水泥熟料煅烧质量的神经网络预报系统研究
|
摘要
水泥熟料的煅烧是一个非线性、多变量、大滞后、强干扰的复杂工业过程，进行水泥熟料煅烧质量的预报对于实现水泥生产过程的优化控制具有重要的作用。由于传统的预报方法难以满足这种复杂系统的预报建模要求，因此，应用人工智能技术，建立水泥熟料煅烧质量的人工神经网络预报系统也就具有重要的意义。
通过对水泥熟料煅烧系统进行分析，提出了由数据预处理系统和神经网络系统组成的水泥熟料煅烧质量的神经网络预报系统方案；研究了神经网络建模所需的数据预处理方法，包括数据清洗、数据变换和神经网络变量选取方法；研究了BP网络模型和BP算法，并对BP算法进行了加入动量项和自适应调节步长的改进。
构筑了一个具有通用性的开发复杂软件的系统框架，应用此软件框架开发了神经网络建模工具软件，进而建立了水泥熟料煅烧质量的神经网络预报模型，并得了较好的预报结果：立升重的预报相对误差为1.8％；f-CaO含量的预报相对误差为2.8％。
The cement clinker calcination is a complicated industry process which is nonlinear, greatly lagged, multi-variable and strongly disturbed. The forecast of cement clinker calcinations quality is very important to the optimization control of the produce process of cement. Because the require of forecast modeling for this complicated system is difficult to be satisfied by the traditional forecast methods, it is significant to apply artificial technology to build the neural network forecast system for the calcination quality of cement clinker.
Through analyzing the calcination system of cement clinker, the scheme of neural network forecast, which include neural network system and data pretreatment system, is proposed; The method of data pretreatment, which is needed for neural network modeling and consist of cleaning data, switching data and selecting the variables of neural network, is researched; BP network model and BP algorithm are also researched and BP algorithm is improved by adding the momentum item and self-adapting the length of train step.
A general system frame which can be used to develop complex software is built. A tool software for modeling neural network is developed on this software frame. A neural network forecast model of cement clinker calcination quality is built by this tool software. A good forecast result is got through this model: the relative error of g/L is 1.8%, the relative error of the content of f-CaO is 2.8%.
通过对水泥熟料煅烧系统进行分析，提出了由数据预处理系统和神经网络系统组成的水泥熟料煅烧质量的神经网络预报系统方案；研究了神经网络建模所需的数据预处理方法，包括数据清洗、数据变换和神经网络变量选取方法；研究了BP网络模型和BP算法，并对BP算法进行了加入动量项和自适应调节步长的改进。
构筑了一个具有通用性的开发复杂软件的系统框架，应用此软件框架开发了神经网络建模工具软件，进而建立了水泥熟料煅烧质量的神经网络预报模型，并得了较好的预报结果：立升重的预报相对误差为1.8％；f-CaO含量的预报相对误差为2.8％。
The cement clinker calcination is a complicated industry process which is nonlinear, greatly lagged, multi-variable and strongly disturbed. The forecast of cement clinker calcinations quality is very important to the optimization control of the produce process of cement. Because the require of forecast modeling for this complicated system is difficult to be satisfied by the traditional forecast methods, it is significant to apply artificial technology to build the neural network forecast system for the calcination quality of cement clinker.
Through analyzing the calcination system of cement clinker, the scheme of neural network forecast, which include neural network system and data pretreatment system, is proposed; The method of data pretreatment, which is needed for neural network modeling and consist of cleaning data, switching data and selecting the variables of neural network, is researched; BP network model and BP algorithm are also researched and BP algorithm is improved by adding the momentum item and self-adapting the length of train step.
A general system frame which can be used to develop complex software is built. A tool software for modeling neural network is developed on this software frame. A neural network forecast model of cement clinker calcination quality is built by this tool software. A good forecast result is got through this model: the relative error of g/L is 1.8%, the relative error of the content of f-CaO is 2.8%.
