Diese Arbeit wird einer Weiterentwicklung der Dichtematrixtheorie und ihrer Anwendung zum Studium ultraschneller laserpulsinduzierter Dynamik in Molekularsystemen in Wechselwirkung mit einem thermischen Bad gewidmet. Zwei grosse Themenkomplexe werden behandelt. Zuerst werden die sogenannten Gedächtniseffekte diskutiert. Diese folgen aus einer reduzierten Beschreibung des Molekularsystems, in der die Umgebungsfreiheitsgrade eliminiert werden. Im zweiten Teil wird die Laserpulssteuerung der dissipativen Molekulardynamik untersucht. Die theoretische Beschreibung von offenen Quantensystemen führt zu einer zeitlich nicht-lokalen Bewegungsgleichung: Die Zeitentwicklung des Molekularsystems hängt von seiner Vergangenheit ab. In dieser Arbeit wird eine numerische Methode zur Lösung der zeitlich nicht-lokalen Bewegungsgleichung entwickelt und mit einem minimalen Modell eines polyatomaren Moleküls unter dissipativem Einfluss der Umgebung getestet. Eine analytische Lösung der Bewegungsgleichung für den speziellen Fall einer sehr langen Gedächtniszeit wurde hergeleitet. Zur Identifizierung solcher Gedächtniseffekte vergleichen wir diese analytische Lösung mit numerischen Rechnungen inklusive Gedächtnis und mit approximativen Rechnungen, die die zeitliche Nicht-Lokalität vernachlässigen. Für eine Anregung mit einem Laserpuls, der kürzer als die Gedächtniszeit des Systems ist, zeigt das Molekularsystem eine erkennbar unterschiedliche Dynamik als ohne Gedächtniss. Die Gedächtniseffekte werden mit abfallender Laserpulslänge deutlich ausgeprägter. Der zweite Teil der Arbeit konzentriert sich auf die Anwendung der Theorie der Optimalen Kontrolle, um die molekulare Dynamik zu steuern. Aus der Theorie der Optimalen Kontrolle erhält man Laserpulse, die bestimmte Aufgaben erfüllen, z.B. die Besetzung gewünschter vibronischer Niveaus des Molekularsystems oder die Platzierung eines Wellenpakets auf einer vorgegebenen Position auf der molekularen Potentialfläche. Als erstes Beispiel haben wir die Kontrolle des dissipativen fotoinduzierten Elektronentransfers in einem Donator-Brückenmolekül-Akzeptor System betrachtet, wobei wir das Gedächtniss vernachlässigt haben. Die Steuerbarkeit des Elektronentransfers wird diskutiert und der Mechanismus, mit dem sie möglich wird, wird identifiziert. Wir haben festgestellt, dass die Steuerung der Elektronentransferreaktionen selbst unter dem Einfluss von Dissipation möglich ist, obwohl die Kontrollausbeute mit steigender Dissipation drastisch abfällt. In Anwesenheit von Dissipation verändert sich auch der Mechanismus der Steuerung. Die experimentelle Ausführbarkeit der Herstellung des aus der Theorie der Optimalen Kontrolle resultierenden Kontrollpulses wird diskutiert und Methoden werden präsentiert, die die Abschätzung der Effizienz ermöglichen, mit der ein Flussigkristall--Laserpulsformer, wie er heute in Experimenten verwendet wird, den gewünschten Puls erzeugen kann. Um zwischen verschiedenen Kontrollaufgaben zu unterscheiden, wird ein quantitatives Mass eingeführt, das die Komplexität der Kontrollaufgabe charakterisiert. Die Theorie der Optimalen Kontrolle wird auch für Molekularsysteme formuliert, die statische Unordnung zeigen, und wird auf ein Ensemble von Molekülen mit zufälligen Orientierungen angewendet. Zum Schluss wird die Bedeutung der Gedächtnisseffekte für die Steuerung der dissipativen Dynamik diskutiert und die Theorie der Optimalen Kontrolle neu formuliert um eine zeitliche Nicht-Lokalität in der Bewegungsgleichung des Molekularsystems zu berücksichtigen. / This work is dedicated to a further development of the density matrix theory and its application to the study of ultrafast laser pulse induced dynamics in molecular systems interacting with a thermal environment. Two topics are considered, first the so-called memory effects are analyzed which result from a reduced description of the molecular system excluding the environmental degrees of freedom. And secondly, the laser pulse control of dissipative molecular dynamics is examined. The theoretical description of open quantum systems results in a time non-local equation of motion so that the evolution of the molecular system depends on its past. In this work a numerical method to solve the time non-local equations of motion has been developed and tested for a minimal model of a polyatomic molecule subject to the dissipative influence of an environment. An analytical solution of the equation of motion for the special case of very long standing memory is also achieved. To identify signatures of such memory effects in general case we compare this analytical solution with numerical calculations involving memory and with approximative computations ignoring time non-locality. For the excitation by a laser pulse shorter than the duration of the memory the molecular systems exhibit noticeably different dynamics than for the absence of the memory. The effects become significantly more pronounced