基于单位时间锌产量预测的密闭鼓风炉炉况研究
|
摘要
密闭鼓风炉铅锌熔炼过程冶金反应机理复杂,具有变量多、非线性、时变、藕合严重的特点,难以用传统的数学模型描述熔炼生产过程。维持炉况稳定是提高锌产量和质量的基本保证。将现有的炉况诊断理论应用于鼓风炉熔炼过程,进行基于炉况诊断的研究,对做好密闭鼓风炉熔炼优化控制是至关重要的,由此可以进一步提高密闭鼓风炉的生产率,减少污染,增加企业的市场竞争力。
单位时间锌产量是反映密闭鼓风炉炉况稳定的主要指标,本文研究了基于单位时间锌产量预测的密闭鼓风炉炉况诊断问题。通过对来自生产现场监测参数的历史数据进行过滤处理,应用机理分析和粗糙集方法选择建模变量,建立了基于BP神经网络的单位时间锌产量预测模型;在运用粗糙集方法数据进行处理的过程中,采用基于高斯混合模型的连续数据离散化方法,用SPSS统计软件验证了此离散化方法是行之有效的;基于机理分析和可拓评价方法建立铅锌熔炼密闭鼓风炉炉况诊断模型。
在理论研究的基础上,开发了铅锌熔炼密闭鼓风炉炉况诊断系统。系统软件采用Visual C++编程,实现了炉况主要参数在线监视、炉况诊断、数据查询和帮助等功能。现场运行情况表明该系统具有较高的故障诊断准确率和工作可靠性,具有很好的实用特性。
Imperial Smelting Furnace (ISF) smelting process is a very complex system with nonlinearity, time variety, coupling and uncertainty and very difficult to describe with conventional mathematics functions. Mainten stability of the Furnace status is basic quarantee of improving production and quality of the Zinc.The ISF status evaluation research based on Zn-output by applying existing evaluation theory into ISFsmelting process is important for the following smelting optimized controlling, Which can improve the productivity, reduce pollution and make plants more competitive.
Unit time Zinc-output is confined the key indicators to reflect the status of the ISF. In the thesis, an evaluation approach of ISF status based on unit time Zinc-output forecasting model is used. Original samples are normalized, noise samples are filtrated, and the mechanism analysis and rough set theories are applied to the historical sample data to select variables used for modeling. BP Artificial Neural Network forecasting model is formed to predict the unit time Zinc output. In rough set theories, a new method for discretization of continuous attributes based on Gaussian mixture model is presented. Another method-the SPSS statistical software, is also used, which improved the discretization mothed is efficient. Applying with extension theory, an estimation model based on extension theory of ISF status is built. the results indicated that the extension method is feasible.
Based on the theory research, a system of ISF smelting process status evaluation was developed. The software was developed with VC++, realized functions of parameter monitoring, ISF status prediction, data query and help. Debugged and tested in the plant, the reliability of fault diagnosis accuracy of the software have been proved.
单位时间锌产量是反映密闭鼓风炉炉况稳定的主要指标,本文研究了基于单位时间锌产量预测的密闭鼓风炉炉况诊断问题。通过对来自生产现场监测参数的历史数据进行过滤处理,应用机理分析和粗糙集方法选择建模变量,建立了基于BP神经网络的单位时间锌产量预测模型;在运用粗糙集方法数据进行处理的过程中,采用基于高斯混合模型的连续数据离散化方法,用SPSS统计软件验证了此离散化方法是行之有效的;基于机理分析和可拓评价方法建立铅锌熔炼密闭鼓风炉炉况诊断模型。
在理论研究的基础上,开发了铅锌熔炼密闭鼓风炉炉况诊断系统。系统软件采用Visual C++编程,实现了炉况主要参数在线监视、炉况诊断、数据查询和帮助等功能。现场运行情况表明该系统具有较高的故障诊断准确率和工作可靠性,具有很好的实用特性。
Imperial Smelting Furnace (ISF) smelting process is a very complex system with nonlinearity, time variety, coupling and uncertainty and very difficult to describe with conventional mathematics functions. Mainten stability of the Furnace status is basic quarantee of improving production and quality of the Zinc.The ISF status evaluation research based on Zn-output by applying existing evaluation theory into ISFsmelting process is important for the following smelting optimized controlling, Which can improve the productivity, reduce pollution and make plants more competitive.