引文
[1] 沈威，王文熙，闵盘荣．水泥工艺学．武汉：武汉工业大学出版社，1995．1-2
[2] 发展改革委，财政部，国土资源部等．关于防止水泥行业盲目投资加快结构调整的若干意见．国务院公报，2004年第6号
[3] 沈威，王文熙，闵盘荣．水泥预分解技术及热工系统工程．武汉：武汉工业大学出版社，1995．1-2
[4] 蒋永惠，陈袁魁，袁林生．建立数学模型预测熟料强度．水泥技术，1995，7(2)：13-18
[5] 贾天运 利用率值进行水泥熟料的水化热的快速预测．水泥，1994，11(3)：29
[6] Semple K E, Cunningham R B, Evans P D. Cement hydration tests using wood flour may not predict the suitability of Acacia mangium and Eucalyptus pellita for the manufacture of wood-wool cement boards. Holzforschung, 1999,53(3):327-332
[7] Thiercelin M, Baumgarte C, Guillot D. Soil mechanics approach to predict cement sheath behavior. SPE/International Society of Rock Mechanics in Petroleum Engineering Conference, 1998,2:329-335
[8] Dean G D, Brennen M A. Unique laboratory gas flow model reveals insight to predict gas migration in cement. Proceedings of the 1992 Western Regional Meeting, 1992:237-248
[9] Sonebi M, Tamimi A K, Bartos P J M. Application of factorial models to predict the effect of anti-washout admixture, super plasticizer and cement on slump, flow time and washout resistance of underwater concrete. Materials and Structures, 2000,33(229):317-323
[10] Eijk R J, Brouwers H J H. Prediction of hydroxyl concentrations in cement pore water using a numerical cement hydration model. Cement and Concrete Research, 2000,30(11):1801-1806
[11] 唐辉红．应用三元线性回归分析方法预报水泥28d强度．水泥，1998，7(2)：13-18
[12] 方真．运用微机对水泥强度的多元线性回归分析．江西建材，1996，6(3)：76-79
[13] 李楷模．回归预测技术运用于生产过程．水泥情报，1989，11(6)：28-30
[14] Suman George O Jr, Wood Edward. Predicting seal effectiveness of cement-inflated packers. World Oil,1995,216(5):7-11
[15] Cobb Steve, Maki Voldi, Sabins Fred.Method predicts foamed-cement compressive strength under temperature, pressure. Oil and Gas Journal, 2002,100(15):48-52
[16] 吴有富．一种新的属性模式聚类识别算法及其在水泥强度预测中的应用．系统工程理论与实践，2000，6(7)：118-122
[17] Amziane S, Andriamanantsilavo N. Prediction of cement paste pore water pressure variations during setting period. Advances in Cement Research, 2004,16(1):23-28
[18] Magni Lalo, Bastin Georges, Wertz Vincent. Multivariable nonlinear predictive control of cement mills. IEEE Transactions on Control Systems Technology, 1999,7(4):502-508
[19] Martin Greg, McGarel Steve. Optimizing the cement production process using model predictive control. Rock Products, 2002,105:16
[20] 高法良 利用模糊数学预测熟料和水泥强度的研究．上海大学学报(自然科学版)，1996，2(2)：229-236
[21] 徐辉，夏虹．水泥强度的模糊预测模型研究．工科数学，2000，16(2)：24-26
[22] Gao Fa-Liang. New way of predicting cement strength - fuzzy logic. Cement and Concrete Research, 1997,27(6):883-888
[23] Hill Eugene D Jr, Frohnsdorff Geoffrey. Portland cement specifications: Performance, prescription, and prediction. Cement, Concrete and Aggregates, 1993,15(2):109-118
