquo / most-probably-unbiased&rsquo / estimator. Its performance properties vis-a-vis the existing estimators are determined and compared when the parameter space of the LDV coefficient is restricted. In addition, the following new results are established: (1) an explicit computable form for the density of the LS estimator is derived for the first time and an efficient method for its numerical evaluation is proposed / (2) the exact bias, mean, median and mode of the distribution of the LS estimator are determined in three specifications of the ARX(1) model / (3) the exact variance and MSE of LS estimator is determined / (4) the standard error associated with the determination of same quantities when simulation rather than numerical integration method is used are established and the methods are compared in terms of computational time and effort / (5) an exact method of evaluating the density of the three estimators is described / (6) their exact bias, mean, variance and MSE are determined and analysed / and finally, (7) a method of obtaining the explicit exact confidence intervals from the distribution functions of the estimators is proposed.</p> <p><br /> The discussion and results show that the estimators are still biased in the usual sense: &lsquo / in expectation&rsquo / . However the bias is substantially reduced compared to that of the LS estimator. The findings are important in the specification of time-series regression models, point and interval estimation, decision theory, and simulation.</p>

Non-stationary signals, that is signals with time-varying (TV) statistical properties, are commonly encountered in engineering practice. The vibration responses of structures, such as traffic-excited bridges, robotic devices, rotating machinery, and so on, constitute typical examples of non-stationary signals. Structures characterized by properties that vary with time are generally referred as TV structures and their vibration-based identification under normal operating conditions is a significant and challenging problem. An important class of parametric methods for the solution of this problem is based on Functional Series Time-dependent AutoRegressive Moving Average (FS-TARMA) models. These models have parameters that explicitly depend on time, with the dependence described by deterministic functions belonging to specific functional sub-spaces. The focus of the present thesis is on the development of complete and advanced FS-TARMA methods that will offer important improvements in overcoming drawbacks of existent methods and will further foster practical use and application of FS-TARMA models in non-stationary vibration analysis. The specific objectives of the thesis are: a) The introduction of a novel class of Adaptable FS-TARMA (AFS-TARMA) models and the development of a method for their effective identification. AFS-TARMA models are adaptable in the sense that they are not based on basis functions of a fixed form, but instead, they use basis functions with a-priori unknown properties that may adapt to the specific random signal characteristics. b) The postulation of a vector FS-TARMA method for output-only structural identification and the development of effective tools for both model parameter estimation and model structure selection. c) The introduction of a statistical method for vibration-based fault diagnosis in TV structures. d) The presentation of a thorough review on FS-TARMA