Konzept zur automatischen Bauteilerkennung innerhalb der FE-SoftwareUmgebung mittels Künstlichen Neuronalen Netzen

DFX 2014: Proceedings of the 24th Symposium Design for X: Bamburg, Germany 1-2 Oktober 2014

Year: 2014
Editor: Dieter Krause, Kristin Paetzold, Sandro Wartzack
Author: Spruegel, Tobias C.; Wartzack, Sandro
Series: DfX
Institution: Friedrich-Alexander-Universit
Section: Wissensbasiertes Konstruieren
Page(s): 47-56
ISBN: 978-3-941492-79-0

Abstract

The Finite Element Method is a very powerful tool to enhance product quality and to reduce unnecessary and costly iterations during the product development process. However numerous errors can occur during the FE simulation. A promising way to realize precise and quick FE simulations is the comparison of the FE solution with analytical machine elements equations. This is only possible if the components in an assembly can be identified automatically. A promising way is to use the presented approach in this paper
for part recognition with artificial neural networks and the use of detection surfaces.

Keywords: automatic part recognition, Artificial Neural Network, ANN, FEM, detection surface