Detail publikace

Small-sample simulation for uncertainties modelling in engineering: Theory, software and applications

Originální název

Small-sample simulation for uncertainties modelling in engineering: Theory, software and applications

Anglický název

Small-sample simulation for uncertainties modelling in engineering: Theory, software and applications

Jazyk

en

Originální abstrakt

The objective of the paper is to present methods for efficient statistical, sensitivity and reliability assessment. The attention is given to the techniques which are developed for an analysis of computationally intensive problems which is typical for a nonlinear FEM analysis. The paper shows the possibility of randomization of computationally intensive problems in the sense of the Monte Carlo type simulation. Latin hypercube sampling is used, in order to keep the number of required simulations at an acceptable level. The technique is used for both random variables and random fields levels. Sensitivity analysis is based on nonparametric rank-order correlation coefficients. Statistical correlation is imposed by the stochastic optimization technique - the simulated annealing. The simulation can be used for preparation of virtual training set for artificial neural network used in inverse analysis. The multipurpose software FReET is briefly described.

Anglický abstrakt

The objective of the paper is to present methods for efficient statistical, sensitivity and reliability assessment. The attention is given to the techniques which are developed for an analysis of computationally intensive problems which is typical for a nonlinear FEM analysis. The paper shows the possibility of randomization of computationally intensive problems in the sense of the Monte Carlo type simulation. Latin hypercube sampling is used, in order to keep the number of required simulations at an acceptable level. The technique is used for both random variables and random fields levels. Sensitivity analysis is based on nonparametric rank-order correlation coefficients. Statistical correlation is imposed by the stochastic optimization technique - the simulated annealing. The simulation can be used for preparation of virtual training set for artificial neural network used in inverse analysis. The multipurpose software FReET is briefly described.

BibTex


@inproceedings{BUT24780,
  author="Drahomír {Novák}",
  title="Small-sample simulation for uncertainties modelling in engineering: Theory, software and applications",
  annote="The objective of the paper is to present methods for efficient statistical, sensitivity and reliability assessment. The attention is given to the techniques which are developed for an analysis of computationally intensive problems which is typical for a nonlinear FEM analysis. The paper shows the possibility of randomization of computationally intensive problems in the sense of the Monte Carlo type simulation. Latin hypercube sampling is used, in order to keep the number of required simulations at an acceptable level. The technique is used for both random variables and random fields levels. Sensitivity analysis is based on nonparametric rank-order correlation coefficients. Statistical correlation is imposed by the stochastic optimization technique - the simulated annealing. The simulation can be used for preparation of virtual training set for artificial neural network used in inverse analysis. The multipurpose software FReET is briefly described.",
  booktitle="Optimization and Stochastic days",
  chapter="24780",
  year="2006",
  month="november",
  pages="1--14",
  type="conference paper"
}