基于数据驱动的烧结处理过程建模和控制
|
摘要
烧结生产是高炉进料的一个重要预处理过程,能够最大程度地减小高炉原料的波动,为高炉的平稳生产提供可靠保障。随着铁矿石价格的攀升和全社会节能环保意识的增强,在钢铁工业中展开面向节能降耗的技术改进势在必行。作为影响高炉产量和质量的重要环节,烧结过程技术改进在整个炼铁过程节能降耗技术改进中必不可少。因此,烧结过程的研究具有重要理论价值和实际意义。
本文研究了烧结点火过程的控制问题和烧结热状态过程的建模控制问题,采用PIDNN (Proportional-Integral-Derivative Neural Network)控制点火炉温度,提出建立时间序列数据研究单元思路,利用多模型模糊加权预测对烧结过程热状态进行建模,再采用模糊控制方法对烧结热状态进行控制。本文的主要工作和贡献如下：
(1)采用PIDNN智能控制算法控制烧结点火炉炉温,在点火炉煤气负压和热值频繁波动的情况下,神经网络的自学习功能能够使控制系统能够实时地跟踪现场对象并加以合适的控制将被控参数准确地稳定在要求范围内。根据现场仪表的特性,将PIDNN控制方法和专家系统技术相结合,实现了对烧结点火炉温进行智能控制的目标,并且在烧结现场实施运行情况良好。
(2)提出建立时间序列数据研究单元思路,采用数据驱动建模思想,针对烧结矿从配料混匀后铺在结台车上点火开始,到破碎冷却前的烧结矿热处理过程,按照台车行进方向进行分段设定研究单元,记录其全过程的数据,包括点火温度,料层厚度,所经过风箱的负压和温度。
(3)提出多模型模糊加权预测方法,利用反向传播神经网络和广义回归神经网络独立对温度预测模型进行辨识,得到两组模型后分别针对同一组输入进行预测输出,将预测结果通过模糊加权得到新的预测输出,从而建立起烧结过程热状态的预测模型。
(4)将烧结机理想工况下的运行数据进行记录,建立数据研究单元,利用多模型模糊加权预测方法建立理想工况下风箱温度预测模型预测风箱温度目标值,然后建立模糊规则,通过控制风箱负压对每个风箱温度进行独立模糊控制。
Sintering is important for iron-making industry. As an essential pre-process of blast furnace materials, sintering could minimize the fluctuations of the materials, and thus guarantee the steady operation of furnace. With the rising price of iron ore and public enhancing consciousness of energy saving and environment protection, it is imperative to improve the technology of iron making in terms of saving energy and reducing cost. In this technical innovation, the sintering process has attracted more and more attention as it plays an important role. Therefore, it is meaningful to study the iron ore sintering process.
The thesis focuses mainly on temperature control of sintering ignition oven and modeling of sintering heating process. We adopt PIDNN (Proportional-Integral-Derivative Neural Network) control algorithm to control ignition oven temperature, based on the idea of establishing time-series data unit, fulfill the modeling of thermal state of sintering process with the help of fuzzy weitghted multi-model and then implement the control of process by fuzzy control algorithm.
In detail, the major contributions of this thesis are summarized as follows:
(1) PIDNN control algorithm is proposed to control ignition oven temperature. Regardless of the frequent fluctuations of pressure and calorific value of gas, with the help of neural network's self-learning behavior, the PIDNN control system turns out that the temperature fluctuates accurately within requirements.
(2) An idea of establishing time-series data unit is employed. We divide the sintering process into data units using data-driven method, from the mixing of ingredients to the crushing of burned sinter. The sintering machine is divided into 16 data units. All the sintering parameters will be recorded for each data unit including ignition temperature, material thickness, pressure and temperature of box while each unit runs from start to end.
(3) A fuzzy weighted multi-model method to build prediction model of sintering process is developed. We model the process by BPNN and GRNN respectively, and combine the results by fuzzy weighted method to build a new predict model.
(4) We extract the recorded parameters of sintering process operated in ideal condition to build the ideal operation data set. With this data set we use fuzzy weighted multi-model algorithm to calculate the ideal setting of box temperature. After we get the goal setting of temperatures, we utilize fuzzy rules to control each box temperature by each box pressure.