models covering both theoretical and practical aspects of the model parameter estimation and structure selection problems with special emphasis being placed on promising recent methods. The methods that are developed in each chapter of this thesis are validated through their application in both numerical and experimental case studies and comparisons with currently available non-stationary signal identification methods. The results of the study demonstrate the new methods' applicability, effectiveness, and high potential for parsimonious and accurate identification and dynamic analysis of TV structures. / Μη-στάσιμα σήματα, δηλαδή σήματα με χρονικά μεταβαλλόμενες (ΧΜ) στατιστικές ιδιότητες, απαντώνται συχνά στην επιστήμη του μηχανικού. Τυπικά παραδείγματα αποτελούν οι ταλαντωτικές αποκρίσεις κατασκευών, όπως γέφυρες με κινούμενα οχήματα, ρομποτικές διατάξεις, περιστρεφόμενες μηχανές και άλλες. Κατασκευές που χαρακτηρίζονται από ιδιότητες οι οποίες μεταβάλλονται με τον χρόνο αναφέρονται ως ΧΜ κατασκευές και η δυναμική αναγνώριση και ανάλυση τους επί τη βάση ταλαντωτικών σημάτων απόκρισης αποτελεί σημαντικό και ταυτόχρονα δύσκολο πρόβλημα. Μια σημαντική τάξη παραμετρικών μεθόδων για την επίλυση αυτού του προβλήματος βασίζεται στα συναρτησιακά χρονικά μεταβαλλόμενα μοντέλα αυτοπαλινδρόμησης κινητού μέσου όρου (FS-TARMA, Functional Series Time-Dependent Auto-Regressive Moving Average). Τα μοντέλα αυτά χαρακτηρίζονται απο ΧΜ παραμέτρους οι οποίες ακολουθούν καθοριστικό πρότυπο και κατά συνέπεια μπορούν να προβληθούν σε κατάλληλα επιλεγμένους συναρτησιακούς υποχώρους. Ως βασικός στόχος της παρούσας διατριβής ορίζεται η ανάπτυξη εξελιγμένων μεθόδων μοντελοποίησης FS-TARMA οι οποίες θα προσφέρουν σημαντικές βελτιώσεις στις υπάρχουσες προσεγγίσεις και θα βοηθήσουν στην αντιμετώπιση πρακτικών προβλημάτων που σχετίζονται τόσο με την αναγνώριση των δυναμικών χαρακτηριστικών όσο και την διάγνωση βλαβών σε ΧΜ κατασκευές. Οι συγκεκριμένοι στόχοι της διατριβής μπορούν να περιγραφούν ως ακολούθως: α) Εισαγωγή καινοτόμων προσαρμόσιμων μοντέλων FS-TARMA και ανάπτυξη κατάλληλης μεθόδου για την αποτελεσματική εκτίμηση τους. Τα νέα μοντέλα είναι προσαρμόσιμα υπό την έννοια ότι δεν βασίζονται σε προκαθορισμένες συναρτήσεις βάσης, αλλά αντιθέτως χρησιμοποιούν συναρτήσεις βάσης με εκ των προτέρων άγνωστες ιδιότητες οι οποίες μπορούν να προσαρμοστούν στα χαρακτηριστικά συγκεκριμένου σήματος. β) Ανάπτυξη διανυσματικής μεθόδου εκτίμησης μοντέλων FS-TARMA για την αναγνώριση κατασκευών μέσα από διανυσματικά σήματα ταλαντωτικής απόκρισης. Ανάπτυξη αποδοτικών εργαλείων τόσο για το πρόβλημα εκτίμησης των παραμέτρων όσο και της επιλογής της δομής του μοντέλου. γ) Εισαγωγή στατιστικής μεθόδου για την διάγνωση βλαβών σε ΧΜ κατασκευές μέσω μοντέλων FS-TAR. δ) Παρουσίαση μιας διεξοδικής επισκόπησης των μοντέλων FS-TARMA η οποία καλύπτει τόσο θεωρητικά όσο και πρακτικά ζητήματα των προβλημάτων εκτίμησης των παραμέτρων και επιλογής της δομής των μοντέλων. Η αποτελεσματικότητα των μοντέλων και των μεθόδων που αναπτύσσονται σε κάθε κεφάλαιο αυτής της διατριβής διερευνάται µέσω της εφαρµογής τους τόσο σε αριθµητικές όσο και πειραµατικές µελέτες και συγκρίσεις µε υπάρχουσες µη-στάσιµες µεθόδους αναγνώρισης σηµάτων. Τα αποτελέσματα της εργασίας αυτής επιδεικνύουν την ικανότητα των νέων μοντέλων να παρέχουν εξαιρετικά ακριβείς αναπαραστάσεις ΧΜ κατασκευών κατάλληλων τόσο για την δυναμική ανάλυση όσο και για την διάγνωση βλαβών σε αυτές.