[24] Bentz D P. Modelling cement microstructure: Pixels, particles, and property prediction. Materials and Structures/Materiaux et Constructions, 1999,32(217):187-195
[25] Koumboulis F N, Kouvakas N D. Model predictive temperature control towards improving cement precalcination. Proceedings of the Institution of Mechanical Engineers. PartⅠ: Journal of Systems and Control Engineering, 2003,217(2):147-153
[26] 王继宗，倪宏光．基于BP网络的水泥抗压强度预测研究．硅酸盐学报，1999．7(2)：13-18
[27] 刘勇健，沈军，刘义健．人工神经网络在水泥加固土力学性能预测中的应用．岩上力学，2001．9(3)：330-333
[28] 郁时炼，高辉．基于神经网络的水泥强度预测．合肥工业大学学报，2002，25(2)：317-320
[29] Basma Adnan A, Barakat Samer A, Al-Oraimi Salim. Prediction of cement degree of hydration using artificial neural networks.ACI Materials Journal,1999,96(2):167-172
[30] Coveney P V, Fletcher P, Hughes T L. Using artificial neural networks to predict the quality and performance of oil-field cements. AI
Magazine, 1996,17(4):41-53
[31] Coveney P V, Hughes T L, Fletcher P. Using artificial neural networks to predict the quality and performance of oilfield cements. Innovative Applications of Artificial Intelligence Conference Proceedings, 1996:1471-1481
[32] 阎平凡，张常水．人工神经网络与模拟进化计算．北京：清华大学出版社，2000．1-60
[33] 袁曾任．人工神经元网络及其应用．北京：清华大学出版社，1996．10-60
[34] Skrzpek T. Neural network simulation environments. Massachusetts: Kluwer Academic Publishers, 1993.43-47
[35] Dan W Patterson. Artificial neural networks: theory and applications. US: Prentice Hall,1996.21-29
[36] Arbib M M. The handbook of brain theory and neural networks. Boston: The MIT Press, 1995.35-39
[37] Kastra M. Forecasting Combining with Neural Networks. Jour of Forec. 1996,15(1):49-61.
[38] Widrow B, et al. Neural Network Application in Industry. business and Science. Communcation of the ACM, 1994,37(4): 93-105
[39] Yan Pingfan, Zhang Changshui. Artificial Neural Network and Computing with the Simulation of Evolution. Beijing China : Tsinghua University Publishers ,2000.13-19
[40] 阎加强，单松高，伍卫琼．ZrO2-SiC材料中原位SiC生成量预报的神经网络模型．硅酸盐学报，1999，7(2)：13-18
[41] 冯利华，金华．神经网络在天气预报中的应用．信息与控制，2001，7(2)：13-18
[42] 王炜，吴耿锋，宋先月．神经网络在地震学方法综合预报中的应用．地震学报，2000，7(2)：13-18
[43] R. C. Eberhart, R. W. Dobbins Neral Network Pc Tools. Academic Press, 1990, 7(2): 13-18
[44] Widrow B, et al. Neural Network Application in Industry, business and Science. Communcation of the ACM, 1994,37(6): 93-105
[45] Hornik K, et al. Multilayer Feedforward Network are Universal Approximators. Neural Network, 1989,23(2): 359-366
[46] 周毓箐等．建筑卫生陶瓷缺陷分析神经网络系统．陶瓷，1998，6(2)：14-18．
[47] 阎加强等．人工神经网络及其在耐火材料研究中的应用．耐火材料，1998，7(4)：238-240．
[48] 唐波，钦佩，陈念贻．人工神经网络在熔盐相图中间相预报中的应用．金属学
报，1994，30(1)：21-26．
[49] 徐驰，陈念贻．人工神经网络对烧结型Nd-Fe-B永磁合金配方的优化．功能材料与器件学报，1995，1(2)：111-114．
[50] 吕允文等．材料设计中专家系统与人工神经网络的应用．材料导报，1994(6)：1—4．
[51] 曾令可等．材料设计中神经网络结构及其有关参数的选取．佛山陶瓷．1993(3)：3-6．
