Neste trabalho buscou-se consolidar aspectos referentes à otimização de problemas envolvidos na mecânica dos meios contínuos, envolvendo diferentes áreas do conhecimento, tais como: otimização matemática, diferenciação automática, análise estrutural, análise aerodinâmica, parametrização de curvas, superfícies e sólidos do tipo B-spline racionais não-uniformes (NURBS, acrônimo do inglês), análise IsoGeométrica (IGA, acrônimo do inglês) e análise por Elementos Finitos (FEA, acrônimo do inglês). Como objetivo final busca-se otimizar formas de cascas estruturais e formas de corpos aerodinâmicos imersos em escoamentos compressíveis. No que concerne à análise estrutural, esta é realizada via análise IsoGeométrica utilizando elementos sólidos para modelar cascas. Uma cinemática co-rotacional abrangente e precisa baseada na exata decomposição polar é desenvolvida, para lidar com problemas estáticos e dinâmicos altamente não lineares. Na análise estática foram implementados o método de Newton-Raphson e controle de deslocamentos generalizado, para problemas dinâmicos foram implementados o método -generalizado (G) e o método energia momento generalizado (GEMM+). A análise aerodinâmica é realizada via análise por Elementos Finitos para modelar escoamentos compressíveis viscosos e não viscosos em regimes transônicos e supersônicos. Um esquema característico baseado na separação da equação de momento (CBS, acrônimo do inglês) é utilizado para obter uma adequada integração temporal. No que concerne à otimização matemática, é utilizado um método baseado em gradientes, conhecido por programação quadrática sequencial (SQP, acrônimo do inglês), onde a avaliação as derivadas de Fréchet são levadas a cabo via diferenciação automática (AD, acrônimo do inglês). No que concerne aos resultados finais é realizada a otimização estrutural de forma de cascas modeladas como sólidos são apresentados, evidenciando um desempenho ótimo com respeito à energia de deformação interna. Os resultados de otimização aerodinâmica bidimensionais apresentam perfis aerodinâmicos ótimos com respeito à relação arrasto/sustentação para uma ampla gama de número de Mach, enquanto um resultado tridimensional é apresentado evidenciando a robustez e eficiência da implementação proposta. Pretendese estabelecer com este trabalho as bases para pesquisas em problemas de otimização aeroelástica. / Consolidation of the link among optimization problems in continuum mechanics, involving different fields, such as mathematical optimization, automatic differentiation, structural analysis, aerodynamic analysis, curves, surfaces and solids parameterization using Non Uniform Rational B-spline (NURBS), IsoGeometric Analysis (IGA), Finite Element Analysis (FEA) is looked for. Structural shape optimization of shell structures and aerodynamic shape optimization of immersed bodies in compressible flows are the main goals of this work. Concerning structural analysis, the so-called IsoGeometric analysis is employed. An accurate and comprehensive corotational kinematic based on the exact polar decomposition is developed in order to study highly nonlinear static and dynamic problems. Static analysis is carried out with Newton-Raphson and Generalized Displacement Control Method, while dynamic analysis is carried out with Generalized- (G) and Generalized Energy-Momentum Method (GEMM+). Aerodynamic analysis is carried out via Finite Element Analysis (FEA) in order to solve compressible flows in transonic and supersonic regimes. A Characteristic Based Split (CBS) method is employed to obtain an accurate time integration, which is based on the splitting of the momentum equation. Concerning mathematical optimization, the so-called Sequential Quadratic Programming (SQP) is employed, which is a gradient-based method, where the Fréchet derivatives are evaluated using Automatic Differentiation (AD). Final results consisting in structural optimization shown an optimal behaviour with respect to internal strain energy. While, results concerning aerodynamic bi-dimensional shape optimization exhibit a optimal behaviour with respect drag/lift ratio, for a large range of Mach number, and a simple result for tri-dimensional case is presented in order to show the efficiency and robustness of the implementation. Bases for future research in aeroelastic optimization problems are established in this work.

