Front-end bending is one of the most common problems in hot rough rolling. Therefore, effective control of front-end bending is a key aspect to ensure the quality of hot-rolled slab. A CNN-LSTM-based model is proposed for the high-precision prediction of front-end bending of hot-rolled slabs. For better data collection and model application, this research introduces an industrial internet of things (IIoT) platform. Our model is built for each pass on the collected production data of 1580 hot rolling mill production lines. To obtain the best model, random search and Bayesian optimisation were used to optimise the hyperparameters. A comparison of the CNN-LSTM model with the XGBoost and LSTM models is made and evaluated. The results show that the CNN-LSTM model is better than other models, which can be well applied to hot rolling production.
DenisATobiasMJensA,et al.A dimensional analysis of front-end bending in plate rolling applications. Mater Process Technol2012; 212: 1387.
2.
HwangYTzouG. An analytical approach to asymmetrical cold strip rolling using the slab method. J Mater Eng Perform1993; 2: 597–606.
3.
HeCChenYChenF,et al.Research on finite element simulation and control method of plate head bending. IOP Conf Series Ear Envir Sci2019; 332: 042030.
4.
RudskoiAKodzhaspirovGKamelinE. Modeling of the recrystallization processes during asymmetric rolling of a high-strength low-alloy steel. Russ Metall2021; 2021: 486–491.
5.
SunXLiuXWangJ,et al.Analysis of asymmetrical rolling of strip considering two deformation region types. Int J Adv Des Manuf Technol2020; 110: 2767–2785.
6.
WangJLiuXSunX. Study on the relationship between asymmetrical rolling deformation zone configuration and rolling parameters. Int J Mech Sci2020; 187: 1.
7.
XuXDengYGuoX,et al.Influence of asymmetric rolling process and thickness reduction on the microstructure and mechanical properties of the Al–Mg-Si alloy. Met Mater Int2022; 28: 1620–1629.
8.
HiroyukiSMasaoMIsaoI. Rolling Process US04365496A. 1982
9.
GreenePJeswietJ. Experimental measurement of curl in rolling. J Mater Process Technol1998; 84: 202.
10.
HartmutHP. Comparison of methods for calculating the influence of asymmetry in strip and plate rolling. Steel Res2000; 71: 490.
11.
PhilippMSchwenzfeierWFischerF, et al.Front end bending in plate rolling influenced by circumferential speed mismatch and geometry. J Mater Process Technol2007; 184: 224–232.
12.
DyjaHKorczakPPilarczykJ,et al.Theoretical and experimental analysis of plates asymmetric rolling. J Mater Process Technol1994; 45: 167–172.
13.
LuJHarrerOSchwenzfeierW,et al.Analysis of the bending of the rolling material in asymmetrical sheet rolling. Int J Mech Sci2000; 42: 49–61.
14.
KnightCHardySLeesA,et al.Investigation into the influence of asymmetric factors and rolling parameters on strip curvature during hot rolling. J Mater Process Technol2003; 2: 134.
15.
PanGWangFShangC, et al.Advances in machine learning- and artificial intelligence-assisted material design of steels. Int J Miner Metall Mater2023; 30: 1003–1024.
16.
CarreónCTolamaJValdezG. Multi-objective optimization of the hot rolling scheduling of steel using a genetic algorithm. MRS Adv2019; 61: 3373.
17.
WangZGongDLiG,et al.Bending force prediction model in hot strip rolling based on artificial neural network optimize by genetic algorithm. Dongbei Daxue Xuebao, Ziran Kexueban2018; 39: 1717.
18.
SongLXuDWangX, et al.Application of machine learning to predict and diagnose for hot-rolled strip crown. Int J Adv Manuf Technol2022; 120: 881–890.
19.
XuDYangQWangX, et al.Application of takagi–sugeno fuzzy model for slab camber control in a hot strip rougher mill. Ironmaking Steelmaking2020; 47: 623–631.
20.
DengJSunJPengW, et al.Application of neural networks for predicting hot-rolled strip crown. Appl Soft Comput2019; 78: 119–131.
21.
AlaeiHSalimiMNouraniA. Online prediction of work roll thermal expansion in a hot rolling process by a neural network. Int J Adv Manuf Technol2016; 85: 1769–1777.
22.
LuXSunJSongZ, et al.Prediction and analysis of cold rolling mill vibration based on a data-driven method. Appl Soft Comput2020; 96: 106706.
23.
WangQSongLHeH, et al.Edge drop prediction of hot-rolled silicon steel based on a mechanism and data model. Ironmaking Steelmaking2024; 51: 195–202.
24.
YanSXuDHeW, et al.Online measurement system of slab front-end bending in hot rough rolling based on line structured light vision. Meas Sci Technol2023; 34: 095201.
25.
SongLXuDLiuP, et al.A novel mechanism fusion data control method for slab camber in hot rolling. J Iron Steel Res Int2023; 30: 960–970.
26.
YanSWangXYangQ, et al. Online deviation measurement system of the strip in the finishing process based on machine vision. Measurement (Mahwah N J)2022; 202: 111735.
HuYChengXLiH,et al.Influence of velocity difference on bending of workpiece front end. J Iron Steel Res Int2004; 16: 37.
29.
BouguettayaAZarzourH. CNN-based hot-rolled steel strip surface defects classification: a comparative study between different pre-trained CNN models. Int J Adv Manuf Technol2024; 132: 399–419.
30.
ElghaziKRamchounHMasrourT. Enhancing CNN structure and learning through NSGA-II-based multi-objective optimization. Evol Syst2024; 15: 1503–1519.
31.
SalemF. Gated RNN: the minimal gated unit (MGU) RNN. Recurrent NN2022: 101–113.
32.
KimKJeongCKimJ, et al.Data-Driven LSTM model and predictive control for vehicle lateral motion. J Electr Eng Technol2024; 19: 3635–3644.
