板带力学性能预报的集成学习模型及其可靠性评价

doi:10.3901/JME.2021.02.239

机械工程学报 ›› 2021, Vol. 57 ›› Issue (2): 239-246.doi: 10.3901/JME.2021.02.239

扫码分享

板带力学性能预报的集成学习模型及其可靠性评价

李飞飞^1,2, 宋勇^1,2, 刘超^1,2, 李博^1,2, 张世伟^1,2

1. 北京科技大学高效轧制国家工程研究中心北京 100083;
2. 北京科技大学国家板带生产先进装备工程技术研究中心北京 100083

收稿日期:2020-05-01 修回日期:2020-10-30 出版日期:2021-01-20 发布日期:2021-03-15
通讯作者: 宋勇(通信作者),男,副研究员,硕士研究生导师。主要研究方向为轧制过程控制模型及智能化。E-mail:songyong@ustb.edu.cn
作者简介:李飞飞,男,1993年出生,博士研究生。主要研究方向为复杂工业过程建模与优化控制。E-mail:15311449327@163.com
基金资助:
国家自然科学基金（51674028）和广西创新驱动发展专项资金计划（GKAA17202008）资助项目。

Ensemble Learning Model for Mechanical Performance Prediction of Strip and Its Reliability Evaluation

LI Feifei^1,2, SONG Yong^1,2, LIU Chao^1,2, LI Bo^1,2, ZHANG Shiwei^1,2

1. National Engineering Research Center for Advanced Rolling Technology, University of Science and Technology Beijing, Beijing 100083;
2. National Engineering Technology Research Center of Flat Rolling Equipment, University of Science and Technology Beijing, Beijing 100083

Received:2020-05-01 Revised:2020-10-30 Online:2021-01-20 Published:2021-03-15

摘要/Abstract

摘要： 热轧板带力学性能预报具有十分广阔的应用前景。但目前的力学性能预报模型精度不高、泛化能力较差，且无法评价预测结果的准确性，限制了其在实际应用中的效果。为了提高力学性能预报模型的精度和实现预测结果的可靠性评价，采用BP神经网络将目前直接预测力学性能结果的建模方法转换为对样本间的力学性能偏差分布预测的建模，并结合模型预测结果分布的离散程度设计可靠性评价指标，同时采用集成学习提高模型的泛化能力。通过试验验证，该集成学习模型具有较高的预测精度，进一步，根据可靠性评价指标分析，除了其中占比约3.5%样本的预测结果具有较大不确定性，剩余样本的屈服强度和抗拉强度在误差±30 MPa的预测准确率达到了98.45%和98.97%，延伸率在误差±5%下的预测准确率达到了99.48%，有效地提高了模型预测准确率，在生产现场应用中具有一定的指导意义。

关键词: 偏差预测, 力学性能预测, 神经网络, 集成学习, 可靠性评价

Abstract: The prediction of mechanical properties of hot rolled strips has a very broad application prospect. However, the current mechanical performance prediction model is not accurate and the generalization ability is poor, and the accuracy of the prediction results cannot be evaluated, which limits its effectiveness in practical applications. In order to improve the accuracy of the mechanical performance prediction model and the reliability evaluation of the prediction results, BP neural network is used to transform the current modeling method of direct prediction of mechanical performance results into the modeling of mechanical property deviation distribution prediction between samples, and combined with the model. The degree of dispersion of the prediction results is designed to evaluate the reliability evaluation index, and ensemble learning is used to improve the generalization ability of the model. Through experimental verification, the ensemble learning model has higher prediction accuracy. Further, according to the reliability evaluation index analysis, except for the prediction results of about 3.5% of the samples, there is greater uncertainty, and the yield strength and tensile strength of the remaining samples. The prediction accuracy of the intensity ±30 MPa is 98.45% and 98.97%, and the prediction accuracy of the elongation under the error of ±5% is 99.48%, which effectively improves the accuracy of model prediction and has certain performance in production field applications. Guiding significance.

