基于Bayesian-LSTM神经网络的热轧轧辊剩余寿命预测及不确定性评估

doi:10.3901/JME.2024.11.181

机械工程学报 ›› 2024, Vol. 60 ›› Issue (11): 181-190.doi: 10.3901/JME.2024.11.181

• 特邀专栏：复杂装备智能设计理论与方法 • 上一篇下一篇

扫码分享

基于Bayesian-LSTM神经网络的热轧轧辊剩余寿命预测及不确定性评估

朱挺¹, 陈兆祥¹, 周笛², 陈震¹, 胡兵³, 潘尔顺¹

1. 上海交通大学机械与动力工程学院上海 200240;
2. 东华大学机械工程学院上海 201620;
3. 上海宝信软件股份有限公司上海 201203

收稿日期:2023-08-16 修回日期:2023-12-12 出版日期:2024-06-05 发布日期:2024-08-02
作者简介:朱挺,男,1999年出生。主要研究方向为退化建模、剩余寿命预测、故障诊断。E-mail:accounter@sjtu.edu.cn
潘尔顺(通信作者),男,1972年出生,博士,教授,博士研究生导师。主要研究方向为可靠性与维护策略、质量控制、故障诊断。E-mail:pes@sjtu.edu.cn
基金资助:
国家重点研发计划资助项目(2020YFB1711100)。

Bayesian-LSTM Neural Network-based Remaining Useful Life Prediction and Uncertainty Estimation of Rollers in A Hot Strip Mill

ZHU Ting¹, CHEN Zhaoxiang¹, ZHOU Di², CHEN Zhen¹, HU Bing³, PAN Ershun¹

1. Institute of Vibration, Shock and Noise, Shanghai Jiao Tong University, Shanghai 200240;
2. College of Mechanical Engineering, Donghua University, Shanghai 201620;
3. Shanghai Baosight Software Co., Ltd., Shanghai 201203

Received:2023-08-16 Revised:2023-12-12 Online:2024-06-05 Published:2024-08-02

摘要/Abstract

摘要： 轧辊性能直接影响钢铁轧制流程的生产效率和生产质量，结合轧辊的复杂运行环境和波动工况条件，精准预测轧辊运行状态的时序变化特征与剩余寿命对生产流程精细化、智能化、高效化尤为重要。考虑轧辊服役过程中的动态不确定性，提出一种结合贝叶斯神经网络的长短期记忆网络(Bayesian long short term memory, Bayesian-LSTM)方法，提取健康指标以评估轧辊健康状态，并智能预测轧辊剩余寿命，量化其分布特征的区间不确定性，进一步讨论Bayesian-LSTM网络结构参数对寿命区间的动态影响。通过某热轧厂的实际运行数据验证了方法的有效性，结果表明：所提出方法相对传统数据驱动方法，预测精度达到96.90%，实现了热轧轧辊寿命智能预测和不确定性评估。

关键词: 热轧, 轧辊, 剩余寿命预测, 不确定性评估, Bayesian-LSTM

Abstract: Roller’s performance directly affects the production efficiency and quality of the steel rolling process. Combined with the complex operating environment and fluctuating working conditions, it is especially important to accurately predict the time-series change characteristics of the roller’s operating status and the remaining useful life (RUL) for the refined, intelligent and efficient production process. Considering the dynamic uncertainty in operation, Bayesian neural network-long short term memory network (Bayesian-LSTM) method is proposed to extract the roller health index to evaluate the roller’s health status and intelligently predict the RUL of the roller and quantify interval uncertainty of the distribution feature. The influence of the constructed Bayesian-LSTM network structure parameters on the lifetime interval is further discussed. And the effectiveness of the method is verified by using the actual industrial data from a hot strip mill. Results show that compared with the traditional data-driven method, the proposed method can not only achieve the prediction accuracy of 96.90%, but also evaluate the uncertainty of roller RUL.

