Root parametrized differential equations

The present thesis is about the inverse problem in differential Galois Theory. Given a differential field, the inverse  problem asks which linear algebraic groups can be realized as differential Galois groups of Picard-Vessiot extensions of this field.   In this thesis we will concentrate on the realization of the classical groups as differential Galois groups. We introduce a method for a very general realization of these groups. This means that we present for the classical groups of Lie rank $l$ explicit linear differential equations where the coefficients are differential polynomials in $l$ differential indeterminates over an algebraically closed field of constants $C$, i.e. our differential ground field is purely differential transcendental over the constants.   For the groups of type $A_l$, $B_l$, $C_l$, $D_l$ and $G_2$ we managed to do these realizations at the same time in terms of Abhyankar's program 'Nice Equations for Nice Groups'. Here the choice of the defining matrix is important. We found out that an educated choice of $l$ negative roots for the parametrization together with the positive simple roots leads to a nice differential equation and at the same time defines a sufficiently general element of the Lie algebra. Unfortunately for the groups of type $F_4$ and $E_6$ the linear differential equations for such elements are of enormous length. Therefore we keep in the case of $F_4$ and $E_6$ the defining matrix differential equation which has also an easy and nice shape.   The basic idea for the realization is the application of an upper and lower bound criterion for the differential Galois group to our parameter equations and to show that both bounds coincide. An upper and lower bound criterion can be found in literature. Here we will only use the upper bound, since for the application of the lower bound criterion an important condition has to be satisfied. If the differential ground field is $C_1$, e.g., $C(z)$ with standard derivation, this condition is automatically satisfied. Since our differential ground field is purely differential transcendental over $C$, we have no information whether this condition holds or not.   The main part of this thesis is the development of an alternative lower bound criterion and its application. We introduce the specialization bound. It states that the differential Galois group of a specialization of the parameter equation is contained in the differential Galois group of the parameter equation. Thus for its application we need a differential equation over $C(z)$ with given differential Galois group. A modification of a result from Mitschi and Singer yields such an equation over $C(z)$ up to differential conjugation, i.e. up to transformation to the required shape. The transformation of their equation to a specialization of our parameter equation is done for each of the above groups in the respective transformation lemma.

Surface Reflectance Recognition and Real-World Illumination Statistics

Humans distinguish materials such as metal, plastic, and paper effortlessly at a glance. Traditional computer vision systems cannot solve this problem at all. Recognizing surface reflectance properties from a single photograph is difficult because the observed image depends heavily on the amount of light incident from every direction. A mirrored sphere, for example, produces a different image in every environment. To make matters worse, two surfaces with different reflectance properties could produce identical images. The mirrored sphere simply reflects its surroundings, so in the right artificial setting, it could mimic the appearance of a matte ping-pong ball. Yet, humans possess an intuitive sense of what materials typically "look like" in the real world. This thesis develops computational algorithms with a similar ability to recognize reflectance properties from photographs under unknown, real-world illumination conditions. Real-world illumination is complex, with light typically incident on a surface from every direction. We find, however, that real-world illumination patterns are not arbitrary. They exhibit highly predictable spatial structure, which we describe largely in the wavelet domain. Although they differ in several respects from the typical photographs, illumination patterns share much of the regularity described in the natural image statistics literature. These properties of real-world illumination lead to predictable image statistics for a surface with given reflectance properties. We construct a system that classifies a surface according to its reflectance from a single photograph under unknown illuminination. Our algorithm learns relationships between surface reflectance and certain statistics computed from the observed image. Like the human visual system, we solve the otherwise underconstrained inverse problem of reflectance estimation by taking advantage of the statistical regularity of illumination. For surfaces with homogeneous reflectance properties and known geometry, our system rivals human performance.

Efficient non-linear data assimilation in geophysical fluid dynamics

New ways of combining observations with numerical models are discussed in which the size of the state space can be very large, and the model can be highly nonlinear. Also the observations of the system can be related to the model variables in highly nonlinear ways, making this data-assimilation (or inverse) problem highly nonlinear. First we discuss the connection between data assimilation and inverse problems, including regularization. We explore the choice of proposal density in a Particle Filter and show how the ’curse of dimensionality’ might be beaten. In the standard Particle Filter ensembles of model runs are propagated forward in time until observations are encountered, rendering it a pure Monte-Carlo method. In large-dimensional systems this is very inefficient and very large numbers of model runs are needed to solve the data-assimilation problem realistically. In our approach we steer all model runs towards the observations resulting in a much more efficient method. By further ’ensuring almost equal weight’ we avoid performing model runs that are useless in the end. Results are shown for the 40 and 1000 dimensional Lorenz 1995 model.