with decreasing laser pulse durations. The second part of the work concentrates on the application of the optimal control theory to guide molecular dynamics. Optimal control theory provides laser pulses which are designed in such a manner to fulfill certain control tasks, e.g. the population of a desired vibrational level of the molecular system or the placement of a wavepacket on a prescribed position on the molecular potential energy surface. As a first example the control of the dissipative photo-induced electron transfer in a donor--bridge--acceptor systems has been particularly considered ignoring the memory. The controllability of the electron transfer has been discussed and the mechanism by which it becomes possible has been identified. We have found the control of electron transfer reactions feasible even under the influence of dissipation although the yield of the control decreases drastically with increasing dissipation. In the presence of dissipation mechanism of the control has been found to change. The feasibility of the reproduction of the control pulses resulting for the optimal control theory in the experiment has been discussed and methods have been presented how to check the efficiency of the reproduction of optimal control pulses by liquid crystal pulse shapers, prevailingly used in modern control experiments. To distinguish different control tasks a quantitative measure has been introduced characterizing complexity of the control task. The optimal control theory has also been formulated for molecular systems showing static disorder and applied on an ensemble of molecules exhibiting random orientations. Finally, the importance of memory effects for the control of dissipative dynamics has been discussed and the optimal control theory has been formulated to account for a time non-locality in the equation of motion for molecular systems.

This thesis is concerned with motion planning for the classical two-phase Stefan problem in level set formulation. The interface separating the fluid phases from the solid phases is represented as the zero level set of a continuous function whose evolution is described by the level set equation. Heat conduction in the two phases is modeled by the heat equation. A quadratic tracking-type cost functional that incorporates temperature tracking terms and a control cost term that expresses the desire to have the interface follow a prescribed trajectory by adjusting the heat flux through part of the boundary of the computational domain. The formal Lagrange approach is used to establish a first-order optimality system by applying shape calculus tools. For the numerical solution, the level set equation and its adjoint are discretized in space by discontinuous Galerkin methods that are combined with suitable explicit Runge-Kutta time stepping schemes, while the temperature and its adjoint are approximated in space by the extended finite element method (which accounts for the weak discontinuity of the temperature by a dynamic local modification of the underlying finite element spaces) combined with the implicit Euler method for the temporal discretization. The curvature of the interface which arises in the adjoint system is discretized by a finite element method as well. The projected gradient method, and, in the absence of control constraints, the limited memory BFGS method are used to solve the arising optimization problems. Several numerical examples highlight the potential of the proposed optimal control approach. In particular, they show that it inherits the geometric flexibility of the level set method. Thus, in addition to unidirectional solidification, closed interfaces and changes of topology can be tracked. Finally, the Moreau-Yosida regularization is applied to transform a state constraint on the position of the interface into a penalty term that is added to the cost functional. The optimality conditions for this penalized optimal control problem and its numerical solution are discussed. An example confirms the efficacy of the state constraint. / Die vorliegende Arbeit beschäftigt sich mit einem Optimalsteuerungsproblem für das klassische Stefan-Problem in zwei Phasen. Die Phasengrenze wird als Niveaulinie einer stetigen Funktion modelliert, was die Lösung der so genannten Level-Set-Gleichung erfordert. Durch Anpassen des Wärmeflusses am Rand des betrachteten Gebiets soll ein gewünschter Verlauf der Phasengrenze angesteuert werden. Zusammen mit dem Wunsch, ein vorgegebenes Temperaturprofil zu approximieren, wird dieses Ziel in einem quadratischen Zielfunktional formuliert. Die notwendigen Optimalitätsbedingungen erster Ordnung werden formal mit Hilfe der entsprechenden Lagrange-Funktion und unter Benutzung von Techniken aus der Formoptimierung hergeleitet. Für die numerische Lösung müssen die auftretenden partiellen Differentialgleichungen diskretisiert werden. Dies geschieht im Falle der Level-Set-Gleichung und ihrer