Certified Rapid Solution of Parametrized Linear Elliptic Equations: Application to Parameter Estimation

We present a technique for the rapid and reliable evaluation of linear-functional output of elliptic partial differential equations with affine parameter dependence. The essential components are (i) rapidly uniformly convergent reduced-basis approximations — Galerkin projection onto a space WN spanned by solutions of the governing partial differential equation at N (optimally) selected points in parameter space; (ii) a posteriori error estimation — relaxations of the residual equation that provide inexpensive yet sharp and rigorous bounds for the error in the outputs; and (iii) offline/online computational procedures — stratagems that exploit affine parameter dependence to de-couple the generation and projection stages of the approximation process. The operation count for the online stage — in which, given a new parameter value, we calculate the output and associated error bound — depends only on N (typically small) and the parametric complexity of the problem. The method is thus ideally suited to the many-query and real-time contexts. In this paper, based on the technique we develop a robust inverse computational method for very fast solution of inverse problems characterized by parametrized partial differential equations. The essential ideas are in three-fold: first, we apply the technique to the forward problem for the rapid certified evaluation of PDE input-output relations and associated rigorous error bounds; second, we incorporate the reduced-basis approximation and error bounds into the inverse problem formulation; and third, rather than regularize the goodness-of-fit objective, we may instead identify all (or almost all, in the probabilistic sense) system configurations consistent with the available experimental data — well-posedness is reflected in a bounded "possibility region" that furthermore shrinks as the experimental error is decreased.

Exploring the spatial relations between soil physical properties and apparent electrical conductivity

Relations between the apparent electrical conductivity of the soil (ECa) and top- and sub-soil physical properties were examined for two arable fields in southern England (Crowmarsh Battle Farms and the Yattendon Estate). The spatial variation of ECa and the soil properties was explored geostatistically. The variogram ranges showed that ECa varied on a similar spatial scale to many of the soil physical properties in both fields. Several features in the map of kriged predictions of ECa were also evident in maps of the soil properties. In addition, the correlation coefficients showed a strong relation between ECa and several soil properties. A moving correlation analysis enabled differences in the relations between ECa and the soil properties to be examined within the fields. The results indicated that relations were inconsistent; they were stronger in some areas than others. A regression of ECa on the principal component scores of the leading components for both fields showed that the first two components accounted for a large proportion of the variance in ECa, whereas the others accounted for little or none. For Crowmarsh topsoil sand and clay, loss on ignition and volumetric water measured in the autumn had large correlations on the first component, and for Yattendon they were large for topsoil sand and clay, and autumn and spring volumetric water. The cross-variograms suggested strong coregionalization between ECa and several soil physical properties; in particular subsoil sand and silt at Crowmarsh, and subsoil sand and clay at Yattendon. The structural correlations from the linear model of coregionalization confirmed the strength of the relations between ECa and the subsoil properties. Nevertheless, no one property was consistently important for both fields. Although a map of ECa can indicate the general patterns of spatial variation in the soil, it is not a substitute for information on soil properties obtained by sampling and analysing the soil. Nevertheless, it could be used to guide further sampling. (c) 2005 Elsevier B.V. All rights reserved.

Approximate factorization constraint preconditioners for saddle-point matrices

We consider the application of the conjugate gradient method to the solution of large, symmetric indefinite linear systems. Special emphasis is put on the use of constraint preconditioners and a new factorization that can reduce the number of flops required by the preconditioning step. Results concerning the eigenvalues of the preconditioned matrix and its minimum polynomial are given. Numerical experiments validate these conclusions.

A new linear initialization in SOM for biomolecular data

In the past decade, the amount of data in biological field has become larger and larger; Bio-techniques for analysis of biological data have been developed and new tools have been introduced. Several computational methods are based on unsupervised neural network algorithms that are widely used for multiple purposes including clustering and visualization, i.e. the Self Organizing Maps (SOM). Unfortunately, even though this method is unsupervised, the performances in terms of quality of result and learning speed are strongly dependent from the neuron weights initialization. In this paper we present a new initialization technique based on a totally connected undirected graph, that report relations among some intersting features of data input. Result of experimental tests, where the proposed algorithm is compared to the original initialization techniques, shows that our technique assures faster learning and better performance in terms of quantization error.

A stable one-step-ahead predictive control of non-linear systems

In this paper stability of one-step ahead predictive controllers based on non-linear models is established. It is shown that, under conditions which can be fulfilled by most industrial plants, the closed-loop system is robustly stable in the presence of plant uncertainties and input–output constraints. There is no requirement that the plant should be open-loop stable and the analysis is valid for general forms of non-linear system representation including the case out when the problem is constraint-free. The effectiveness of controllers designed according to the algorithm analyzed in this paper is demonstrated on a recognized benchmark problem and on a simulation of a continuous-stirred tank reactor (CSTR). In both examples a radial basis function neural network is employed as the non-linear system model.

