Logo des Repositoriums
Zur Startseite
  • English
  • Deutsch
Anmelden
  1. Startseite
  2. SuUB
  3. Dissertationen
  4. Korrekturverfahren zur numerischen Lösung nichtlinearer Optimierungsprobleme mittels Methoden der parametrischen Sensitivitätsanalyse
 
Zitierlink URN
https://nbn-resolving.de/urn:nbn:de:gbv:46-00102497-19

Korrekturverfahren zur numerischen Lösung nichtlinearer Optimierungsprobleme mittels Methoden der parametrischen Sensitivitätsanalyse

Veröffentlichungsdatum
2012-02-24
Autoren
Nikolayzik, Tim  
Betreuer
Büskens, Christof  
Gutachter
Gerdts, Matthias  
Zusammenfassung
The nonlinear optimization has becoming of the key technologies of the field of applied mathematics. There are various numbers of applications in different fields of science, e.g. economics, medicine and engineering. Also industry is using mathematical algorithms for solving nonlinear optimization problems. Especially in the aerospace industry nonlinear optimization is frequently used for optimizing trajectories. The aim of this thesis is to extend an existing method for solving nonlinear optimization problems by using parametric sensitivity analysis. At first the iterative SQP-method WORHP for solving nonlinear optimization problems is described and introduced. Then it is shown that the sensitivity derivatives of a quadratic subproblem within WORHP can be computed efficiently with almost no extra costs. The subproblem is now treated as a perturbed problem. In the thesis several different perturbations are introduced. With the sensitivity derivatives it is now possible to compute an approximation of the solution of the perturbed quadratic subproblem. This solution is used within the overall nonlinear optimization method for obtaining a new iterate. The numerical results show that it is possible to reduce of number of iterations and sometimes it also possible to reduce the time needed for solving a problem.
Schlagwörter
nonlinear optimization

; 

correction steps

; 

parametric sensitivity analysis

; 

WORHP

; 

quadratic subproblem
Institution
Universität Bremen  
Fachbereich
Fachbereich 03: Mathematik/Informatik (FB 03)  
Dokumenttyp
Dissertation
Zweitveröffentlichung
Nein
Sprache
Deutsch
Dateien
Lade...
Vorschaubild
Name

00102497-1.pdf

Size

2.97 MB

Format

Adobe PDF

Checksum

(MD5):8182d1597eb088a1fdeda9754a998809

Built with DSpace-CRIS software - Extension maintained and optimized by 4Science

  • Datenschutzbestimmungen
  • Endnutzervereinbarung
  • Feedback schicken