Key words: deviation prediction, mechanical performance prediction, neural network, ensemble learning, reliability evaluation

中图分类号:

TG156

李飞飞, 宋勇, 刘超, 李博, 张世伟. 板带力学性能预报的集成学习模型及其可靠性评价[J]. 机械工程学报, 2021, 57(2): 239-246.

LI Feifei, SONG Yong, LIU Chao, LI Bo, ZHANG Shiwei. Ensemble Learning Model for Mechanical Performance Prediction of Strip and Its Reliability Evaluation[J]. Journal of Mechanical Engineering, 2021, 57(2): 239-246.

参考文献

[1] 干勇. 薄板坯连铸连轧(TSCR)热轧过程组织性能预报技术的开发[J]. 钢铁,2003,38(8):10-15. GAN Yong. The R&D of process modeling in thin slab hot strip rolling[J]. Iron & Steel,2003,38(8):10-15.
[2] 余万华,韩静涛,佘广夫,等. HSMM软件在攀枝花钢铁公司热轧板厂的应用[J]. 钢铁研究学报,2006(11):60-62. YU Wanhua,HAN Jingtao,SHE Guangfu,et al. Application of HSMM software in hot strip mill of panzhihua steel company[J]. Journal of Iron and Steel Research,2006(11):60-62.
[3] GHRIBERNIG J,GHUBMER D. For the first time ever:Full metallurgical control of the mechanical properties of hot-rolled strip with VAI-Q strip[J]. Iron & Steel,2001(01):42-46.
[4] 王蕾. 热轧带钢的相变和力学性能模型研究及应用[D]. 北京:北京科技大学,2017. WANG Lei. Research and application of phase transformation and mechanical properties models in hot strip rolling[D]. Beijing:University of Science and Technology Beijing,2017.
[5] 郑坚,唐广波,程杰锋,等. 珠钢CSP热轧组织性能预报软件设计[J]. 轧钢,2003(1):15-17. ZHENG Jian,TANG Guangbo,CHENG Jiefeng,et al. Software design of microstructure and property prediction system for CSP line of Zhujiang Steel Co[J]. Steel Rolling,2003(1):15-17.
[6] 刘正东,董瀚,干勇. 热连轧过程中组织性能预报系统的应用[J]. 钢铁,2003(2):68-71. LIU Zhengdong,DONG Han,GAN Yong. Application of process modeling to hot strip rolling[J]. Iron & Steel,2003(2):68-71..
[7] 郭朝晖,张群亮,苏异才,等. 关于热轧带钢力学性能预报技术的思考[J]. 冶金自动化,2009,33(2):1-6. GUO Zhaohui,ZHANG Qunliang,SU Yicai,et al. Thoughts on mechanical property prediction of hot rolled strip[J]. Metallurgical Industry Automation,2009,33(2):1-6.
[8] 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 Yong,WANG Xiaopeng,MA Shasha. Prediction of mechanical properties of hot rolled strip based on BP neural network under Keras deep learning framework[J]. Metallurgical Industry Automation,2019,43(2):6-10.
[9] 贾涛,刘振宇,胡恒法,等. 基于贝叶斯神经网络的SPA-H热轧板力学性能预测[J]. 东北大学学报,2008(4):521-524. JIA Tao,LIU Zhenyu,HU Hengfa,et al. Mechanical property prediction for hot rolled SPA-H steel using bayesian neural network[J]. Journal of Northeastern University,2008(4):521-524.
[10] 李维刚,杨威,赵云涛,等. 融合大数据与冶金机理的热轧带钢力学性能预报模型[J]. 钢铁研究学报,2018,30(4):302-308. LI Weigang,YANG Wei,ZHAO Yuntao,et al. Mechanical property prediction model of hot-rolled strip via big data and metallurgical mechanism analysis[J]. Journal of Iron and Steel Research,2018,30(4):302-308.
[11] 李维刚,刘超,杨威,等. 热轧微合金钢力学性能预报模型研究[C]//第十一届中国钢铁年会论文集——S18:冶金自动化与智能管控. 北京:中国金属学会,2017:1-8. LI Weigang,LIU Chao,YANG Wei,et al. Prediction model of mechanical properties of hot-rolled micro-alloyed steel[C]//Proceedings of the 11th CSM Steel Congress:Metallurgical Automation and Intelligent Control. Beijing:The Chinese Society for Metals,2017:1-8.
