Bounded-But-Unknown uncertainty optimization using design sensitivities and parallel computing: Application to MEMS

In the present paper uncertainty-based design optimization of structures is carried out. A description of uncertainties via bounds on the uncertainty variables is adopted. An anti-optimization technique, which searches for the combinations of uncertainties yielding the worst responses, is used to ta...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Computers & structures Jg. 83; H. 14; S. 1134 - 1149
Hauptverfasser: Gurav, S.P., Goosen, J.F.L., vanKeulen, F.
Format: Journal Article Tagungsbericht
Sprache:Englisch
Veröffentlicht: Oxford Elsevier Ltd 01.05.2005
Elsevier Science
Schlagworte:
Anti-optimization
Bounded-But-Unknown
Design sensitivities
Exact sciences and technology
Fundamental areas of phenomenology (including applications)
Instruments, apparatus, components and techniques common to several branches of physics and astronomy
Mechanical instruments, equipment and techniques
MEMS
Micromechanical devices and systems
Optimization
Parallel computing
Physics
Uncertainties
Parallel computing
Bounded-But-Unknown
Design sensitivities
MEMS
Anti-optimization
Uncertainties
Optimization
Uncertain systems
Sensitivity analysis
Worst case method
Parallel algorithms
Design process
Database
Parallel computation
Microelectromechanical device
Non convex analysis
ISSN:0045-7949, 1879-2243
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:In the present paper uncertainty-based design optimization of structures is carried out. A description of uncertainties via bounds on the uncertainty variables is adopted. An anti-optimization technique, which searches for the combinations of uncertainties yielding the worst responses, is used to tackle these Bounded-But-Unknown uncertainties of non-convex or discontinuous nature. This anti-optimization technique is computationally very expensive and can become impractical for real world applications, in particularly when expensive numerical response evaluations are involved. In order to reduce the number of expensive numerical response evaluations, a modified anti-optimization technique is proposed in the present paper. This enhanced anti-optimization technique incorporates design sensitivities and database technique and is further modified to use parallel computing in order to increase the computational efficiency. The enhanced anti-optimization technique is studied on the basis of test examples from literature and a Microelectromechanical Systems (MEMS) structure. A comparison between results for the examples, clearly shows an improvement in computational efficiency for the anti-optimization technique, due to the use of sensitivities, database and parallel computing. The enhanced anti-optimization technique can be applied efficiently to general problems involving uncertainties of non-convex or discontinuous nature.
Bibliographie:ObjectType-Article-2
SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 23
ISSN:0045-7949
1879-2243
DOI:10.1016/j.compstruc.2004.11.021