Detail publikace

Combinatorial reliability-based optimization of nonlinear finite element model using an artificial neural network-based approximation

SLOWIK, O. LEHKÝ, D. NOVÁK, D.

Originální název

Combinatorial reliability-based optimization of nonlinear finite element model using an artificial neural network-based approximation

Typ

článek ve sborníku ve WoS nebo Scopus

Jazyk

angličtina

Originální abstrakt

The paper describes the reliability-based optimization of TT shaped precast roof girder produced in Austria. Extensive experimental studies on small specimens and small and full-scale beams have been performed to gain information on fracture mechanical behaviour of utilized concrete. Subsequently, the destructive shear tests under laboratory conditions were performed. Experiments helped to develop an accurate numerical model of the girder. The developed model was consequently used for advanced stochastic analysis of structural response followed by reliability-based optimization to maximize shear and bending capacity of the beam and minimize production cost under defined reliability constraints. The enormous computational requirements were significantly reduced by the utilization of artificial neural network-based approximations of the original nonlinear finite element model of optimized structure.

Klíčová slova

Reliability-based optimization, combinatorial optimization, heuristic optimization, artificial neural network, double-loop reliability-based optimization, prestressed concrete girder optimization, stochastic analysis.

Autoři

SLOWIK, O.; LEHKÝ, D.; NOVÁK, D.

Vydáno

8. 1. 2021

Místo

Siena, Italy

ISBN

978-3-030-64583-0

Kniha

Lecture Notes in Computer Science

Strany od

359

Strany do

371

Strany počet

13

BibTex

@inproceedings{BUT169080,
  author="Ondřej {Slowik} and David {Lehký} and Drahomír {Novák}",
  title="Combinatorial reliability-based optimization of nonlinear finite element model using an artificial neural network-based approximation",
  booktitle="Lecture Notes in Computer Science",
  year="2021",
  pages="359--371",
  address="Siena, Italy",
  doi="10.1007/978-3-030-64583-0\{_}33",
  isbn="978-3-030-64583-0"
}