Adjungierten auf Basis von unstetigen Galerkin-Verfahren und expliziten Runge-Kutta-Methoden. Die Wärmeleitungsgleichung und die entsprechende Gleichung im adjungierten System werden mit einer erweiterten Finite-Elemente-Methode im Ort sowie dem impliziten Euler-Verfahren in der Zeit diskretisiert. Dieser Zugang umgeht die aufwändige Adaption des Gitters, die normalerweise bei der FE-Diskretisierung von Phasenübergangsproblemen unvermeidbar ist. Auch die Krümmung der Phasengrenze wird numerisch mit Hilfe der Methode der finiten Elemente angenähert. Zur Lösung der auftretenden Optimierungsprobleme werden ein Gradienten-Projektionsverfahren und, im Fall dass keine Kontrollschranken vorliegen, die BFGS-Methode mit beschränktem Speicherbedarf eingesetzt. Numerische Beispiele beleuchten die Stärken des vorgeschlagenen Zugangs. Es stellt sich insbesondere heraus, dass sich die geometrische Flexibilität der Level-Set-Methode auf den vorgeschlagenen Zugang zur optimalen Steuerung vererbt. Zusätzlich zur gerichteten Bewegung einer flachen Phasengrenze können somit auch geschlossene Phasengrenzen sowie topologische Veränderungen angesteuert werden. Exemplarisch, und zwar an Hand einer Beschränkung an die Lage der Phasengrenze, wird auch noch die Behandlung von Zustandsbeschränkungen mittels der Moreau-Yosida-Regularisierung diskutiert. Ein numerisches Beispiel demonstriert die Wirkung der Zustandsbeschränkung.

Motivation and background The enormous amount of capabilities that every human learns throughout his life, is probably among the most remarkable and fascinating aspects of life. Learning has therefore drawn lots of interest from scientists working in very different fields like philosophy, biology, sociology, educational sciences, computer sciences and mathematics. This thesis focuses on the information theoretical and mathematical aspects of learning. We are interested in the learning process of an agent (which can be for example a human, an animal, a robot, an economical institution or a state) that interacts with its environment. Common models for this interaction are Markov decision processes (MDPs) and partially observable Markov decision processes (POMDPs). Learning is then considered to be the maximization of the expectation of a predefined reward function. In order to formulate general principles (like a formal definition of curiosity-driven learning or avoidance of unpleasant situation) in a rigorous way, it might be desirable to have a theoretical framework for the optimization of more complex functionals of the underlying process law. This might include the entropy of certain sensor values or their mutual information. An optimization of the latter quantity (also known as predictive information) has been investigated intensively both theoretically and experimentally using computer simulations by N. Ay, R. Der, K Zahedi and G. Martius. In this thesis, we develop a mathematical theory for learning in the sensorimotor loop beyond expected reward maximization. Approaches and results This thesis covers four different topics related to the theory of learning in the sensorimotor loop. First of all, we need to specify the model of an agent interacting with the environment, either with learning or without learning. This interaction naturally results in complex causal dependencies. Since we are interested in asymptotic properties of learning algorithms, it is necessary to consider infinite time horizons. It turns out that the well-understood theory of causal networks known from the machine learning literature is not powerful enough for our purpose. Therefore we extend important theorems on causal networks to infinite graphs and general state spaces using analytical methods from measure theoretic probability theory and the theory of discrete time stochastic processes. Furthermore, we prove a generalization of the strong Markov property from Markov processes to infinite causal networks. Secondly, we develop a new idea for a projected stochastic constraint optimization algorithm. Generally a discrete gradient ascent algorithm can be used to generate an iterative sequence that converges to the stationary points of a given optimization problem. Whenever the optimization takes place over a compact subset of a vector space, it is possible that the iterative sequence leaves the constraint set. One possibility to cope with this problem is to project all points to the constraint set using Euclidean best-approximation. The latter is sometimes difficult to calculate. A concrete example is an optimization over the unit ball in a matrix space equipped with operator norm. Our idea consists of a back-projection using quasi-projectors different from the Euclidean best-approximation. In the matrix example, there is another canonical way to force the iterative sequence to stay in the constraint set: Whenever a point leaves the unit ball, it is divided by its norm. For a given target function, this