| タイトル | Response Surface Modeling Using Multivariate Orthogonal Functions |
| 本文（外部サイト） | http://hdl.handle.net/2060/20010023440 |
| 著者（英） | Morelli, Eugene A.; DeLoach, Richard |
| 著者所属（英） | NASA Langley Research Center |
| 発行日 | 2001-01-01 |
| 言語 | eng |
| 内容記述 | A nonlinear modeling technique was used to characterize response surfaces for non-dimensional longitudinal aerodynamic force and moment coefficients, based on wind tunnel data from a commercial jet transport model. Data were collected using two experimental procedures - one based on modem design of experiments (MDOE), and one using a classical one factor at a time (OFAT) approach. The nonlinear modeling technique used multivariate orthogonal functions generated from the independent variable data as modeling functions in a least squares context to characterize the response surfaces. Model terms were selected automatically using a prediction error metric. Prediction error bounds computed from the modeling data alone were found to be- a good measure of actual prediction error for prediction points within the inference space. Root-mean-square model fit error and prediction error were less than 4 percent of the mean response value in all cases. Efficacy and prediction performance of the response surface models identified from both MDOE and OFAT experiments were investigated. |
| NASA分類 | Aerodynamics |
| レポートNO | AIAA Paper 2001-0168 |
| 権利 | Copyright, Distribution as joint owner in the copyright |
| URI | https://repository.exst.jaxa.jp/dspace/handle/a-is/93281 |
