Dissertations / Theses on the topic 'Weighted residual methods'

A general global approximation scheme is developed and its generality is demonstrated by the derivation of classical Lagrange and Hermite interpolation and finite difference and finite element approximations as its special cases. It is also shown that previously reported general approximation techniques which use the idea of moving least square are also special cases of the present method. The combination of the developed general global approximation technique with the weighted residual methods provides a very powerful scheme for the solution of the boundary value problems formulated in terms of differential equations. Although this application is the main purpose of the this project, nevertheless, the power and flexibility of the developed approximation allows it to be used in many other areas. In this study the following applications of the described approximation are developed: 1- data fitting (including curve and surface fitting) 2- plane mapping (both in cases where a conformal mapping exists and for non-conformal mapping) 3- solution of eigenvalue problems with particular application to spectral expansions used in the modal representation of shallow water equations 4- solution of ordinary differential equations (including Sturm-Liouville equations, non-homogeneous equations with non-smooth right hand sides and 4th order equations) 5- elliptic partial differential equations (including Poisson equations with non-smooth right hand sides) A computer program which can handle the above applications is developed. This program utilises symbolic, numerical and graphical and the programming language provided by the Mathematica package.

The problem of using the α,0 and the α, β-quasi periodic transformations within a finite element method in studying electromagnetic waves in a periodic space is addressed. We investigate an a priori error estimate for both transformations which allows us to solve our problem numerically on a uniform mesh. We also analyse the Dual Weighted Residual (DWR) method with the α,0-quasi periodic transformation to derive an a posteriori error estimate. This error estimate is later used to compute efficiently the numerical solution using an adaptive method. We then implement the above finite element methods. It is shown numerically that our numerical results are in good agreement with those in the literature, the α, β-quasi periodic method converges at a far lower number of degrees of freedom than the α,0-quasi periodic method and the DWR method converges faster and requires fewer degrees of freedom than the global a posteriori error estimate or the uniform mesh. We also explore the geometrical freedom given by the finite element method and examine wave scattering by the Morpho butterfly wing.