Unit time Zinc-output is confined the key indicators to reflect the status of the ISF. In the thesis, an evaluation approach of ISF status based on unit time Zinc-output forecasting model is used. Original samples are normalized, noise samples are filtrated, and the mechanism analysis and rough set theories are applied to the historical sample data to select variables used for modeling. BP Artificial Neural Network forecasting model is formed to predict the unit time Zinc output. In rough set theories, a new method for discretization of continuous attributes based on Gaussian mixture model is presented. Another method-the SPSS statistical software, is also used, which improved the discretization mothed is efficient. Applying with extension theory, an estimation model based on extension theory of ISF status is built. the results indicated that the extension method is feasible.
Based on the theory research, a system of ISF smelting process status evaluation was developed. The software was developed with VC++, realized functions of parameter monitoring, ISF status prediction, data query and help. Debugged and tested in the plant, the reliability of fault diagnosis accuracy of the software have been proved.
引文
[1]韶关冶炼厂.韶冶培训教程.1994(12)
[2]刘晓颖.复杂过程的智能故障诊断技术及其在大型工业窑炉中的应用研究:[博士学位论文].长沙:中南大学,2003
[3]李瑞娟.基于锌产量预测模型的密闭鼓风炉炉况优化研究:[硕士学位论文].长沙:中南大学,2003
[4]吴湘华.密闭鼓风炉炉况智能优化操作指导系统:[硕士学位论文].长沙:中南大学,2004
[5]胡志坤,彭小奇,桂卫华.密闭鼓风炉锌产量的支持向量机实时预报模型.计算计工程,2004(12):16-18
[6]Cai Wen.New development of the basic theory of extenics.Engineering Science,1995,5(2):80-86.(in Chinese)
[7]Cai Wen.Extension set and non-compatible problems.Advances in Applied Mathematics and Mechanics in China.1983(1):83-97.(in Chinese)
[8]CaiWen.Extension Engineering Method.Beijing:Science Press,1997,68-80.(in Chinese)
[9]赵智国.密闭鼓风炉熔炼过程烧结块软化点智能集成预测模型的研究:[硕士学位论文].长沙:中南大学,2002
[10]铜铅锌冶炼参考设计资料编写组.铜铅锌冶炼参考设计资料.北京:冶金工业出版社,1979
[11]刘晓颖.复杂过程的智能故障诊断技术及其在大型工业窑炉中的应用研究:[博士学位论文].长沙:中南大学,2003
[12]王国胤.Rough集理论与知识获取.西安:西安交通大学出版社,2001
[13]王加阳.面向海量数据的粗糙集理论与方法研究:[博士学位论文].长沙:中南大学,2005
[14]Wang Guo-Yin.Rough set theory and knowledge acquisition.Xi'an.Jiaotong University Press,2001(in Chinese)
[15]PawlakZ.Rough Set Theory and Its Applications to Data Analysis.Cybernetics and Systems,1998,29(7):661-688
[16]陈淑珍.基于粗集的几种属性约简算法分析.武汉工业学院学报,2005,24(3)
[17]张文修,吴伟志,梁吉业,李德玉编著.粗糙集理论与方法.北京:科学出版社,2001
[18]陈湘辉,朱善君,吉吟东.基于嫡和变精度粗糙集的规则不确定性度量.清华 大学学报,2001,41(3):109-111
[19]Hu XiaoHua,Cercone N.Learning in relational databases:a rough set approach.Computational InteHigence,1995,11(2):323-337
[20]Tsumoto S.Automated discovery of positive and negative knowledge in clinical databased based on rough set model.IEEE EMB Magazine,2000,19(4):56-62