Neste trabalho buscou-se consolidar aspectos referentes à otimização de problemas envolvidos na mecânica dos meios contínuos, envolvendo diferentes áreas do conhecimento, tais como: otimização matemática, diferenciação automática, análise estrutural, análise aerodinâmica, parametrização de curvas, superfícies e sólidos do tipo B-spline racionais não-uniformes (NURBS, acrônimo do inglês), análise IsoGeométrica (IGA, acrônimo do inglês) e análise por Elementos Finitos (FEA, acrônimo do inglês). Como objetivo final busca-se otimizar formas de cascas estruturais e formas de corpos aerodinâmicos imersos em escoamentos compressíveis. No que concerne à análise estrutural, esta é realizada via análise IsoGeométrica utilizando elementos sólidos para modelar cascas. Uma cinemática co-rotacional abrangente e precisa baseada na exata decomposição polar é desenvolvida, para lidar com problemas estáticos e dinâmicos altamente não lineares. Na análise estática foram implementados o método de Newton-Raphson e controle de deslocamentos generalizado, para problemas dinâmicos foram implementados o método -generalizado (G) e o método energia momento generalizado (GEMM+). A análise aerodinâmica é realizada via análise por Elementos Finitos para modelar escoamentos compressíveis viscosos e não viscosos em regimes transônicos e supersônicos. Um esquema característico baseado na separação da equação de momento (CBS, acrônimo do inglês) é utilizado para obter uma adequada integração temporal. No que concerne à otimização matemática, é utilizado um método baseado em gradientes, conhecido por programação quadrática sequencial (SQP, acrônimo do inglês), onde a avaliação as derivadas de Fréchet são levadas a cabo via diferenciação automática (AD, acrônimo do inglês). No que concerne aos resultados finais é realizada a otimização estrutural de forma de cascas modeladas como sólidos são apresentados, evidenciando um desempenho ótimo com respeito à energia de deformação interna. Os resultados de otimização aerodinâmica bidimensionais apresentam perfis aerodinâmicos ótimos com respeito à relação arrasto/sustentação para uma ampla gama de número de Mach, enquanto um resultado tridimensional é apresentado evidenciando a robustez e eficiência da implementação proposta. Pretendese estabelecer com este trabalho as bases para pesquisas em problemas de otimização aeroelástica. / Consolidation of the link among optimization problems in continuum mechanics, involving different fields, such as mathematical optimization, automatic differentiation, structural analysis, aerodynamic analysis, curves, surfaces and solids parameterization using Non Uniform Rational B-spline (NURBS), IsoGeometric Analysis (IGA), Finite Element Analysis (FEA) is looked for. Structural shape optimization of shell structures and aerodynamic shape optimization of immersed bodies in compressible flows are the main goals of this work. Concerning structural analysis, the so-called IsoGeometric analysis is employed. An accurate and comprehensive corotational kinematic based on the exact polar decomposition is developed in order to study highly nonlinear static and dynamic problems. Static analysis is carried out with Newton-Raphson and Generalized Displacement Control Method, while dynamic analysis is carried out with Generalized- (G) and Generalized Energy-Momentum Method (GEMM+). Aerodynamic analysis is carried out via Finite Element Analysis (FEA) in order to solve compressible flows in transonic and supersonic regimes. A Characteristic Based Split (CBS) method is employed to obtain an accurate time integration, which is based on the splitting of the momentum equation. Concerning mathematical optimization, the so-called Sequential Quadratic Programming (SQP) is employed, which is a gradient-based method, where the Fréchet derivatives are evaluated using Automatic Differentiation (AD). Final results consisting in structural optimization shown an optimal behaviour with respect to internal strain energy. While, results concerning aerodynamic bi-dimensional shape optimization exhibit a optimal behaviour with respect drag/lift ratio, for a large range of Mach number, and a simple result for tri-dimensional case is presented in order to show the efficiency and robustness of the implementation. Bases for future research in aeroelastic optimization problems are established in this work.

