Qz. Shi et al., Global optimization method to locate multiple local optima with response surface approximation methodology, JSME A, 44(1), 2001, pp. 175-184
Citations number
16
Categorie Soggetti
Mechanical Engineering
Journal title
JSME INTERNATIONAL JOURNAL SERIES A-SOLID MECHANICS AND MATERIAL ENGINEERING
A considerable number of functional evaluations may be required in the proc
ess of optimization. Although approximation models constructed by response
surface methodology can significantly reduce functional evaluations, the de
sign accuracy may be strongly dependent on the type of activation functions
and designs used. In this paper, we propose techniques to search the desig
n space containing the global optimal design using designs by conditioned r
andom seeds, and techniques for determining more accurate approximations by
employing a sequential approach called the most probable optimal design (M
POD) method. The MPOD method is a response surface methodology based on the
holographic neural network, which uses the exponential function as an acti
vation function. In the MPOD method, extrapolation is employed to make the
technique available for general application in structural optimization. The
formula to estimate the necessary functional evaluations under certain con
ditions is expressed. Application examples show that the MPOD method is an
effective methodology, which is expected to determine the optimal design wi
th multiple local optima.