NASA Logo

NTRS

NTRS - NASA Technical Reports Server

Back to Results
New approaches to optimization in aerospace conceptual designAerospace design can be viewed as an optimization process, but conceptual studies are rarely performed using formal search algorithms. Three issues that restrict the success of automatic search are identified in this work. New approaches are introduced to address the integration of analyses and optimizers, to avoid the need for accurate gradient information and a smooth search space (required for calculus-based optimization), and to remove the restrictions imposed by fixed complexity problem formulations. (1) Optimization should be performed in a flexible environment. A quasi-procedural architecture is used to conveniently link analysis modules and automatically coordinate their execution. It efficiently controls a large-scale design tasks. (2) Genetic algorithms provide a search method for discontinuous or noisy domains. The utility of genetic optimization is demonstrated here, but parameter encodings and constraint-handling schemes must be carefully chosen to avoid premature convergence to suboptimal designs. The relationship between genetic and calculus-based methods is explored. (3) A variable-complexity genetic algorithm is created to permit flexible parameterization, so that the level of description can change during optimization. This new optimizer automatically discovers novel designs in structural and aerodynamic tasks.
Document ID
19950018016
Acquisition Source
Legacy CDMS
Document Type
Contractor Report (CR)
Authors
Gage, Peter J.
(Stanford Univ. CA, United States)
Date Acquired
September 6, 2013
Publication Date
March 1, 1995
Subject Category
Aircraft Design, Testing And Performance
Report/Patent Number
NASA-CR-196695
A-950044
NAS 1.26:196695
Accession Number
95N24436
Funding Number(s)
CONTRACT_GRANT: NAG1-1494
PROJECT: RTOP 505-69-50
CONTRACT_GRANT: NAG2-640
CONTRACT_GRANT: NAG1-1558
Distribution Limits
Public
Copyright
Work of the US Gov. Public Use Permitted.
No Preview Available