Abstract
Surrogate models play an important role in improving design productivity and discovering knowledge at early design stage. In this paper, a preliminary version of an E_volutionary M_odeling A_pproach (EMA) is presented to generate surrogate models for highly nonlinear system with a large design space and limited resources. Through an evolutionary process, less accurate surrogate models gradually evolve into more accurate ones as the quality of the sampling data set is improved. Its use is demonstrated through an application at the early stage of automotive bumper system design and analysis. The results are verified by both FEA and physical test data.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