Evidence of dispersion relations for the nonlinear response of the Lorenz 63 system

Along the lines of the nonlinear response theory developed by Ruelle, in a previous paper we have proved under rather general conditions that Kramers-Kronig dispersion relations and sum rules apply for a class of susceptibilities describing at any order of perturbation the response of Axiom A non equilibrium steady state systems to weak monochromatic forcings. We present here the first evidence of the validity of these integral relations for the linear and the second harmonic response for the perturbed Lorenz 63 system, by showing that numerical simulations agree up to high degree of accuracy with the theoretical predictions. Some new theoretical results, showing how to derive asymptotic behaviors and how to obtain recursively harmonic generation susceptibilities for general observables, are also presented. Our findings confirm the conceptual validity of the nonlinear response theory, suggest that the theory can be extended for more general non equilibrium steady state systems, and shed new light on the applicability of very general tools, based only upon the principle of causality, for diagnosing the behavior of perturbed chaotic systems and reconstructing their output signals, in situations where the fluctuation-dissipation relation is not of great help.

Conditioning of the weak-constraint variational data assimilation problem for numerical weather prediction

4-Dimensional Variational Data Assimilation (4DVAR) assimilates observations through the minimisation of a least-squares objective function, which is constrained by the model flow. We refer to 4DVAR as strong-constraint 4DVAR (sc4DVAR) in this thesis as it assumes the model is perfect. Relaxing this assumption gives rise to weak-constraint 4DVAR (wc4DVAR), leading to a different minimisation problem with more degrees of freedom. We consider two wc4DVAR formulations in this thesis, the model error formulation and state estimation formulation. The 4DVAR objective function is traditionally solved using gradient-based iterative methods. The principle method used in Numerical Weather Prediction today is the Gauss-Newton approach. This method introduces a linearised `inner-loop' objective function, which upon convergence, updates the solution of the non-linear `outer-loop' objective function. This requires many evaluations of the objective function and its gradient, which emphasises the importance of the Hessian. The eigenvalues and eigenvectors of the Hessian provide insight into the degree of convexity of the objective function, while also indicating the difficulty one may encounter while iterative solving 4DVAR. The condition number of the Hessian is an appropriate measure for the sensitivity of the problem to input data. The condition number can also indicate the rate of convergence and solution accuracy of the minimisation algorithm. This thesis investigates the sensitivity of the solution process minimising both wc4DVAR objective functions to the internal assimilation parameters composing the problem. We gain insight into these sensitivities by bounding the condition number of the Hessians of both objective functions. We also precondition the model error objective function and show improved convergence. We show that both formulations' sensitivities are related to error variance balance, assimilation window length and correlation length-scales using the bounds. We further demonstrate this through numerical experiments on the condition number and data assimilation experiments using linear and non-linear chaotic toy models.

A new adaptive multiple modelling approach for non-linear and non-stationary systems

This paper proposes a novel adaptive multiple modelling algorithm for non-linear and non-stationary systems. This simple modelling paradigm comprises K candidate sub-models which are all linear. With data available in an online fashion, the performance of all candidate sub-models are monitored based on the most recent data window, and M best sub-models are selected from the K candidates. The weight coefficients of the selected sub-model are adapted via the recursive least square (RLS) algorithm, while the coefficients of the remaining sub-models are unchanged. These M model predictions are then optimally combined to produce the multi-model output. We propose to minimise the mean square error based on a recent data window, and apply the sum to one constraint to the combination parameters, leading to a closed-form solution, so that maximal computational efficiency can be achieved. In addition, at each time step, the model prediction is chosen from either the resultant multiple model or the best sub-model, whichever is the best. Simulation results are given in comparison with some typical alternatives, including the linear RLS algorithm and a number of online non-linear approaches, in terms of modelling performance and time consumption.