[52] 康德山，王学业，陈念贻．人工神经网络多目标法用于辅助V-PTC材料研制．功能材料与器件学报，1995，1(2)：115-118．
[53] 吕允文，夏宗宁，赖树纲．材料设计专家系统与人工神经网络的应用．材料导报，1996，(6)：1-4．
[54] 郭进，刘霖，黄铁生等．计算机辅助材料设计的偏最小二乘法—人工神经网络研究．计算机与应用化学，1996，13(4)：253-256．
[55] 夏宗宁，雷永明．封接合金材料设计的知识获取的研究．机械工程材料，1996，20(3)：33—35．
[56] 张建民，陈宏力，魏秀兰．一种动态神经网络在水泥回转窑温度控制中的应用．吉林化工学院学报，2001，7(2)：13-18
[57] 方千山，黄天东．基于神经网络的立窑水泥煅烧过程的模糊控制．水泥技术，2000，7(2)：13-18
[58] 王学民．应用多元统计分析．上海：上海财经大学出版社，1999．1-60
[59] 任若恩，王惠文．多元统计数据分析——理论、方法、实例．长沙：国防工业出版社，1997．1-60
[60] Jedward Jackson. Principal Components and Factor Analysis:Part1 Principal Components. Journal of Quality Technology. 1980,12(4):45-48
[61] Michael J, Piovoso and Karlene A. Kosanovich. Applications of Multivariate Statistical Methods to Process Monitoring and Controller Design. INT. J. CONTROL, 1994,59(3):743-765
[62] Pal N R, Bezdek J C. On cluster validity for the fuzzy c-mean model. IEEE Trans Fuzzy System, 1995, 3(3):370-379.
[63] Fang Kaitai, Pan Enpei. Cluster Analysis. Beijing China : Geology Publishers ,1982.12-34
[64] Donald G F. Object-oriented Requirements Analysis and Logical Design. New York: John Wiley &Sons, Inc, 1993.5-33
[65] Graham I. Object Oriented Methods. London: Addison Welseley Publishers Ltd.,1993.3-20
[2] 发展改革委，财政部，国土资源部等．关于防止水泥行业盲目投资加快结构调整的若干意见．国务院公报，2004年第6号
[3] 沈威，王文熙，闵盘荣．水泥预分解技术及热工系统工程．武汉：武汉工业大学出版社，1995．1-2
[4] 蒋永惠，陈袁魁，袁林生．建立数学模型预测熟料强度．水泥技术，1995，7(2)：13-18
[5] 贾天运 利用率值进行水泥熟料的水化热的快速预测．水泥，1994，11(3)：29
[6] Semple K E, Cunningham R B, Evans P D. Cement hydration tests using wood flour may not predict the suitability of Acacia mangium and Eucalyptus pellita for the manufacture of wood-wool cement boards. Holzforschung, 1999,53(3):327-332
[7] Thiercelin M, Baumgarte C, Guillot D. Soil mechanics approach to predict cement sheath behavior. SPE/International Society of Rock Mechanics in Petroleum Engineering Conference, 1998,2:329-335
[8] Dean G D, Brennen M A. Unique laboratory gas flow model reveals insight to predict gas migration in cement. Proceedings of the 1992 Western Regional Meeting, 1992:237-248
[9] Sonebi M, Tamimi A K, Bartos P J M. Application of factorial models to predict the effect of anti-washout admixture, super plasticizer and cement on slump, flow time and washout resistance of underwater concrete. Materials and Structures, 2000,33(229):317-323
[10] Eijk R J, Brouwers H J H. Prediction of hydroxyl concentrations in cement pore water using a numerical cement hydration model. Cement and Concrete Research, 2000,30(11):1801-1806
[11] 唐辉红．应用三元线性回归分析方法预报水泥28d强度．水泥，1998，7(2)：13-18
[12] 方真．运用微机对水泥强度的多元线性回归分析．江西建材，1996，6(3)：76-79
[13] 李楷模．回归预测技术运用于生产过程．水泥情报，1989，11(6)：28-30
[14] Suman George O Jr, Wood Edward. Predicting seal effectiveness of cement-inflated packers. World Oil,1995,216(5):7-11
