02947 PDE constrained optimization
2016/2017
Kurset udbydes kun i ulige år. Næste gang i
2017.
Overordnede kursusmål
Students following the course should be able to recognise problems
from the technical sciences that can be modelled as PDE constrained
optimization problems. They should be able to formulate such
problems in a mathematically satisfactory way, derive and prove
properties of the models. Finally students should be able to
approximate solutions to the problems using numerical algorithms.
Læringsmål
En studerende, der fuldt ud har opfyldt kursets mål, vil kunne:
- Recognise and formulate problems involving optimization with
PDE constraints
- Understand and apply theory for optimization problems in Banach
spaces
- Derive the adjoint PDE based on the Lagrangian formulation of
PDE constraints
- Derive necessary optimitality conditions for PDE constrained
problems
- Understand and apply gradient descent methods for the numerical
solution of PDE constrained problems
- Understand and apply Newton methods for the numerical solution
of PDE constrained problems
- Derive and understand the structure of saddle point systems
related to PDE constrained optimization problems and apply suitable
preconditioners
- Practice presentation and discussion of scientific
material
- Reflect upon seminars delivered by leading experts in PDE
constrained optimization
Kursusindhold
Basic Theory of Partial Differential Equations and Their
Discretization; Theory of PDE-Constrained Optimization; Numerical
Optimization Methods; Box-Constrained Problems; Nonsmooth
PDE-Constrained Optimization; Saddle Point Systems; ODE-Constrained
Optimization.
Litteraturhenvisninger
1. Numerical PDE-constrained optimization, Juan Carlos De los
Reyes, Springer 2015 (available as ebook)
2. Optimal Control of PDEs, Fredi Tröltzsch, AMS, 2010.
Sidst opdateret
25. april, 2016