procedure might introduce spurious stationary points on the boundary. We show that this problem can be circumvented by using a gradient that is tailored to the quasi-projector used for back-projection. We state a general technical compatibility condition between a quasi-projector and a metric used for gradient ascent, prove convergence of stochastic iterative sequences and provide an appropriate metric for the unit-ball example. Thirdly, a class of learning problems in the sensorimotor loop is defined and motivated. This class of problems is more general than the usual expected reward maximization and is illustrated by numerous examples (like expected reward maximization, maximization of the predictive information, maximization of the entropy and minimization of the variance of a given reward function). We also provide stationarity conditions together with appropriate gradient formulas. Last but not least, we prove convergence of a stochastic optimization algorithm (as considered in the second topic) applied to a general learning problem (as considered in the third topic). It is shown that the learning algorithm converges to the set of stationary points. Among others, the proof covers the convergence of an improved version of an algorithm for the maximization of the predictive information as proposed by N. Ay, R. Der and K. Zahedi. We also investigate an application to a linear Gaussian dynamic, where the policies are encoded by the unit-ball in a space of matrices equipped with operator norm.

Motivation and background The enormous amount of capabilities that every human learns throughout his life, is probably among the most remarkable and fascinating aspects of life. Learning has therefore drawn lots of interest from scientists working in very different fields like philosophy, biology, sociology, educational sciences, computer sciences and mathematics. This thesis focuses on the information theoretical and mathematical aspects of learning. We are interested in the learning process of an agent (which can be for example a human, an animal, a robot, an economical institution or a state) that interacts with its environment. Common models for this interaction are Markov decision processes (MDPs) and partially observable Markov decision processes (POMDPs). Learning is then considered to be the maximization of the expectation of a predefined reward function. In order to formulate general principles (like a formal definition of curiosity-driven learning or avoidance of unpleasant situation) in a rigorous way, it might be desirable to have a theoretical framework for the optimization of more complex functionals of the underlying process law. This might include the entropy of certain sensor values or their mutual information. An optimization of the latter quantity (also known as predictive information) has been investigated intensively both theoretically and experimentally using computer simulations by N. Ay, R. Der, K Zahedi and G. Martius. In this thesis, we develop a mathematical theory for learning in the sensorimotor loop beyond expected reward maximization. Approaches and results This thesis covers four different topics related to the theory of learning in the sensorimotor loop. First of all, we need to specify the model of an agent interacting with the environment, either with learning or without learning. This interaction naturally results in complex causal dependencies. Since we are interested in asymptotic properties of learning algorithms, it is necessary to consider infinite time horizons. It turns out that the well-understood theory of causal networks known from the machine learning literature is not powerful enough for our purpose. Therefore we extend important theorems on causal networks to infinite graphs and general state spaces using analytical methods from measure theoretic probability theory and the theory of discrete time stochastic processes. Furthermore, we prove a generalization of the strong Markov property from Markov processes to infinite causal networks. Secondly, we develop a new idea for a projected stochastic constraint optimization algorithm. Generally a discrete gradient ascent algorithm can be used to generate an iterative sequence that converges to the stationary points of a given optimization problem. Whenever the optimization takes place over a compact subset of a vector space, it is possible that the iterative sequence leaves the constraint set. One possibility to cope with this problem is to project all points to the constraint set using Euclidean best-approximation. The latter is sometimes difficult to calculate. A concrete example is an optimization over the unit ball in a matrix space equipped with operator norm. Our idea consists of a back-projection using quasi-projectors different from the Euclidean best-approximation. In the matrix example, there is another canonical way to force the iterative sequence to stay in the constraint