Search for dissertations about: "Convergence Acceleration"
Showing result 1 - 5 of 20 swedish dissertations containing the words Convergence Acceleration.
-
1. Convergence Acceleration for Flow Problems
Abstract : Convergence acceleration techniques for the iterative solution of system of equations arising in the discretisations of compressible flow problems governed by the steady state Euler or Navier-Stokes equations is considered. The system of PDE is discretised using a finite difference or finite volume method yielding a large sparse system of equations. READ MORE
-
2. Acceleration of Compressible Flow Simulations with Edge using Implicit Time Stepping
Abstract : Computational fluid dynamics (CFD) has become a significant tool routinely used in design and optimization in aerospace industry. Typical flows may be characterized by high-speed and compressible flow features and, in many cases, by massive flow separation and unsteadiness. READ MORE
-
3. Acceleration of Compressible Flow Simulations with Edge Using Implicit Time Stepping
Abstract : Computational fluid dynamics (CFD) is a significant tool routinely used indesign and optimization in aerospace industry. Often cases with unsteadyflows must be computed, and the long compute times of standard methods hasmotivated the present work on new implicit methods to replace the standardexplicit schemes. READ MORE
-
4. Accuracy and Convergence Studies of the Numerical Solution of Compressible Flow Problems
Abstract : The numerical solution of compressible flow problems governed by the Navier-Stokes equations is considered. A finite volume method is used for the discretization in space. Different techniques to accelerate the convergence to a steady state are suggested, and the accuracy of the spatial difference operator is analyzed. READ MORE
-
5. Scalable Optimization Methods for Machine Learning : Acceleration, Adaptivity and Structured Non-Convexity
Abstract : This thesis aims at developing efficient optimization algorithms for solving large-scale machine learning problems. To cope with the increasing scale and complexity of such models, we focus on first-order and stochastic methods in which updates are carried out using only (noisy) information about function values and (sub)gradients. READ MORE