Publication detail

Higher-order Approximations Methods for Global Sensitivity Analysis of Nonlinear Model Outputs

KALA, Z.

Original Title

Higher-order Approximations Methods for Global Sensitivity Analysis of Nonlinear Model Outputs

Type

journal article in Scopus

Language

English

Original Abstract

The article deals with the use of higher-order approximation polynomials and Latin Hypercube Sampling numerical simulation for the assessment of global sensitivity analysis of nonlinear model output. Highly effective computer evaluation of first and higher order sensitivity indices in non-additive systems is shown. The presented example of a non-additive system of a load-carrying structure was modelled using the nonlinear finite element method and assessed via variance-based sensitivity analysis. The presented sensitivity assessment allows exploration of all regions of the input space, accounting for interactions as well as nonlinear model responses.

Keywords

Sensitivity analysis, mathematical modelling, simulation, buckling, imperfection, reliability, stochastic.

Authors

KALA, Z.

Released

15. 10. 2016

ISBN

1998-0159

Periodical

International Journal of Mathematics and Computers in Simulation

Year of study

2016

Number

10

State

United States of America

Pages from

260

Pages to

264

Pages count

5

BibTex

@article{BUT133162,
  author="Zdeněk {Kala}",
  title="Higher-order Approximations Methods for Global Sensitivity Analysis of Nonlinear Model Outputs",
  journal="International Journal of Mathematics and Computers in Simulation",
  year="2016",
  volume="2016",
  number="10",
  pages="260--264",
  issn="1998-0159"
}