Cette thèse est consacrée au développement d'outils pour FreeFem++ destinés à faciliter la résolution des problèmes d'optimisation. Ce travail se compose de deux parties principales. La première consiste en la programmation, la validation et l'exploitation d'interfaces permettant l¿utilisation de routines d'optimisation directement dans le logiciel. La seconde comprend le développement de solutions pour le calcul automatisé des dérivées, toujours au sein de FreeFem++, en exploitant les paradigmes de la différentiation automatique. FreeFem++ est un environnement de développement intégré dédié à la résolution numérique d¿équations aux dérivées partielles en dimension 2 et 3. Son langage ergonomique permet à l'utilisateur d'exploiter aisément ses nombreux outils de création de maillages, de résolution de systèmes linéaires, ainsi que ses bibliothèques d'éléments finis, etc... Nous introduisons les nouvelles routines d'optimisation désormais accessibles depuis la bibliothèque de modules du logiciel. En particulier, le logiciel libre d'optimisation sous contraintes IPOPT, qui implémente une méthode de points intérieurs très robuste pour l¿optimisation en grande dimension. Nous appliquons avec succès ces algorithmes à une série de problèmes concrets parmi lesquels la résolution numérique de problèmes de sur- faces minimales, la simulation de condensats de Bose-Einstein, ou encore un problème de positionnement inverse en mécanique des fluides. Une version prototypique de FreeFem++ contenant les outils de différentiation automatique est présentée, après avoir exposé les principes fondamentaux de cette méthode de calcul de dérivées pour le calcul scientifique. / The goal of this Ph.D. thesis was the development of tools for the FreeFem++ software in order to make optimization problems easier to deal with. This has been accomplished following two main directions. Firstly, a set of optimization softwares is interfaced and validated before making use of them. Then, we analyse the field of automatic differentiation as a potential mean of simplification for the users. FreeFem++ is an integrated development environment dedicated to numerically solving partial differential equations. Its high level language allows the user for a comfortable experience while using its mesh generation capabilities, linear system solvers, as well as finite elements capabilities. We describe the newly available optimization features, with a certain emphasis on the open source software IPOPT, which implements a state of the art interior points method for large scale optimization. These optimization tools are then used in a set of quite successful applications, among which minimal surfaces, Bose-Einstein condensate simulation, and an inverse positioning problem in the context of computational fluid dynamics. Finally, after an introduction to the techniques of algorithmic differentiation, we also present an unstable prototype version of FreeFem++ including automatic differentiation features.

Zur Bestimmung singulärer Punkte eines bestimmten Typs muss eine zugehörige reduzierte Funktion und deren Ableitungen bestimmte Bedingungen erfüllen. Dabei ist diese reduzierte Funktion implizit durch ein nichtlineares Gleichungssystem definiert. Man erhält letztendlich ein minimal erweitertes System, das auch Ableitungen der reduzierten Funktion enthält,und den singulären Punkt als reguläre Lösung besitzt. In der vorliegenden Arbeit wird die Technik des automatischen Differenzierens für die Vorwärtsmethode dargestellt, insbesondere wird die Differentiation iterativer Verfahren untersucht. Es wird ein Überblick über die Theorie von singulären Punkten gegeben und das Erkennungsproblem definiert. Ein zweistufiges Verfahren zur Bestimmung singulärer Punkte wird auf Basis der Vorwärtsmethode und des Newton-Verfahrens beschrieben und wurde an verschiedenen Typen von singulären Punkten getestet.