set: Whenever a point leaves the unit ball, it is divided by its norm. For a given target function, this procedure might introduce spurious stationary points on the boundary. We show that this problem can be circumvented by using a gradient that is tailored to the quasi-projector used for back-projection. We state a general technical compatibility condition between a quasi-projector and a metric used for gradient ascent, prove convergence of stochastic iterative sequences and provide an appropriate metric for the unit-ball example. Thirdly, a class of learning problems in the sensorimotor loop is defined and motivated. This class of problems is more general than the usual expected reward maximization and is illustrated by numerous examples (like expected reward maximization, maximization of the predictive information, maximization of the entropy and minimization of the variance of a given reward function). We also provide stationarity conditions together with appropriate gradient formulas. Last but not least, we prove convergence of a stochastic optimization algorithm (as considered in the second topic) applied to a general learning problem (as considered in the third topic). It is shown that the learning algorithm converges to the set of stationary points. Among others, the proof covers the convergence of an improved version of an algorithm for the maximization of the predictive information as proposed by N. Ay, R. Der and K. Zahedi. We also investigate an application to a linear Gaussian dynamic, where the policies are encoded by the unit-ball in a space of matrices equipped with operator norm.

This thesis is concerned with motion planning for the classical two-phase Stefan problem in level set formulation. The interface separating the fluid phases from the solid phases is represented as the zero level set of a continuous function whose evolution is described by the level set equation. Heat conduction in the two phases is modeled by the heat equation. A quadratic tracking-type cost functional that incorporates temperature tracking terms and a control cost term that expresses the desire to have the interface follow a prescribed trajectory by adjusting the heat flux through part of the boundary of the computational domain. The formal Lagrange approach is used to establish a first-order optimality system by applying shape calculus tools. For the numerical solution, the level set equation and its adjoint are discretized in space by discontinuous Galerkin methods that are combined with suitable explicit Runge-Kutta time stepping schemes, while the temperature and its adjoint are approximated in space by the extended finite element method (which accounts for the weak discontinuity of the temperature by a dynamic local modification of the underlying finite element spaces) combined with the implicit Euler method for the temporal discretization. The curvature of the interface which arises in the adjoint system is discretized by a finite element method as well. The projected gradient method, and, in the absence of control constraints, the limited memory BFGS method are used to solve the arising optimization problems. Several numerical examples highlight the potential of the proposed optimal control approach. In particular, they show that it inherits the geometric flexibility of the level set method. Thus, in addition to unidirectional solidification, closed interfaces and changes of topology can be tracked. Finally, the Moreau-Yosida regularization is applied to transform a state constraint on the position of the interface into a penalty term that is added to the cost functional. The optimality conditions for this penalized optimal control problem and its numerical solution are discussed. An example confirms the efficacy of the state constraint. / Die vorliegende Arbeit beschäftigt sich mit einem Optimalsteuerungsproblem für das klassische Stefan-Problem in zwei Phasen. Die Phasengrenze wird als Niveaulinie einer stetigen Funktion modelliert, was die Lösung der so genannten Level-Set-Gleichung erfordert. Durch Anpassen des Wärmeflusses am Rand des betrachteten Gebiets soll ein gewünschter Verlauf der Phasengrenze angesteuert werden. Zusammen mit dem Wunsch, ein vorgegebenes Temperaturprofil zu approximieren, wird dieses Ziel in einem quadratischen Zielfunktional formuliert. Die notwendigen Optimalitätsbedingungen erster Ordnung werden formal mit Hilfe der entsprechenden Lagrange-Funktion und unter Benutzung von Techniken aus der Formoptimierung hergeleitet. Für die numerische Lösung müssen die auftretenden partiellen Differentialgleichungen diskretisiert werden. Dies geschieht im Falle der Level-Set-Gleichung und ihrer Adjungierten auf Basis von unstetigen Galerkin-Verfahren und expliziten Runge-Kutta-Methoden. Die Wärmeleitungsgleichung und die entsprechende Gleichung im adjungierten System werden mit einer erweiterten Finite-Elemente-Methode im Ort sowie dem impliziten