正在加载图片...
第11期 梁辉等:基于减法聚类的带钢厚度数据驱动建模 ·1345· 充分利用了生产过程中的大量历史数据,又可以根 neural network for prediction for rolling force in hot-rolling mill. 据工况变化在线调整模型的结构和参数. Mater Process Technol,2005,164/165:1612 (2)通过数字轧机仿真实验中几种模型的性能 [10]Jeong S Y,Lee M,Lee S Y,et al.Improving lookup table con- trol of a hot coil strip process with online retrainable RBF net- 对比可以看出,在线多支持向量机厚度模型比最优 work.IEEE Trans Ind Electron,2000,47(3):679 的偏最小二乘统计模型的RMSE提高45.3%, 01] Portmann N F,Lindhoff D.Sorgel G,et al.Application of neu- MAPE提高43.2%,TP提高1.3%.工业现场数据 ral networks in rolling mill automation.Iron Steel Eng,1995,72 仿真分析表明了本文算法具有较高的预测精度,能 (2):33 满足工业在线应用要求. [12]Ma J S,Theiler J,Perkins S.Accurate on-ine support vector re- gression.Neural Comput,2003,15(11)2683 (3)通过基于数据驱动的热轧带钢厚度预测模 13] Mukherjee S.Osuna E,Girosi F.Nonlinear prediction of chaotic 型探索性研究,在忽略精轧机组间的相互影响以及 time series using a support vector machine//Proceedings of1997 厚度与板形、宽带的耦合影响因素的前提下,建立了 IEEE Workshop.New York,1997:511 单机架厚度相关因素模型. ū4] Suykens J A K.Non-inear modelling and support vector ma- chines /Proceedings of IEEE Instrumentation and Measurement Technology Conference.Budapest,2001:287 参考文献 [15]Li W,Yang Y P,Wang N.Multi-model LSSVM regression mod- eling based on kernel fuzzy clustering.Control Decis,2008,23 [1]Hou Z S,Xu J X.On data-driven control theory:the state of the (5):560 art and perspective.Acta Autom Sin,2009,35(6):650 (李卫,杨煜普,王娜.基于核模糊聚类的多模型LSSVM回 (侯忠生,许建新.数据驱动控制理论及方法的回顾和展望 归建模.控制与决策,2008,23(5):560) 自动化学报,2009,35(6):650) [16]Yu Z H,Fu X,Li Y L.Online support vector regression for sys- Gui W H,Yang C H,Li Y G,et al.Data-driven operational-pat- tem identification.Lect Notes Comput Sci,2005,3611:627 tem optimization for copper flash smelting process.Acta Auom 17] Nemmour H,Chibani Y.Multiple support vector machines for Sin,2009,35(6):717 land cover change detection:an application for mapping urban (桂卫华,阳春华,李勇刚,等。基于数据驱动的铜闪速熔炼 extensions.ISPRS J Photogramm Remote Sens,2006,61(2): 过程操作模式优化及应用.自动化学报,2009,35(6):717) 125 B]Gao C H,Jian L,Chen J M,et al.Data-driven modeling and pre- 08] Chiu S L.Fuzzy model identification based on cluster estimation dictive algorithm for complex blast fumace ironmaking process.Ac- J Intell Fuzzy Syst,1994,2(3):267 ta Autom Sin,2009,35(6):725 [19]Li X L,Su H Y,Chu J.Multiple models soft-ensing technique (郜传厚,渐令,陈积明,等.复杂高炉炼铁过程的数据驱动 based on online clustering arithmetic.J Chem Ind Eng China, 建模及预测算法.自动化学报,2009,35(6):725) 2007,58(11):2834 4]Rhodes C,Morari M,Tsimring LS.Data-based control trajectory (李修亮,苏宏业,褚键.基于在线聚类的多模型软测量建 planning for nonlinear systems.Phys Rev E,1997,56(3):2398 模方法.