.png){kind=icon}
.png){kind=icon}
Return
A high-dimensional optimization method combining projection correlation-based Kriging and multimodal parallel computing
.png){kind=icon}
PRE
AI
DOI：10.1007/s00158-022-03450-3.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In surrogate-based optimization (SBO), the recognized issues associated with the high-dimensional surrogate models focus on the prohibitive computational costs and the low model accuracy. However, there is a lack of effective solutions in the face of the 'curse of dimensionality'. In this paper, we propose a novel Kriging metamodel to remedy this deficiency. The Kriging model based on projection correlation (KPC) introduces the projection correlation into the Kriging modeling process as prior information, taking into account the nature of hyperparameters. The effectiveness and accuracy of the KPC are illustrated through 10-70-dimensional numerical examples. Furthermore, a parallel computing strategy that combines the multi-peak characteristics of expected improvement and minimizing prediction (MEI & MP) is proposed to further improve high-dimensional optimization efficiency and potential. The global performance and optimization efficiency of our method are validated via typical test functions and structural optimization problems.
Keywords:
Kriging model
Projection correlation
Fast modeling
Parallel adaptive sampling
High-dimensional optimization
Journal
IF:
4
Papers:
4.8K
Citations:
1.7W