利用深度学习对深冲钢机械性能的预测

徐钢; 何金珊; 吕志民; 黎敏; 徐金梧

doi:10.1007/s12613-022-2547-8

Volume 30 Issue 1

Jan. 2023

Turn off MathJax

图(8) / 表(4)

数据统计

计量

文章访问数: 910
HTML全文浏览量: 297
PDF下载量: 64
被引次数: 0

文章导航 > 矿物冶金与材料学报（英文） > 2023 > 30(1): 156-165

Gang Xu, Jinshan He, Zhimin Lü, Min Li, and Jinwu Xu, Prediction of mechanical properties for deep drawing steel by deep learning, Int. J. Miner. Metall. Mater., 30(2023), No. 1, pp. 156-165. https://doi.org/10.1007/s12613-022-2547-8

Cite this article as:

Gang Xu, Jinshan He, Zhimin Lü, Min Li, and Jinwu Xu, Prediction of mechanical properties for deep drawing steel by deep learning, Int. J. Miner. Metall. Mater., 30(2023), No. 1, pp. 156-165. https://doi.org/10.1007/s12613-022-2547-8

Citation:

PDF( 2150 KB)

引用本文 PDF XML SpringerLink

研究论文

利用深度学习对深冲钢机械性能的预测

1)
北京科技大学钢铁共性技术协同创新中心，北京 100083，中国
2)
苏州宝联重工有限公司，苏州 215131，中国

通讯作者:
徐钢 E-mail: watermoon2012@gmail.com

文章亮点

(1) 采用机器学习方法建立数字孪生模型应用在IF钢在线力学性能预测。

(2) 采用GRU方法建立的数字孪生模型有更好的预测精度。

(3) 挖掘出的成分–工艺–性能映射关系可用于新材料研发和产品质量提升。

中文摘要

当前钢铁企业主要采用“事后抽检”方式来控制最终的产品质量。但因无法基于传统方法对所有产品实现在线质量预测，常发生索赔和退货，这也是导致企业经济损失的一大因素。在生产过程中为实现对深冲钢在线质量预测，引入了基于深度学习的机械性能预测模型。首先利用LSTM（长短时记忆）、GRU（门控循环单元）网络和GPR（高斯过程回归）模型去预测深冲钢的机械性能，并讨论了三种模式的预测精度和学习效率，其次提出了在线迁移学习模型。从结果来看不仅预测精度得到改善，而且预测耗时缩短能更好满足在线实时预测的要求。

关键词:

Research Article

Prediction of mechanical properties for deep drawing steel by deep learning

+ Author Affiliations

Abstract

Abstract

At present, iron and steel enterprises mainly use “after spot test ward” to control final product quality. However, it is impossible to realize on-line quality predetermining for all products by this traditional approach, hence claims and returns often occur, resulting in major economic losses of enterprises. In order to realize the on-line quality predetermining for steel products during manufacturing process, the prediction models of mechanical properties based on deep learning have been proposed in this work. First, the mechanical properties of deep drawing steels were predicted by using LSTM (long short team memory), GRU (gated recurrent unit) network, and GPR (Gaussian process regression) model, and prediction accuracy and learning efficiency for different models were also discussed. Then, on-line re-learning methods for transfer learning models and model parameters were proposed. The experimental results show that not only the prediction accuracy of optimized transfer learning models has been improved, but also predetermining time was shortened to meet real time requirements of on-line property predetermining. The industrial production data of interstitial-free (IF) steel was used to demonstrate that R² value of GRU model in training stage reaches more than 0.99, and R² value in testing stage is more than 0.96.
- machine learning;
- recurrent natural network;
- transfer learning;
- on-line prediction;
- deep drawing steel;
- mechanical properties

FullText(HTML)

Supplements(0)

References(38)

