锻压技术

基于深度神经网络的斜轧穿孔机调整参数预测

英文标题：Prediction on adjusting parameters for skew rolling puncher based on deep neural network

单位：1.太原科技大学电子信息工程学院 2.太原科技大学 3.太原重工股份有限公司

分类号：TG355

出版年,卷(期):页码：2023,48(11):73-78

摘要：

针对无缝钢管二辊斜轧穿孔生产工艺中轧机调整参数对钢管质量影响较大，且传统机理公式计算的设定值精度不高的问题，提出了基于深度神经网络的斜轧穿孔机调整参数预测模型。首先，综合分析了传统的调整参数的数学模型，并在此基础上确定了调整参数的主要影响因素。依据现场收集的数据集，训练了二辊斜轧穿孔时轧机参数的深度神经网络预测模型。在训练深度神经网络时，运用小批量梯度下降法和Adam算法相结合的方法进行了梯度估计修正，优化了训练速度。仿真结果表明：经深度神经网络模型预测的轧机调整参数与实测数据比较，预测模型的R-squared值控制在0.98左右，调整参数的相对误差均可控制在5%以内。相比于传统数学模型，该预测模型具有更高的预测精度，能够实现轧机调整参数高精度预测并用于指导生产。

Aiming at the problems that the adjustment parameters of rolling mill has a great influence on the quality of steel pipe in the two-roll skew rolling and punching production process of seamless steel pipes, and the accuracy of the set value calculated by the traditional mechanism formula is not high, a prediction model for adjustment parameters of skew rolling puncher based on deep neural network was proposed. Firstly, the traditional mathematical model of adjustment parameters was analyzed comprehensively, and the main influencing factors were determined on this basis. Then, based on the data set collected in the field, the deep neural network prediction model of rolling mill parameters during the two-roll skew rolling and punching was trained, and in the deep neural network training, the gradient estimation correction was conducted by using the combination of mini-batch gradient descent method and Adam algorithm to optimize the training speed. The simulation results show that the adjustment parameters of rolling mill predicted by the deep neural network model are compared with the measured data, the R-squared value of the prediction model is controlled at about 0.98, and the relative error of the adjustment parameters can be controlled within 5%. Compared with the traditional mathematical model, this prediction model has higher prediction accuracy, and can realize the high-precision prediction of rolling mill adjustment parameters and be used to guide production.

基金项目：

山西省科技重大专项(20191102009)

作者简介：王清华（1980-），女，博士，讲师，E-mail：2002043@tyust.edu.cn；通信作者：孙继芸（1996-），女，硕士研究生，E-mail：s18835387178@163.com

参考文献：

[1]李连诗，韩观昌.小型无缝钢管生产（上册）[M].北京：冶金工业出版社，1989.

Li L S，Han G C. Production of Small Seamless Steel Pipe(Volume 1)[M].Beijing: Metallurgical Industry Press,1989.

[2]Lee D, Lee Y. Application of neural-network for improving accuracy of roll force model in hot-rolling mill[J]. IFAC Proceedings Volumes, 2000, 33(22):227-231.

[3]刘欣玉,潘露,帅美荣.基于Matlab的BP神经网络轧制力预报模型及应用[J].重庆科技学院学报：自然科学版,2016,18(6):96-98,103.

Liu X Y, Pan L, Shuai M R. Prediction model and its application of BP neural network rolling force based on MATLAB [J]. Journal of Chongqing University of Science and Technology: Natural Science Edition,2016,18(6):96-98,103.

[4]杨景明,闫晓莹,顾佳琪,等.基于改进粒子群优化RBF神经网络的轧制力预报[J].矿冶工程,2014,34(6):110-113，118.

Yang J M, Yan X Y, Gu J Q, et al. Rolling force prediction based on improved particle swarm optimization-RBF neural network [J]. Mining and Metallurgy Engineering,2014,34(6): 110-113,118.

[5]陈鑫, 朱明杰, 吴敏,等. 结合机理计算与神经网络预测的无缝钢管轧制力建模[J]. 冶金自动化, 2015, 39(4):32-37.

Chen X, Zhu M J, Wu M, et al. Rolling force modeling for seamless steel pipe combining mechanism model and neural network prediction [J]. Metallurgical Industry Automation, 2015, 39(4):32-37.

[6]王清华，加世滢，胡建华，等.基于GRA的PSO-BP神经网络斜轧穿孔管形预测[J].锻压技术，2022，47（8）：88-94.

Wang Q H, Jia S Y, Hu J H. Prediction of pipe shape in cross-rolled piercing by PSO-BP neural network based on GRA [J]. Forging & Stamping Technology, 2022,47 (8)：88-94.

[7]何垚东, 李旭, 丁敬国,等. 融合轧制机理和深度学习的带钢精轧宽度预测[J]. 轧钢, 2022,(2):76-81,86.

He Y D, Li X, Ding J G, et al. Hot finishing rolling strip width predicting model based on rolling mechanism and deep learning [J]. Steel Rolling, 2022,(2):76-81,86.

[8]马威, 李维刚, 赵云涛,等. 基于深度学习的热连轧轧制力预测[J]. 钢铁研究学报, 2019, 31(9):805-815.

Ma W, Li W G, Zhao Y T, et al. Prediction of hot-rolled roll force based on deep learning [J]. Journal of Iron and Steel Research, 2019, 31(9):805-815.

[9]孙士秀.鞍钢140自动轧管机组穿孔机调整参数的确定[J].鞍钢技术,1982,(4):180-184.

Sun S X. Determination of adjustment parameters of puncher in Angang 140 automatic tube rolling unit [J]. Angang Technology,1982,(4):180-184.

[10]李连诗. 钢管塑性变形原理(上册)[M]. 北京：冶金工业出版社, 1985.

Li L S. Principle of Plastic Deformation of Steel Pipe (Volume 1)[M]. Beijing: Metallurgical Industry Press, 1985.

[11]马湧, 王晓鹏, 马莎莎. 基于Keras深度学习框架下BP神经网络的热轧带钢力学性能预测[J]. 冶金自动化, 2019,43(2):6-10.

Ma Y, Wang X P, Ma S S. Prediction of mechanical properties of hot rolled strip steel based on BP neural network under Keras deep learning framework [J]. Metallurgical Industry Automation, 2019,43(2):6-10.

[12]魏立新, 魏新宇, 孙浩,等. 基于深度网络训练的铝热轧轧制力预报[J]. 中国有色金属学报, 2018, 28(10):128-134.

Wei L X, Wei X Y, Sun H, et al. Prediction of aluminum hot rolling force based on deep network [J]. The Chinese Journal of Nonferrous Metals, 2018, 28(10):128-134.

[13]姬壮伟. 基于Pytorch 的神经网络优化算法研究 [J]. 山西大同大学学报: 自然科版,2020,36(6):51-53,58.

Ji Z W. Research on neural network optimization algorithm based on Pytorch [J]. Journal of Shanxi Datong University: Natural Science Edition, 2019,36(6):51-53,58.

[14]胡石雄. 基于深度学习的热轧带钢力学性能预报[D]. 武汉：武汉科技大学, 2019.

Hu S X.Mechanical Property Prediction of Hot Rolled Strip Based on Deep Learning[D].Wuhan：Wuhan University of Science and Technology, 2019.

服务与反馈：

【本网站尚未开通全文下载服务】【加入收藏】