[21]Wojcik Z M.Detecting spots for nasa space programs using rough sets.In:Proc the 2 nd International Conference on Rough Sets and Current Trends in Computing.RSCTC2000,Canada,2000,531-537
[22]Wang G Y.Algebra view and information view of rough sets theory.Proceedings of SPIE,2001,4384:200-207
[23]MARQUARDT D.An algorithm for least squares estimation of nonlinear parameters.J Soc Ind AppI Math,1963,11:431-441
[24]Masa-aki SATO and Shin ISHH.Hon-line EM Algorithm for the Norm alized Gaussian Network.neural computation,2002,12:407-432
[25]Miao Duo-Qian,Hu Gui-Rong.A heuristic algorithm for reduction of knowledge.Joutnal of Computer Research and Development,1999,36(6):681-684(in Chinese)
[26]许东,吴铮.基于MATLAB6.X的系统分析与设计-神经网络.西安:西安电子科技大学出版社,2002
[27]朱志勇,刘伟铭,伍友龙.于Bayes法则和BP神经网络的高速动态情形下车型识别.计算机测量与控制,2005,13(7)
[28]L.F.M Zorzetto,R.M.Filho,M.R.Wolf-Maciel.Processing modeling development through artificial neural networks and hybrid models.Computers and Chemical Engineering,2000,24(2):1355-1360
[29]D.S.Lee,C.O.Jeon,J.M.Park.Hybrid neural network modeling of a full-scale industrial wastewater treatment process.Biotechnology and Bioengineering,2002,78(6):670-682
[30]Liu Zilong,Liu Guozhong,Liu Jie.Adaptive tracking controller using BP neural networks for a class of nonlinear systems.Journal of Systems Engineering and Electronics(English Edition).2004(04)
[31]EXu,Liangshan Shao,Xuedong Gao,Baofeng Zha.An Algorithm for Predicting Customer Chum via BP Neural Network Based on Rough Set.Journal of Proceedings of the 2006 IEEE Asia-Pacific Conference on Services Computing,2006(10)
[32]Chun Hua Ju,PYing Liang,Dong Sheng Liu.A kind of E-government website evaluation method based on Neural Network.Proceedings of the IEEE International Conference on e-Business Engineering.2006(09)
[33]米红,张文璋.实用现代统计分析方法与SPSS应用.当代中国出版社,2000
[34]Wang HQ,Song HZ and Li P.improved PCA with application to process monitoring and fault diagnosis.Journal of Chemical Industry and Engineering (China),2001,52(6),471-475
[35]ZHAO Xu,WEN Xiang-jun,SHAO Hui-he,Combination Method of Principal Component Analysis and Support Vector Machine for On-line Process Monitoring and Fault Diagnosis Journal of Donghua University(English Edition),2006(01)
[36]阳春华,沈德耀,吴敏等.焦炉配煤专家系统的定性定量综合设计方法.自动化学报,2000,26(2):226-232
[37]Cai Wen.New development of the basic theory of extenics.Engineering Science,1995,5(2):80-86.(in Chinese)
[38]Cai Wen.Extension set and non-compatible problems.Advances in Applied Mathematics and Mechanics in China,1983,(1):83-97(in Chinese)
[39]CaiWen.Extension Engineering Method.Beijing:Science Press,1997,68-80.(in Chinese)