Nonlinear error dynamics for cycled data assimilation methods

We investigate the error dynamics for cycled data assimilation systems, such that the inverse problem of state determination is solved at tk, k = 1, 2, 3, ..., with a first guess given by the state propagated via a dynamical system model from time tk − 1 to time tk. In particular, for nonlinear dynamical systems that are Lipschitz continuous with respect to their initial states, we provide deterministic estimates for the development of the error ||ek|| := ||x(a)k − x(t)k|| between the estimated state x(a) and the true state x(t) over time. Clearly, observation error of size δ > 0 leads to an estimation error in every assimilation step. These errors can accumulate, if they are not (a) controlled in the reconstruction and (b) damped by the dynamical system under consideration. A data assimilation method is called stable, if the error in the estimate is bounded in time by some constant C. The key task of this work is to provide estimates for the error ||ek||, depending on the size δ of the observation error, the reconstruction operator Rα, the observation operator H and the Lipschitz constants K(1) and K(2) on the lower and higher modes of controlling the damping behaviour of the dynamics. We show that systems can be stabilized by choosing α sufficiently small, but the bound C will then depend on the data error δ in the form c||Rα||δ with some constant c. Since ||Rα|| → ∞ for α → 0, the constant might be large. Numerical examples for this behaviour in the nonlinear case are provided using a (low-dimensional) Lorenz '63 system.

A Variance-Minimizing Filter for Large-Scale Applications

A truly variance-minimizing filter is introduced and its per for mance is demonstrated with the Korteweg– DeV ries (KdV) equation and with a multilayer quasigeostrophic model of the ocean area around South Africa. It is recalled that Kalman-like filters are not variance minimizing for nonlinear model dynamics and that four - dimensional variational data assimilation (4DV AR)-like methods relying on per fect model dynamics have dif- ficulty with providing error estimates. The new method does not have these drawbacks. In fact, it combines advantages from both methods in that it does provide error estimates while automatically having balanced states after analysis, without extra computations. It is based on ensemble or Monte Carlo integrations to simulate the probability density of the model evolution. When obser vations are available, the so-called importance resampling algorithm is applied. From Bayes’ s theorem it follows that each ensemble member receives a new weight dependent on its ‘ ‘distance’ ’ t o the obser vations. Because the weights are strongly var ying, a resampling of the ensemble is necessar y. This resampling is done such that members with high weights are duplicated according to their weights, while low-weight members are largely ignored. In passing, it is noted that data assimilation is not an inverse problem by nature, although it can be for mulated that way . Also, it is shown that the posterior variance can be larger than the prior if the usual Gaussian framework is set aside. However , i n the examples presented here, the entropy of the probability densities is decreasing. The application to the ocean area around South Africa, gover ned by strongly nonlinear dynamics, shows that the method is working satisfactorily . The strong and weak points of the method are discussed and possible improvements are proposed.

Birkhoffian Systems in Infinite Dimensional Manifolds

We generalize the theory of Kobayashi and Oliva (On the Birkhoff Approach to Classical Mechanics. Resenhas do Instituto de Matematica e Estatistica da Universidade de Sao Paulo, 2003) to infinite dimensional Banach manifolds with a view towards applications in partial differential equations.

Affine processes, arbitrage-Free Term structures of legendre polynomials,and option pricing

Multivariate Affine term structure models have been increasingly used for pricing derivatives in fixed income markets. In these models, uncertainty of the term structure is driven by a state vector, while the short rate is an affine function of this vector. The model is characterized by a specific form for the stochastic differential equation (SDE) for the evolution of the state vector. This SDE presents restrictions on its drift term which rule out arbitrages in the market. In this paper we solve the following inverse problem: Suppose the term structure of interest rates is modeled by a linear combination of Legendre polynomials with random coefficients. Is there any SDE for these coefficients which rules out arbitrages? This problem is of particular empirical interest because the Legendre model is an example of factor model with clear interpretation for each factor, in which regards movements of the term structure. Moreover, the Affine structure of the Legendre model implies knowledge of its conditional characteristic function. From the econometric perspective, we propose arbitrage-free Legendre models to describe the evolution of the term structure. From the pricing perspective, we follow Duffie et al. (2000) in exploring Legendre conditional characteristic functions to obtain a computational tractable method to price fixed income derivatives. Closing the article, the empirical section presents precise evidence on the reward of implementing arbitrage-free parametric term structure models: The ability of obtaining a good approximation for the state vector by simply using cross sectional data.