本文研究了烧结点火过程的控制问题和烧结热状态过程的建模控制问题,采用PIDNN (Proportional-Integral-Derivative Neural Network)控制点火炉温度,提出建立时间序列数据研究单元思路,利用多模型模糊加权预测对烧结过程热状态进行建模,再采用模糊控制方法对烧结热状态进行控制。本文的主要工作和贡献如下：
(1)采用PIDNN智能控制算法控制烧结点火炉炉温,在点火炉煤气负压和热值频繁波动的情况下,神经网络的自学习功能能够使控制系统能够实时地跟踪现场对象并加以合适的控制将被控参数准确地稳定在要求范围内。根据现场仪表的特性,将PIDNN控制方法和专家系统技术相结合,实现了对烧结点火炉温进行智能控制的目标,并且在烧结现场实施运行情况良好。
(2)提出建立时间序列数据研究单元思路,采用数据驱动建模思想,针对烧结矿从配料混匀后铺在结台车上点火开始,到破碎冷却前的烧结矿热处理过程,按照台车行进方向进行分段设定研究单元,记录其全过程的数据,包括点火温度,料层厚度,所经过风箱的负压和温度。
(3)提出多模型模糊加权预测方法,利用反向传播神经网络和广义回归神经网络独立对温度预测模型进行辨识,得到两组模型后分别针对同一组输入进行预测输出,将预测结果通过模糊加权得到新的预测输出,从而建立起烧结过程热状态的预测模型。
(4)将烧结机理想工况下的运行数据进行记录,建立数据研究单元,利用多模型模糊加权预测方法建立理想工况下风箱温度预测模型预测风箱温度目标值,然后建立模糊规则,通过控制风箱负压对每个风箱温度进行独立模糊控制。
Sintering is important for iron-making industry. As an essential pre-process of blast furnace materials, sintering could minimize the fluctuations of the materials, and thus guarantee the steady operation of furnace. With the rising price of iron ore and public enhancing consciousness of energy saving and environment protection, it is imperative to improve the technology of iron making in terms of saving energy and reducing cost. In this technical innovation, the sintering process has attracted more and more attention as it plays an important role. Therefore, it is meaningful to study the iron ore sintering process.
The thesis focuses mainly on temperature control of sintering ignition oven and modeling of sintering heating process. We adopt PIDNN (Proportional-Integral-Derivative Neural Network) control algorithm to control ignition oven temperature, based on the idea of establishing time-series data unit, fulfill the modeling of thermal state of sintering process with the help of fuzzy weitghted multi-model and then implement the control of process by fuzzy control algorithm.
In detail, the major contributions of this thesis are summarized as follows:
(1) PIDNN control algorithm is proposed to control ignition oven temperature. Regardless of the frequent fluctuations of pressure and calorific value of gas, with the help of neural network's self-learning behavior, the PIDNN control system turns out that the temperature fluctuates accurately within requirements.
(2) An idea of establishing time-series data unit is employed. We divide the sintering process into data units using data-driven method, from the mixing of ingredients to the crushing of burned sinter. The sintering machine is divided into 16 data units. All the sintering parameters will be recorded for each data unit including ignition temperature, material thickness, pressure and temperature of box while each unit runs from start to end.
(3) A fuzzy weighted multi-model method to build prediction model of sintering process is developed. We model the process by BPNN and GRNN respectively, and combine the results by fuzzy weighted method to build a new predict model.
(4) We extract the recorded parameters of sintering process operated in ideal condition to build the ideal operation data set. With this data set we use fuzzy weighted multi-model algorithm to calculate the ideal setting of box temperature. After we get the goal setting of temperatures, we utilize fuzzy rules to control each box temperature by each box pressure.
引文
[1]中国产业信息网.2010年1-12月中国钢材产量分省市统计[M].2011.
[2]郭利杰.钢铁工业发展周期及中国钢产量饱和点预测[J].科技和产业,2011,3)：5-8.
[3]范晓慧,王海东.烧结过程数学模型与人工智能[M].长沙：中南大学出版社,2002.