[40]GuoIaaizhong The form of relevant function Brainpower and Matter-elernentAnalyse,1985,(1):48-55.(in Chinese)
[41]CaiWen,Yang Chunyar Linweic;hu Exte}a.siwaalE} aguaeer-uag M ethcxl[M]Beijing Science Press 2000 16-22 95-97,202-209(in Chinese)
[42]刘金琨.人工智能在高炉炉况判断中的理论与应用研究:[博士学位论文].沈阳:东北大学,1996
[43]CaiWen,Yang Chunyar Linweic;hu Exte}a.siwaalE}aguaeer-uag M ethcxl Beijing Science Press 2000 16-22 95-97,202-209(in Chinese)
[44]李克钢,许江,李树春,陶云奇.基于可拓理论的边坡稳定性评价研究.重庆建筑大学学报,2007,29(4):75-78
[45]孟繁晶,邓家提.合作伙伴的可拓综合评价方法.计算机集成制造系统,2005,11(6):869-874
[46]彭军龙.基于可拓学的生态影响评价方法在铁路选线中的应用.中国铁道科学,2007,28(3):1-5
[47]傅新平,胡伶俐.基于物元分析法的综合交通系统服务水平评价*.武汉理工大学学报(交通科学与工程版),2007,31(4):580-584
[48]盛晨兴,严新平,徐泰富.可拓理论在某柴油机磨损状态监测中的应用.润滑与密封,2007,32(6):27-30
[49]原国红,陈剑平,马琳.可拓评判方法在岩体质量分类中的应用.岩石力学与工程学报,2005,24(9):1539-1544
[50]张甫仁,杨昭,郁文红.室内环境评价物元模型及可拓评价方法.天津人学学报,2005,38(4):307-312
[51]陈增强等.基于Windows NT环境的工业锅炉监控系统设计.计算机工程与应用,2000,36(8):163-165
[52]唐朝晖,桂卫华.密闭鼓风炉铅锌熔炼的统计过程监测系统设计,2007,24(2)
[53]重有色金属冶炼设计手册(铅锌铋卷).北京:冶金工业出版社,2002(03)
[54]陈白玮,吴俊彬,韶冶锌精馏12~#塔数据采集系统(DAS)运行.韶关冶炼厂技术资料,1994,4:42-46
[55]唐朝晖,桂卫华.铅锌密闭鼓风炉故障诊断知识获取的粗糙集方法,小型微型计算机系统.2006,27(12)
[56]许东,吴铮.基于MATLAB6.X的系统分析与设计-神经网络.西安:西安电子科技大学出版社,2002
[57]王雅琳.智能集成建模理论及其在有色冶炼过程优化控制中的应用研究:[博士学位论文].长沙:中南大学,2001
[58]梅炽,胡志坤,彭小奇,姚俊峰,胡军.基于神经网络和自适应残差补偿德炼铜转炉吹炼终点预报模型.中国有色金属学报,2000(05)
[59]LAURITZEN S L The EM algorithm for graphical association and Data Analysis,1995,19:191-201
[60]Nomikos P.,Macgregor J.F.Monitoring Batch Process Using Multiway Principal Component Analysis.RICHE Journal,1994,40(8):1361-1375
[61]刘元扬,张凤兮有色金属生产过程检测长沙:中南工业大学出版社,1992
[62]Pawlak Z.Rough Set.International Journal of Computer Information Sciences,1982,11(5):342-356
[63]MacGregor,J.F.,Marlin,T.E.,Kresta,J.etal.Multivariate statistical methods in process analysis and control.Chemical process control.1991.79-100
[64]WILAMOWSKIB,IPLIKCI S,KAYANK O,et al.An algorithm for fast convergenece in training neural networks.International Joint Conference on Neural Networks(IJCNN'01).Washington DC:2001:1778-1782
[2]刘晓颖.复杂过程的智能故障诊断技术及其在大型工业窑炉中的应用研究:[博士学位论文].长沙:中南大学,2003
[3]李瑞娟.基于锌产量预测模型的密闭鼓风炉炉况优化研究:[硕士学位论文].长沙:中南大学,2003
[4]吴湘华.密闭鼓风炉炉况智能优化操作指导系统:[硕士学位论文].长沙:中南大学,2004
[5]胡志坤,彭小奇,桂卫华.密闭鼓风炉锌产量的支持向量机实时预报模型.计算计工程,2004(12):16-18
[6]Cai Wen.New development of the basic theory of extenics.Engineering Science,1995,5(2):80-86.(in Chinese)
[7]Cai Wen.Extension set and non-compatible problems.Advances in Applied Mathematics and Mechanics in China.1983(1):83-97.(in Chinese)
[8]CaiWen.Extension Engineering Method.Beijing:Science Press,1997,68-80.(in Chinese)
[9]赵智国.密闭鼓风炉熔炼过程烧结块软化点智能集成预测模型的研究:[硕士学位论文].长沙:中南大学,2002
[10]铜铅锌冶炼参考设计资料编写组.铜铅锌冶炼参考设计资料.北京:冶金工业出版社,1979
[11]刘晓颖.复杂过程的智能故障诊断技术及其在大型工业窑炉中的应用研究:[博士学位论文].长沙:中南大学,2003
