Welcome to the upgraded MacSphere! We're putting the finishing touches on it; if you notice anything amiss, email macsphere@mcmaster.ca

On the convergence of space mapping optimization algorithms

Abstract

In this paper, sufficient conditions for the convergence of typical Space Mapping optimization algorithms are considered. It follows that convergence of the algorithms as well as convergence rate depend essentially on the proximity between the fine and coarse model of the object of interest. This proximity can be formulated using some natural analytical conditions involving the model responses and first order derivatives.

Description

Slides for a presentation given as part of the “Space mapping: a knowledge-based engineering modeling and optimization methodology exploiting surrogates” session co-organized by Bandler and Madsen at the Eighth SIAM Conference on Optimization. The conference was held May 15 through 19, 2005 in Stockholm, Sweden.

Citation

Koziel, S., J.W. Bandler, and K. Madsen, “On the convergence of space mapping optimization algorithms,” SIAM Conference on Optimization, Stockholm, Sweden, May 18, 2005.

Endorsement

Review

Supplemented By

Referenced By