An Efficient Uncertainty Quantification Method Using Non-Intrusive Polynomial Chaos Approach

DOI Open Access

Bibliographic Information

Other Title
  • Non-Intrusive Polynomial Chaos 法を用いた効率的な不確かさ定量化法

Search this article

Description

<p>Uncertainty quantification is an evaluation of the probabilistic effect of uncertainty in the input parameters of a system on its output. Monte Carlo methods are familiar as statistical methods for this purpose. However, they have a problem of slow convergence. On the other hand, non-statistical methods expand the input and output random fields into functions in probability space, respectively, and therefore have high computational efficiency and accuracy. In this paper, we integrate the Non-Intrusive Polynomial Chaos (NIPC) method, one of non-statistical methods, with ADVENTURE_Thermal, a heat conduction parallel analysis tool based on Finite Element Method, and solve a steady-state heat conduction equation with thermal conductivity as a stochastic field. For this problem, we evaluate the computational complexity of the proposed method both theoretically and numerically, and highlight the bottleneck processes for larger scale of the problems. In addition, we propose a new method that can be applied to larger scale problems by reducing the computational complexity of the bottleneck process.</p>

Journal

Related Projects

See more

Keywords

Details 詳細情報について

  • CRID
    1390293568073491712
  • DOI
    10.11421/jsces.2022.20220013
  • ISSN
    13478826
    13449443
  • Text Lang
    ja
  • Article Type
    journal article
  • Data Source
    • JaLC
    • KAKEN
  • Abstract License Flag
    Disallowed

Report a problem

Back to top