化工学报,2007,58(11):2834) 5]Arif M,Ishihara T.Inooka H.Incorporation of experience in iter- [20]Xiong S W.Niu XX,Liu H B.Support vector machines based ative learning controllers using locally weighted learning.Automat- on subtractive clustering /Proceeding of the Fourth International ica,2001,37(6):881 Conference on Machine Learning and Cybernetics.Guangzhou, [6]Luo Y J,Li Z F,Cao J M,et al.Application of GA to rolling 2005:4345 forces re-distribution in hot strip mills.J Uni Sci Technol Beijing, 1]Tang H S,Xue S T,Chen R,et al.Sequential LS-SVM for 2004,26(4):420 structural systems identification.J Vib Eng,2006,19(3):382 (罗永军,李忠富,曹军民,等。基于遗传算法的热连轧机负 (唐和生,薛松涛,陈镕,等。序贯最小二乘支持向量机的结 荷再分配.北京科技大学学报,2004,26(4):420) 构系统识别.振动工程学报,2006,19(3):382) Peng K X,Yu SZ.Prediction and control of strip thickness based 22]Li LJ,Su H Y,Chu J.Generalized predictive control with online on Bayesianneural networks.J Unir Sci Technol Beijing,2010,32 least squares support vector machines.Acta Autom Sin,2007,33 (2):256 (11):1182 (彭开香,余尚志.基于贝叶斯神经网络的带钢厚度预测与控 [23]Suykens J A K,Lukas L,Vandewalle J.Sparse approximation 制.北京科技大学学报,2010,32(2):256) using least squares support vector machines//Proceeding of IEEE 8]Hong Y,Tang L X.Data-driven rolling force calculation for cold International Symposium on Circuits and Systems.Geneva,2000: rolling mill.Control Eng China,2009,16(Suppl 3):146 757 (洪悦,唐立新.基于数据驱动的冷连轧过程轧制力设定方 24]Cauwenberghs G,Poggio T.Incremental and decremental support 法.控制工程,2009,16(增刊3):146) vector machine learning.Adr Neural Inf Process Syst,2001,13: [9]Son J S,Lee D M.Kilm I S,et al.A study on on-ine learning 409第 11 期 梁 辉等: 基于减法聚类的带钢厚度数据驱动建模 充分利用了生产过程中的大量历史数据,又可以根 据工况变化在线调整模型的结构和参数. ( 2) 通过数字轧机仿真实验中几种模型的性能 对比可以看出,在线多支持向量机厚度模型比最优 的偏 最 小 二 乘 统 计 模 型 的 RMSE 提 高 45. 3% , MAPE 提高 43. 2% ,TP 提高 1. 3% . 工业现场数据 仿真分析表明了本文算法具有较高的预测精度,能 满足工业在线应用要求. ( 3) 通过基于数据驱动的热轧带钢厚度预测模 型探索性研究,在忽略精轧机组间的相互影响以及 厚度与板形、宽带的耦合影响因素的前提下,建立了 单机架厚度相关因素模型. 参 考 文 献 [1] Hou Z S,Xu J X. On data-driven control theory: the state of the art and perspective. Acta Autom Sin,2009,35( 6) : 650 ( 侯忠生,许建新. 数据驱动控制理论及方法的回顾和展望. 自动化学报,2009,35( 6) : 650) [2] Gui W H,Yang C H,Li Y G,et al. Data-driven operational-pat￾tern optimization for copper flash smelting process. Acta Autom Sin,2009,35( 6) : 717 ( 桂卫华,阳春华,李勇刚,等. 基于数据驱动的铜闪速熔炼 过程操作模式优化及应用. 自动化学报,2009,35( 6) : 717) [3] Gao C H,Jian L,Chen J M,et al. Data-driven modeling and pre￾dictive algorithm for complex blast furnace ironmaking process. Ac￾ta Autom Sin,2009,35( 6) : 725 ( 郜传厚,渐令,陈积明,等. 复杂高炉炼铁过程的数据驱动 建模及预测算法. 自动化学报,2009,35( 6) : 725) [4] Rhodes C,Morari M,Tsimring L S. Data-based control trajectory planning for nonlinear systems. Phys Rev E,1997,56( 3) : 2398 [5] Arif M,Ishihara T,Inooka H. Incorporation of experience in iter￾ative learning controllers using locally weighted learning. Automat￾ica,2001,37( 6) : 881 [6] Luo Y J,Li Z F,Cao J M,et al. Application of GA to rolling forces re-distribution in hot strip mills. J Univ Sci Technol Beijing, 2004,26( 4) : 420 ( 罗永军,李忠富,曹军民,等. 基于遗传算法的热连轧机负 荷再分配. 北京科技大学学报,2004,26( 4) : 420) [7] Peng K X,Yu S Z. Prediction and control of strip thickness based on Bayesianneural networks. J Univ Sci Technol Beijing,2010,32 ( 2) : 256 ( 彭开香,余尚志. 