References

[1]	T. Xie and J.C. Grossman, Crystal graph convolutional neural networks for an accurate and interpretable prediction of material properties, Phys. Rev. Lett., 120(2018), No. 14, art. No. 145301. doi: 10.1103/PhysRevLett.120.145301
[2]	L. Ward, S.C. O’Keeffe, J. Stevick, et al., A machine learning approach for engineering bulk metallic glass alloys, Acta Mater., 159(2018), p. 102. doi: 10.1016/j.actamat.2018.08.002
[3]	D. Shin, Y. Yamamoto, M.P. Brady, S. Lee, and J.A. Haynes, Modern data analytics approach to predict creep of high-temperature alloys, Acta Mater., 168(2019), p. 321. doi: 10.1016/j.actamat.2019.02.017
[4]	C. Wang, D.Q. Shi, and S.L. Li, A study on establishing a microstructure-related hardness model with precipitate segmentation using deep learning method, Materials (Basel), 13(2020), No. 5, art. No. 1256.
[5]	K. Rajan, Materials informatics: The materials “gene” and big data, Annu. Rev. Mater. Res., 45(2015), p. 153. doi: 10.1146/annurev-matsci-070214-021132
[6]	S. Kalidindi, Hierarchical Materials Informatics: Novel Analytics for Materials Data, Butterworth-Heinemann, 2015.
[7]	K. Kim, L. Ward, J.G. He, et al., Machine-learning-accelerated high-throughput materials screening: Discovery of novel quaternary Heusler compounds, Phys. Rev. Mater., 2(2018), No. 12, art. No. 123801. doi: 10.1103/PhysRevMaterials.2.123801
[8]	C.W. Rosenbrock, K. Gubaev, A.V. Shapeev, et al., Machine-learned interatomic potentials for alloys and alloy phase diagrams, npj Comput. Mater., 7(2021), No. 1, p. 1. doi: 10.1038/s41524-020-00473-6
[9]	C. Nyshadham, M. Rupp, B. Bekker, et al., Machine-learned multi-system surrogate models for materials prediction, npj Comput. Mater., 5(2019), art. No. 51. doi: 10.1038/s41524-019-0189-9
[10]	Y.S. Fan, X.G. Yang, D.Q. Shi, L. Tan, and W.Q. Huang, Quantitative mapping of service process-microstructural degradation-property deterioration for a Ni-based superalloy based on chord length distribution imaging process, Mater. Des., 203(2021), art. No. 109561. doi: 10.1016/j.matdes.2021.109561
[11]	B. Yucel, S. Yucel, A. Ray, L. Duprez, and S.R. Kalidindi, Mining the correlations between optical micrographs and mechanical properties of cold-rolled HSLA steels using machine learning approaches, Integrating Mater. Manuf. Innov., 9(2020), No. 3, p. 240. doi: 10.1007/s40192-020-00183-3
[12]	N.H. Paulson, M.W. Priddy, D.L. McDowell, and S.R. Kalidindi, Reduced-order structure-property linkages for polycrystalline microstructures based on 2-point statistics, Acta Mater., 129(2017), p. 428. doi: 10.1016/j.actamat.2017.03.009
[13]	A. Mangal and E.A. Holm, Applied machine learning to predict stress hotspots I: Face centered cubic materials, Int. J. Plast., 111(2018), p. 122. doi: 10.1016/j.ijplas.2018.07.013
[14]	A. Mangal and E.A. Holm, A comparative study of feature selection methods for stress hotspot classification in materials, Integrating Mater. Manuf. Innov., 7(2018), No. 3, p. 87. doi: 10.1007/s40192-018-0109-8
[15]	P. Fernandez-Zelaia, V. Roshan Joseph, S.R. Kalidindi, and S.N. Melkote, Estimating mechanical properties from spherical indentation using Bayesian approaches, Mater. Des., 147(2018), p. 92. doi: 10.1016/j.matdes.2018.03.037
[16]	A. Solomou, G. Zhao, S. Boluki, et al., Multi-objective Bayesian materials discovery: Application on the discovery of precipitation strengthened NiTi shape memory alloys through micromechanical modeling, Mater. Des., 160(2018), p. 810. doi: 10.1016/j.matdes.2018.10.014
[17]	A. Rovinelli, M.D. Sangid, H. Proudhon, et al., Predicting the 3D fatigue crack growth rate of small cracks using multimodal data via Bayesian networks: In-situ experiments and crystal plasticity simulations, J. Mech. Phys. Solids, 115(2018), p. 208. doi: 10.1016/j.jmps.2018.03.007
[18]	G. Tapia, S. Khairallah, M. Matthews, W.E. King, and A. Elwany, Gaussian process-based surrogate modeling framework for process planning in laser powder-bed fusion additive manufacturing of 316L stainless steel, Int. J. Adv. Manuf. Technol., 94(2018), No. 9-12, p. 3591. doi: 10.1007/s00170-017-1045-z
[19]	L.B. Meng and J. Zhang, Process design of laser powder bed fusion of stainless steel using a Gaussian process-based machine learning model, JOM, 72(2020), No. 1, p. 420. doi: 10.1007/s11837-019-03792-2
