A Machine Learning Approach to Determine Abundance of Inclusions in Stainless Steel

  1. Héctor Mesa 1
  2. Daniel Urda 1
  3. Ruiz-Aguilar, Juan J. 1
  4. Moscoso-López, José A. 1
  5. Juan Almagro 2
  6. Patricia Acosta 2
  7. Turias, Ignacio J. 1
  1. 1 Universidad de Cádiz

    Universidad de Cádiz

    Cádiz, España

    ROR https://ror.org/04mxxkb11

  2. 2 ACERINOX Europa. Dpto. Técnico. (Los Barrios, Cádiz)
Hybrid Artificial Intelligent Systems. 14th International Conference, HAIS 2019: León, Spain, September 4–6, 2019. Proceedings
  1. Hilde Pérez García (coord.)
  2. Lidia Sánchez González (coord.)
  3. Manuel Castejón Limas (coord.)
  4. Héctor Quintián Pardo (coord.)
  5. Emilio Corchado Rodríguez (coord.)

Publisher: Springer Suiza

ISBN: 978-3-030-29859-3 978-3-030-29858-6

Year of publication: 2019

Pages: 504-513

Congress: Hybrid Artificial Intelligent Systems (14. 2019. León)

Type: Conference paper


Steel-making process is a complex procedure involving the presence of exogenous materials which could potentially lead to nonmetallic inclusions. Determining the abundance of inclusions in the earliest stage possible may help to reduce costs and avoid further postprocessing manufacturing steps to alleviate undesired effects. This paper presents a data analysis and machine learning approach to analyze data related to austenitic stainless steel (Type 304L) in order to develop a decision-support tool helping to minimize the inclusion content present in the final product. Several machine learning models (generalized linear models with regularization, random forest, artificial neural networks and support vector machines) were tested in this analysis. Moreover, two different outcomes were analyzed (average and maximum abundance of inclusions per steel cast) and two different settings were considered within the analysis based on the input features used to train the models (full set of features and more relevant ones). The results showed that the average abundance of inclusions can be predicted more accurately than the maximum abundance of inclusions using linear models and the reduced set of features. A list of the more relevant features linked to the abundance of inclusions based on the data and models used in this study is additionally provided.