Mechanical property testing of wide and heavy plates is delayed, and changes in composition and process parameters are eventually manifested as differences in the mechanical properties of different plates, bringing difficulties to the stable control of product quality. To reduce mechanical property fluctuations caused by composition and process parameter changes, a feedforward control strategy for rolling combining deep learning and big data analysis was proposed. The coupling relationship among alloy composition, rolling process parameters, and mechanical properties was analyzed through big data mining. On this basis, a deep learning model was constructed to map composition deviations to required process compensations. With the help of this model, dynamic adjustment of downstream rolling parameters could be realized by using real-time composition monitoring. As shown by a case study on X80 pipeline steel, performance deviations caused by composition fluctuations can be effectively suppressed by the proposed method. At the same time, the quality stability of wide and heavy plates is improved by this method, and technical support is provided for precise control in the wide and heavy plate production process.
4) 基于GWO的影响因子求解.2014年,澳大利亚学者Mirjalili等[23]提出了灰狼优化(grey wolf optimizer,GWO)算法,这是一种群智能优化算法.其灵感来源于灰狼的捕食行为,通过模拟狼群的群体狩猎策略构建而成.灰狼具有群居特性,群体内存在严格的社会等级结构:处于最高层级的是α狼,作为群体的领导者,承担着制定各项重要决策的职责,对应算法中的最优解;β狼处于第二层级,主要协助α狼进行决策与群体管理,代表次优解;δ狼位于第三层级,需服从α狼和β狼的指令,负责放哨、看护等事务;ω狼处于最底层,无决策权力,作用是维持种群内部的关系平衡,对应其余候选解.目前,灰狼算法在调度优化等多个领域得到了广泛应用[24-25].所建立的PSO-DNN模型可构建合金成分-轧制工艺-机械性能的非线性映射关系:首先,输入合金成分,以DNN的预测输出(屈服强度、抗拉强度)为优化目标,通过GWO算法找出宽厚板的机械性能满足生产标准时的最优轧制工艺参数前馈量,优化过程如图7a所示.GWO算法优化500条样本的屈服强度和抗拉强度如图7b,7c所示,可以看出,算法优化后的屈服强度和抗拉强度处于机械性能要求的区间内,并且分布更加集中,保留终轧温度前馈优化量用于后续计算.
WangGuo-dong. Status and prospects of research and development of key common technologies for high-quality heavy and medium plate production[J]. Steel Rolling, 2019, 36(1): 1-8, 30.
GuoQiang. Mechanism and data fusion methods and applications for energy consumption accounting and prediction of plates mills[D]. Beijing: University of Science and Technology Beijing, 2025.
WangGuo-dong, ZhangDian-hua, SunJie. Creating data-driven innovation infrastructure for iron and steel material to accelerate digital transformation of iron and steel industry[J]. Metallurgical Industry Automation, 2023, 47(1): 2-9.
[7]
王国栋. 数字钢铁白皮书[M]. 北京: 冶金工业出版社, 2022.
[8]
WangGuo-dong. White paper on digital steel[M]. Beijing: Metallurgical Industry Press, 2022.
[9]
ZhangH, ZhaoJ H, YongX Y, et al. Genetic algorithm based approaches for medium-thick plate stress analysis feature extraction and product defect prediction[C]//2017 29th Chinese Control and Decision Conference (CCDC). Chongqing, 2017: 6366-6369.
ZhangTian, ZhangZi-hao, TianYong, et al. Research and application of deep learning method for plate ultra fast cooling[J]. Journal of Northeastern University (Natural Science), 2019, 40(5): 635-640.
[12]
GuoQ, ZhouZ M, LiJ, et al. Mechanism-and data-driven algorithms of electrical energy consumption accounting and prediction for medium and heavy plate rolling[J]. Electronic Research Archive, 2025, 33(1): 381-408.
[13]
WangX L, AiY B, ZhangW D, et al. Predicting the medium-temperature thermal stability of impact-strengthened medium-thick plate aluminum alloy using a back propagation artificial neural network[J]. Heliyon, 2023, 9(12): 23018.
[14]
XiuX C, MiaoZ H. Robust sparse canonical correlation analysis: new formulation and application to fault detection[J]. IEEE Sensors Letters, 2022, 6(8): 7002804.
[15]
WangY J, WangJ D, CaoM, et al. Prediction method of wind farm power generation capacity based on feature clustering and correlation analysis[J]. Electric Power Systems Research, 2022, 212: 108634.
[16]
YuanX F, WangY C, WangC, et al. Variable correlation analysis-based convolutional neural network for far topological feature extraction and industrial predictive modeling[J]. IEEE Transactions on Instrumentation and Measurement, 2024, 73: 3001110.
[17]
ShewhartW A. Economic quality control of manufactured product[J]. Bell System Technical Journal, 1930, 9(2): 364-389.
[18]
ShantalM, OthmanZ, BakarA A. A novel approach for data feature weighting using correlation coefficients and min-max normalization[J]. Symmetry, 2023, 15(12): 2185.
[19]
KassamS. The mean-absolute-error criterion for quantization[C]//IEEE International Conference on Acoustics, Speech, and Signal Processing. Hartford, 2003: 632-635.
[20]
RodríguezS A, SalmerónG R, GarcíaC. The coefficient of determination in the ridge regression[J]. Communications in Statistics—Simulation and Computation, 2022, 51(1): 201-219.
[21]
YuanS C. Review of root-mean-square error calculation methods for large deployable mesh reflectors[J]. International Journal of Aerospace Engineering, 2022, 2022(1): 5352146.
[22]
AkherF B, ShuY N, VargaZ, et al. Parametrically managed activation function for fitting a neural network potential with physical behavior enforced by a low-dimensional potential[J]. The Journal of Physical Chemistry A, 2023, 127(24): 5287-5297.
[23]
ShenS L, ZhangN, ZhouA N, et al. Enhancement of neural networks with an alternative activation function tanhLU[J]. Expert Systems with Applications, 2022, 199: 117181.
[24]
KouY W, ChenZ X, GuQ Q. Implicit bias of gradient descent for two-layer ReLU and leaky ReLU networks on nearly-orthogonal data[J]. Advances in Neural Information Processing Systems, 2023, 36: 30167-30221.
[25]
LiY, WangJ S, WangH R, et al. Optimization of tandem cold rolling schedule based on collaborative optimized PSO[C]//International Conference on Information Computing and Applications. Berlin: Springer, 2012: 399-407.
(Li Jia-lin, He Wei-hua, Qu Yong-zhi. Diagnosis of gear early pitting faults using PSO optimized deep neural network[J]. Journal of Northeastern University (Natural Science), 2019, 40(7): 974-979.)
[30]
Mirjalili S, Mirjalili S M, Lewis A. Grey wolf optimizer[J]. Advances in Engineering Software, 2014, 69: 46-61.
[31]
Gao Y F, Li X Y, Zhang C J, et al. An improved gray wolf optimizer for wafer probing scheduling problem[C]//2025 28th International Conference on Computer Supported Cooperative Work in Design (CSCWD). Compiegne, 2025: 819-824.
(Wang Shu-hong, Wei Wei, Han Wen-shuai, et al. Global optimization search method for minimum safety factor of slope based on chaotic grey wolf optimization algorithm[J]. Journal of Northeastern University (Natural Science), 2022, 43(7): 1033-1042.)