Euler-Verfahren in der Zeit diskretisiert. Dieser Zugang umgeht die aufwändige Adaption des Gitters, die normalerweise bei der FE-Diskretisierung von Phasenübergangsproblemen unvermeidbar ist. Auch die Krümmung der Phasengrenze wird numerisch mit Hilfe der Methode der finiten Elemente angenähert. Zur Lösung der auftretenden Optimierungsprobleme werden ein Gradienten-Projektionsverfahren und, im Fall dass keine Kontrollschranken vorliegen, die BFGS-Methode mit beschränktem Speicherbedarf eingesetzt. Numerische Beispiele beleuchten die Stärken des vorgeschlagenen Zugangs. Es stellt sich insbesondere heraus, dass sich die geometrische Flexibilität der Level-Set-Methode auf den vorgeschlagenen Zugang zur optimalen Steuerung vererbt. Zusätzlich zur gerichteten Bewegung einer flachen Phasengrenze können somit auch geschlossene Phasengrenzen sowie topologische Veränderungen angesteuert werden. Exemplarisch, und zwar an Hand einer Beschränkung an die Lage der Phasengrenze, wird auch noch die Behandlung von Zustandsbeschränkungen mittels der Moreau-Yosida-Regularisierung diskutiert. Ein numerisches Beispiel demonstriert die Wirkung der Zustandsbeschränkung.

It is well-known that the solution of a so-called singularly perturbed differential equation exhibits layers. These are small regions in the domain where the solution changes drastically. These layers deteriorate the convergence of standard numerical algorithms, such as the finite element method on a uniform mesh. In the past many approaches were developed to overcome this difficulty. In this context it was very helpful to understand the structure of the solution - especially to know where the layers can occur. Therefore, we have a lot of analysis in the literature concerning the properties of solutions of such problems. Nevertheless, this field is far from being understood conclusively. More recently, there is an increasing interest in the numerics of optimal control problems subject to a singularly perturbed convection-diffusion equation and box constraints for the control. However, it is not much known about the solutions of such optimal control problems. The proposed solution methods are based on the experience one has from scalar singularly perturbed differential equations, but so far, the analysis presented does not use the structure of the solution and in fact, the provided bounds are rather meaningless for solutions which exhibit boundary layers, since these bounds scale like epsilon^(-1.5) as epsilon converges to 0. In this thesis we strive to prove bounds for the solution and its derivatives of the optimal control problem. These bounds show that there is an additional layer that is weaker than the layers one expects knowing the results for scalar differential equation problems, but that weak layer deteriorates the convergence of the proposed methods. In Chapter 1 and 2 we discuss the optimal control problem for the one-dimensional case. We consider the case without control constraints and the case with control constraints separately. For the case without control constraints we develop a method to prove bounds for arbitrary derivatives of the solution, given the data is smooth enough. For the latter case we prove bounds for the derivatives up to the second order. Subsequently, we discuss several discretization methods. In this context we use special Shishkin meshes. These meshes are piecewise equidistant, but have a very fine subdivision in the region of the layers. Additionally, we consider different ways of discretizing the control constraints. The first one enforces the compliance of the constraints everywhere and the other one enforces it only in the mesh nodes. For each proposed algorithm we prove convergence estimates that are independent of the parameter epsilon. Hence, they are meaningful even for small values of epsilon. As a next step we turn to the two-dimensional case. To be able to adapt the proofs of Chapter 2 to this case we require bounds for the solution of the scalar differential equation problem for a right hand side f only in W^(1,infty). Although, a lot of results for this problem can be found in the literature but we can not apply any of them, because they require a smooth right hand side f in C^(2,alpha) for some alpha in (0,1). Therefore, we dedicate Chapter 3 to the analysis of the scalar differential equations problem only using a right hand side f that is not very smooth. In Chapter 4 we strive to prove bounds for the solution of the optimal control problem in the two dimensional case. The analysis for this problem is not complete. Especially, the characteristic layers induce subproblems that are not understood completely. Hence, we can not prove sharp bounds for all terms in the solution decomposition we construct. Nevertheless, we propose a solution method. Numerical results indicate an epsilon-independent convergence for the considered examples - although we are not able to prove this.