[4]叶匡吾.关于我国球团矿质量问题的探讨[J].烧结球团,2005,30(5)：
[5]商秀芹.新型进化计算方法在炼铁烧结过程建模与优化中的应用[D].杭州；浙江大学,2010.
[6]VENKATARAMANA R G S S, KAPUR P C.. A combined model for granule size distribution and cold bed permeability in the wet stage of iron ore sintering process [J]. International Journal of Mineral Processing,1999,57(1):43-58.
[7]ASSOCIATION W S. Worldsteel Top Producers 2010 [M].2011.
[8]刘俊.烧结过程智能优化控制系统研究与开发[D].杭州；浙江大学,2009.
[9]王介生.烧结过程软测量建模综述[J].烧结球团,2007,32(4)：31-5.
[10]王海东,邱冠周,黄圣生.烧结过程控制技术的发展[J].矿冶工程,1999,19(3)：3-6.
[11]华优基.烧结点火炉燃烧控制系统[J].自动化仪表,1993,14(9)：22-4.
[12]陈洪魁,李庆东.烧结机点火炉燃烧控制系统[J].黑龙江冶金,2007,1(25-7.
[13]宋迎,徐志伟,王福利.烧结点火模糊控制[J].基础自动化,1997,04)：11-3+7.
[14]杨英华,郭文军,王福利,et al.烧结机点火器燃烧模糊控制及其应用[J].钢铁,1998,02)：3-5+11.
[15]李福东.基于点火强度优化设定的烧结点火燃烧智能控制方法研究[D]；中南大学,2009.
[16]舒怀林.神经元网络及其控制系统[M].北京：国防工业出版社,2006.
[17]吴晓峰.基于支持向量机的烧结终点预报与控制研究[D]；上海大学,2009.
[18]李桃.烧结过程智能实时操作指导系统的研究[D]；中南大学,2000.
[19]WANG B, YANG B, SCIENCE J S S O I, et al. An Improved Neural Network Algorithm and its Application in Sinter Cost Prediction,俄罗斯莫斯科,F,2009[C].
[20]周敏,李世玲.广义回归神经网络在非线性系统建模中的应用[J].计算机测量与控制,2007,09)：1189-91.
[21]向婕.铁矿石烧结过程智能集成优化控制技术及其应用研究[D]；中南大学,2010.
[22]向齐良.基于烧结终点预测的烧结过程智能控制系统及应用研究[D]；中南大学,2008.
[23]HOU Z-S, XU J-X. On Data-driven Control Theory:the State of the Art and Perspective [J]. Acta Automatica Sinica,2009,35(6):650-67.
[24]WU W, DING S-Y. Model Predictive Control of Nonlinear Distributed Parameter Systems Using Spatial Neural-Network Architectures [J]. Industrial & Engineering Chemistry Research,2008,47(19): 7264-73.
[25]AGGELOGIANNAKI E, SARIMVEIS H. Nonlinear model predictive control for distributed parameter systems using data driven artificial neural network models [J]. Computers & Chemical Engineering,2008,32(6):1225-37.
[26]王国强,王志新.粒子群与PIDNN控制器在VSC-HVDC中的应用[J].中国电机工程学报,2011,03)：8-13.
[27]顾海杰,荣冈.数据驱动方法在流程工业中的应用命题综述[J].化工自动化及仪表,2009,05)：1-6.
[28]JANES K A. YAFFE M B. Data-driven modelling of signal-transduction networks [J]. Nature Reviews Molecular Cell Biology,2006,7(11):820-8.
[29]KADLEC P, GABRYS B, STRANDT S. Data-driven Soft Sensors in the process industry [J]. Computers & Chemical Engineering,2009,33(4):795-814.
[30]JIANG T, CHEN B, HE X, et al. Application of steady-state detection method based on wavelet transform [J]. Computers & Chemical Engineering,2003,27(4):569-78.
[31]郑法平.杭钢2×90m^2烧结机工程特点[J].烧结球团,2006,31(3)：11.
[32]王珂,赵相轩,张惠,et al莱钢烧结厂265m2烧结机点火温度优化控制[J].烧结球团,2008,33(5)：25-7.
