Talks

Spectral deferred corrections (SDC) is an iterative method for the numerical solution of ordinary differential equations. It can be interpreted as a Picard iteration for the collocation problem, preconditioned with a low order method. SDC has been studied for first order problems, using explicit, implicit or implicit-explicit Euler as preconditioner. It has been shown that SDC can achieve arbitrary high order of accuracy and possesses good stability properties.

In this talk, we will present an analysis of the convergence and stability properties of the SDC method when applied to second-order ODEs and using velocity-Verlet as preconditioner. While a variant of this method called Boris-SDC for the Lorentz equation has been investigated, no general analysis of its properties for general second order problems exists.

We will show that the order of convergence depends on whether the force on the right hand side of the system depends on velocity (like in the Lorentz equation) or not (like in the undamped harmonic oscillator). Moreover, we also show that the SDC iteration is stable under certain conditions. We compare its stability domain with that of the Picard iteration and validate our theoretical analysis in numerical examples.

In this talk, I would like to present the main results of my PhD thesis.
We study a general method to obtain observability estimates for control systems in infinite dimensional spaces by combining an uncertainty principle and a dissipation estimate. Contrary to previous results obtained in the context of Hilbert spaces, we obtain conditions for observability in Banach spaces, allow for more general asymptotic behavior in the assumptions, and retain explicit estimates on the observability constant.
Our approach has applications, e.g., to control systems on non-reflexive spaces and anomalous diffusion operators.
Further, we derive duality results that connect observability estimates to controllability and stabilizability properties. As an application, we study controllability properties of systems given by fractional powers of elliptic differential operators with constant coefficients in $L_p(\mathbb{R}^d)$ for $p\in [1,\infty)$ and thick control sets.

Zoomlink:
https://tuhh.zoom.us/j/84729171896?pwd=ODArbForaUxMM3Q3VTJsNG1kaVNYQT09

Advancement in computational speed is nowadays gained by using more processing units rather than faster ones.
Faults in the processing units caused by numerous sources including radiation and aging have been neglected in the past.
However, the increasing size of HPC machines makes them more susceptible and it is important to develop a resilience strategy to avoid losing millions of CPU hours.
Parallel-in-time methods target the very largest of computers and are hence required to come with algorithm-based fault tolerance.
We look here at spectral deferred corrections (SDC), which is a time marching scheme that is at the heart of parallel-in-time methods such as PFASST.
Due to its iterative nature, there is ample opportunity to plug in computationally inexpensive fault tolerance schemes, many of which are also easy to implement.
We experimentally examine the capability of various strategies to recover from single bit flips in time serial SDC, which will later be applied to parallel-in-time methods.

Zoomlink:
https://tuhh.zoom.us/j/84729171896?pwd=ODArbForaUxMM3Q3VTJsNG1kaVNYQT09

Robots have the potential to disrupt many aspects of our lives, from healthcare to manufacturing. To realize this potential, a key challenge is real-time robot manipulation. Given a task, how can a robot quickly generate a motion plan to successfully complete it? How can the robot react in real-time to potential uncertainties in the real-world as it executes its plan? In this talk, we will overview recent developments at the University of Leeds, to realize real-time robot manipulation. These will include parallel-in-time integration methods that leverage parallel computing to significantly speed-up physics predictions for various robot manipulation tasks. It will also include learning-based and optimal control-based methods for robots to handle real-world uncertainties in object pose estimation and model parameters. We hope these recent advances will help accelerate the next generation of intelligent robots.

Zoomlink: https://tuhh.zoom.us/j/85353626407?pwd=MEIzeTEvY3dRTmtYZjFWUHJaVll4UT09

Meeting ID: 853 5362 6407
Passcode: 045209

The Norros-Reittu model is an inhomogeneous random multigraph that exhibits the so-called scale-free or power-law behaviour, which is observed in real-world complex networks. We study the component sizes of the Norros-Reittu model in the subcritical regime, i.e. in the abscence of a giant component, and show convergence of the point process of the component sizes to a Poisson process. It is planned to derive similar results for other models such as the random connection model.

The thesis deals with boundary layer enrichment of convection dominated flow problems using the Hybrid Discontinuous Galerkin Method. It aims to introduce an appropriate and computationally efficient Hybrid Discontinuous Galerkin formulation for the most important model problems of incompressible fluid flow, namely the convection-diffusion equation.The main contribution is the derivation, discussion and analysis of the Enriched Finite elementSpace using non-polynomial spaces (specifically boundary layer functions) for both the Discontinuous Galerkin Methods and the Hybrid Discontinuous Galerkin Method. We evaluate the robustness (i.e linear stability as well as reasonable linear systems) and accuracy of this method using various analytical and realistic problems and compare the results to those obtained using the standard (H)DG method. Numerical results are provided to contrast the Enriched (H)DG methods with standard (H)DG approaches.