Die vorliegende Arbeit zum Thema der numerischen Analysis von Eigenwertproblemen befasst sich mit fünf wesentlichen Aspekten der numerischen Analysis von Eigenwertproblemen. Der erste Teil präsentiert einen Algorithmus von asymptotisch quasi-optimaler Rechenlaufzeit, der die adaptive Finite Elemente Methode mit einem iterativen algebraischen Eigenwertlöser kombiniert. Der zweite Teil präsentiert explizite beidseitige Schranken für die Eigenwerte des Laplace Operators auf beliebig groben Gittern basierend auf einer Approximation der zugehörigen Eigenfunktion in dem nicht konformen Finite Elemente Raum von Crouzeix und Raviart und einem Postprocessing. Die Effizienz der garantierten Schranke des Eigenwertfehlers hängt von der globalen Gitterweite ab. Der dritte Teil betrachtet eine adaptive Finite Elemente Methode basierend auf Verfeinerungen von Knoten-Patchen. Dieser Algorithmus zeigt eine asymptotische Fehlerreduktion der adaptiven Sequenz von einfachen Eigenwerten und Eigenfunktionen des Laplace Operators. Die hier erstmals bewiesene Eigenschaft der Saturation des Eigenwertfehlers zeigt Zuverlässigkeit und Effizienz für eine Klasse von hierarchischen a posteriori Fehlerschätzern. Der vierte Teil betrachtet a posteriori Fehlerschätzer für Konvektion-Diffusion Eigenwertprobleme, wie sie von Heuveline und Rannacher (2001) im Kontext der dual-gewichteten residualen Methode (DWR) diskutiert wurden. Zwei neue dual-gewichtete a posteriori Fehlerschätzer werden vorgestellt. Der letzte Teil beschäftigt sich mit drei adaptiven Algorithmen für Eigenwertprobleme von nicht selbst-adjungierten Operatoren partieller Differentialgleichungen. Alle drei Algorithmen basieren auf einer Homotopie-Methode die vom einfacheren selbst-adjungierten Problem startet. Neben der Gitterverfeinerung wird der Prozess der Homotopie sowie die Anzahl der Iterationen des algebraischen Löser adaptiv gesteuert und die verschiedenen Anteile am gesamten Fehler ausbalanciert.
This thesis "on the numerical analysis of eigenvalue problems" consists of five major aspects of the numerical analysis of adaptive finite element methods for eigenvalue problems. The first part presents a combined adaptive finite element method with an iterative algebraic eigenvalue solver for a symmetric eigenvalue problem of asymptotic quasi-optimal computational complexity. The second part introduces fully computable two-sided bounds on the eigenvalues of the Laplace operator on arbitrarily coarse meshes based on some approximation of the corresponding eigenfunction in the nonconforming Crouzeix-Raviart finite element space plus some postprocessing. The efficiency of the guaranteed error bounds involves the global mesh-size and is proven for the large class of graded meshes. The third part presents an adaptive finite element method (AFEM) based on nodal-patch refinement that leads to an asymptotic error reduction property for the adaptive sequence of simple eigenvalues and eigenfunctions of the Laplace operator. The proven saturation property yields reliability and efficiency for a class of hierarchical a posteriori error estimators. The fourth part considers a posteriori error estimators for convection-diffusion eigenvalue problems as discussed by Heuveline and Rannacher (2001) in the context of the dual-weighted residual method (DWR). Two new dual-weighted a posteriori error estimators are presented. The last part presents three adaptive algorithms for eigenvalue problems associated with non-selfadjoint partial differential operators. The basis for the developed algorithms is a homotopy method which departs from a well-understood selfadjoint problem. Apart from the adaptive grid refinement, the progress of the homotopy as well as the solution of the iterative method are adapted to balance the contributions of the different error sources.

O presente trabalho trata da aplicação do Método de Resíduos Ponderados, mais especificamente do Método dos Mínimos Quadrados, na obtenção de soluções aproximadas de problemas de estruturas em casca, em especial os reservatórios cilíndricos submetidos a carregamento hidrostático em regime de comportamento linear. Os meios empregados para a obtenção das soluções aproximadas referem-se à adoção de bases aproximativas lineares, polinomiais, além da possibilidade de enriquecimento da aproximação mediante a adição de funções com características similares à própria solução exata. Uma outra alternativa utilizada refere-se à aplicação do Método dos Mínimos Quadrados com divisão do domínio de integração. Tais procedimentos podem ser úteis na análise de estruturas por evitar, de modo significativo, a elevação do esforço computacional, mediante a utilização de uma base aproximativa que corresponda às características requeridas pela solução analítica do problema.
The present dissertation deals with the application of the Weighed Residual Method to analysis of cilindrical shells, more specifically of the Least Squared Method, in the attainment of approach solutions of shells structural problems, in special the cylindrical reservoirs submitted the hydrostatic shipment in regimen of linear behavior. The half employees for obtention of the approach solutions refer to adotion of linear, polynomial approaches bases, beyond the possibility of enrichment of the approach by means of the addition of functions with similar characteristics to the proper accurate solution.One another used alternative mentions the application to Least Squared Method with division of the integration domain. Such procedures can be useful in the analysis of structures for preventing, in significant way, the rise of the computational effort, by means of the use of a aproximativa base that correspond to the characteristics required for the analytical solution of the problem.

