Comparison of models created for the prediction of the mechanical properties of galvanized steel coils

  1. Ordieres-Meré, J. 1
  2. Martínez-De-Pisón-Ascacibar, F.J. 2
  3. González-Marcos, A. 2
  4. Ortiz-Marcos, I. 1
  1. 1 Universidad Politécnica de Madrid
    info

    Universidad Politécnica de Madrid

    Madrid, España

    ROR https://ror.org/03n6nwv02

  2. 2 Universidad de La Rioja
    info

    Universidad de La Rioja

    Logroño, España

    ROR https://ror.org/0553yr311

Revista:
Journal of Intelligent Manufacturing

ISSN: 0956-5515

Año de publicación: 2010

Volumen: 21

Número: 4

Páginas: 403-421

Tipo: Artículo

beta Ver similares en nube de resultados
DOI: 10.1007/S10845-008-0189-Y SCOPUS: 2-s2.0-77955660677 WoS: WOS:000279714600006 GOOGLE SCHOLAR

Otras publicaciones en: Journal of Intelligent Manufacturing

Resumen

The mechanical properties of steel coils can be predicted prior to the galvanizing process, thereby preventing the production of galvanized steel coils that are not suitable for the customer's purpose. At the same time, this will reduce the waste of zinc and unfriendly effects on the environment. The purpose of this paper is to present a comparative assessment of the different techniques currently being used to estimate the mechanical properties of steel coils before being processed on the hot dip galvanizing line (HDGL). The authors have evaluated the models established by linear and non-linear, bagging and other more aggregated construction techniques. Using a total of 30 models, predictions were made for each parameter in order to obtain valuable information about the capabilities of these models and techniques. Each model was tested against each data set ten times. Also, models were constructed to determine the relevance of steel grade to the accuracy of the models' predictions by excluding from the models any consideration of steel grade. As a result of our tests of these models, it is possible to recommend the model to choose for each parameter. © 2008 Springer Science+Business Media, LLC.