Detail publikace

Finite-Element Mesh Optimization Based on the Non-Linear Random Search

Originální název

Finite-Element Mesh Optimization Based on the Non-Linear Random Search

Anglický název

Finite-Element Mesh Optimization Based on the Non-Linear Random Search

Jazyk

en

Originální abstrakt

In the presented submission, an original adaptive optimization of a finite-element mesh for the analysis of shielded planar transmission lines is described. The optimization algorithm is based on the Linear Random Search which is completed by a set of non-linear limitations to eliminate both the extreme sizes of finite elements and a mismatch of the mesh with respect to the analysed structure. From the theoretical point of view, the optimization algorithm can be interpreted as an artificial neural network. From the practical point of view, the algorithm can be shown to converge to the very small error of the finite-element solution.

Anglický abstrakt

In the presented submission, an original adaptive optimization of a finite-element mesh for the analysis of shielded planar transmission lines is described. The optimization algorithm is based on the Linear Random Search which is completed by a set of non-linear limitations to eliminate both the extreme sizes of finite elements and a mismatch of the mesh with respect to the analysed structure. From the theoretical point of view, the optimization algorithm can be interpreted as an artificial neural network. From the practical point of view, the algorithm can be shown to converge to the very small error of the finite-element solution.

Dokumenty

BibTex


@inproceedings{BUT1338,
  author="Zbyněk {Raida}",
  title="Finite-Element Mesh Optimization Based on the Non-Linear Random Search",
  annote="In the presented submission, an original adaptive optimization of a finite-element mesh for the analysis of shielded planar transmission lines is described. The optimization algorithm is based on the Linear Random Search which is completed by a set of non-linear limitations to eliminate both the extreme sizes of finite elements and a mismatch of the mesh with respect to the analysed structure. From the theoretical point of view, the optimization algorithm can be interpreted as an artificial neural network. From the practical point of view, the algorithm can be shown to converge to the very small error of the finite-element solution.",
  address="Int.Conference RADIOELEKTRONIKA 96, VUT v Brně",
  booktitle="Proc. of the 8th Int. Conference Radioelektronika 98",
  chapter="1338",
  edition="2",
  institution="Int.Conference RADIOELEKTRONIKA 96, VUT v Brně",
  year="1998",
  month="march",
  pages="262",
  publisher="Int.Conference RADIOELEKTRONIKA 96, VUT v Brně",
  type="conference paper"
}