The technique of automatic differentiation provides directional derivatives and discrete adjoints with working accuracy. A complete complexity analysis of the basic modes of automatic differentiation is available. Therefore, the research activities are focused now on different aspects of the derivative calculation, as for example the efficient implementation by exploitation of structural information, studies of the theoretical properties of the provided derivatives in the context of optimization problems, and the development and analysis of new mathematical algorithms based on discrete adjoint information. According to this motivation, this habilitation presents an analysis of different checkpointing strategies to reduce the memory requirement of the discrete adjoint computation. Additionally, a new algorithm for computing sparse Hessian matrices is presented including a complexity analysis and a report on practical experiments. Hence, the first two contributions of this thesis are dedicated to an efficient computation of discrete adjoints. The analysis of discrete adjoints with respect to their theoretical properties is another important research topic. The third and fourth contribution of this thesis focus on the relation of discrete adjoint information and continuous adjoint information for optimal control problems. Here, differences resulting from different discretization strategies as well as convergence properties of the discrete adjoints are analyzed comprehensively. In the fifth contribution, checkpointing approaches that are successfully applied for the computation of discrete adjoints, are adapted such that they can be used also for the computation of continuous adjoints. Additionally, the fifth contributions presents a new proof of optimality for the binomial checkpointing that is based on new theoretical results. Discrete adjoint information can be applied for example for the approximation of dense Jacobian matrices. The development and analysis of new mathematical algorithms based on these approximate Jacobians is the topic of the sixth contribution. Is was possible to show global convergence to first-order critical points for a whole class of trust-region methods. Here, the usage of inexact Jacobian matrices allows a considerable reduction of the computational complexity.

Mathematical optimization has many applications in operations research, image processing, and machine learning, demanding not only computational efficiency but also convenience and correctness in constructing complex models. In this work, we introduce HashedExpression, an open-source algebraic modeling lan- guage (AML) that allows users to express unconstrained, box-constrained, and scalar-expressions-constrained optimization problems, aimed at embeddability, type-safety, and high-performance through symbolic transformation and code generation. Written in Haskell, a statically-typed, purely functional program- ming language, HashedExpression places a great emphasis on modeling correct- ness by providing users with a type-safe, correct-by-construction interface that uses Haskell type-level programming to express constraints on correctness which the compiler uses to flag many modelling errors as type errors (at compile time). We show how type-safety can be added in steps, first matching expressions’ shape and then associated physical units. The library implements symbolic ex- pressions with a hashed indexing scheme to implement common subexpression elimination (CSE). It abstracts away details of the underlying lookup table via monadic type class instances. We explain how using symbolic expressions with CSE enables performance-enhance transformations and automatic computation of derivatives without the issue of “expression swelling”. For high-performance purposes, we generate low-level C/C++ code for symbolic expressions and pro- vide bindings to open-source optimization solvers such as Ipopt or L-BFGS-B. We explain how this architecture lays the groundwork for future work on par- allelization including SIMDization and targetting multi-core CPUs and GPUs, and other hardware acceleration. / Thesis / Master of Science (MSc)

Design optimization is a method that can be used to automate the design process to obtain better results. When applied to aeroelastic structures, design optimization often leads to the creation of highly flexible aeroelastic structures. There are, however, a number of conventional design procedures that must be modified when dealing with highly flexible aeroelastic structures. First, the deformed geometry must be the baseline for weight, structural, and stability analyses. Second, potential couplings between aeroelasticity and rigid-body dynamics must be considered. Third, dynamic analyses must be modified to handle large nonlinear displacements. These modifications to the conventional design process significantly increase the difficulty of developing an optimization framework appropriate for highly flexible aeroelastic structures. As a result, when designing these structures, often either gradient-free optimization is performed (which limits the optimization to relatively few design variables) or optimization is simply omitted from the design process. Both