Metamodels for Computer-based Engineering Design: Survey and recommendations Export

@article{citeulike:3443864, abstract = {Abstract.\ \  The use of statistical techniques to build approximations of expensive computer analysis codes pervades much of today's engineering design. These statistical approximations, or metamodels, are used to replace the actual expensive computer analyses, facilitating multidisciplinary, multiobjective optimization and concept exploration. In this paper, we review several of these techniques, including design of experiments, response surface methodology, Taguchi methods, neural networks, inductive learning and kriging. We survey their existing application in engineering design, and then address the dangers of applying traditional statistical techniques to approximate deterministic computer analysis codes. We conclude with recommendations for the appropriate use of statistical approximation techniques in given situations, and how common pitfalls can be avoided.}, author = {Simpson, T. W. and Poplinski, J. D. and Koch, P. N. and Allen, J. K.}, citeulike-article-id = {3443864}, citeulike-linkout-0 = {http://dx.doi.org/10.1007/PL00007198}, citeulike-linkout-1 = {http://www.springerlink.com/content/74h70gye73ay4amd}, day = {20}, doi = {10.1007/PL00007198}, journal = {Engineering with Computers}, keywords = {computer, doe, model}, month = {July}, number = {2}, pages = {129--150}, posted-at = {2009-08-09 14:52:09}, priority = {2}, title = {Metamodels for Computer-based Engineering Design: Survey and recommendations}, url = {http://dx.doi.org/10.1007/PL00007198}, volume = {17}, year = {2001} }

TY - JOUR ID - citeulike:3443864 L3 - citeulike-article-id:3443864 N2 - Abstract.   The use of statistical techniques to build approximations of expensive computer analysis codes pervades much of today’s engineering design. These statistical approximations, or metamodels, are used to replace the actual expensive computer analyses, facilitating multidisciplinary, multiobjective optimization and concept exploration. In this paper, we review several of these techniques, including design of experiments, response surface methodology, Taguchi methods, neural networks, inductive learning and kriging. We survey their existing application in engineering design, and then address the dangers of applying traditional statistical techniques to approximate deterministic computer analysis codes. We conclude with recommendations for the appropriate use of statistical approximation techniques in given situations, and how common pitfalls can be avoided. IS - 2 JF - Engineering with Computers EP - 150 TI - Metamodels for Computer-based Engineering Design: Survey and recommendations VL - 17 SP - 129 KW - computer KW - doe KW - model AU - Simpson, TW AU - Poplinski, JD AU - Koch, PN AU - Allen, JK PY - 2001/07/20/ UR - http://dx.doi.org/10.1007/PL00007198 ER -