[12] 郭朝晖,苏异才,张群亮,等. 热轧带钢性能预报技术研究中的几个误区[J]. 轧钢,2013,30(1):29-32. GUO Zhaohui,SU Yicai,ZHANG Qunliang,et al. Wrong cognitions on research of hot-rolled strip property prediction[J]. Steel Rolling,2013,30(1):29-32.
[13] BREIMAN L. Bagging predictors[J]. Machine Learning,1996,24(2):123-140.
[14] 闫飞宇,李伟卓,杨卫卫,等. 基于Bagging神经网络集成的燃料电池性能预测方法[J]. 中国科学:技术科学,2019,49(4):391-401. YAN Feiyu,LI Weizhuo,YANG Weiwei,et al. Prediction of fuel cell performance based on Bagging neural network ensemble model[J]. Scientia Sinica Technologica,2019,49(4):391-401.
[15] 谢琪,程耕国,徐旭. 基于神经网络集成学习股票预测模型的研究[J]. 计算机工程与应用,2019,55(8):238-243. XIE Qi,CHENG Gengguo,XU Xu. Research based on stock predicting model of neural networks ensemble learning[J]. Computer Engineering and Applications,2019,55(8):238-243.
[16] 王征宇. 神经网络集成分类方法及其在并行计算环境中的应用研究[D]. 广州:华南理工大学,2015. WANG Zhengyu. Research on neural network ensemble classification methods and their applications in parallel computing environment[D]. Guangzhou:South China University of Technology,2015.
[17] 于彬,李珊,陈成,等. 基于集成学习的人类LncRNA大数据基因预测[J]. 青岛科技大学学报,2018,39(1):106-113. YU Bin,LI Shan,CHEN Cheng,et al. Prediction of human lncRNA big data genes based on ensemble learning[J]. Journal of Qingdao University of Science and Technology,2018,39(1):106-113.
[18] 张祝亭. 基于数据挖掘的马钢CSP热轧板卷的性能预测[D]. 马鞍山:安徽工业大学,2012. ZHANG Zhuting. The prediction of masteel CSP hot rolled strip based on data mining[D]. Maanshan:Anhui University of Technology,2012.
[19] 刘贵立,张国英,曾梅光. 用人工神经网络模型研究微量元素对钢力学性能的影响[J]. 钢铁研究,2000(1):48-50. LIU Guili,ZHANG Guoying,ZENG Meiguang. Studies on effects of trace elements on mechanical properties of steel by artificial nerve network model[J]. Research on Iron & Steel,2000(1):48-50.
[20] LIU Feitong,TING Kaiming,ZHOU Zhihua. Isolation forest[C]//2008 Eighth IEEE International Conference on Data Mining,December 15-19,2008. Pisa:IEEE,2009:413-422.
[21] LIU Feitong,TING Kaiming,ZHOU Zhihua. Isolation-based anomaly detection[J]. ACM Transactions on Knowledge Discovery From Data,2012,6(1):1-39.
[22] 赵帅,黄亦翔,王浩任,等. 基于随机森林与主成分分析的刀具磨损评估[J]. 机械工程学报,2017,53(21):181-189. ZHAO Shuai,HUANG Yixiang,WANG Haoren,et al. Random forest and principle components analysis based on health assessment methodology for tool wear[J]. Journal of Mechanical Engineering,2017,53(21):181-189.
[23] HORNIK KM,STINCHCOMB M,WHITE H. Multilayer feedforward networks are universal approximators[J]. Neural Networks,1989,2(5):359-366.
[24] 从威. 非均衡分类的集成学习应用研究[D]. 南京:南京信息工程大学,2017. CONG Wei. Application research on ensemble learning of unbalanced classification[D]. Nanjing:Nanjing University of Information Science & Technology,2017.
[25] KHOSRAVI A,NAHAVANDI S,CREIGHTON D. Construction of optimal prediction intervals for load forecasting problems[J]. IEEE Transactions on Power Systems,2010,25(3):1496-1503.