options significantly decrease the design exploration capabilities of a designer compared to a scenario in which gradient-based optimization is used. This dissertation therefore presents various contributions that allow gradient-based optimization to be more easily used to optimize highly flexible aeroelastic structures. One of our primary motivations for developing these capabilities is to accurately capture the design constraints of solar-regenerative high-altitude long-endurance (SR-HALE) aircraft. In this dissertation, we therefore present a SR-HALE aircraft optimization framework which accounts for the peculiarities of structurally flexible aircraft while remaining suitable for use with gradient-based optimization. These aircraft tend to be extremely large and light, which often leads to significant amounts of structural flexibility. Using this optimization framework, we design an aircraft that is capable of flying year-round at \SI{35}{\degree} latitude at \SI{18}{\kilo\meter} above sea level. We subject this aircraft to a number of constraints including energy capture, energy storage, material failure, local buckling, stall, static stability, and dynamic stability constraints. Critically, these constraints were designed to accurately model the actual design requirements of SR-HALE aircraft, rather than to provide a rough approximation of them. To demonstrate the design exploration capabilities of this framework, we also performed several parameters sweeps to determine optimal design sensitivities to altitude, latitude, battery specific energy, solar efficiency, avionics and payload power requirements, and minimum design velocity. Through this optimization framework, we demonstrate both the potential of gradient-based optimization applied to highly flexible aeroelastic structures and the challenges associated with it. One challenge associated with the gradient-based optimization of highly flexible aeroelastic structures, is the ability to accurately, efficiently, and reliably model the large deflections of these structures in gradient-based optimization frameworks. To enable large-scale optimization involving structural models with large deflections to be performed more easily, we present a finite-element implementation of geometrically exact beam theory which is designed specifically for gradient-based optimization. A key feature of this implementation of geometrically exact beam theory is its compatibility with forward and reverse-mode automatic differentiation, which allows accurate design sensitivities to be obtained with minimal development effort. Another key feature is its native support for unsteady adjoint sensitivity analysis, which allows design sensitivities to be obtained efficiently from time-marching simulations. Other features are also presented that build upon previous implementations of geometrically exact beam theory, including a singularity-free rotation parameterization based on Wiener-Milenkovi\'c parameters, an implementation of stiffness-proportional structural damping using a discretized form of the compatibility equations, and a reformulation of the equations of motion for geometrically exact beam theory from a fully implicit index-1 differential algebraic equation to a semi-explicit index-1 differential algebraic equation. Several examples are presented which verify the utility and validity of each of these features. Another challenge associated with the gradient-based optimization of highly flexible aeroelastic structures is the ability to reliably track and constrain individual dynamic stability modes across the design iterations of an optimization framework. To facilitate the development of mode-specific dynamic stability constraints in gradient-based optimization frameworks we develop a mode tracking method that uses an adaptive step size in order to maintain an arbitrarily high degree of confidence in mode correlations. This mode tracking method is then applied to track the modes of a linear two-dimensional aeroelastic system and a nonlinear three-dimensional aeroelastic system as velocity is increased. When used in a gradient-based optimization framework, this mode tracking method has the potential to allow continuous dynamic stability constraints to be constructed without constraint aggregation. It also has the potential to allow the stability and shape of specific modes to be constrained independently. Finally, to facilitate the development and use of highly flexible aeroelastic systems for use in gradient-based optimization frameworks, we introduce a general methodology for coupling aerodynamic and structural models together to form modular monolithic aeroelastic systems. We also propose efficient methods for computing the Jacobians of these coupled systems without significantly increasing the amount of time necessary to construct these systems. For completeness we also discuss how to ensure that the resulting system of equations constitutes a set of first-order index-1 differential algebraic equations. We then derive direct and adjoint sensitivities for these systems which are compatible with automatic differentiation so that derivatives for gradient-based optimization can be obtained with minimal development effort.