[12]王国胤.Rough集理论与知识获取.西安:西安交通大学出版社,2001
[13]王加阳.面向海量数据的粗糙集理论与方法研究:[博士学位论文].长沙:中南大学,2005
[14]Wang Guo-Yin.Rough set theory and knowledge acquisition.Xi'an.Jiaotong University Press,2001(in Chinese)
[15]PawlakZ.Rough Set Theory and Its Applications to Data Analysis.Cybernetics and Systems,1998,29(7):661-688
[16]陈淑珍.基于粗集的几种属性约简算法分析.武汉工业学院学报,2005,24(3)
[17]张文修,吴伟志,梁吉业,李德玉编著.粗糙集理论与方法.北京:科学出版社,2001
[18]陈湘辉,朱善君,吉吟东.基于嫡和变精度粗糙集的规则不确定性度量.清华 大学学报,2001,41(3):109-111
[19]Hu XiaoHua,Cercone N.Learning in relational databases:a rough set approach.Computational InteHigence,1995,11(2):323-337
[20]Tsumoto S.Automated discovery of positive and negative knowledge in clinical databased based on rough set model.IEEE EMB Magazine,2000,19(4):56-62
[21]Wojcik Z M.Detecting spots for nasa space programs using rough sets.In:Proc the 2 nd International Conference on Rough Sets and Current Trends in Computing.RSCTC2000,Canada,2000,531-537
[22]Wang G Y.Algebra view and information view of rough sets theory.Proceedings of SPIE,2001,4384:200-207
[23]MARQUARDT D.An algorithm for least squares estimation of nonlinear parameters.J Soc Ind AppI Math,1963,11:431-441
[24]Masa-aki SATO and Shin ISHH.Hon-line EM Algorithm for the Norm alized Gaussian Network.neural computation,2002,12:407-432
[25]Miao Duo-Qian,Hu Gui-Rong.A heuristic algorithm for reduction of knowledge.Joutnal of Computer Research and Development,1999,36(6):681-684(in Chinese)
[26]许东,吴铮.基于MATLAB6.X的系统分析与设计-神经网络.西安:西安电子科技大学出版社,2002
[27]朱志勇,刘伟铭,伍友龙.于Bayes法则和BP神经网络的高速动态情形下车型识别.计算机测量与控制,2005,13(7)
[28]L.F.M Zorzetto,R.M.Filho,M.R.Wolf-Maciel.Processing modeling development through artificial neural networks and hybrid models.Computers and Chemical Engineering,2000,24(2):1355-1360
[29]D.S.Lee,C.O.Jeon,J.M.Park.Hybrid neural network modeling of a full-scale industrial wastewater treatment process.Biotechnology and Bioengineering,2002,78(6):670-682
[30]Liu Zilong,Liu Guozhong,Liu Jie.Adaptive tracking controller using BP neural networks for a class of nonlinear systems.Journal of Systems Engineering and Electronics(English Edition).2004(04)
[31]EXu,Liangshan Shao,Xuedong Gao,Baofeng Zha.An Algorithm for Predicting Customer Chum via BP Neural Network Based on Rough Set.Journal of Proceedings of the 2006 IEEE Asia-Pacific Conference on Services Computing,2006(10)
[32]Chun Hua Ju,PYing Liang,Dong Sheng Liu.A kind of E-government website evaluation method based on Neural Network.Proceedings of the IEEE International Conference on e-Business Engineering.2006(09)
[33]米红,张文璋.实用现代统计分析方法与SPSS应用.当代中国出版社,2000
[34]Wang HQ,Song HZ and Li P.improved PCA with application to process monitoring and fault diagnosis.Journal of Chemical Industry and Engineering (China),2001,52(6),471-475
