| タイトル | Convergence Estimates for Multidisciplinary Analysis and Optimization |
| 本文（外部サイト） | http://hdl.handle.net/2060/19970041147 |
| 著者（英） | Arian, Eyal |
| 著者所属（英） | Institute for Computer Applications in Science and Engineering |
| 発行日 | 1997-10-01 |
| 言語 | eng |
| 内容記述 | A quantitative analysis of coupling between systems of equations is introduced. This analysis is then applied to problems in multidisciplinary analysis, sensitivity, and optimization. For the sensitivity and optimization problems both multidisciplinary and single discipline feasibility schemes are considered. In all these cases a "convergence factor" is estimated in terms of the Jacobians and Hessians of the system, thus it can also be approximated by existing disciplinary analysis and optimization codes. The convergence factor is identified with the measure for the "coupling" between the disciplines in the system. Applications to algorithm development are discussed. Demonstration of the convergence estimates and numerical results are given for a system composed of two non-linear algebraic equations, and for a system composed of two PDEs modeling aeroelasticity. |
| NASA分類 | Numerical Analysis |
| レポートNO | 97N32231
NASA/CR-97-201752
NAS 1.26:201752
ICASE-97-57 |
| 権利 | No Copyright |
|