Este trabalho aborda os conceitos básicos da teoria de segunda ordem das placas elásticas delgadas, utilizando o Método dos Elementos Finitos (introduzido através do Método dos Resíduos Ponderados, na variante de Galerkin). São deduzidas as matrizes de rigidez geométrica, de rigidez secante e de rigidez tangente, relativas ao problema em consideração. É proposta ainda uma conduta notavelmente simplificada, que facilita sobremaneira a construção da matriz de rigidez tangente.
This paper delas with the basic concepts of the secondf order theory of thin elastic plates, through the use of the Finite Element Method 9introcuced through the Weighted Residual Method, in Galerkin\'s approach). The matrices of geometric stiffness, secant stiffness, and tangent stiffness for the problem under consideration are deduced. It is also proposed an outstandingly simplified conduct, which will greatly easen the construction of the tangent stiffness matrix.

The presented master’s thesis deals with the application of the gradient elasticity in fracture mechanics problems. Specifically, the displacement and stress field around the crack tip is a matter of interest. The influence of a material microstructure is considered. Introductory chapters are devoted to a brief historical overview of gradient models and definition of basic equations of dipolar gradient elasticity derived from Mindlin gradient theory form II. For comparison, relations of classical elasticity are introduced. Then a derivation of asymptotic displacement field using the Williams asymptotic technique follows. In the case of gradient elasticity, also the calculation of the J-integral is included. The mathematical formulation is reduced due to the singular nature of the problem to singular integral equations. The methods for solving integral equations in Cauchy principal value and Hadamard finite part sense are briefly introduced. For the evaluation of regular kernel, a Gauss-Chebyshev quadrature is used. There also mentioned approximate methods for solving systems of integral equations such as the weighted residual method, especially the least square method with collocation points. In the main part of the thesis the system of integral equations is derived using the Fourier transform for straight crack in an infinite body. This system is then solved numerically in the software Mathematica and the results are compared with the finite element model of ceramic foam.

In the past three decades, numerous methods have been proposed to transcribe optimal control problems (OCP) into nonlinear programming problems (NLP). In this dissertation work, a unifying weighted residual framework is developed under which most of the existing transcription methods can be derived by judiciously choosing test and trial functions. This greatly simplifies the derivation of optimality conditions and costate estimation results for direct transcription methods. Under the same framework, three new transcription methods are devised which are particularly suitable for implementation in an adaptive refinement setting. The method of Hilbert space projection, the least square method for optimal control and generalized moment method for optimal control are developed and their optimality conditions are derived. It is shown that under a set of equivalence conditions, costates can be estimated from the Lagrange multipliers of the associated NLP for all three methods. Numerical implementation of these methods is described using B-Splines and global interpolating polynomials as approximating functions. It is shown that the existing pseudospectral methods for optimal control can be formulated and analyzed under the proposed weighted residual framework. Performance of Legendre, Gauss and Radau pseudospectral methods is compared with the methods proposed in this research. Based on the variational analysis of first-order optimality conditions for the optimal control problem, an posteriori error estimation procedure is developed. Using these error estimates, an h-adaptive scheme is outlined for the implementation of least square method in an adaptive manner. A time-scaling technique is described to handle problems with discontinuous control or multiple phases. Several real-life examples were solved to show the efficacy of the h-adaptive and time-scaling algorithm.

博士
國立成功大學
機械工程學系
88
The error analyses are difficult to do in traditional numerical methods. But for MWR, the maximum principles of differential equations offer the mathematical tools to analysis the error bounds by using the MWR. By using the maximum principles, a double inequality equations can be constructed for engineering problems, i.e. the error bounds of the solutions. A mathematical programming problem is then be constructed by the MWR to find the upper and lower bounds of the engineering problems. Therefore, the error bounds can be found by solving the mathematical programming problems. The approximate solutions and error bounds can be solve simultaneously by using the proposed approach which is the major different between the traditional method such as finite elements method or boundary elements. For nonlinear mathematical programming problems, a genetic algorithm(GAs) is proposed in this dissertation. The properties and performance of GA is also discussed in this dissertation. The parameters properties of GAs could be find in the discussion in chapter 4. In chapter 5, a solving procedure is proposed. There are 5 engineering problems of heat transfer and solid mechanics are solved by the proposed approach. The results show the efficiency, accuracy, and simplicity by combining the GAs and MWR.