Die Arbeit abstrahiert bekannte Konzepte der endlichdimensionalen Optimierung im Hinblick auf deren Anwendung in Banachräumen. Hierfür werden zunächst grundlegende Elemente der Funktionalanalysis wie schwache Konvergenz, Dualräume und Reflexivität vorgestellt. Anschließend erfolgt eine kurze Einführung in die Thematik der Fréchet-Differenzierbarkeit und eine Abstraktion des Begriffs der partiellen Ordnungsrelation in normierten Räumen. Nach der Formulierung eines allgemeinen Existenzsatzes für globale Optimallösungen von abstrakten Optimierungsaufgaben werden notwendige Optimalitätsbedingungen vom Karush-Kuhn-Tucker-Typ hergeleitet. Abschließend wird eine hinreichende Optimalitätsbedingung vom Karush-Kuhn-Tucker-Typ unter verallgemeinerten Konvexitätsvoraussetzungen verifiziert.

This thesis is devoted to the study of optimal control problems governed by a quasistatic, thermoviscoplastic model at small strains with linear kinematic hardening, von Mises yield condition and mixed boundary conditions. Mathematically, the thermoviscoplastic equations are given by nonlinear partial differential equations and a variational inequality of second kind in order to represent the elastic, plastic and thermal effects. Taking into account thermal effects we have to handle numerous mathematical challenges during the analysis of the thermoviscoplastic model, mainly due to the low integrability of the nonlinear terms on the right-hand side of the heat equation. One of our main results is the existence of a unique weak solution, which is proved by means of a fixed-point argument and by employing maximal parabolic regularity theory. Furthermore, we define the related control-to-state mapping and investigate properties of this mapping such as boundedness, weak continuity and local Lipschitz continuity. Another major result is the finding that the mapping is Hadamard differentiable; a main ingredient is the reformulation of the variational inequality, the so called viscoplastic flow rule, as a Banach space-valued ordinary differential equation with non-differentiable right-hand side. Subsequently, we consider an optimal control problem governed by thermoviscoplasticity and show the existence of a minimizer. Finally, close this thesis with numerical examples. / Diese Arbeit ist der Untersuchung von Optimalsteuerproblemen gewidmet, denen ein quasistatisches, thermoviskoplastisches Model mit kleinen Deformationen, mit linearem kinematischen Hardening, von Mises Fließbedingung und gemischten Randbedingungen zu Grunde liegt. Mathematisch werden thermoviskoplastische Systeme durch nichtlineare partielle Differentialgleichungen und eine variationelle Ungleichung der zweiten Art beschrieben, um die elastischen, plastischen und thermischen Effekte abzubilden. Durch die Miteinbeziehung thermischer Effekte, treten verschiedene mathematische Schwierigkeiten während der Analysis des thermoviskoplastischen Systems auf, die ihren Ursprung hauptsächlich in der schlechten Regularität der nichtlinearen Terme auf der rechten Seite der Wärmeleitungsgleichung haben. Eines unserer Hauptresultate ist die Existenz einer eindeutigen schwachen Lösung, welches wir mit Hilfe von einem Fixpunktargument und unter Anwendung von maximaler parabolischer Regularitätstheorie beweisen. Zudem definieren wir die entsprechende Steuerungs-Zustands-Abbildung und untersuchen Eigenschaften dieser Abbildung wie die Beschränktheit, schwache Stetigkeit und lokale Lipschitz Stetigkeit. Ein weiteres wichtiges Resultat ist, dass die Abbildung Hadamard differenzierbar ist; Hauptbestandteil des Beweises ist die Umformulierung der variationellen Ungleichung, der sogenannten viskoplastischen Fließregel, als eine Banachraum-wertige gewöhnliche Differentialgleichung mit nichtdifferenzierbarer rechter Seite. Schließlich runden wir diese Arbeit mit numerischen Beispielen ab.