In dieser Arbeit werden Techniken beschrieben, die es erlauben (höhere) Ableitungen und Taylorapproximationen solcher Computerprogramme effizient zu berechnen. Auch inbesondere dann, wenn die Programme Algorithmen der numerischen linearen Algebra (NLA) enthalten. Im Gegensatz zur traditionellen algorithmischen Differentiation (AD), bei der die zugrunde liegenden Algorithmen um zusätzliche Befehlere erweitert werden, sind in dieser Arbeit die Zerlegungen durch definierende Gleichungen charakterisiert. Basierend auf den definierenden Gleichungen werden Strukturausnutzende Algorithmen hergeleitet. Genauer, neuartige Algorithmen für die Propagation von Taylorpolynomen durch die QR, Cholesky und reell-symmetrischen Eigenwertzerlegung werden präsentiert. Desweiteren werden Algorithmen für den Rückwärtsmodus der AD hergeleitet, welche im Wesentlichen nur die Faktoren der Zerlegungen benötigen. Im Vergleich zum traditionellen Ansatz, bei dem alle Zwischenergebnisse gespeichert werden, ist dies eine Reduktion von O(N^3) zu O(N^2) für Algorithmen mit O(N^3) Komplexität. N ist hier die Größe der Matrix. Zusätzlich kann bestehende, hoch-optimierte Software verwendet werden. Ein Laufzeitvergleich zeigt, dass dies im Vergleich zum traditionellen Ansatz zu einer Beschleunigung in der Größenordnung 100 führen kann. Da die NLA Funktionen als Black Box betrachtet werden, ist desweiteren auch der Berechnungsgraph um Größenordnungen kleiner. Dies bedeutet, dass Software, welche Operator Overloading benutzt, weniger Overhead hervorruft und auch weniger Speicher benötigt. / This thesis provides a framework for the evaluation of first and higher-order derivatives and Taylor series expansions through large computer programs that contain numerical linear algebra (NLA) functions. It is a generalization of traditional algorithmic differentiation (AD) techniques in that NLA functions are regarded as black boxes where the inputs and outputs are related by defining equations. Based on the defining equations, structure-exploiting algorithms are derived. More precisely, novel algorithms for the propagation of Taylor polynomials through the QR, Cholesky,- and real-symmetric eigenvalue decomposition are shown. Recurrences for the reverse mode of AD, which require essentially only the returned factors of the decomposition, are also derived. Compared to the traditional approach where all intermediates of an algorithm are stored, this is a reduction from O(N^3) to O(N^2) for algorithms with O( N^3) complexity. N denotes the matrix size. The derived algorithms make it possible to use existing high-performance implementations. A runtime comparison shows that the treatment of NLA functions as atomic can be more than one order of magnitude faster than an automatic differentiation of the underlying algorithm. Furthermore, the computational graph is orders of magnitudes smaller. This reduces the additional memory requirements, as well as the overhead, of operator overloading techniques to a fraction.

Mit Hilfe der Technik des Automatischen Differenzierens (AD) lassen sich für Funktionen, die als Programmquellcode gegeben sind, Ableitungsinformationen rechentechnisch effizient und mit geringem Aufwand für den Nutzer bereitstellen. Eine Variante der Implementierung von AD basiert auf der Überladung von Operatoren und Funktionen, die von vielen modernen Programmiersprachen ermöglicht wird. Durch Ausnutzung des Konzepts der Überladung wird eine interne Funktions-Repräsentation (Tape) generiert, die anschließend für die Ableitungsberechnung herangezogen wird. In der Dissertation werden neue Techniken erarbeitet, die eine effizientere Tape-Erstellung und die parallele Tape-Auswertung ermöglichen. Anhand von Laufzeituntersuchungen für numerische Beispiele werden die Möglichkeiten der neuen Techniken verdeutlicht. / Using the technique of Automatic Differentiation (AD), derivative information can be computed efficiently for any function that is given as source code in a supported programming languages. One basic implementation strategy is based on the concept of operator overloading that is available for many programming languages. Due the overloading of operators, an internal representation of the function can be generated at runtime. This so-called tape can then be used for computing derivatives. In the thesis, new techniques are introduced that allow a more efficient tape creation and the parallel evaluation of tapes. Advantages of the new techniques are demonstrated by means of runtime analyses for numerical examples.

This thesis is concerned with the efficient computation of Jacobian matrices of nonlinear vector maps using automatic differentiation (AD). Specifically, we propose the use of two directed edge separator methods, the weighted minimum separator and natural order separator methods, to exploit the structure of the computational graph of the nonlinear system.This allows for the efficient determination of the Jacobian matrix using AD software. We will illustrate the promise of this approach with computational experiments.