[15] Cobb Steve, Maki Voldi, Sabins Fred.Method predicts foamed-cement compressive strength under temperature, pressure. Oil and Gas Journal, 2002,100(15):48-52
[16] 吴有富．一种新的属性模式聚类识别算法及其在水泥强度预测中的应用．系统工程理论与实践，2000，6(7)：118-122
[17] Amziane S, Andriamanantsilavo N. Prediction of cement paste pore water pressure variations during setting period. Advances in Cement Research, 2004,16(1):23-28
[18] Magni Lalo, Bastin Georges, Wertz Vincent. Multivariable nonlinear predictive control of cement mills. IEEE Transactions on Control Systems Technology, 1999,7(4):502-508
[19] Martin Greg, McGarel Steve. Optimizing the cement production process using model predictive control. Rock Products, 2002,105:16
[20] 高法良 利用模糊数学预测熟料和水泥强度的研究．上海大学学报(自然科学版)，1996，2(2)：229-236
[21] 徐辉，夏虹．水泥强度的模糊预测模型研究．工科数学，2000，16(2)：24-26
[22] Gao Fa-Liang. New way of predicting cement strength - fuzzy logic. Cement and Concrete Research, 1997,27(6):883-888
[23] Hill Eugene D Jr, Frohnsdorff Geoffrey. Portland cement specifications: Performance, prescription, and prediction. Cement, Concrete and Aggregates, 1993,15(2):109-118
[24] Bentz D P. Modelling cement microstructure: Pixels, particles, and property prediction. Materials and Structures/Materiaux et Constructions, 1999,32(217):187-195
[25] Koumboulis F N, Kouvakas N D. Model predictive temperature control towards improving cement precalcination. Proceedings of the Institution of Mechanical Engineers. PartⅠ: Journal of Systems and Control Engineering, 2003,217(2):147-153
[26] 王继宗，倪宏光．基于BP网络的水泥抗压强度预测研究．硅酸盐学报，1999．7(2)：13-18
[27] 刘勇健，沈军，刘义健．人工神经网络在水泥加固土力学性能预测中的应用．岩上力学，2001．9(3)：330-333
[28] 郁时炼，高辉．基于神经网络的水泥强度预测．合肥工业大学学报，2002，25(2)：317-320
[29] Basma Adnan A, Barakat Samer A, Al-Oraimi Salim. Prediction of cement degree of hydration using artificial neural networks.ACI Materials Journal,1999,96(2):167-172
[30] Coveney P V, Fletcher P, Hughes T L. Using artificial neural networks to predict the quality and performance of oil-field cements. AI
Magazine, 1996,17(4):41-53
[31] Coveney P V, Hughes T L, Fletcher P. Using artificial neural networks to predict the quality and performance of oilfield cements. Innovative Applications of Artificial Intelligence Conference Proceedings, 1996:1471-1481
[32] 阎平凡，张常水．人工神经网络与模拟进化计算．北京：清华大学出版社，2000．1-60
[33] 袁曾任．人工神经元网络及其应用．北京：清华大学出版社，1996．10-60
[34] Skrzpek T. Neural network simulation environments. Massachusetts: Kluwer Academic Publishers, 1993.43-47
[35] Dan W Patterson. Artificial neural networks: theory and applications. US: Prentice Hall,1996.21-29
[36] Arbib M M. The handbook of brain theory and neural networks. Boston: The MIT Press, 1995.35-39
[37] Kastra M. Forecasting Combining with Neural Networks. Jour of Forec. 1996,15(1):49-61.