博士
國立成功大學
機械工程學系
89
This paper presents a double side approximate method, which combines the Method of Weighted Residual, mathematical programming and maximum principle theory, and which may be used to solve differential equations found within engineering problems. The proposed method may be readily extended to solve a wide range of nonlinear engineering problems. As stated above, a double side approximate method combines mathematical programming with the Method of Weighted Residual (MWR). If the solution Z(x) exists in the defining domain, V, of a problem, then it is bound by two limits, represented by the functions and . These functions satisfy the definite condition that if >0> , then in , where R is the residual operator. It is possible to obtain the values of minimum and maximum which satisfy the above inequality by using the Genetic Algorithms (GAs) optimization method. This paper considers the use of a double side approximate method to solve differential equations and monotone problems, using the vector of residuals as given by the Method of Weighted Residuals. The paper considers the application of the proposed method to several nonlinear differential equation problems. In this way the efficiency and simplicity of this method are illustrated, indicating that the proposed method can be easily extended to tackle other nonlinear engineering problems. It is possible to use different Methods of Weighted Residual to solve the bilateral inequality. As has been mentioned previously, by using the GAs optimization method it is possible to determine the values of the minimum and maximum functions which satisfy the inequality. The Laplace transform is well known as a powerful tool in the analysis of time independent problems. In this paper, a method which combines the use of Laplace transformation and double side approach method, has been applied to the solution of transient nonlinear heat conduction problems. A double side approach method is then used to solve the generalized physical engineering problems. The proposed method demonstrates efficiency, accuracy, simplicity, no convergence problems, and requires less computer processing time, and as such, represents a major step forward from the traditional problem solving techniques.

Because current full‐core neutronic‐calculations use two‐group neutron diffusion and rely on homogenizing fuel assemblies, reconstructing pin powers from such a calculation is an elaborate and not very accurate process; one which becomes more difficult with increased core heterogeneity. A three‐dimensional Heterogeneous Finite Element Method (HFEM) is developed to address the limitations of current methods by offering fine‐group energy representation and fuel‐pin‐level spatial detail at modest computational cost. The calculational cost of the method is roughly equal to the calculational cost of the Finite Differences Method (FDM) using one mesh box per fuel assembly and a comparable number of energy groups. Pin‐level fluxes are directly obtained from the method’s results without the need for reconstruction schemes.
UOIT

The Generalized Weighted Residual Method (GWRM) is a recently developed time- spectral method for parabolic or hyperbolic initial-value partial differential equations. In this paper, spatial subdomains, used in this method, are analyzed. Subdomains are used to enhance efficiency by dividing entire domains into smaller parts that can be independently solved for and then combined to get the final solution. An automatic grid mapping algorithm for spatial subdomains, called "Compressive Method", is presented and applied to Burgers' viscous equation. The error of the solution, as compared to the analytic solution, is compared for this compressive Method and the uniform grid case. Results show that accuracy can be gained at a small extra cost, using this compressive Method. Conclusions are that this adaptive algorithm shows great potential for further development.