Análise da modelagem experimental da perda de injetividade

Injectivity decline, which can be caused by particle retention, generally occurs during water injection or reinjection in oil fields. Several mechanisms, including straining, are responsible for particle retention and pore blocking causing formation damage and injectivity decline. Predicting formation damage and injectivity decline is essential in waterflooding projects. The Classic Model (CM), which incorporates filtration coefficients and formation damage functions, has been widely used to predict injectivity decline. However, various authors have reported significant discrepancies between Classical Model and experimental results, motivating the development of deep bed filtration models considering multiple particle retention mechanisms (Santos & Barros, 2010; SBM). In this dissertation, inverse problem solution was studied and a software for experimental data treatment was developed. Finally, experimental data were fitted using both the CM and SBM. The results showed that, depending on the formation damage function, the predictions for injectivity decline using CM and SBM models can be significantly different

Efficient robust optimization based on polynomial chaos and tabu search algorithm

The study of robust design methodologies and techniques has become a new topical area in design optimizations in nearly all engineering and applied science disciplines in the last 10 years due to inevitable and unavoidable imprecision or uncertainty which is existed in real word design problems. To develop a fast optimizer for robust designs, a methodology based on polynomial chaos and tabu search algorithm is proposed. In the methodology, the polynomial chaos is employed as a stochastic response surface model of the objective function to efficiently evaluate the robust performance parameter while a mechanism to assign expected fitness only to promising solutions is introduced in tabu search algorithm to minimize the requirement for determining robust metrics of intermediate solutions. The proposed methodology is applied to the robust design of a practical inverse problem with satisfactory results.

Detecção de falhas em estruturas inteligentes usando otimização por nuvem de partículas: fundamentos e estudo de casos

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Nonlinear parameter estimation in laminar forced convection within a circular sector tube

In this article we examine an inverse heat convection problem of estimating unknown parameters of a parameterized variable boundary heat flux. The physical problem is a hydrodynamically developed, thermally developing, three-dimensional steady state laminar flow of a Newtonian fluid inside a circular sector duct, insulated in the flat walls and subject to unknown wall heat flux at the curved wall. Results are presented for polynomial and sinusoidal trial functions, and the unknown parameters as well as surface heat fluxes are determined. Depending on the nature of the flow, on the position of experimental points the inverse problem sometimes could not be solved. Therefore, an identification condition is defined to specify a condition under which the inverse problem can be solved. Once the parameters have been computed it is possible to obtain the statistical significance of the inverse problem solution. Therefore, approximate confidence bounds based on standard statistical linear procedure, for the estimated parameters, are analyzed and presented.

A new implementation of population based incremental learning method for optimizations in electromagnetics

To enhance the global search ability of population based incremental learning (PBIL) methods, it is proposed that multiple probability vectors are to be included on available PBIL algorithms. The strategy for updating those probability vectors and the negative learning and mutation operators are thus re-defined correspondingly. Moreover, to strike the best tradeoff between exploration and exploitation searches, an adaptive updating strategy for the learning rate is designed. Numerical examples are reported to demonstrate the pros and cons of the newly implemented algorithm.

Optimization procedures in the design of continuous induction hardening and tempering installations for magnetic steel bars

An inverse problem concerning the industrial process of steel bars hardening and tempering is considered. The associated optimization problem is formulated in terms of membership functions and, for the sake of comparison, also in terms of quadratic residuals; both geometric and electromagnetic design variables have been considered. The numerical solution is achieved by coupling a finite difference procedure for the calculation of the electromagnetic and thermal fields to a deterministic strategy of minimization based on modified Flctcher and Reeves method. © 1998 IEEE.

Detecção de falhas estruturais usando sensores e atuadores piezelétricos e algoritmos genéticos

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Mapeamento do relevo do embasamento de bacias sedimentares através da inversão gravimétrica vinculada