[33]宋云霞,朱学峰.大时滞过程控制方法及应用[J].化工自动化及仪表,2001,28(4)：6-15.
[34]杨智,高靖.时变大时滞神经元自适应预测PID控制器[J].仪器仪表学报,2000,21(3)：300-3.
[35]靳其兵,李鸿儒,刘子静,et al.带神经网络补偿的Smith预估极点配置自校正控制[J].东北大学学报(自然科学版),1999,20(2)：133.
[36]丁晓贵,刘桂江.基于PIDNN的污水处理系统参数辨识研究[J].计算机技术与发展,2008,05)：200-2.
[37]董伟杰,刘长华,宋华.基于PIDNN控制的飞行模拟器人感系统[J].北京航空航天大学学报,2008,02)：153-7.
[38]董伟杰,刘长华,宋华.RBF和PIDNN在伺服电机模型中的应用比较[J].控制工程,2008,S1)：113-5+8.
[39]李宇杰,陈强.基于PID神经网络的电液比例速度控制系统的研究[J].科技广场,2008,05)：68-9.
[40]汤积仁,雷玉勇,聂光伟,et al神经网络P1D控制在柴油机调速系统中的应用[J].煤矿机械,2010,01)：201-4.
[41]周岸,高键.融合FLC和PIDNN在船舶航向控制中的应用[J].微计算机信息,2008,22)：60-2.
[42]田红霞.基于PIDNN的加热炉炉温控制系统的研究[J].福建电脑,2011,01)：91-2.
[43]王海龙,黄进,訾斌.基于DSP温度PIDNN控制系统设计[J].电气传动,2006,10)：40-3.
[44]吕晓云,郝俊宇,郑青栋.AS215自控系统在烧结点火器温度控制中的应用[J].冶金自动化,1993,17(6)：19-23.
[45]张强.烧结终点模糊控制研究与应用[D]；山东大学,2010.
[46]RUMELHART D E, HINTON G E, WILLIAMS R J. LEARNING REPRESENTATIONS BY BACK-PROPAGATING ERRORS [J]. Nature,1986,323(6088):533-6.
[47]MCCLELLAND J L, D.E. RUMELHART AND THE PDP RESEARCH GROUP. Parallel Distributed Processing:Explorations in the Microstructure of Cognition [M]. Cambridge, MA:MIT Press,1986.
[48]史峰,王辉,胡斐,et al. MATLAB智能算法30个案例分析[M].北京：北京航空航天大学出版社,2011.
[49]SPECHT D F. A general regression neural network [J]. Neural Networks, IEEE Transactions on, 1991,2(6):568-76.
[50]SU Z-G, WANG P-H, SHEN J, et al. Multi-model strategy based evidential soft sensor model for predicting evaluation of variables with uncertainty [J]. Applied Soft Computing,2011,11(2):2595-610.
[51]XU X, SANTEE W R. Sweat loss prediction using a multi-model approach [J]. International Journal of Biometeorology,2010,55(4):501-8.
[52]tAWRYNCZUK M, TATJEWSKI P. Nonlinear predictive control based on neural multi-models [J]. International Journal of Applied Mathematics and Computer Science,2010,20(1):7-21.
[53]RAVINDRANATHAN M, LEITCH R. Model switching in intelligent control systems [J]. Artificial Intelligence in Engineering,1999,13(2):175-87.
[54]PETRIDIS V, KEHAGIAS A. A Multi-model Algorithm for Parameter Estimation of Time-varying Nonlinear Systems [J]. Automatica,1998,34(4):469-75.
[55]JYH-SHING ROGER JANG C-T S, EIJI MIZUTANI. Neuro-Fuzzy and Soft Computing:A Computational Approach to Learning and Machine Intelligence [M]. Englewood Cliffs NJ:Prentice Hall,1997.
[56]胡宝清.模糊理论基础[M].武汉：武汉大学出版社,2004.
[571皮道映,孙优贤.多模型MIMO系统的模糊加权控制策略[J].自动化学报,1998,03)：101-4.
[58]易武,李力争.模糊控制器的一种加权平均算法[J].控制与决策,1992,7(5)：382-5.
[2]郭利杰.钢铁工业发展周期及中国钢产量饱和点预测[J].科技和产业,2011,3)：5-8.