基于贝叶斯神经网络的带钢厚度预测与控 制. 北京科技大学学报,2010,32( 2) : 256) [8] Hong Y,Tang L X. Data-driven rolling force calculation for cold rolling mill. Control Eng China,2009,16( Suppl 3) : 146 ( 洪悦,唐立新. 基于数据驱动的冷连轧过程轧制力设定方 法. 控制工程,2009,16( 增刊 3) : 146) [9] Son J S,Lee D M,Kilm I S,et al. A study on on-line learning neural network for prediction for rolling force in hot-rolling mill. J Mater Process Technol,2005,164 /165: 1612 [10] Jeong S Y,Lee M,Lee S Y,et al. Improving lookup table con￾trol of a hot coil strip process with online retrainable RBF net￾work. IEEE Trans Ind Electron,2000,47( 3) : 679 [11] Portmann N F,Lindhoff D,Sorgel G,et al. Application of neu￾ral networks in rolling mill automation. Iron Steel Eng,1995,72 ( 2) : 33 [12] Ma J S,Theiler J,Perkins S. Accurate on-line support vector re￾gression. Neural Comput,2003,15( 11) : 2683 [13] Mukherjee S,Osuna E,Girosi F. Nonlinear prediction of chaotic time series using a support vector machine / / Proceedings of 1997 IEEE Workshop. New York,1997: 511 [14] Suykens J A K. Non-linear modelling and support vector ma￾chines / / Proceedings of IEEE Instrumentation and Measurement Technology Conference. Budapest,2001: 287 [15] Li W,Yang Y P,Wang N. Multi-model LSSVM regression mod￾eling based on kernel fuzzy clustering. Control Decis,2008,23 ( 5) : 560 ( 李卫,杨煜普,王娜. 基于核模糊聚类的多模型 LSSVM 回 归建模. 控制与决策,2008,23( 5) : 560) [16] Yu Z H,Fu X,Li Y L. Online support vector regression for sys￾tem identification. Lect Notes Comput Sci,2005,3611: 627 [17] Nemmour H,Chibani Y. Multiple support vector machines for land cover change detection: an application for mapping urban extensions. ISPRS J Photogramm Remote Sens,2006,61 ( 2) : 125 [18] Chiu S L. Fuzzy model identification based on cluster estimation. J Intell Fuzzy Syst,1994,2( 3) : 267 [19] Li X L,Su H Y,Chu J. Multiple models soft-sensing technique based on online clustering arithmetic. J Chem Ind Eng China, 2007,58( 11) : 2834 ( 李修亮,苏宏业,褚键. 基于在线聚类的多模型软测量建 模方法. 化工学报,2007,58( 11) : 2834) [20] Xiong S W,Niu X X,Liu H B. Support vector machines based on subtractive clustering / / Proceeding of the Fourth International Conference on Machine Learning and Cybernetics. Guangzhou, 2005: 4345 [21] Tang H S,Xue S T,Chen R,et al. Sequential LS-SVM for structural systems identification. J Vib Eng,2006,19( 3) : 382 ( 唐和生,薛松涛,陈镕,等. 序贯最小二乘支持向量机的结 构系统识别. 振动工程学报,2006,19( 3) : 382) [22] Li L J,Su H Y,Chu J. Generalized predictive control with online least squares support vector machines. Acta Autom Sin,2007,33 ( 11) : 1182 [23] Suykens J A K,Lukas L,Vandewalle J. Sparse approximation using least squares support vector machines / /Proceeding of IEEE International Symposium on Circuits and Systems. Geneva,2000: 757 [24] Cauwenberghs G,Poggio T. Incremental and decremental support vector machine learning. Adv Neural Inf Process Syst,2001,13: 409 ·1345·
<<向上翻页
©2008-现在 cucdc.com 高等教育资讯网 版权所有