Machine learning mechanical properties of steel sheets from an industrial production route

Research output: Contribution to journalArticleResearchpeer-review

Standard

Machine learning mechanical properties of steel sheets from an industrial production route. / Millner, Gerfried; Mücke, Manfred; Romaner, Lorenz et al.
In: Materials, Vol. 30.2023, No. August, 101810, 08.2023.

Research output: Contribution to journalArticleResearchpeer-review

Vancouver

Millner G, Mücke M, Romaner L, Scheiber D. Machine learning mechanical properties of steel sheets from an industrial production route. Materials. 2023 Aug;30.2023(August):101810. doi: 10.1016/j.mtla.2023.101810

Author

Millner, Gerfried ; Mücke, Manfred ; Romaner, Lorenz et al. / Machine learning mechanical properties of steel sheets from an industrial production route. In: Materials. 2023 ; Vol. 30.2023, No. August.

Bibtex - Download

@article{2818d7024a314659babd2111dac9f436,
title = "Machine learning mechanical properties of steel sheets from an industrial production route",
abstract = "In this work we apply AI regression models for predicting r-value, tensile strength, yield stress and elongation at fracture of steel coils from chemical composition and process parameters. The data from steel production includes a full chemical analysis, as well as many parameters measured during production of the process and the resulting mechanical properties from tensile tests. As a prerequisite for training AI models, the data needs to be understood, analyzed, checked, and unreasonable data be removed. The result of this data cleaning is a machine-readable dataset fit for various modeling tasks, using AI models such as Random Forest Regression, Support Vector Regression, Artificial Neural Networks and Extreme Gradient Boost. We document the effort for hyperparameter tuning and training for each model type and compare their prediction accuracy. Additionally we apply feature importance determination techniques to reveal the influence of information from different working steps in our models.",
keywords = "Explainable AI, Machine learning, Mechanical properties, Shapley values, Steel coils",
author = "Gerfried Millner and Manfred M{\"u}cke and Lorenz Romaner and Daniel Scheiber",
note = "Publisher Copyright: {\textcopyright} 2023 Acta Materialia Inc.",
year = "2023",
month = aug,
doi = "10.1016/j.mtla.2023.101810",
language = "English",
volume = "30.2023",
journal = "Materials",
issn = "1996-1944",
publisher = "Multidisciplinary Digital Publishing Institute (MDPI)",
number = "August",

}

RIS (suitable for import to EndNote) - Download

TY - JOUR

T1 - Machine learning mechanical properties of steel sheets from an industrial production route

AU - Millner, Gerfried

AU - Mücke, Manfred

AU - Romaner, Lorenz

AU - Scheiber, Daniel

N1 - Publisher Copyright: © 2023 Acta Materialia Inc.

PY - 2023/8

Y1 - 2023/8

N2 - In this work we apply AI regression models for predicting r-value, tensile strength, yield stress and elongation at fracture of steel coils from chemical composition and process parameters. The data from steel production includes a full chemical analysis, as well as many parameters measured during production of the process and the resulting mechanical properties from tensile tests. As a prerequisite for training AI models, the data needs to be understood, analyzed, checked, and unreasonable data be removed. The result of this data cleaning is a machine-readable dataset fit for various modeling tasks, using AI models such as Random Forest Regression, Support Vector Regression, Artificial Neural Networks and Extreme Gradient Boost. We document the effort for hyperparameter tuning and training for each model type and compare their prediction accuracy. Additionally we apply feature importance determination techniques to reveal the influence of information from different working steps in our models.

AB - In this work we apply AI regression models for predicting r-value, tensile strength, yield stress and elongation at fracture of steel coils from chemical composition and process parameters. The data from steel production includes a full chemical analysis, as well as many parameters measured during production of the process and the resulting mechanical properties from tensile tests. As a prerequisite for training AI models, the data needs to be understood, analyzed, checked, and unreasonable data be removed. The result of this data cleaning is a machine-readable dataset fit for various modeling tasks, using AI models such as Random Forest Regression, Support Vector Regression, Artificial Neural Networks and Extreme Gradient Boost. We document the effort for hyperparameter tuning and training for each model type and compare their prediction accuracy. Additionally we apply feature importance determination techniques to reveal the influence of information from different working steps in our models.

KW - Explainable AI

KW - Machine learning

KW - Mechanical properties

KW - Shapley values

KW - Steel coils

UR - http://www.scopus.com/inward/record.url?scp=85162164928&partnerID=8YFLogxK

U2 - 10.1016/j.mtla.2023.101810

DO - 10.1016/j.mtla.2023.101810

M3 - Article

VL - 30.2023

JO - Materials

JF - Materials

SN - 1996-1944

IS - August

M1 - 101810

ER -