Filters

2016 - 201942020 - 20212

More filters

2019 - 201932020 - 20223
Reset filters

Settings

Author: Hintermüller, Michael × Author: Surowiec, Thomas M. × Start date: 2010 × Type: Text × WGL Institute: WIAS ×

Search Results

Now showing 1 - 6 of 6
  • Thumbnail Image
    Item
    Uncertainty Quantification in Image Segmentation Using the Ambrosio–Tortorelli Approximation of the Mumford–Shah Energy
    (Dordrecht [u.a.] : Springer Science + Business Media B.V, 2021) Hintermüller, Michael; Stengl, Steven-Marian; Surowiec, Thomas M.
    The quantification of uncertainties in image segmentation based on the Mumford–Shah model is studied. The aim is to address the error propagation of noise and other error types in the original image to the restoration result and especially the reconstructed edges (sharp image contrasts). Analytically, we rely on the Ambrosio–Tortorelli approximation and discuss the existence of measurable selections of its solutions as well as sampling-based methods and the limitations of other popular methods. Numerical examples illustrate the theoretical findings.
  • Thumbnail Image
    Item
    Optimization of a multiphysics problem in semiconductor laser design
    (Berlin : Weierstraß-Institut für Angewandte Analysis und Stochastik, 2018) Adam, Lukáš; Hintermüller, Michael; Peschka, Dirk; Surowiec, Thomas M.
    A @multimaterial topology optimization framework is suggested for the simultaneous optimization of mechanical and optical properties to be used in the development of optoelectronic devices. Based on the physical aspects of the underlying device, a nonlinear multiphysics model for the elastic and optical properties is proposed. Rigorous proofs are provided for the sensitivity of the fundamental mode of the device with respect to the changes in the underlying topology. After proving existence and optimality results, numerical experiments leading to an optimal material distribution for maximizing the strain in a Ge-on-Si microbridge are given. The highly favorable electronic properties of this design are demonstrated by steady-state simulations of the corresponding van Roosbroeck (drift-diffusion) system.
  • Thumbnail Image
    Item
    A PDE-constrained optimization approach for topology optimization of strained photonic devices
    (Berlin : Weierstraß-Institut für Angewandte Analysis und Stochastik, 2017) Adam, Lukáš; Hintermüller, Michael; Surowiec, Thomas M.
    Recent studies have demonstrated the potential of using tensile-strained, doped Germanium as a means of developing an integrated light source for (amongst other things) future microprocessors. In this work, a multi-material phase-field approach to determine the optimal material configuration within a so-called Germanium-on-Silicon microbridge is considered. Here, an optimal configuration is one in which the strain in a predetermined minimal optical cavity within the Germanium is maximized according to an appropriately chosen objective functional. Due to manufacturing requirements, the emphasis here is on the cross-section of the device; i.e. a socalled aperture design. Here, the optimization is modeled as a non-linear optimization problem with partial differential equation (PDE) and manufacturing constraints. The resulting problem is analyzed and solved numerically. The theory portion includes a proof of existence of an optimal topology, differential sensitivity analysis of the displacement with respect to the topology, and the derivation of first and second-order optimality conditions. For the numerical experiments, an array of first and second-order solution algorithms in function-space are adapted to the current setting, tested, and compared. The numerical examples yield designs for which a significant increase in strain (as compared to an intuitive empirical design) is observed.
  • Thumbnail Image
    Item
    A semismooth Newton method with analytical path-following for the H1-projection onto the Gibbs simplex
    (Berlin : Weierstraß-Institut für Angewandte Analysis und Stochastik, 2016) Adam, Lukáš; Hintermüller, Michael; Surowiec, Thomas M.
    An efficient, function-space-based second-order method for the H1-projection onto the Gibbs-simplex is presented. The method makes use of the theory of semismooth Newton methods in function spaces as well as Moreau-Yosida regularization and techniques from parametric optimization. A path-following technique is considered for the regularization parameter updates. A rigorous first and second-order sensitivity analysis of the value function for the regularized problem is provided to justify the update scheme. The viability of the algorithm is then demonstrated for two applications found in the literature: binary image inpainting and labeled data classification. In both cases, the algorithm exhibits meshindependent behavior.
  • Thumbnail Image
    Item
    Generalized Nash equilibrium problems with partial differential operators: Theory, algorithms, and risk aversion
    (Berlin : Weierstraß-Institut für Angewandte Analysis und Stochastik, 2019) Gahururu, Deborah; Hintermüller, Michael; Stengl, Steven-Marian; Surowiec, Thomas M.
    PDE-constrained (generalized) Nash equilibrium problems (GNEPs) are considered in a deterministic setting as well as under uncertainty. This includes a study of deterministic GNEPs with nonlinear and/or multivalued operator equations as forward problems and PDE-constrained GNEPs with uncertain data. The deterministic nonlinear problems are analyzed using the theory of generalized convexity for set-valued operators, and a variational approximation approach is proposed. The stochastic setting includes a detailed overview of the recently developed theory and algorithms for risk-averse PDE-constrained optimization problems. These new results open the way to a rigorous study of stochastic PDE-constrained GNEPs.
  • Thumbnail Image
    Item
    Uncertainty quantification in image segmentation using the Ambrosio--Tortorelli approximation of the Mumford--Shah energy
    (Berlin : Weierstraß-Institut für Angewandte Analysis und Stochastik, 2020) Hintermüller, Michael; Stengl, Steven-Marian; Surowiec, Thomas M.
    The quantification of uncertainties in image segmentation based on the Mumford-Shah model is studied. The aim is to address the error propagation of noise and other error types in the original image to the restoration result and especially the reconstructed edges (sharp image contrasts). Analytically, we rely on the Ambrosio-Tortorelli approximation and discuss the existence of measurable selections of its solutions as well as sampling-based methods and the limitations of other popular methods. Numerical examples illustrate the theoretical findings.