Le problème inverse en électroencéphalographie (EEG) est la localisation de sources de courant dans le cerveau utilisant les potentiels de surface sur le cuir chevelu générés par ces sources. Une solution inverse implique typiquement de multiples calculs de potentiels de surface sur le cuir chevelu, soit le problème direct en EEG. Pour résoudre le problème direct, des modèles sont requis à la fois pour la configuration de source sous-jacente, soit le modèle de source, et pour les tissues environnants, soit le modèle de la tête. Cette thèse traite deux approches bien distinctes pour la résolution du problème direct et inverse en EEG en utilisant la méthode des éléments de frontières (BEM): l’approche conventionnelle et l’approche réciproque. L’approche conventionnelle pour le problème direct comporte le calcul des potentiels de surface en partant de sources de courant dipolaires. D’un autre côté, l’approche réciproque détermine d’abord le champ électrique aux sites des sources dipolaires quand les électrodes de surfaces sont utilisées pour injecter et retirer un courant unitaire. Le produit scalaire de ce champ électrique avec les sources dipolaires donne ensuite les potentiels de surface. L’approche réciproque promet un nombre d’avantages par rapport à l’approche conventionnelle dont la possibilité d’augmenter la précision des potentiels de surface et de réduire les exigences informatiques pour les solutions inverses. Dans cette thèse, les équations BEM pour les approches conventionnelle et réciproque sont développées en utilisant une formulation courante, la méthode des résidus pondérés. La réalisation numérique des deux approches pour le problème direct est décrite pour un seul modèle de source dipolaire. Un modèle de tête de trois sphères concentriques pour lequel des solutions analytiques sont disponibles est utilisé. Les potentiels de surfaces sont calculés aux centroïdes ou aux sommets des éléments de discrétisation BEM utilisés. La performance des approches conventionnelle et réciproque pour le problème direct est évaluée pour des dipôles radiaux et tangentiels d’excentricité variable et deux valeurs très différentes pour la conductivité du crâne. On détermine ensuite si les avantages potentiels de l’approche réciproquesuggérés par les simulations du problème direct peuvent êtres exploités pour donner des solutions inverses plus précises. Des solutions inverses à un seul dipôle sont obtenues en utilisant la minimisation par méthode du simplexe pour à la fois l’approche conventionnelle et réciproque, chacun avec des versions aux centroïdes et aux sommets. Encore une fois, les simulations numériques sont effectuées sur un modèle à trois sphères concentriques pour des dipôles radiaux et tangentiels d’excentricité variable. La précision des solutions inverses des deux approches est comparée pour les deux conductivités différentes du crâne, et leurs sensibilités relatives aux erreurs de conductivité du crâne et au bruit sont évaluées. Tandis que l’approche conventionnelle aux sommets donne les solutions directes les plus précises pour une conductivité du crâne supposément plus réaliste, les deux approches, conventionnelle et réciproque, produisent de grandes erreurs dans les potentiels du cuir chevelu pour des dipôles très excentriques. Les approches réciproques produisent le moins de variations en précision des solutions directes pour différentes valeurs de conductivité du crâne. En termes de solutions inverses pour un seul dipôle, les approches conventionnelle et réciproque sont de précision semblable. Les erreurs de localisation sont petites, même pour des dipôles très excentriques qui produisent des grandes erreurs dans les potentiels du cuir chevelu, à cause de la nature non linéaire des solutions inverses pour un dipôle. Les deux approches se sont démontrées également robustes aux erreurs de conductivité du crâne quand du bruit est présent. Finalement, un modèle plus réaliste de la tête est obtenu en utilisant des images par resonace magnétique (IRM) à partir desquelles les surfaces du cuir chevelu, du crâne et du cerveau/liquide céphalorachidien (LCR) sont extraites. Les deux approches sont validées sur ce type de modèle en utilisant des véritables potentiels évoqués somatosensoriels enregistrés à la suite de stimulation du nerf médian chez des sujets sains. La précision des solutions inverses pour les approches conventionnelle et réciproque et leurs variantes, en les comparant à des sites anatomiques connus sur IRM, est encore une fois évaluée pour les deux conductivités différentes du crâne. Leurs avantages et inconvénients incluant leurs exigences informatiques sont également évalués. Encore une fois, les approches conventionnelle et réciproque produisent des petites erreurs de position dipolaire. En effet, les erreurs de position pour des solutions inverses à un seul dipôle sont robustes de manière inhérente au manque de précision dans les solutions directes, mais dépendent de l’activité superposée d’autres sources neurales. Contrairement aux attentes, les approches réciproques n’améliorent pas la précision des positions dipolaires comparativement aux approches conventionnelles. Cependant, des exigences informatiques réduites en temps et en espace sont les avantages principaux des approches réciproques. Ce type de localisation est potentiellement utile dans la planification d’interventions neurochirurgicales, par exemple, chez des patients souffrant d’épilepsie focale réfractaire qui ont souvent déjà fait un EEG et IRM.
