基于主成分分析协同随机森林算法的热连轧带钢宽度预测

doi:10.12068/j.issn.1005-3026.2021.09.008

东北大学学报（自然科学版） ›› 2021, Vol. 42 ›› Issue (9): 1268-1275.DOI: 10.12068/j.issn.1005-3026.2021.09.008

基于主成分分析协同随机森林算法的热连轧带钢宽度预测

丁敬国，郭锦华

(东北大学轧制技术及连轧自动化国家重点实验室，辽宁沈阳110819)

修回日期:2020-12-21 接受日期:2020-12-21 发布日期:2021-09-16
通讯作者: 丁敬国
作者简介:丁敬国(1981-)，男，吉林白山人，东北大学副教授.
基金资助:
国家重点研发计划项目(2018YFB1308705，2017YFB0304100).

Prediction of Rough Rolling Width Based on Principal Component Analysis Collaborated with Random Forest Algorithm

DING Jing-guo， GUO Jin-hua

State Key Laboratory of Rolling and Automation， Northeastern University， Shenyang 110819， China.

Revised:2020-12-21 Accepted:2020-12-21 Published:2021-09-16
Contact: DING Jing-guo
About author:-
Supported by:
-

摘要/Abstract

摘要： 为提高热连轧粗轧带钢生产过程中换钢种、换规格及换辊后的首块带钢宽度设定模型精度，本文提出一种基于主成分分析协同随机森林(PCA-RF)算法的宽度预测模型.采用主成分分析法对数据样本合理分析，通过计算特征值、主成分贡献度及累计贡献度进行特征选择.在PCA筛选的变量数据集上训练最佳随机森林宽度预测模型.同时，使用支持向量机回归(SVR)、K-最近邻(KNN)模型进行对比验证.通过实际应用表明，PCA-RF各道次宽度模型R-squared值控制在99.9%~1，且96%以上样本点预测误差在-5~5mm，从而证明该模型实现了换钢种、换规格及换辊后的首块钢宽度的高精度预测.

关键词: 热连轧粗轧;主成分分析;特征选择;宽度预测;随机森林算法

Abstract: To improve the accuracy of the predicted width of the first piece of steel after changing the steel type， the steel specification and the roll in the process of hot continuous rough rolling strip production， a new width prediction model based on the principal component analysis collaborated with random forest (PCA-RF) algorithm is proposed in this work. The PCA method is used to analyze the reasonability of data samples and the feature selection is carried out by calculating the eigenvalue， and principal component and cumulative contribution degrees. The best RF model is trained on variant dataset selected from the PCA. At the same time， support vector machine regression(SVR)and K-nearest neighbor(KNN)models are used for comparison and verification. The practical applications show that the R-squared value from the each pass width predicted by the PCA-RF model is controlled within the range of 0.999~1， and the prediction error of more than 96% samples is -5~5mm， which proves that the model can predict the steel width with a high precision.

Key words: hot continuous rough rolling; principal component analysis(PCA); feature selection; width prediction; random forest(RF) algorithm

中图分类号:

TG335.5

丁敬国，郭锦华. 基于主成分分析协同随机森林算法的热连轧带钢宽度预测[J]. 东北大学学报（自然科学版）, 2021, 42(9): 1268-1275.

DING Jing-guo， GUO Jin-hua. Prediction of Rough Rolling Width Based on Principal Component Analysis Collaborated with Random Forest Algorithm[J]. Journal of Northeastern University(Natural Science), 2021, 42(9): 1268-1275.