[3]范晓慧,王海东.烧结过程数学模型与人工智能[M].长沙：中南大学出版社,2002.
[4]叶匡吾.关于我国球团矿质量问题的探讨[J].烧结球团,2005,30(5)：
[5]商秀芹.新型进化计算方法在炼铁烧结过程建模与优化中的应用[D].杭州；浙江大学,2010.
[6]VENKATARAMANA R G S S, KAPUR P C.. A combined model for granule size distribution and cold bed permeability in the wet stage of iron ore sintering process [J]. International Journal of Mineral Processing,1999,57(1):43-58.
[7]ASSOCIATION W S. Worldsteel Top Producers 2010 [M].2011.
[8]刘俊.烧结过程智能优化控制系统研究与开发[D].杭州；浙江大学,2009.
[9]王介生.烧结过程软测量建模综述[J].烧结球团,2007,32(4)：31-5.
[10]王海东,邱冠周,黄圣生.烧结过程控制技术的发展[J].矿冶工程,1999,19(3)：3-6.
[11]华优基.烧结点火炉燃烧控制系统[J].自动化仪表,1993,14(9)：22-4.
[12]陈洪魁,李庆东.烧结机点火炉燃烧控制系统[J].黑龙江冶金,2007,1(25-7.
[13]宋迎,徐志伟,王福利.烧结点火模糊控制[J].基础自动化,1997,04)：11-3+7.
[14]杨英华,郭文军,王福利,et al.烧结机点火器燃烧模糊控制及其应用[J].钢铁,1998,02)：3-5+11.
[15]李福东.基于点火强度优化设定的烧结点火燃烧智能控制方法研究[D]；中南大学,2009.
[16]舒怀林.神经元网络及其控制系统[M].北京：国防工业出版社,2006.
[17]吴晓峰.基于支持向量机的烧结终点预报与控制研究[D]；上海大学,2009.
[18]李桃.烧结过程智能实时操作指导系统的研究[D]；中南大学,2000.
[19]WANG B, YANG B, SCIENCE J S S O I, et al. An Improved Neural Network Algorithm and its Application in Sinter Cost Prediction,俄罗斯莫斯科,F,2009[C].
[20]周敏,李世玲.广义回归神经网络在非线性系统建模中的应用[J].计算机测量与控制,2007,09)：1189-91.
[21]向婕.铁矿石烧结过程智能集成优化控制技术及其应用研究[D]；中南大学,2010.
[22]向齐良.基于烧结终点预测的烧结过程智能控制系统及应用研究[D]；中南大学,2008.
[23]HOU Z-S, XU J-X. On Data-driven Control Theory:the State of the Art and Perspective [J]. Acta Automatica Sinica,2009,35(6):650-67.
[24]WU W, DING S-Y. Model Predictive Control of Nonlinear Distributed Parameter Systems Using Spatial Neural-Network Architectures [J]. Industrial & Engineering Chemistry Research,2008,47(19): 7264-73.
[25]AGGELOGIANNAKI E, SARIMVEIS H. Nonlinear model predictive control for distributed parameter systems using data driven artificial neural network models [J]. Computers & Chemical Engineering,2008,32(6):1225-37.
[26]王国强,王志新.粒子群与PIDNN控制器在VSC-HVDC中的应用[J].中国电机工程学报,2011,03)：8-13.
[27]顾海杰,荣冈.数据驱动方法在流程工业中的应用命题综述[J].化工自动化及仪表,2009,05)：1-6.
[28]JANES K A. YAFFE M B. Data-driven modelling of signal-transduction networks [J]. Nature Reviews Molecular Cell Biology,2006,7(11):820-8.
[29]KADLEC P, GABRYS B, STRANDT S. Data-driven Soft Sensors in the process industry [J]. Computers & Chemical Engineering,2009,33(4):795-814.
[30]JIANG T, CHEN B, HE X, et al. Application of steady-state detection method based on wavelet transform [J]. Computers & Chemical Engineering,2003,27(4):569-78.
[31]郑法平.杭钢2×90m^2烧结机工程特点[J].烧结球团,2006,31(3)：11.