Key words: hot strip mill, rollers, remaining useful life prediction, uncertainty estimation, Bayesian-LSTM

中图分类号:

TH17

朱挺, 陈兆祥, 周笛, 陈震, 胡兵, 潘尔顺. 基于Bayesian-LSTM神经网络的热轧轧辊剩余寿命预测及不确定性评估[J]. 机械工程学报, 2024, 60(11): 181-190.

ZHU Ting, CHEN Zhaoxiang, ZHOU Di, CHEN Zhen, HU Bing, PAN Ershun. Bayesian-LSTM Neural Network-based Remaining Useful Life Prediction and Uncertainty Estimation of Rollers in A Hot Strip Mill[J]. Journal of Mechanical Engineering, 2024, 60(11): 181-190.

参考文献

[1] JIAO Ruihua,PENG Kaixiang,DONG Jie.Remaining useful life prediction for a roller in a hot strip mill based on deep recurrent neural networks[J].IEEE/CAA Journal of Automatica Sinica,2021,8(7):1345-1354.
[2] JARDINE A,LIN D,BANJEVIC D.A review on machinery diagnostics and prognostics implementing condition-based maintenance-ScienceDirect[J].Mechanical Systems and Signal Processing,2006,20(7):1483-1510.
[3] LI Zhe,WANG Yi,WANG Kesheng.A deep learning driven method for fault classification and degradation assessment in mechanical equipment[J].Computers in Industry,2019,104(1):1-10.
[4] SOUALHI M,NGUYEN K T P,SOUALHI A,et al.Health monitoring of bearing and gear faults by using a new health indicator extracted from current signals[J].Measurement,2019,141(7):37-51.
[5] AN D,KIM N H,CHOI J H.Practical options for selecting data-driven or physics-based prognostics algorithms with reviews[J].Reliability Engineering&System Safety,2015,133(1):223-236.
[6] 王嘉,张云安,韩旭.基于互依关系的退化与随机冲击建模研究[J].机械工程学报,2021,57(2):230-238.WANG Jia,ZHANG Yunan,HAN Xu.Research on the degradation process and random shocks modeling based on their interdependency[J].Journal of Mechanical Engineering,2021,57(2):230-238.
[7] 张义民,孙志礼.机械产品的可靠性大纲[J].机械工程学报,2014,50(14):14-20.ZHANG Yimin,SUN Zhili.The reliability syllabus of mechanical products[J].Journal of Mechanical Engineering,2014,50(14):14-20.
[8] WANG Jinjiang,GAO R,YUAN Zhuang,et al.A joint particle filter and expectation maximization approach to machine condition prognosis[J].Journal of Intelligent Manufacturing,2019,30(2):605-621.
[9] EL-THALJI I,JANTUNEN E.A summary of fault modeling and predictive health monitoring of rolling element bearings[J].Mechanical Systems and Signal Processing,2015,60-61(8):252-272.
[10] 潘尔顺,陈震.高可靠性产品退化建模研究综述[J].工业工程与管理,2015,20(6):1-6.PAN Ershun,CHEN Zhen.Review of degradation model for high reliability products[J].Industrial Engineering and Management,2015,20(6):1-6.
[11] 裴洪,胡昌华,司小胜,等.基于机器学习的设备剩余寿命预测方法综述[J].机械工程学报,2019,55(8):1-13.PEI Hong,HU Changhua,SI Xiaosheng,et al.Review of machine learning based remaining useful life prediction methods for equipment[J].Journal of Mechanical Engineering,2019,55(8):1-13.
[12] ZHANG Zhengxin,SI Xiaosheng,HU Changhua,et al.Degradation data analysis and remaining useful life estimation:A review on wiener-process-based methods[J].European Journal of Operational Research,2018,271(3):775-796.
[13] 许焕卫,黄鑫,黄洪钟,等.多退化指标条件下卫星动量轮可靠性建模与评估[J].机械工程学报,2022,58(17):67-74.XU Huanwei,HUANG Xin,HUANG Hongzhong,et al.Reliability modeling and evaluation of inertial actuators based on inverse gaussian process of individual differences[J].Journal of Mechanical Engineering,2022,58(17):67-74.
[14] 王新刚,张鑫垚,杨禄杰,等.竞争失效条件下针对磨损退化数据的刀具可靠性分析[J].中国机械工程,2020,31(14):1672-1677,1746.WANG Xingang,ZHANG Xinyao,YANG Lujie,et al.Tool reliability analysis for wear degradation data under competitive failure conditions[J].China Mechanical Engineering,2020,31(14):1672-1677,1746.