Análise não linear física e geométrica de pórticos espaciais

Apresenta-se uma formulação do tipo incrementaliterativa destinada a análise não linear de pórticos espaciais. Considera-se os efeitos não lineares introduzidos pelas mudanças de configuração geométrica da estrutura e também pela combinação destes efeitos com aqueles inerentes ao comportamento plástico exibido pelo material. As relações cinemáticas empregadas permitem a consideração de deslocamentos arbitrariamente grandes, acompanhadas de pequenas deformações . A modelagem do comportamento plástico do material é efetuada através do conceito de rótula plástica, estabelecido a partir de um critério de plastificação generalizado. Adota-se uma matriz de rigidez geométrica de barra baseada em momentos semitangenciais. Para elementos com extremos plastificados, é deduzida uma matriz de rigidez elasto-plástica. Emprega-se um método numérico do tipo incremental-iterativo, que utiliza como condição básica de controle da análise a constância do trabalho realizado pelos incrementos de cargas, em cada passo incremental (Método de Controle por Trabalho).A formulação permite uma descricão completa do desempenho mecânico da estrutura, inclusive em estágio de deformação pós-crítico em que ocorre regressão do carregamento com aumento de deslocamentos, ou vice-versa. A formulação foi implementada em um programa computacional elaborado em linguagem FORTRAN. Vários exemplos numéricos são apresentados para mostrar a eficiência das procedimentos propostos.

Joint dynamic probabilistic constraints with projected linear decision rules

We consider multistage stochastic linear optimization problems combining joint dynamic probabilistic constraints with hard constraints. We develop a method for projecting decision rules onto hard constraints of wait-and-see type. We establish the relation between the original (in nite dimensional) problem and approximating problems working with projections from di erent subclasses of decision policies. Considering the subclass of linear decision rules and a generalized linear model for the underlying stochastic process with noises that are Gaussian or truncated Gaussian, we show that the value and gradient of the objective and constraint functions of the approximating problems can be computed analytically.

Graphical constraints: a graphical user interface for constraint problems

A constraint satisfaction problem is a classical artificial intelligence paradigm characterized by a set of variables (each variable with an associated domain of possible values), and a set of constraints that specify relations among subsets of these variables. Solutions are assignments of values to all variables that satisfy all the constraints. Many real world problems may be modelled by means of constraints. The range of problems that can use this representation is very diverse and embraces areas like resource allocation, scheduling, timetabling or vehicle routing. Constraint programming is a form of declarative programming in the sense that instead of specifying a sequence of steps to execute, it relies on properties of the solutions to be found, which are explicitly defined by constraints. The idea of constraint programming is to solve problems by stating constraints which must be satisfied by the solutions. Constraint programming is based on specialized constraint solvers that take advantage of constraints to search for solutions. The success and popularity of complex problem solving tools can be greatly enhanced by the availability of friendly user interfaces. User interfaces cover two fundamental areas: receiving information from the user and communicating it to the system; and getting information from the system and deliver it to the user. Despite its potential impact, adequate user interfaces are uncommon in constraint programming in general. The main goal of this project is to develop a graphical user interface that allows to, intuitively, represent constraint satisfaction problems. The idea is to visually represent the variables of the problem, their domains and the problem constraints and enable the user to interact with an adequate constraint solver to process the constraints and compute the solutions. Moreover, the graphical interface should be capable of configure the solver’s parameters and present solutions in an appealing interactive way. As a proof of concept, the developed application – GraphicalConstraints – focus on continuous constraint programming, which deals with real valued variables and numerical constraints (equations and inequalities). RealPaver, a state-of-the-art solver in continuous domains, was used in the application. The graphical interface supports all stages of constraint processing, from the design of the constraint network to the presentation of the end feasible space solutions as 2D or 3D boxes.

Influência dos intermediários de distribuição de produtos turísticos no processo de decisão do meio de hospedagem pelo consumidor

This study examines the complex hotel buyer decision process in front of the tourism distribution channels. Its objective is to describe the influence level of the tourism marketing intermediaries, mainly the travel agents and tour operators, over the hotel decision process by the buyer-tourist. The data collection process was done trough a survey with three hundred brazilian tourists hosted in nineteen hotels of Natal, capital of Rio Grande do Norte, Brazil. The data analysis was done using some multivariate statistic techniques as correlation analysis, multiple regression analysis, factor analysis and multiple discriminant analysis. The research characterizes the hotel services consumers profile and his trip, and identifying the distribution channels used by them. Furthermore, the research verifies the intermediaries influence exercised over hotel buyer decision process, looking for identify causality relations between the influence level and the buyer profile. Verifies that information about hotels available on internet reduces the probability that this influence can be practiced; however it was possible identifying those consumers considers this information complementary and non-substitutes than the information from intermediaries. The characteristics of the data do not allow indentifying the factors that constraint the intermediaries influence neither identifying discriminant functions of the specific distribution channel choice by consumers. The study concludes that consumers don t agree in have been influenced by intermediaries or don t know if they have, still considering important to consult them and internet doesn t substitute their function as information source

A Geração de inovação na carcinicultura do RN: Uma análise a partir da imersão social

