The boundary element method was applied to polymer analysis. The comparison of two existing BEM approaches was carried out solving a benchmark viscoelastic problem numerically and comparing with the analytical solutions. The fundamental solutions due to both Heaviside and Dirac impulse were obtained for a generalised Maxwell SLS material model. A new time-domain BEM formulation for viscoelasticity was derived, and the computer program was implemented and validated. A mixed method for quasi-static viscoelasticity was proposed. Several viscoelastic problems were solved for the purpose of validating this formulation. Numerical results were compared with analytical solutions, and good agreement was achieved. The BEM was applied to viscoelastic fracture problems. The effectiveness of the adopted BEM modelling was tested on an elastic fracture problem. The time-dependent strain energy release rate and J-integral in viscoelasticity were evaluated under different loading conditions. The crack propagation velocity under constant strain loading was also obtained. Adopting BE methodology, an integral equation for nonlinear viscoelastic problems was derived. The method to remove the high singularity in the irreducible domain integral was proposed. A computer program for this nonlinear viscoelastic formulation was developed. A central-crack problem was solved and the expected effect of non-linearity on stress field was obtained.

In this thesis, we are interested in the stochastic differential equation with jumps under regime switching. Firstly, we investigate a continuous-time version of the mean-variance portfolio selection model with jumps under regime switching. The portfolio selection proposed and analyzed for a market consisting of one bank account an d multiple stocks. The random regime switching is assumed to be independent of the underlying Brownian motion and jump processes. Secondly, we consider the problem of pricing contigent claims on a stock whose price process is modeled by a Levy process. Since the market is incomplete and there is not a unique equivalent martingale measure. We study approaches to pricing options. Finally, we investigate a continuous-time version Markowitz's mean-variance portfolio selection problem which is studied in a market with one bank account, one stock and proportional transaction costs. This is a singular stochastic control problem. Via a series of transformations, the problem is turned into a double obstacle problem.

This thesis explores higher order numerical methods for solving fractional differential equations.

This thesis introduces and analyses a numerical method for solving time-dependent partial differential equations (PDEs) on surfaces. This method is based on the closest point method, and solves the surface PDE by solving a suitably chosen equation in a band surrounding the surface. As it uses an implicit closest point representation of the surface, the method has the advantages of being simple to implement for very general surfaces, and amenable to discretization with a broad class of numerical schemes. The method proposed in this work introduces a new equation in the embedding space, which satisfies a key consistency property with the surface PDE. Rather than alternating between explicit time-steps and re-extensions of the surface function as in the original closest point method, we investigate an alternative approach, in which a single equation can be solved throughout the embedding space, without separate extension steps. This is achieved by creating a modified embedding equation with a penalty term, which enforces a constraint on the solution. The resulting equation admits a method of lines discretization, and can therefore be discretized with implicit or explicit time-stepping schemes, and analysed with standard techniques. The method can be formulated in a straightforward way for a large class of problems, including equations featuring variable coefficients, higher-order terms or nonlinearities. The effectiveness of the method is demonstrated with a range of examples, drawing from applications involving curvature-dependent diffusion and systems of reaction-diffusion equations, as well as equations arising in PDE-based image processing on surfaces.

No description available.