The production and transportation of hydrocarbon fluids in multiphase pipelines could be severely hindered by particulate solids deposit such as produced sand particles which accompany hydrocarbon production. Knowledge of the flow characteristics of solid particles in fluids transported in pipelines is important in order to accurately predict solid particles deposition in pipelines. This research thesis presents the development of a three-dimensional (3D) Computational Fluids Dynamics (CFD) modelling technique for the prediction of liquid-solids multiphase flow in pipes, with special emphasis on the flow in V-inclined pipe bends. The Euler-Euler (two-fluid) multiphase modelling methodology has been adopted and the multiphase model equations and closure models describing the liquid-solids flow have been implemented and calculated using the finite volume method in a CFD code software. The liquid phase turbulence has been modelled using a two-equation k−ε turbulence model which contains additional terms to account for the effects of the solid-particles phase on the multiphase turbulence structure. The developed CFD numerical framework has been verified for the relevant forces and all the possible interaction mechanisms of the liquid-solids multiphase flow by investigating four different numerical frameworks, in order to determine the optimum numerical framework that captures the underlying physics and covers the interaction mechanisms that lead to sand deposition and the range of sand transport flow regimes in pipes. The flow of liquid-sand in pipe has been studied extensively and the numerical results of sand concentration distribution across pipe and other flow properties are in good agreement with published experimental data on validation. The numerical framework has been employed to investigate the multiphase flow in V-inclined pipe bends of ±4o−6o, seemingly small inclined bend angles. The predicted results which include the sand segregation, deposition velocity and flow turbulence modulation in the pipe bend show that the seemingly small pipe bends have significant effect on the flow differently from that of horizontal pipes. The pipe bend causes abrupt local change in the multiphase flow characteristic and formation of stationary sand deposit in the pipe at a relatively high flow velocity. The threshold velocity to keep sand entrained in liquid in pipe bends is significantly higher than that required for flow horizontal pipes. A critical implication of this is that the correlations for predicting sand deposition in pipelines must account for the effect of pipe bend on flow characteristics in order to provide accurate predictions of the critical sand transport velocity (MTV) in subsea petroleum flowlines, which V-inclined pipe bends are inevitable due to seabed topology.

Cette thèse propose trois méthodes de traitement du signal orientées vers la surveillance d’état et le diagnostic. Les techniques proposées sont surtout adaptées pour la surveillance d’état, effectuée à la base de vibrations, des machines tournantes qui fonctionnent dans des conditions d’opération non-stationnaires comme par exemple les éoliennes mais elles ne sont pas limitées à un tel usage. Toutes les méthodes proposées sont des algorithmes automatiques et gérés par les données.La première technique proposée permet de sélectionner la partie la plus stationnaire d’un signal en cadrant la représentation temps-fréquence d’un signal.La deuxième méthode est un algorithme pour l’association des dispositions spectrales, des séries harmoniques et des séries à bandes latérales avec des fréquences caractéristiques provennant du cinématique d'un système analysé. Cette méthode propose une approche unique dédiée à l’élément roulant du roulement qui permet de surmonter les difficultés causées par le phénomène de glissement.La troisième technique est un algorithme de démodulation de bande latérale entière. Elle fonctionne à la base d’un filtre multiple et propose des indicateurs de santé pour faciliter une évaluation d'état du système sous l’analyse.Dans cette thèse, les méthodes proposées sont validées sur les signaux simulés et réels. Les résultats présentés montrent une bonne performance de toutes les méthodes. / This thesis proposes a three signal-processing methods oriented towards the condition monitoring and diagnosis. In particular the proposed techniques are suited for vibration-based condition monitoring of rotating machinery which works under highly non-stationary operational condition as wind turbines, but it is not limited to such a usage. All the proposed methods are automatic and data-driven algorithms.The first proposed technique enables a selection of the most stationary part of signal by cropping time-frequency representation of the signal.The second method is an algorithm for association of spectral patterns, harmonics and sidebands series, with characteristic frequencies arising from kinematic of a system under inspection. This method features in a unique approach dedicated for rolling-element bearing which enables to overcome difficulties caused by a slippage phenomenon.The third technique is an all-sideband demodulation algorithm. It features in a multi-rate filter and proposes health indicators to facilitate an evaluation of the condition of the investigated system.In this thesis the proposed methods are validated on both, simulated and real-world signals. The presented results show good performance of all the methods.

This thesis includes solving the basic examples of differential calculus (especially two and three variables). The work covers examples of the solutions of the domain, the first and second partial derivatives, determining the functions of local extremes (both explicitly and implicitly given) and find the equation of the tangent plane to a point in the graphs of functions. The examples are sorted by difficulty.