The inverse problem of electroencephalography (EEG) is the localization of current sources within the brain using surface potentials on the scalp generated by these sources. An inverse solution typically involves multiple calculations of scalp surface potentials, i.e., the EEG forward problem. To solve the forward problem, models are needed for both the underlying source configuration, the source model, and the surrounding tissues, the head model. This thesis treats two distinct approaches for the resolution of the EEG forward and inverse problems using the boundary-element method (BEM): the conventional approach and the reciprocal approach. The conventional approach to the forward problem entails calculating the surface potentials starting from source current dipoles. The reciprocal approach, on the other hand, first solves for the electric field at the source dipole locations when the surface electrodes are reciprocally energized with a unit current. A scalar product of this electric field with the source dipoles then yields the surface potentials. The reciprocal approach promises a number of advantages over the conventional approach, including the possibility of increased surface potential accuracy and decreased computational requirements for inverse solutions. In this thesis, the BEM equations for the conventional and reciprocal approaches are developed using a common weighted-residual formulation. The numerical implementation of both approaches to the forward problem is described for a single-dipole source model. A three-concentric-spheres head model is used for which analytic solutions are available. Scalp potentials are calculated at either the centroids or the vertices of the BEM discretization elements used. The performance of the conventional and reciprocal approaches to the forward problem is evaluated for radial and tangential dipoles of varying eccentricities and two widely different skull conductivities. We then determine whether the potential advantages of the reciprocal approach suggested by forward problem simulations can be exploited to yield more accurate inverse solutions. Single-dipole inverse solutions are obtained using simplex minimization for both the conventional and reciprocal approaches, each with centroid and vertex options. Again, numerical simulations are performed on a three-concentric-spheres model for radial and tangential dipoles of varying eccentricities. The inverse solution accuracy of both approaches is compared for the two different skull conductivities and their relative sensitivity to skull conductivity errors and noise is assessed. While the conventional vertex approach yields the most accurate forward solutions for a presumably more realistic skull conductivity value, both conventional and reciprocal approaches exhibit large errors in scalp potentials for highly eccentric dipoles. The reciprocal approaches produce the least variation in forward solution accuracy for different skull conductivity values. In terms of single-dipole inverse solutions, conventional and reciprocal approaches demonstrate comparable accuracy. Localization errors are low even for highly eccentric dipoles that produce large errors in scalp potentials on account of the nonlinear nature of the single-dipole inverse solution. Both approaches are also found to be equally robust to skull conductivity errors in the presence of noise. Finally, a more realistic head model is obtained using magnetic resonance imaging (MRI) from which the scalp, skull, and brain/cerebrospinal fluid (CSF) surfaces are extracted. The two approaches are validated on this type of model using actual somatosensory evoked potentials (SEPs) recorded following median nerve stimulation in healthy subjects. The inverse solution accuracy of the conventional and reciprocal approaches and their variants, when compared to known anatomical landmarks on MRI, is again evaluated for the two different skull conductivities. Their respective advantages and disadvantages including computational requirements are also assessed. Once again, conventional and reciprocal approaches produce similarly small dipole position errors. Indeed, position errors for single-dipole inverse solutions are inherently robust to inaccuracies in forward solutions, but dependent on the overlapping activity of other neural sources. Against expectations, the reciprocal approaches do not improve dipole position accuracy when compared to the conventional approaches. However, significantly smaller time and storage requirements are the principal advantages of the reciprocal approaches. This type of localization is potentially useful in the planning of neurosurgical interventions, for example, in patients with refractory focal epilepsy in whom EEG and MRI are often already performed.