In these lecture notes attention is focused on the use of Response Surface Models as approximation models in engineering optimization. Common strategies are discussed for efficient model construction, based on principles from statistical experimental design theory. Furthermore, modifications of the experimental design theory will be treated, which are necessary and useful on behalf of numerical experimental designs. Finally, guidelines are presented for building and application of response surface models, based on numerical computations.
|Title of host publication||Emerging Methods for Multidisciplinary Optimization|
|Editors||J. Blachut, H.A. Eschenauer|
|Place of Publication||Wien|
|Number of pages||318|
|Publication status||Published - 2001|