This paper proposes a new methodology for selecting a pension fund base on the probability of loss for its customers. The methodology includes the historical performance of the portfolio between, March 2006 and May 2013, and projects its profitability either to maturity or to the customer’s retirement age. / El presente trabajo desarrolla una nueva metodología para la selección de fondos gestionados por las Administradoras de Fondos de Pensiones (AFP), basada en la probabilidad de pérdida del afiliado. La metodología incorpora el efecto del desempeño histórico de la gestión del portafolio, en el período marzo 2006-mayo 2013, y lo hace en la proyección de la rentabilidad esperada al vencimiento o edad de jubilación del afiliado. / Este trabalho desenvolve uma nova metodologia para a seleção dos fundos geridos pe las Administradoras deFundos de Pensão (AFP), com base na probabilidade de perda do afiliado. A metodologia considera o efeito do desempenho histórico na gestão das carteiras de março de 2006 até maio de 2013, e o inclui na projeção da taxa derendimento esperada ao vencimento ou até a idade para aposentadoria do afiliado.

Desde os trabalhos de Clausius, Boltzmann e Gibbs, sabe-se que partículas que interagem através de potenciais de curto alcance alcançam, após um processo de relaxação, o estado final estacionário que corresponde ao equilíbrio termodinâmico [I]. Embora nenhuma prova exata exista para isso, na prática, verifica-se que os sistemas não-integráveis com uma energia fixa e um número finito de partículas (ensemble microcanônico, por exemplo) sempre relaxam para um estado estacionário que só depende de quantidades globais conservadas pela dinâmica: energia, momentum e momentum angular. Este estado estacionário corresponde ao estado de equilíbrio termodinâmico e não depende das especificidades da distribuição inicial de partículas. Este cenário muda drasticamente quando a interação entre as partículas passa a ser de longo alcance [2]. A descrição estatística e termodinâmica desses sistemas ainda é objeto de estudo. Contudo, o que se sabe é que esses sistemas têm como propriedade fundamental o fato de que, no limite termodinâmico o tempo de colisão diverge e o equilíbrio termodinâmico nunca é atingido [3]. Nesse trabalho analisamos do ponto de vista teórico e por simulação de dinâmica molecular o estado estacionário atingido por sistemas auto-gravitantes em uma, duas e três dimensões e plasmas não-neutros na dinâmica de um feixe de partículas carregadas. Analisamos ainda um modelo com transição de fases para o estado fora do equilíbrio (HMF). Em todos os casos a teoria proposta na tese mostrou-se consistente com os simulações numéricas empregadas. / Since the work of Clausius, Boltzmann and Gibbs, it is known that particles interacting by a short-range potential, after a relaxation process, reach a final stationary state that corresponds to thermodynamic equilibrium. Although no exact proof exists, in practice non-integrable systems with fixed energy and a finite number of particles (i.e., microcanonical ensemble) always relax to a stationary state that depends only on global quantities conserved by the dynamics: energy, momentum and angular momentum. This stationary state corresponds to the state of thermodynamic equilibrium and does not depend on the specifics of the initial particle distribution. This scenario changes drastically when the interaction between particles is longranged [2] The statistical and thermodynamic description of these systems is still an object of study. However, a fundamental property of these systems is the fact that, in the thermodynamic limit, the collision time diverges and thermodynamic equilibrium is never achieved [3].. In this thesis we analyse, from a theoretical point of view and using molecular dynamics simulations, the stationary state achieved by self-gravitating systems in one, two and three dimensions and non-neutral plasmas in the dynamics of charged particle beams. We also analyse a model with out-of-equilibrium phase transitions (HMF). In all these cases, the theory proposed in this thesis is shown to be consistent with the numerical simulations applied.