Finite volume methods for compressible Euler equations suffer from an excessive diffusion in the limit of low Mach numbers. This PhD thesis explores new approaches to overcome this. The analysis of a simpler set of equations that also possess a low Mach number limit is found to give valuable insights. These equations are the acoustic equations obtained as a linearization of the Euler equations. For both systems the limit is characterized by a divergencefree velocity. This constraint is nontrivial only in multiple spatial dimensions. As the Jacobians of the acoustic system do not commute, acoustics cannot be reduced to some kind of multi-dimensional advection. Therefore first an exact solution in multiple spatial dimensions is obtained. It is shown that the low Mach number limit can be interpreted as a limit of long times. It is found that the origin of the inability of a scheme to resolve the low Mach number limit is the lack a discrete counterpart to the limit of long times. Numerical schemes whose discrete stationary states discretize all the analytic stationary states of the PDE are called stationarity preserving. It is shown that for the acoustic equations, stationarity preserving schemes are vorticity preserving and are those that are able to resolve the low Mach limit (low Mach compliant). This establishes a new link between these three concepts. Stationarity preservation is studied in detail for both dimensionally split and multi-dimensional schemes for linear acoustics. In particular it is explained why the same multi-dimensional stencils appear in literature in very different contexts: These stencils are unique discretizations of the divergence that allow for stabilizing stationarity preserving diffusion. Stationarity preservation can also be generalized to nonlinear systems such as the Euler equations. Several ways how such numerical schemes can be constructed for the Euler equations are presented. In particular a low Mach compliant numerical scheme is derived that uses a novel construction idea. Its diffusion is chosen such that it depends on the velocity divergence rather than just derivatives of the different velocity components. This is demonstrated to overcome the low Mach number problem. The scheme shows satisfactory results in numerical simulations and has been found to be stable under explicit time integration. / Finite Volumen Methoden für die kompressiblen Euler-Gleichungen zeigen übermäßige Diffusion im Grenzwert kleiner Machzahlen. Diese Dissertation beschäftigt sich mit neuen Ansätzen, um dieses Problem zu beheben. Die Analyse eines Systems einfacherer Gleichungen, die ebenso einen Grenzwert niedriger Machzahlen haben, liefert wichtige Einsichten. Diese Gleichungen sind die als Linearisierung der Euler-Gleichungen erhaltenen akustischen Gleichungen. Für beide Gleichungssysteme ist der Grenzwert durch ein divergenzfreies Geschwindigkeitsfeld charakterisiert, was nur in mehreren Raumdimensionen nichttrivial ist. Da die Jacobi-Matrizen des akustischen Systems nicht vertauschen, kann Akustik nicht auf irgendeine Art mehrdimensionaler Advektion zurückgeführt werden. Deswegen wird zunächst eine exakte Lösung in mehreren Raumdimensionen gefunden. Es wird gezeigt, dass sich der Grenzwert kleiner Machzahlen als Grenzwert langer Zeiten interpretieren lässt. Als der Ursprung des Versagens eines Schemas im Grenzwert kleiner Machzahlen wird das Fehlen einer diskreten Entsprechung zum Grenzwert langer Zeiten identifiziert. Numerische Schemata, deren diskrete stationäre Zustände alle analytischen stationären Zustände diskretisieren, werden stationaritätserhaltend genannt. Es zeigt sich, dass für die akustischen Gleichungen stationaritätserhaltende Schemata vortizitätserhaltend sind, und gerade diejenigen sind, die auch den Grenzwert kleiner Machzahlen aufzulösen vermögen. Das zeigt eine neue Verbindung zwischen diesen drei Konzepten auf. Erhaltung der Stationarität wird für lineare Akustik im Detail für Schemata studiert, die nach Raumdimensionen aufgeteilt sind, und auch für multi-dimensionale Schemata. Insbesondere wird ein Grund geliefert, warum die gleichen multi-dimensionalen diskreten Operatoren in der Literatur in sehr unterschiedlichen Kontexten auftauchen: Sie sind Diskretisierungen der Divergenz, für die eine stabilisierende, stationaritätserhaltende Diffusion gefunden werden kann. Auch für nichtlineare Gleichungen, wie die Euler-Gleichungen, kann die Erhaltung der Stationarität verallgemeinert werden. Es werden dazu mehrere Wege der Konstruktion numerischer Schemata gezeigt. Insbesondere im Hinblick auf den Grenzwert kleiner Machzahlen wird ein neuartiges Schema hergeleitet, dessen Diffusion so gewählt ist, dass es von der Divergenz der Geschwindigkeit, und nicht bloß von irgendswelchen Ableitungen der Geschwindigkeitskomponenten abhängt. Es wird gezeigt, dass dieses Schema in der Lage ist, den Grenzwert kleiner Machzahlen aufzulösen. Das Schema zeigt zufriedenstellende Resultate in Simulationen und ist stabil unter Verwendung eines expliziten Zeitintegrators.

Es erfolgte die Untersuchung der Eigenschaften der gegebenen Zeitreihen der Wirkleistung von zwei Windenergieanlagen sowie des gesamten Windparks. Außerdem analysiert wurden die entsprechenden Eigenschaften der auf Viertelstundendaten konvertierten Reihen. Diese Eigenschaften, wie z.B. Mittelwert, Varianz o.a. Maximum wurden als Informationen angesehen und so ermöglichte die Gegenüberstellung einen Rückschluss auf dem Informationsverlust. Da die gegebene Einspeiseleistung keine Periodizitäten aufwies, erfolgte anschließend die Erzeugung synthetischer Daten.:Einleitung 1. Mathematische Grundlagen 2. Datenanalyse 3. Erzeugung und Untersuchung synthetischer Daten 4. Zusammenfassung und Ausblick A Anhang Literaturverzeichnis Ehrenwörtliche Erklärung

This cumulative habilitation thesis is concerned with eigenvalues of compactly perturbed operators in Banach and Hilbert spaces. A general theory for studying such eigenvalues is developed and applied to the study of some concrete operators of mathematical physics.

The spectral theory of the Laplace operator has long been studied in connection with physics. It appears in the wave equation, the heat equation, Schroedinger's equation and in the expression of quantum effects such as the Casimir force. The Casimir effect can be studied in terms of spectral invariants computed entirely from the spectrum of the Laplace operator. It is these spectral invariants and their computation that are the object of study in the present work. The objective of this thesis is to present a computational framework for the spectral zeta function $\zeta(s)$ and its derivative on a Euclidean domain in $\mathbb{R}^2$, with rigorous theoretical error bounds when this domain is polygonal. To obtain error bounds that remain practical in applications an improvement to existing heat trace estimates is necessary. Our main result is an original estimate and proof of a heat trace estimate for polygons that improves the one of van den Berg and Srisatkunarajah, using finite propagation speed of the corresponding wave kernel. We then use this heat trace estimate to obtain a rigorous error bound for $\zeta(s)$ computations. We will provide numerous examples of our computational framework being used to calculate $\zeta(s)$ for a variety of situations involving a polygonal domain, including examples involving cutouts and extrusions that are interesting in applications. Our second result is the development a new eigenvalue solver for a planar polygonal domain using a partition of unity decomposition technique. Its advantages include multiple precision and ease of use, as well as reduced complexity compared to Finite Elemement Method. While we hoped that it would be able to contend with existing packages in terms of speed, our implementation was many times slower than MPSPack when dealing with the same problem (obtaining the first 5 digits of the principal eigenvalue of the regular unit hexagon). Finally, we present a collection of numerical examples where we compute the spectral determinant and Casimir energy of various polygonal domains. We also use our numerical tools to investigate extremal properties of these spectral invariants. For example, we consider a square with a small square cut out of the interior, which is allowed to rotate freely about its center.

In this dissertation we study the Laplace operator acting on functions on a smooth, compact Riemannian manifold. Our approach is based on the study of the spectrum of the aforementioned operator. The main objects of our interest are the counting function of the Laplacian and its Riesz means. We discuss the asymptotics of aforementioned functions when the argument approaches infinity.