[35]ZHAO Xu,WEN Xiang-jun,SHAO Hui-he,Combination Method of Principal Component Analysis and Support Vector Machine for On-line Process Monitoring and Fault Diagnosis Journal of Donghua University(English Edition),2006(01)
[36]阳春华,沈德耀,吴敏等.焦炉配煤专家系统的定性定量综合设计方法.自动化学报,2000,26(2):226-232
[37]Cai Wen.New development of the basic theory of extenics.Engineering Science,1995,5(2):80-86.(in Chinese)
[38]Cai Wen.Extension set and non-compatible problems.Advances in Applied Mathematics and Mechanics in China,1983,(1):83-97(in Chinese)
[39]CaiWen.Extension Engineering Method.Beijing:Science Press,1997,68-80.(in Chinese)
[40]GuoIaaizhong The form of relevant function Brainpower and Matter-elernentAnalyse,1985,(1):48-55.(in Chinese)
[41]CaiWen,Yang Chunyar Linweic;hu Exte}a.siwaalE} aguaeer-uag M ethcxl[M]Beijing Science Press 2000 16-22 95-97,202-209(in Chinese)
[42]刘金琨.人工智能在高炉炉况判断中的理论与应用研究:[博士学位论文].沈阳:东北大学,1996
[43]CaiWen,Yang Chunyar Linweic;hu Exte}a.siwaalE}aguaeer-uag M ethcxl Beijing Science Press 2000 16-22 95-97,202-209(in Chinese)
[44]李克钢,许江,李树春,陶云奇.基于可拓理论的边坡稳定性评价研究.重庆建筑大学学报,2007,29(4):75-78
[45]孟繁晶,邓家提.合作伙伴的可拓综合评价方法.计算机集成制造系统,2005,11(6):869-874
[46]彭军龙.基于可拓学的生态影响评价方法在铁路选线中的应用.中国铁道科学,2007,28(3):1-5
[47]傅新平,胡伶俐.基于物元分析法的综合交通系统服务水平评价*.武汉理工大学学报(交通科学与工程版),2007,31(4):580-584
[48]盛晨兴,严新平,徐泰富.可拓理论在某柴油机磨损状态监测中的应用.润滑与密封,2007,32(6):27-30
[49]原国红,陈剑平,马琳.可拓评判方法在岩体质量分类中的应用.岩石力学与工程学报,2005,24(9):1539-1544
[50]张甫仁,杨昭,郁文红.室内环境评价物元模型及可拓评价方法.天津人学学报,2005,38(4):307-312
[51]陈增强等.基于Windows NT环境的工业锅炉监控系统设计.计算机工程与应用,2000,36(8):163-165
[52]唐朝晖,桂卫华.密闭鼓风炉铅锌熔炼的统计过程监测系统设计,2007,24(2)
[53]重有色金属冶炼设计手册(铅锌铋卷).北京:冶金工业出版社,2002(03)
[54]陈白玮,吴俊彬,韶冶锌精馏12~#塔数据采集系统(DAS)运行.韶关冶炼厂技术资料,1994,4:42-46
[55]唐朝晖,桂卫华.铅锌密闭鼓风炉故障诊断知识获取的粗糙集方法,小型微型计算机系统.2006,27(12)
[56]许东,吴铮.基于MATLAB6.X的系统分析与设计-神经网络.西安:西安电子科技大学出版社,2002
[57]王雅琳.智能集成建模理论及其在有色冶炼过程优化控制中的应用研究:[博士学位论文].长沙:中南大学,2001
[58]梅炽,胡志坤,彭小奇,姚俊峰,胡军.基于神经网络和自适应残差补偿德炼铜转炉吹炼终点预报模型.中国有色金属学报,2000(05)
[59]LAURITZEN S L The EM algorithm for graphical association and Data Analysis,1995,19:191-201
[60]Nomikos P.,Macgregor J.F.Monitoring Batch Process Using Multiway Principal Component Analysis.RICHE Journal,1994,40(8):1361-1375
[61]刘元扬,张凤兮有色金属生产过程检测长沙:中南工业大学出版社,1992
[62]Pawlak Z.Rough Set.International Journal of Computer Information Sciences,1982,11(5):342-356
[63]MacGregor,J.F.,Marlin,T.E.,Kresta,J.etal.Multivariate statistical methods in process analysis and control.Chemical process control.1991.79-100
[64]WILAMOWSKIB,IPLIKCI S,KAYANK O,et al.An algorithm for fast convergenece in training neural networks.International Joint Conference on Neural Networks(IJCNN'01).Washington DC:2001:1778-1782