This dissertation examines the organizational innovation as a nonlinear process, which occurs in a social and political context and, therefore, socially immersed. Examines the case of shrimp in the state of RN, starting from the following problem: although the norteriograndense shrimp occupies the largest producer of farmed shrimp from Brazil, has a series of bottlenecks concerning the generation of industry innovation, concerning the social relationships and policies between the various actors in the network, whether private or public, and its consequences in terms of opportunity and limits generated for the innovative dynamics. The objective of the research is to understand how the social embeddedness of political actors affects norteriograndense shrimp within the context of structural relations, the industry generation of innovation, throughout its technological trajectory . The approach of social embeddedness balances atomised perspectives, undersocialized and oversocialized, of economic action, considering both the human capacity to act as sources of constraint, whose mechanisms are analyzed the structural and political. In methodological terms this is a case study, analyzed from the research literature, documentary and experimental. Primary data were collected through semi-structured interviews and analyzed in depth by the technique of content analysis. Was adopted a longitudinal approach, seeking to understand the phenomenon from the perspective of the subjects, describing it in an inductive process of investigation. After characterizing the sector and defining their technological trajectory, the analysis of the results followed its four stages: (1) Introduction of Technology: 1973-1980, (2) Intensification of Research: 1981-1991, (3) Technological Adaptation, 1992 -2003, (4) Technological Crisis: 2004-2009. A cross-sectional analysis along the evolutionary trajectory revealed the character of structural changes and policies over time, and implications on the generating process of innovation. Note that, the technological limit to which the sector reached requires changes in technology standards, but is more likely that the potiguar shrimp is entering a new phase of his career in technology rather than a new technological paradigm

Transferência de calor transiente na agitação linear intermitente de latas

Foi estudada a transferência de calor transiente na agitação linear e intermitente (ALI) de embalagens metálicas contendo simulantes de alimentos, objetivando-se sua aplicação em processos de pasteurização ou esterilização e conseqüentes tratamentos térmicos mais eficientes, homogêneos e com produto de melhor qualidade. Foram utilizados quatro meios fluidos simulantes de alimentos de diferentes viscosidades e massas específicas: três óleos e água. Foram combinados efeitos de cinco tratamentos, sendo: meio simulante (4 níveis), espaço livre (3 níveis), freqüência de agitação (4 níveis), amplitude de agitação (2 níveis) e posição das latas (4 níveis). Os ensaios de aquecimento e resfriamento foram feitos em tanque com água à temperatura de 98 °C e 17-20 °C, respectivamente. Com os dados de penetração de calor em cada experimento, foram calculados os parâmetros de penetração de calor fh, jh, fc e jc. Os resultados foram modelados utilizando-se grupos de números adimensionais e expressos em termos de Nusselt, Prandtl, Reynolds e funções trigonométricas (com medidas de amplitude e freqüência de agitação, espaço livre e dimensões da embalagem). Foram estabelecidas as duas Equações gerais para as fases de aquecimento e resfriamento: Nu = ReA 0,199.Pr 0,288.sen(xa/AM)0,406.cos(xf/FA) 1,039.cos((xf/FA).(EL/H).p) 4,556 Aquecimento Nu = 0,1295.ReA 0,047.Pr 0,193.sen(xa/AM)0,114.cos(xf/FA) 0,641.cos((xf/FA).(EL/H).p) 2,476 Resfriamento O processo de ALI pode ser aplicado em pasteurizadores ou autoclaves estáticas horizontais e verticais, com modificações simples. Concluiu-se que a ALI aumenta significativamente a taxa de transferência de calor, tanto no aquecimento como no resfriamento.

Sistema de navegação para robôs móveis autônomos

The main task and one of the major mobile robotics problems is its navigation process. Conceptualy, this process means drive the robot from an initial position and orientation to a goal position and orientation, along an admissible path respecting the temporal and velocity constraints. This task must be accomplished by some subtasks like robot localization in the workspace, admissible path planning, trajectory generation and motion control. Moreover, autonomous wheeled mobile robots have kinematics constraints, also called nonholonomic constraints, that impose the robot can not move everywhere freely in its workspace, reducing the number of feasible paths between two distinct positions. This work mainly approaches the path planning and trajectory generation problems applied to wheeled mobile robots acting on a robot soccer environment. The major dificulty in this process is to find a smooth function that respects the imposed robot kinematic constraints. This work proposes a path generation strategy based on parametric polynomials of third degree for the 'x' and 'y' axis. The 'theta' orientation is derived from the 'y' and 'x' relations in such a way that the generated path respects the kinematic constraint. To execute the trajectory, this work also shows a simple control strategy acting on the robot linear and angular velocities