Technical Briefs

Numerical Strategies to Reduce the Effect of Ill-Conditioned Correlation Matrices and Underflow Errors in Kriging

[+] Author and Article Information
Lukas J. Haarhoff

Powertech Transformers,
P.O. Box 691,
Pretoria, Gauteng, 0001, South Africa
e-mail: johan.haarhoff@pttransformers.co.za

Schalk Kok

Advanced Mathematical Modelling,
CSIR Modelling and Digital Science,
P.O. Box 395,
Pretoria, Gauteng, 0001, South Africa
e-mail: skok@csir.co.za

Daniel N. Wilke

Department of Mechanical Engineering,
University of Pretoria,
Private bag X20,
Hatfield, Pretoria, Gauteng, 0028, South Africa
e-mail: nico.wilke@up.ac.za

Contributed by the Design Automation Committee of ASME for publication in the JOURNAL OF MECHANICAL DESIGN. Manuscript received October 31, 2011; final manuscript received February 6, 2013; published online March 26, 2013. Assoc. Editor: Timothy W. Simpson.

J. Mech. Des 135(4), 044502 (Mar 26, 2013) (4 pages) Paper No: MD-11-1442; doi: 10.1115/1.4023631 History: Received October 31, 2011; Revised February 06, 2013

Kriging is used extensively as a metamodel in multidisciplinary design optimization. The correlation matrix used in Kriging metamodeling frequently becomes ill-conditioned. Therefore different numerical methods used to solve the Kriging equations affect the search for the optimum Kriging parameters and the ability of the Kriging surface to accurately interpolate known data points. We illustrate this by firstly computing the inverse of the correlation matrix in the Kriging equations, and secondly by solving the systems of equations using decomposition and back substitution, thereby avoiding the inversion of the correlation matrix. Our results clearly show that by decomposing and back substituting, the interpolation accuracy is maintained for significantly higher condition numbers. We then show that computing the natural logarithm of the determinant using additive calculations as opposed to multiplicative calculations significantly reduces numerical underflow errors encountered when searching for the optimum Kriging parameters. Although the effect of decomposition and back substitution are known, and the underflow difficulties when computing the natural logarithm of the determinant of the correlation matrix has been mentioned in passing in Kriging literature, this work clearly quantifies and reinforces these methods, hopefully for the benefit of researchers entering the field.

Copyright © 2013 by ASME
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Fig. 1

Computing the Kriging interpolation errors using either an explicit inverse for R or by using decomposition and back substitution for test problem

Grahic Jump Location
Fig. 2

Using R1 and decomposition and back substitution for Hock 2, DOE 2

Grahic Jump Location
Fig. 3

Using R1 and decomposition and back substitution for Hock 10, DOE 1

Grahic Jump Location
Fig. 4

Using R1 and decomposition and back substitution for Hock 10, DOE 2

Grahic Jump Location
Fig. 5

Using R−1 and decomposition and back substitution for Hock 26, DOE 3

Grahic Jump Location
Fig. 6

Comparing the numerical computation of ln|R| following the multiplicative computation given by Eq. (13), against the additive computation given by Eq. (14), with arbitrary precision results provided as a reference




Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In