Desarrollo de un modelo de red neuronal para la predicción de distorsión durante el proceso de formado metálico utilizando líneas de calentamiento
DOI:
https://doi.org/10.25043/19098642.77Palabras clave:
modelo de red, formación de placa, predicción de la distorsión, line heating, retropropagaciónResumen
Con el fin de lograr la automatización del proceso de formado metálico por medio de líneas de calentamiento, es necesario conocer de antemano la deformación que se obtendrá bajo condiciones de calentamiento específicos. En la actualidad, hay diferentes métodos para predecir la deformación, pero, éstos se limitan a aplicaciones específicas, y la mayoría de ellos dependen de la capacidad computacional existente, de modo que sólo estructuras simples pueden ser analizadas. En este artículo, un modelo de red neuronal que puede predecir con precisión las distorsiones producidas durante el proceso de formado de placas curvas mediante líneas de calentamiento, para una amplia gama de condiciones iniciales, incluyendo estructuras de gran tamaño es presentado. Los resultados del modelo de red neuronal fueron compararon con datos existentes en la literatura y estos muestran una excelente precisión. Para aquellos casos que están fuera del rango de datos de entrenamiento de la red también se obtuvieron excelentes resultados.
Descargas
Referencias bibliográficas
CHANG, C.W.; LIU, C.S. AND CHANG, J.R. (2005): A Group Preserving Scheme for Inverse Heat Conduction Problems. CMES: Computer Modeling in Engineering & Sciences, 10, 1, pp.13-38.
CHENG, P.; YAO, Y. L.; LIU, C.; PRATT, D.; FAN, Y. (2005): Analysis and Prediction of Size Effect on Laser Forming of Sheet Metal. Journal of Manufacturing Process, SME Vol. 7/No.1; 28-40.
VEGA, A. (2009): Development of Inherent Deformation Database for Automatic Forming of Thick Steel Plates by Line Heating Considering Complex Heating Patterns. Doctoral Thesis. Osaka University, Japan, 2009.
LIU, C.S. (2006): An Efficient Simultaneous Estimation of Temperature-Dependent Thermophysical Properties. CMES: Computer Modeling in Engineering & Sciences, 14, 2, pp.77-90.
MOSHAIOV, A. AND LATORRE, R. (1985): Temperature Distribution during Plate Bending by Torch Flame Heating. Journal of Ship Research, 29, 1, pp.1-11.
LIU, C.S. (2006): An Efficient Simultaneous Estimation of Temperature-Dependent Thermophysical Properties. CMES: Computer Modeling in Engineering & Sciences, 14, 2, pp.77-90.
C. D. JANG, J. S. KIM AND Y. S. HA AND Y. H LEE (2005). Prediction of Plate Deformation Considering Film Boiling in Water Cooling Process after Line Heating. ISOPE’2005, pages 191-197.
OSAWA, N.; HASHIMOTO, K.; SAWAMURA, J.; KIKUCHI, J.; DEGUCHI, Y. AND YAMAURA, T. (2007): Development of Heat Input Estimation Technique for Simulation of Shell Forming by Line-Heating. CMES: Computer Modeling in Engineering & Sciences, 20, 1, pp.45-53.
TANGO, Y, ISHIYAMA, M, NAGAHARA, S, NAGASHIMA, T, AND KOBAYASHI, J, (2003). "Automated Line Heating for Plate Forming by IHI-ALPHA System and its Application to Construction of Actual Vessels-System Outline and Application Record to date," Journal of the Society of Naval Architects of Japan. Vol. 193. pp. 85-95.
TERASAKI, T.; KITAMURA, N. AND NAKAI, M. (1999): Predictive Equation for Thermal Cycle Generated by Line Heating Method. Trans. The West-Japan Soc. Naval Architects, 99, pp.321-329 (in Japanese).
VEGA, A., RASHED, S., TANGO, Y., ISHIYAMA, M., MURAKAWA, H. (2008): Analysis and prediction of multi-heating lines effect on plate forming by line heating. CMES Journal: Computer Modeling in Engineering & Sciences, CMES, Vol. 28, No. 1, pp. 1-14, 2008.
JAIN A., MAO J., MOHIUDDIN A. (1996), “Artificial Neural Networks: A tutorial”, IEEE Computer Society. 1996.
GURVEY K. (1997), “An introduction to Neural Networks”. UCI Press Limited, 1997.
HAGAN M., BEALE M., DEMUTH H. (1996), “Neural Network Design”. PWS Publishing Company, Boston, MA, 1996.
FAUSETT L. (1994), “Fundamentals of neural networks, architectures, algorithms and applications”. Englegood Cliff. NJ; Prentice – Hall. 1994.
Descargas
Publicado
Cómo citar
Número
Sección
Licencia
The authors who publish in this Journal certify that:
- The work submitted for publication in The Ship Science and Technology journal, was written by the author, given that its content is the product of his/her direct intellectual contribution.
- All data and references to material already published are duly identified with their respective credits and are included in the bibliographic notes and quotations highlighted as such.
- All materials submitted for publication are completely free of copyrights; consequently, the author accepts responsibility for any lawsuit or claim related with Intellectual Property Rights thereof, Exonerating of responsibility to The Science and Technology for the Development of Naval, Maritime, and Riverine Industry Corporation, COTECMAR.
- In the event that the article is chosen for publication by The Ship Science and Technology journal, the author state that he/she totally transfers reproduction rights of such to The Science and Technology for the Development of Naval, Maritime, and Riverine Industry Corporation, COTECMAR.
- The authors retain the copyright and transfer to COTECMAR the right of publication and reproduction of the work which will be simultaneously subject to the Creative Commons Attribution License (CC - BY), which allows the license to copy, distribute, display and represent the work and to make derivative works as long as it recognizes and cites the work in the manner specified by the author or licensor.
- For more information about the Creative Commons Attribution License (CC -BY) and his use and scope, please visit the following web page https://creativecommons.org/licenses/by-sa/4.0/legalcode