[15] SIKORSKA J Z,HODKIEWICZ M,MA L.Prognostic modelling options for remaining useful life estimation by industry[J].Mechanical Systems&Signal Processing,2011,25(5):1803-1836.
[16] 吴斌,苏红伟,潘尔顺,等.基于改进预测算法优化的风机齿轮箱维护策略[J].工业工程与管理,2016,21(1):132-135.WU bin,SU Hongwei,PAN Ershun,et al.Maintenance policy of wind turbine gearbox based on improved prediction algorithm optimization[J].Industrial Engineering and Management,2016,21(1):132-135.
[17] 雷亚国,贾峰,周昕,等.基于深度学习理论的机械装备大数据健康监测方法[J].机械工程学报,2015,51(21):49-56.LEI Yaguo,JIA Feng,ZHOU Xin,et al.A deep learning-based method for machinery health monitoring with big data[J].Journal of Mechanical Engineering,2015,51(21):49-56.
[18] LI Xiang,DING Qian,SUN Jianqiao.Remaining useful life estimation in prognostics using deep convolution neural networks[J].Reliability Engineering&System Safety,2018,172(C):1-11.
[19] DEUTSCH J,D HE.Using deep learning-based approach to predict remaining useful life of rotating components[J].IEEE Transactions on Systems,Man,and Cybernetics:Systems,2017,48(99):11-20.
[20] MALHI A,YAN Ruqiang,GAO R X.Prognosis of defect propagation based on recurrent neural networks[J].IEEE Transactions on Instrumentation and Measurement,2011,60(3):703-711.
[21] KUN ZHANG,NING CHEN,JIAN LIU,et al.A GRU-based ensemble learning method for time-variant uncertain structural response analysis[J].Computer Methods in Applied Mechanics and Engineering,2022,391(3):114516.
[22] LI Gaoyang,YANG Li,LEE C G,et al.A bayesian deep learning rul framework integrating epistemic and aleatoric uncertainties[J].IEEE Transactions on Industrial Electronics,2020,68(6):8829-8841.
[23] 彭艳,李浩然,刘洋,等.基于均布临界域本征损伤耗散的疲劳极限等量关系[J].机械工程学报,2019,55(10):54-61.PENG Yan,LI Haoran,LIU Yang,et al.Fatigue limit equivalent relation based on uniform intrinsic damage dissipation in critical domain[J].Journal of Mechanical Engineering,2019,55(10):54-61.
[24] PENG Weiwen,YE Zhisheng,CHEN Nan.Bayesian deep learning based health prognostics towards prognostics uncertainty[J].IEEE Transactions on Industrial Electronics,2019,67(4):2283-2293.
[25] 赵申坤,姜潮,龙湘云.一种基于数据驱动和贝叶斯理论的机械系统剩余寿命预测方法[J].机械工程学报,2018,54(12):115-124.ZHAO Shenkun,JIANG Chao,LONG Xiangyun.Remaining useful life estimation of mechanical systems based on the data-driven method and bayesian theory[J].Journal of Mechanical Engineering,2018,54(12):115-124.
[26] MAZAEV T,CREVECOEUR G,HOECKE S V.Bayesian convolutional neural networks for rul prognostics of solenoid valves with uncertainty estimations[J].IEEE Transactions on Industrial Informatics,2021,17(12):8418-8428.
[27] ZHOU Haoxuan,HUANG Xin,WEN Guangrui,et al.Construction of health indicators for condition monitoring of rotating machinery:A review of the research[J].Expert Systems with Applications,2022,70(2):1-13.
[28] MA Meng,MAO Zhu.Deep-convolution-based LSTM network for remaining useful life prediction[J].IEEE Transactions on Industrial Informatics,2021,17(3):1658-1667.
[29] ELSHEIKH A,YACOUT S,OUALI M S.Bidirectional handshaking LSTM for remaining useful life prediction[J].Neurocomputing,2019,323(5):148-156.
[30] QIN Yi,CHEN Dingliang,XIANG Sheng,et al.Gated dual attention unit neural networks for remaining useful life prediction of rolling bearings[J].IEEE Transactions on Industrial Informatics,2021,17(9):6438-6447.

