Development of a neural network model to predict distortion during the metal forming process by line heating

 

Authors
Pinzón, César; Plazaola, Carlos; Banfield, Ilka; Fong, Amaly; Vega, Adán
Format
Article
Status
publishedVersion
Description

In order to achieve automation of the plate forming process by line heating, it is necessary to know in advance the deformation to be obtained under specific heating conditions. Currently, different methods exist to predict deformation, but these are limited to specific applications and most of them depend on the computational capacity so that only simple structures can be analyzed. In this paper, a neural network model that can accurately predict distortions produced during the plate forming process by line heating, for a wide range of initial conditions including large structures, is presented. Results were compared with data existing in the literature showing excellent performance. Excellent results were obtained for those cases out of the range of the training data.
In order to achieve automation of the plate forming process by line heating, it is necessary to know in advance the deformation to be obtained under specific heating conditions. Currently, different methods exist to predict deformation, but these are limited to specific applications and most of them depend on the computational capacity so that only simple structures can be analyzed. In this paper, a neural network model that can accurately predict distortions produced during the plate forming process by line heating, for a wide range of initial conditions including large structures, is presented. Results were compared with data existing in the literature showing excellent performance. Excellent results were obtained for those cases out of the range of the training data.

Publication Year
2013
Language
eng
Topic
network model
plate forming
distortion prediction
line heating
back propagation
network model
plate forming
distortion prediction
line heating
back propagation
Repository
RI de Documento Digitales de Acceso Abierto de la UTP
Get full text
http://ridda2.utp.ac.pa/handle/123456789/2864
Rights
openAccess
License
https://creativecommons.org/licenses/by-nc-sa/4.0/