[38] Widrow B, et al. Neural Network Application in Industry. business and Science. Communcation of the ACM, 1994,37(4): 93-105
[39] Yan Pingfan, Zhang Changshui. Artificial Neural Network and Computing with the Simulation of Evolution. Beijing China : Tsinghua University Publishers ,2000.13-19
[40] 阎加强，单松高，伍卫琼．ZrO2-SiC材料中原位SiC生成量预报的神经网络模型．硅酸盐学报，1999，7(2)：13-18
[41] 冯利华，金华．神经网络在天气预报中的应用．信息与控制，2001，7(2)：13-18
[42] 王炜，吴耿锋，宋先月．神经网络在地震学方法综合预报中的应用．地震学报，2000，7(2)：13-18
[43] R. C. Eberhart, R. W. Dobbins Neral Network Pc Tools. Academic Press, 1990, 7(2): 13-18
[44] Widrow B, et al. Neural Network Application in Industry, business and Science. Communcation of the ACM, 1994,37(6): 93-105
[45] Hornik K, et al. Multilayer Feedforward Network are Universal Approximators. Neural Network, 1989,23(2): 359-366
[46] 周毓箐等．建筑卫生陶瓷缺陷分析神经网络系统．陶瓷，1998，6(2)：14-18．
[47] 阎加强等．人工神经网络及其在耐火材料研究中的应用．耐火材料，1998，7(4)：238-240．
[48] 唐波，钦佩，陈念贻．人工神经网络在熔盐相图中间相预报中的应用．金属学
报，1994，30(1)：21-26．
[49] 徐驰，陈念贻．人工神经网络对烧结型Nd-Fe-B永磁合金配方的优化．功能材料与器件学报，1995，1(2)：111-114．
[50] 吕允文等．材料设计中专家系统与人工神经网络的应用．材料导报，1994(6)：1—4．
[51] 曾令可等．材料设计中神经网络结构及其有关参数的选取．佛山陶瓷．1993(3)：3-6．
[52] 康德山，王学业，陈念贻．人工神经网络多目标法用于辅助V-PTC材料研制．功能材料与器件学报，1995，1(2)：115-118．
[53] 吕允文，夏宗宁，赖树纲．材料设计专家系统与人工神经网络的应用．材料导报，1996，(6)：1-4．
[54] 郭进，刘霖，黄铁生等．计算机辅助材料设计的偏最小二乘法—人工神经网络研究．计算机与应用化学，1996，13(4)：253-256．
[55] 夏宗宁，雷永明．封接合金材料设计的知识获取的研究．机械工程材料，1996，20(3)：33—35．
[56] 张建民，陈宏力，魏秀兰．一种动态神经网络在水泥回转窑温度控制中的应用．吉林化工学院学报，2001，7(2)：13-18
[57] 方千山，黄天东．基于神经网络的立窑水泥煅烧过程的模糊控制．水泥技术，2000，7(2)：13-18
[58] 王学民．应用多元统计分析．上海：上海财经大学出版社，1999．1-60
[59] 任若恩，王惠文．多元统计数据分析——理论、方法、实例．长沙：国防工业出版社，1997．1-60
[60] Jedward Jackson. Principal Components and Factor Analysis:Part1 Principal Components. Journal of Quality Technology. 1980,12(4):45-48
[61] Michael J, Piovoso and Karlene A. Kosanovich. Applications of Multivariate Statistical Methods to Process Monitoring and Controller Design. INT. J. CONTROL, 1994,59(3):743-765
[62] Pal N R, Bezdek J C. On cluster validity for the fuzzy c-mean model. IEEE Trans Fuzzy System, 1995, 3(3):370-379.
[63] Fang Kaitai, Pan Enpei. Cluster Analysis. Beijing China : Geology Publishers ,1982.12-34
[64] Donald G F. Object-oriented Requirements Analysis and Logical Design. New York: John Wiley &Sons, Inc, 1993.5-33
[65] Graham I. Object Oriented Methods. London: Addison Welseley Publishers Ltd.,1993.3-20