参考文献

[1]Ding J G，Li Q Y，Ma G S，et al.Analytical approach for compulsory broadsiding of continuous rough rolling process［J］.Proceedings of the Institution of Mechanical Engineers.Part C:Journal of Mechanical Engineering Science，2018，232(20):3685-3695.
[2]Park C J，Han S H，Lee D M，et al.Direct width control systems based on width prediction models in hot strip mill［J］.ISIJ International，2007，47(1):105-113.
[3]Song X R，Hu Z H，He X Q，et al.The prediction of the hot strip's width based on Bayesian neural network［J］.Advanced Materials Research，2013，823:489-493.
[4]吕程，王国栋，刘相华，等.利用人工神经网络预测热连轧精轧机组带钢宽度变化［J］.上海金属，1998，20(4):36-39.(Lyu Cheng，Wang Guo-dong，Liu Xiang-hua，et al.Predicting the width variation of strip in the finishing stand by artificial neural networks ［J］.Shanghai Metal，1998，20(4):36-39.)
[5]丁敬国，焦景民，昝培，等.基于模糊聚类的PSO-神经网络预测热连轧粗轧宽度［J］.东北大学学报(自然科学版)，2007，28(9):1282-1284.(Ding Jing-guo，Jiao Jing-min，Zan Pei，et al.Hot strip width prediction during rough rolling with PSO neural network based on fuzzy clustering［J］.Journal of Northeastern University(Natural Science)，2007，28(9):1282-1284.)
[6]李文婷.基于改进型粒子群算法的热轧带钢宽度神经网络预报模型的研究［D］.太原:太原理工大学，2011.(Li Wen-ting.Research on neural network prediction model of hot strip width based on improved particle swarm algorithm［D］.Taiyuan:Taiyuan University of Technology，2011.)
[7]Li X，Wang H Y，Ding J G，et al.Analysis and prediction of fishtail during V-H hot rolling process［J］.Journal of Central South University，2015，22(4):1184-1190.
[8]Kim B，Lee K，Jeon J.Short stroke control model for improving width precision at head and tail of slab in hot vertical-horizontal rolling process［J］.International Journal of Precision Engineering and Manufacturing，2020，21(4):699-710.
[9]Lee K，Han J，Park J，et al.Prediction and control of front-end curvature in hot finish rolling process［J］.Advances in Mechanical Engineering，2015，7(11):1-10.
[10]He D，Liu R.Ultra-short-term wind power prediction using ANN ensemble based on PCA［C］ // IEEE 7th International Power Electronics and Motion Control Conference—ECCE Asia.Harbin，2012:2108-2112.
[11]Breiman L.Random forests［J］.Machine Learning，2001，45(1):5-32.

[1]	李强，李瑶坤，夏书月，康雁. 一种改进的医疗文本分类模型:LS-GRU[J]. 东北大学学报:自然科学版, 2020, 41(7): 938-943.
[2]	李艳梅，杨梦奇，张舒展，张田. 高渗透轧制对SA738 Gr.B 厚板变形渗透性的影响[J]. 东北大学学报:自然科学版, 2019, 40(12): 1711-1715.
[3]	韩东序，杜林秀，吴红艳，齐敏. 轧后冷却速率对GCr15轴承钢球化组织的影响[J]. 东北大学学报:自然科学版, 2019, 40(10): 1392-1397.
[4]	彭文，姬亚锋，陈小睿，张殿华. 热轧非稳态过程轧制力自学习模型优化[J]. 东北大学学报:自然科学版, 2019, 40(10): 1408-1412.
[5]	姚春霞，蓝慧芳，杜林秀. 等温淬火工艺对超高强贝氏体钢组织性能影响[J]. 东北大学学报:自然科学版, 2019, 40(7): 948-952.
[6]	矫志杰，闫恩波，李建平. 难变形金属薄带温轧卷取加热温度计算[J]. 东北大学学报:自然科学版, 2019, 40(7): 952-955.
[7]	陈敬琪，刘相华，闫述，于庆波. 铜/钢/铜冷轧复合薄带弯曲性能实验研究[J]. 东北大学学报:自然科学版, 2019, 40(5): 647-652.
[8]	王坤，刘金霖，于庆波，栾伟鹏. 铜锰复合载氧体吸释氧反应性实验研究[J]. 东北大学学报:自然科学版, 2019, 40(1): 48-52.
[9]	李小琳，邓想涛，李艳梅，王昭东. Nb-V低碳微合金钢等温淬火过程析出行为[J]. 东北大学学报:自然科学版, 2018, 39(12): 1708-1712.
[10]	王振华，龚殿尧，李广焘，张殿华. 遗传算法优化神经网络的热轧带钢弯辊力预报模型[J]. 东北大学学报:自然科学版, 2018, 39(12): 1717-1722.
[11]	杨德翰，骆宗安，谢广明，余焕. 表面处理方式对钛/钢复合板界面结合性能的影响[J]. 东北大学学报:自然科学版, 2018, 39(10): 1396-1402.
[12]	贾庆仁，车德福，李佳徐，陈凯. 动态精化的煤层三维建模方法[J]. 东北大学学报:自然科学版, 2018, 39(5): 726-730.
[13]	王青龙，孙杰，王振华，张殿华. UCM轧机板形调控机构对轧制压力分布影响[J]. 东北大学学报:自然科学版, 2018, 39(3): 345-350.
[14]	汤德林，刘相华，宋孟. 铜极薄带轧制过程中尺寸效应的研究[J]. 东北大学学报:自然科学版, 2017, 38(9): 1239-1242.
[15]	彭文，陈庆安，马更生，张殿华. 热连轧宽度自适应模型优化[J]. 东北大学学报:自然科学版, 2017, 38(9): 1243-1246.

基于主成分分析协同随机森林算法的热连轧带钢宽度预测

Prediction of Rough Rolling Width Based on Principal Component Analysis Collaborated with Random Forest Algorithm

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价