La montée en puissance des systèmes de production de l'énergie électrique, à partir de centrales éoliennes ou photovoltaïques, a fait apparaître un besoin aigu en moyens de stockage de cette énergie. Les technologies d'accumulateurs électrochimiques sont à même de répondre à cette problématique, en particulier les batteries sodium chlorure de nickel, une technologie d'accumulateur à haute température. Mais afin de gérer au mieux les performances de ces batteries, il est nécessaire d'avoir une connaissance approfondie du comportement électrothermique des cellules unitaires les composant. Ce travail de thèse présente en détail le développement d'un modèle multiphysique 2D d'une cellule Na-MCl2 commerciale, à même de simuler son fonctionnement en décharge à courant constant. Ce modèle a été construit sur une étude approfondie des mécanismes électrochimiques à l'œuvre dans ce type de technologie, et des effets thermiques associés. Il repose également sur la synthèse et la critique des modèles proposés précédemment dans la littérature. Le modèle inédit proposé dans ce manuscrit permet de prendre en compte deux réactions électrochimiques simultanées, ainsi que le comportement thermique de la cellule. Le modèle a été validé en confrontant les résultats de simulations aux performances électriques réelles d'une cellule commerciale Na-MCl2 (incluant une caractérisation des phénomènes électrochimiques par voltammetrie cyclique). Ce modèle permet de simuler avec succès des décharges à courant constant à différents régimes, et à partir de différentes températures. Il est également capable de prévoir les effets électrothermiques qui seraient consécutifs à un changement de design de la cellule (design structurel et de constitution). / The growing production capacity of wind turbines or photo-voltaic plants has revealed an acute need for electric energy storage systems. Electrochemical accumulator technologies are able to address this issue, in particular sodium nickel chloride batteries, a high temperature battery system. But to enhance the energy management of these batteries, it is necessary to have a deep knowledge of the electro-thermal behavior of its unit cell component. This thesis presents the development of a comprehensive multiphysics 2D model of a commercial Na-MCl2 cell, able to simulate its operation in constant current discharge. This model was built on a thorough study of electrochemical mechanisms at work with this type of technology, and its associated thermal effects. It is also based on the synthesis of existing models, and their critique. The new model proposed in this manuscript can take into account two simultaneous electrochemical reactions, as well as the thermal behavior of the cell. The model was validated by comparison between simulation results and the actual electrical performances of a commercial Na-MCl2 cell (including the characterization of electrochemical phenomena by cyclic voltammetry). This model successfully simulates constant current discharges at different rates, and from different temperatures. It is also capable of predicting the electro-thermal effects consecutive to design changes of the cell (structural and relevant to the active material).

Made available in DSpace on 2016-06-02T20:27:39Z (GMT). No. of bitstreams: 1 3744.pdf: 1323892 bytes, checksum: 71a7a7180d61db167b8cbec4db2bbe8b (MD5) Previous issue date: 2011-05-30 / Universidade Federal de Sao Carlos / We address the question of existence of stationary stable solutions to a class of reaction-diffusion equations with spatial dependence in 2 and 3-dimensional bounded domains. The approach consists of proving the existence of local minimizer of the corres-ponding energy functional. For existence, it was enough to give sufficient conditions on the diffusion coefficient and on the reaction term to ensure the existence of isolated mi¬nima of the &#915;limit functional of the energy functional family. In the second part we take the techniques developed in the first part to minimize functional in 2 and 3-dimensional rectangles, with and without constraint, solving in a more general form this problem, which was originaly proposed in 1989 by Robert Kohn and Peter Sternberg. / Na primeira parte deste trabalho, abordamos a existência de soluções estacioná-rias estáveis para uma classe de equações de reação-difusão com dependência espacial em domínios limitados 2 e 3-dimensionais. Esta abordagem foi feita via existência de míni¬mos locais dos funcionais de energia correspondentes. Para tal, foi suficiente encontrar condições no coeficiente de difusão e no termo de reação que garantam existência de míni¬mos isolados do funcional &#915;limite da família de funcionais de energia. Na segunda parte, aproveitamos as técnicas desenvolvidas na primeira parte para minimizar funcionais em retângulos e paralelepípedos, com e sem vínculo, resolvendo de forma bem mais geral este problema, originalmente proposto em 1989 por Robert Kohn e Peter Sternberg.

[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.