Apresentamos três novos métodos estáveis de inversão gravimétrica para estimar o relevo de uma interface arbitrária separando dois meios. Para a garantia da estabilidade da solução, introduzimos informações a priori sobre a interface a ser mapeada, através da minimização de um (ou mais) funcional estabilizante. Portanto, estes três métodos se diferenciam pelos tipos de informação físico-geológica incorporados. No primeiro método, denominado suavidade global, as profundidades da interface são estimadas em pontos discretos, presumindo-se o conhecimento a priori sobre o contraste de densidade entre os meios. Para a estabilização do problema inverso introduzimos dois vínculos: (a) proximidade entre as profundidades estimadas e verdadeiras da interface em alguns pontos fornecidas por furos de sondagem; e (b) proximidade entre as profundidades estimadas em pontos adjacentes. A combinação destes dois vínculos impõe uma suavidade uniforme a toda interface estimada, minimizando, simultaneamente em alguns pontos, os desajustes entre as profundidades conhecidas pelas sondagens e as estimadas nos mesmos pontos. O segundo método, denominado suavidade ponderada, estima as profundidades da interface em pontos discretos, admitindo o conhecimento a priori do contraste de densidade. Neste método, incorpora-se a informação geológica que a interface é suave, exceto em regiões de descontinuidades produzidas por falhas, ou seja, a interface é predominantemente suave porém localmente descontínua. Para a incorporação desta informação, desenvolvemos um processo iterativo em que três tipos de vínculos são impostos aos parâmetros: (a) ponderação da proximidade entre as profundidades estimadas em pontos adjacentes; (b) limites inferior e superior para as profundidades; e (c) proximidade entre todas as profundidades estimadas e um valor numérico conhecido. Inicializando com a solução estimada pelo método da suavidade global, este segundo método, iterativamente, acentua as feições geométricas presentes na solução inicial; ou seja, regiões suaves da interface tendem a tornar-se mais suaves e regiões abruptas tendem a tornar-se mais abruptas. Para tanto, este método atribui diferentes pesos ao vínculo de proximidade entre as profundidades adjacentes. Estes pesos são automaticamente atualizados de modo a acentuar as descontinuidades sutilmente detectadas pela solução da suavidade global. Os vínculos (b) e (c) são usados para compensar a perda da estabilidade, devida à introdução de pesos próximos a zero em alguns dos vínculos de proximidade entre parâmetros adjacentes, e incorporar a informação a priori que a região mais profunda da interface apresenta-se plana e horizontal. O vínculo (b) impõe, de modo estrito, que qualquer profundidade estimada é não negativa e menor que o valor de máxima profundidade da interface conhecido a priori; o vínculo (c) impõe que todas as profundidades estimadas são próximas a um valor que deliberadamente viola a profundidade máxima da interface. O compromisso entre os vínculos conflitantes (b) e (c) resulta na tendenciosidade da solução final em acentuar descontinuidades verticais e apresentar uma estimativa suave e achatada da região mais profunda. O terceiro método, denominado mínimo momento de inércia, estima os contrastes de densidade de uma região da subsuperfície discretizada em volumes elementares prismáticos. Este método incorpora a informação geológica que a interface a ser mapeada delimita uma fonte anômala que apresenta dimensões horizontais maiores que sua maior dimensão vertical, com bordas mergulhando verticalmente ou em direção ao centro de massa e que toda a massa (ou deficiência de massa) anômala está concentrada, de modo compacto, em torno de um nível de referência. Conceitualmente, estas informações são introduzidas pela minimização do momento de inércia das fontes em relação ao nível de referência conhecido a priori. Esta minimização é efetuada em um subespaço de parâmetros consistindo de fontes compactas e apresentando bordas mergulhando verticalmente ou em direção ao centro de massa. Efetivamente, estas informações são introduzidas através de um processo iterativo inicializando com uma solução cujo momento de inércia é próximo a zero, acrescentando, em cada iteração, uma contribuição com mínimo momento de inércia em relação ao nível de referência, de modo que a nova estimativa obedeça a limites mínimo e máximo do contraste de densidade, e minimize, simultaneamente, os desajustes entre os dados gravimétricos observados e ajustados. Adicionalmente, o processo iterativo tende a "congelar" as estimativas em um dos limites (mínimo ou máximo). O resultado final é uma fonte anômala compactada em torno do nível de referência cuja distribuição de constraste de densidade tende ao limite superior (em valor absoluto) estabelecido a priori. Estes três métodos foram aplicados a dados sintéticos e reais produzidos pelo relevo do embasamento de bacias sedimentares. A suavidade global produziu uma boa reconstrução do arcabouço de bacias que violam a condição de suavidade, tanto em dados sintéticos como em dados da Bacia do Recôncavo. Este método, apresenta a menor resolução quando comparado com os outros dois métodos. A suavidade ponderada produziu uma melhoria na resolução de relevos de embasamentos que apresentam falhamentos com grandes rejeitos e altos ângulos de mergulho, indicando uma grande potencialidade na interpretação do arcabouço de bacias extensionais, como mostramos em testes com dados sintéticos e dados do Steptoe Valley, Nevada, EUA, e da Bacia do Recôncavo. No método do mínimo momento de inércia, tomou-se como nível de referência o nível médio do terreno. As aplicações a dados sintéticos e às anomalias Bouguer do Graben de San Jacinto, California, EUA, e da Bacia do Recôncavo mostraram que, em comparação com os métodos da suavidade global e ponderada, este método estima com excelente resolução falhamentos com pequenos rejeitos sem impor a restrição da interface apresentar poucas descontinuidades locais, como no método da suavidade ponderada.