Title: A multi-criteria adaptive sequential sampling method for radial basis function

Authors: Haiyang Hu; Zhansi Jiang; Yanxue Wang; Shuilong He

Addresses: Department of Mechanical and Electrical Engineering, Guilin University of Electronic Technology, Guilin, China ' Department of Mechanical and Electrical Engineering, Guilin University of Electronic Technology, Guilin, China ' Beijing Key Laboratory of Performance Guarantee on Urban Rail Transit Vehicles, Beijing University of Civil Engineering and Architecture, Beijing, China ' Department of Mechanical and Electrical Engineering, Guilin University of Electronic Technology, Guilin, China

Abstract: A multi-criteria adaptive sequential sampling method is proposed for radial basis function metamodel and a new global approximation method is developed in this paper. In this new sampling method, objective, curvature and distance are considered as sampling criteria. With the three criteria, it guarantees that the entire domain will be covered by samples, and more sampling points will be gathered in the peak and valley regions, which is useful for enhance accuracy and efficiency of approximation model. Intensive testing shows that the efficiency of method and accuracy of metamodel are satisfactory by this new global approximation method.

Keywords: multi-criteria adaptive sequential sampling; global approximation; metamodel; radial basis function.

DOI: 10.1504/IJCSM.2020.10029254

International Journal of Computing Science and Mathematics, 2020 Vol.11 No.4, pp.305 - 315

Received: 02 Nov 2017
Accepted: 11 Jan 2018

Published online: 02 Jun 2020 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article