基于Bayesian-LSTM神经网络的热轧轧辊剩余寿命预测及不确定性评估

Bayesian-LSTM Neural Network-based Remaining Useful Life Prediction and Uncertainty Estimation of Rollers in A Hot Strip Mill

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]	王宇, 刘秋发, 彭一真. 非均匀监测条件下滚动轴承剩余寿命预测方法[J]. 机械工程学报, 2023, 59(23): 96-104.
[2]	沈天浩, 丁康, 黎杰, 黄如意, 李巍华. 图结构联合时序数据驱动的装备剩余使用寿命预测方法[J]. 机械工程学报, 2023, 59(12): 183-194.
[3]	杨超颖, 刘颉, 周凯波. 基于路图注意力网络的轴承剩余寿命预测方法[J]. 机械工程学报, 2023, 59(12): 195-201.
[4]	孙文权, 何安瑞, 刘超, 徐日石, 袁铁衡. 宽幅薄规格热轧带钢纵向屈曲变形行为解析研究[J]. 机械工程学报, 2022, 58(14): 203-211.
[5]	宋守许, 郁炯, 蔚辰, 张志旭. 考虑多辊修磨关系的轧辊主动再制造设计方法[J]. 机械工程学报, 2022, 58(13): 203-212.
[6]	付洋, 曹宏瑞, 郜伟强, 高文辉. 数字孪生驱动的航空发动机涡轮盘剩余寿命预测[J]. 机械工程学报, 2021, 57(22): 106-113.
[7]	王久健, 杨绍普, 刘永强, 文桂林. 一种基于空间卷积长短时记忆神经网络的轴承剩余寿命预测方法[J]. 机械工程学报, 2021, 57(21): 88-95.
[8]	李乃鹏, 蔡潇, 雷亚国, 徐鹏程, 王文廷, 王彪. 一种融合多传感器数据的数模联动机械剩余寿命预测方法[J]. 机械工程学报, 2021, 57(20): 29-37,46.
[9]	王泽洲, 陈云翔, 蔡忠义, 项华春, 王莉莉. 基于复合非齐次泊松过程的不完美维修设备剩余寿命预测[J]. 机械工程学报, 2020, 56(22): 14-23.
[10]	裴洪, 胡昌华, 司小胜, 张建勋, 庞哲楠, 张鹏. 基于机器学习的设备剩余寿命预测方法综述[J]. 机械工程学报, 2019, 55(8): 1-13.
[11]	黄华贵, 赵阳, 王超, 燕猛. 界面涂层对厚规格热轧钢/铝复合板界面结构与力学性能的影响[J]. 机械工程学报, 2019, 55(14): 30-36.
[12]	韩飞, 牛丽丽, 王允, 杨亚辉. 超高强钢辊弯成形回弹机理分析及控制研究[J]. 机械工程学报, 2018, 54(2): 131-137.
[13]	邓超, 陶志奎, 吴军, 钱有胜, 夏爽. 基于性能退化的数控机床剩余寿命预测[J]. 机械工程学报, 2018, 54(17): 181-189.
[14]	赵申坤, 姜潮, 龙湘云. 一种基于数据驱动和贝叶斯理论的机械系统剩余寿命预测方法[J]. 机械工程学报, 2018, 54(12): 115-124.
[15]	靳淇超, 汪文虎, 蒋睿嵩, 赵德中, 崔康. 压气机叶片辊轧模具型腔回弹补偿方法研究[J]. 机械工程学报, 2017, 53(16): 148-155.