Automated Response Surface Model Generation with Sequential Design (original) (raw)
The increasing use of expensive computer simulations in engineering places a serious computational burden on associated optimization problems. Surrogate modelling becomes standard practice in analyzing such expensive blackbox problems. Moreover, surrogate based optimization (SBO) is able to drastically reduce the number of needed function evaluations with respect to traditional methods. This paper briefly discusses several approaches available which use surrogate models for optimization and highlights one sequential design approach in particular, i.e., expected improvement. Expected improvement is described in detail, along with recent related work. The approach has been implemented in a readily available research platform for surrogate modelling and demonstrated on a concrete application from Electro-Magnetics (EM). The results hold competitive designs and one optimum is even able to outperform the reference optimum obtained using extensive domain specific knowledge.
Sign up for access to the world's latest research.
checkGet notified about relevant papers
checkSave papers to use in your research
checkJoin the discussion with peers
checkTrack your impact
Loading Preview
Sorry, preview is currently unavailable. You can download the paper by clicking the button above.