[20]	B.L. DeCost, B. Lei, T. Francis, and E.A. Holm, High throughput quantitative metallography for complex microstructures using deep learning: A case study in ultrahigh carbon steel, Microsc. Microanal., 25(2019), No. 1, p. 21. doi: 10.1017/S1431927618015635
[21]	D. Morgan and R. Jacobs, Opportunities and challenges for machine learning in materials science, Annu. Rev. Mater. Res., 50(2020), No. 1, p. 71. doi: 10.1146/annurev-matsci-070218-010015
[22]	T. Thankachan, K.S. Prakash, C. David Pleass, et al., Artificial neural network to predict the degraded mechanical properties of metallic materials due to the presence of hydrogen, Int. J. Hydrog. Energy, 42(2017), No. 47, p. 28612. doi: 10.1016/j.ijhydene.2017.09.149
[23]	S. Feng, H.Y. Zhou, and H.B. Dong, Using deep neural network with small dataset to predict material defects, Mater. Des., 162(2019), p. 300. doi: 10.1016/j.matdes.2018.11.060
[24]	F. Brun, T. Yoshida, J.D. Robson, et al., Theoretical design of ferritic creep resistant steels using neural network, kinetic, and thermodynamic models, Mater. Sci. Technol., 15(1999), No. 5, p. 547. doi: 10.1179/026708399101506085
[25]	G.L.W. Hart, T. Mueller, C. Toher, and S. Curtarolo, Machine learning for alloys, Nat. Rev. Mater., 6(2021), No. 8, p. 730. doi: 10.1038/s41578-021-00340-w
[26]	R. Kondo, S. Yamakawa, Y. Masuoka, S. Tajima, and R. Asahi, Microstructure recognition using convolutional neural networks for prediction of ionic conductivity in ceramics, Acta Mater., 141(2017), p. 29. doi: 10.1016/j.actamat.2017.09.004
[27]	A. Cecen, H.J. Dai, Y.C. Yabansu, S.R. Kalidindi, and L. Song, Material structure-property linkages using three-dimensional convolutional neural networks, Acta Mater., 146(2018), p. 76. doi: 10.1016/j.actamat.2017.11.053
[28]	A. Nouira, J. Crivello, and N. Sokolovska, CrystalGAN: learning to discover crystallographic structures with generative adversarial networks, 2018, arXiv:1810.11203. DOI: 10.48550/arXiv.1810.11203
[29]	T. Xie and J.C. Grossman, Hierarchical visualization of materials space with graph convolutional neural networks, J. Chem. Phys., 149(2018), No. 17, art. No. 174111. doi: 10.1063/1.5047803
[30]	C. Chen, W.K. Ye, Y.X. Zuo, C. Zheng, and S.P. Ong, Graph networks as a universal machine learning framework for molecules and crystals, Chem. Mater., 31(2019), No. 9, p. 3564. doi: 10.1021/acs.chemmater.9b01294
[31]	V. Korolev, A. Mitrofanov, A. Korotcov, and V. Tkachenko, Graph convolutional neural networks as “general-purpose” property predictors: The universality and limits of applicability, J. Chem. Inf. Model., 60(2020), No. 1, p. 22. doi: 10.1021/acs.jcim.9b00587
[32]	C.W. Park and C. Wolverton, Developing an improved crystal graph convolutional neural network framework for accelerated materials discovery, Phys. Rev. Mater., 4(2020), No. 6, art. No. 063801. doi: 10.1103/PhysRevMaterials.4.063801
[33]	B. Ma, X. Wei, C. Liu, et al., Data augmentation in microscopic images for material data mining, npj Comput. Mater., 6(2020), art. No. 125. doi: 10.1038/s41524-020-00392-6
[34]	S. Curtarolo, G.L. Hart, M.B. Nardelli, et al., The high-throughput highway to computational materials design, Nat. Mater., 12(2013), No. 3, p. 191. doi: 10.1038/nmat3568
[35]	I. Tanaka, K. Rajan, and C. Wolverton, Data-centric science for materials innovation, MRS Bull., 43(2018), No. 9, p. 659. doi: 10.1557/mrs.2018.205
[36]	R. Arróyave and D.L. McDowell, Systems approaches to materials design: Past, present, and future, Annu. Rev. Mater. Res., 49(2019), p. 103. doi: 10.1146/annurev-matsci-070218-125955
[37]	B. Nenchev, Q. Tao, Z. Dong, C. Panwisawas, H. Li, B. Tao, and H. Dong, Evaluating data-driven algorithms for predicting mechanical properties with small datasets: A case study on gear steel hardenability, Int. J. Miner. Metall. Mater., 29(2022), No. 4, p. 836. doi: 10.1007/s12613-022-2437-0
[38]	S.W. Wu, J. Yang, and G.M. Cao, Prediction of the Charpy V-notch impact energy of low carbon steel using a shallow neural network and deep learning, Int. J. Miner. Metall. Mater., 28(2021), No. 8, p. 1309. doi: 10.1007/s12613-020-2168-z