板带力学性能预报的集成学习模型及其可靠性评价

Ensemble Learning Model for Mechanical Performance Prediction of Strip and Its Reliability Evaluation

PDF

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

[1]	郑杰, 权伟, 刘洋, 卢学民, 刘晓红, 岳雅楠, 周腾, 蔡振兵. Cr/Zr涂层环向压缩表面裂纹识别研究[J]. 机械工程学报, 2024, 60(8): 1-10.
[2]	刘占广, 张云, 刘晴雨. 基于差分融合长短期记忆神经网络数控机床热误差建模[J]. 机械工程学报, 2024, 60(7): 249-257.
[3]	狄子钧, 袁东风, 李东阳, 梁道君, 周晓天, 信苗苗, 曹凤, 雷腾飞. 基于多尺度-高效通道注意力网络的刀具故障诊断方法[J]. 机械工程学报, 2024, 60(6): 82-90.
[4]	王俊玖, 刘金玉, 侯秀娟, 齐振国, 李志敏, 刘涛. 列车车体柔性装配精度预测的力学-数据混合建模方法[J]. 机械工程学报, 2024, 60(6): 177-186.
[5]	李春凯, 王嘉昕, 石玗, 代悦. GTAW熔池激光条纹动态行为与熔透预测方法研究[J]. 机械工程学报, 2024, 60(6): 236-244.
[6]	周宁, 姚建勇, 邓文翔. 基于神经网络的比例伺服阀阀芯液动力补偿鲁棒智能控制[J]. 机械工程学报, 2024, 60(4): 126-133.
[7]	王丹丹, 黄伟迪, 张军辉, 赵守军, 于斌, 刘施镐, 吕飞, 苏琦, 徐兵. 基于边缘计算的轴向柱塞泵磨损状态辨识方法研究[J]. 机械工程学报, 2024, 60(4): 189-199.
[8]	黄浩, 单忠德, 张丽娇, 孙正, 郭子桐, 刘检华, 金鹏. 异形截面复合材料构件成形及力学性能预测方法研究[J]. 机械工程学报, 2024, 60(2): 107-118.
[9]	张昊, 魏超, 胡纪滨, 陈泳丹. 基于转向模式切换的三轴独立转向车辆路径跟踪控制研究[J]. 机械工程学报, 2024, 60(2): 243-251.
[10]	段书用, 杨建华, 韩旭, 刘桂荣. 高维数据自适应降维方法[J]. 机械工程学报, 2024, 60(17): 283-296.
[11]	刘建琴, 胡桂菘, 乔金丽, 郭晓. 基于双路神经网络的TBM实时性能预测研究[J]. 机械工程学报, 2024, 60(16): 43-53.
[12]	胡伟飞, 廖家乐, 郭云飞, 鄢继铨, 李光, 岳海峰, 谭建荣. 基于物理信息神经网络的时变可靠性分析方法[J]. 机械工程学报, 2024, 60(13): 141-153.
[13]	张德权, 李星奥, 张宁, 宁国松, 韩旭. 工业机器人全域误差场精细化建模方法及其误差补偿策略[J]. 机械工程学报, 2024, 60(13): 316-329.
[14]	王俊, 杨轶青, 刘金朝, 沈长青, 黄伟国, 朱忠奎. 小波核编码的脉冲卷积神经网络在可解释性智能诊断中的应用研究[J]. 机械工程学报, 2024, 60(12): 41-50.
[15]	袁静, 任港星, 蒋会明, 赵倩, 魏臣隽, 朱骏. 基于多元提升核神经网络的机械故障诊断方法及其特征提取可解释性研究[J]. 机械工程学报, 2024, 60(12): 51-64.