Advanced search options

Advanced Search Options 🞨

Browse by author name (“Author name starts with…”).

Find ETDs with:

in
/  
in
/  
in
/  
in

Written in Published in Earliest date Latest date

Sorted by

Results per page:

You searched for id:"oai:tuprints.ulb.tu-darmstadt.de:6781". One record found.

Search Limiters

Last 2 Years | English Only

No search limiters apply to these results.

▼ Search Limiters


Technische Universität Darmstadt

1. Ludovici, Francesco. Numerical analysis of parabolic optimal control problems with restrictions on the state and its first derivative.

Degree: PhD, 04 Department of MathematicsNumerical Analysis and Scientific ComputingOptimizationNonlinear Optimization, 2017, Technische Universität Darmstadt

The aim of this thesis is the numerical analysis of optimal control problems governed by parabolic PDEs and subject to constraints on the state variable and its first derivative. The control is acting distributed in time only while the state constraints are considered point-wise in time and global in space; this setting generates an optimization problem of semi-infinite type. The consideration of a space-time discretization of the problem requires the analysis of the convergence of the discretized solution toward the continuous one, as temporal and space mesh size tend to zero. This is based, at any level of discretization, on a priori error estimates for the solution of the parabolic differential equation which are obtained within this thesis. One of the main challenge for state-constrained problem consists in the presence of a Lagrange multiplier appearing as a Borel measure in the system of first-order optimality conditions. In particular, such measure enters the optimality system as data in the adjoint equation affecting the regularity of the adjoint variable itself. Therefore, in the derivation of the convergence rates the use of adjoint information has to be avoided. When considering non-convex problems, the presence of local solutions and the need for second-order optimality conditions require a different strategy compared to the convex case, making the analysis more involved. In particular, the convergence of the discretized solution toward the continuous one is based on a so-called quadratic growth-condition, which arises from the second-order optimality conditions. The a priori error estimates for the PDEs are verified numerically. Advisors/Committee Members: Wollner, Winnifried (advisor), Vexler, Boris (advisor).

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Ludovici, F. (2017). Numerical analysis of parabolic optimal control problems with restrictions on the state and its first derivative. (Doctoral Dissertation). Technische Universität Darmstadt. Retrieved from http://tuprints.ulb.tu-darmstadt.de/6781/

Chicago Manual of Style (16th Edition):

Ludovici, Francesco. “Numerical analysis of parabolic optimal control problems with restrictions on the state and its first derivative.” 2017. Doctoral Dissertation, Technische Universität Darmstadt. Accessed September 22, 2017. http://tuprints.ulb.tu-darmstadt.de/6781/.

MLA Handbook (7th Edition):

Ludovici, Francesco. “Numerical analysis of parabolic optimal control problems with restrictions on the state and its first derivative.” 2017. Web. 22 Sep 2017.

Vancouver:

Ludovici F. Numerical analysis of parabolic optimal control problems with restrictions on the state and its first derivative. [Internet] [Doctoral dissertation]. Technische Universität Darmstadt; 2017. [cited 2017 Sep 22]. Available from: http://tuprints.ulb.tu-darmstadt.de/6781/.

Council of Science Editors:

Ludovici F. Numerical analysis of parabolic optimal control problems with restrictions on the state and its first derivative. [Doctoral Dissertation]. Technische Universität Darmstadt; 2017. Available from: http://tuprints.ulb.tu-darmstadt.de/6781/

.