[32]王珂,赵相轩,张惠,et al莱钢烧结厂265m2烧结机点火温度优化控制[J].烧结球团,2008,33(5)：25-7.
[33]宋云霞,朱学峰.大时滞过程控制方法及应用[J].化工自动化及仪表,2001,28(4)：6-15.
[34]杨智,高靖.时变大时滞神经元自适应预测PID控制器[J].仪器仪表学报,2000,21(3)：300-3.
[35]靳其兵,李鸿儒,刘子静,et al.带神经网络补偿的Smith预估极点配置自校正控制[J].东北大学学报(自然科学版),1999,20(2)：133.
[36]丁晓贵,刘桂江.基于PIDNN的污水处理系统参数辨识研究[J].计算机技术与发展,2008,05)：200-2.
[37]董伟杰,刘长华,宋华.基于PIDNN控制的飞行模拟器人感系统[J].北京航空航天大学学报,2008,02)：153-7.
[38]董伟杰,刘长华,宋华.RBF和PIDNN在伺服电机模型中的应用比较[J].控制工程,2008,S1)：113-5+8.
[39]李宇杰,陈强.基于PID神经网络的电液比例速度控制系统的研究[J].科技广场,2008,05)：68-9.
[40]汤积仁,雷玉勇,聂光伟,et al神经网络P1D控制在柴油机调速系统中的应用[J].煤矿机械,2010,01)：201-4.
[41]周岸,高键.融合FLC和PIDNN在船舶航向控制中的应用[J].微计算机信息,2008,22)：60-2.
[42]田红霞.基于PIDNN的加热炉炉温控制系统的研究[J].福建电脑,2011,01)：91-2.
[43]王海龙,黄进,訾斌.基于DSP温度PIDNN控制系统设计[J].电气传动,2006,10)：40-3.
[44]吕晓云,郝俊宇,郑青栋.AS215自控系统在烧结点火器温度控制中的应用[J].冶金自动化,1993,17(6)：19-23.
[45]张强.烧结终点模糊控制研究与应用[D]；山东大学,2010.
[46]RUMELHART D E, HINTON G E, WILLIAMS R J. LEARNING REPRESENTATIONS BY BACK-PROPAGATING ERRORS [J]. Nature,1986,323(6088):533-6.
[47]MCCLELLAND J L, D.E. RUMELHART AND THE PDP RESEARCH GROUP. Parallel Distributed Processing:Explorations in the Microstructure of Cognition [M]. Cambridge, MA:MIT Press,1986.
[48]史峰,王辉,胡斐,et al. MATLAB智能算法30个案例分析[M].北京：北京航空航天大学出版社,2011.
[49]SPECHT D F. A general regression neural network [J]. Neural Networks, IEEE Transactions on, 1991,2(6):568-76.
[50]SU Z-G, WANG P-H, SHEN J, et al. Multi-model strategy based evidential soft sensor model for predicting evaluation of variables with uncertainty [J]. Applied Soft Computing,2011,11(2):2595-610.
[51]XU X, SANTEE W R. Sweat loss prediction using a multi-model approach [J]. International Journal of Biometeorology,2010,55(4):501-8.
[52]tAWRYNCZUK M, TATJEWSKI P. Nonlinear predictive control based on neural multi-models [J]. International Journal of Applied Mathematics and Computer Science,2010,20(1):7-21.
[53]RAVINDRANATHAN M, LEITCH R. Model switching in intelligent control systems [J]. Artificial Intelligence in Engineering,1999,13(2):175-87.
[54]PETRIDIS V, KEHAGIAS A. A Multi-model Algorithm for Parameter Estimation of Time-varying Nonlinear Systems [J]. Automatica,1998,34(4):469-75.
[55]JYH-SHING ROGER JANG C-T S, EIJI MIZUTANI. Neuro-Fuzzy and Soft Computing:A Computational Approach to Learning and Machine Intelligence [M]. Englewood Cliffs NJ:Prentice Hall,1997.
[56]胡宝清.模糊理论基础[M].武汉：武汉大学出版社,2004.
[571皮道映,孙优贤.多模型MIMO系统的模糊加权控制策略[J].自动化学报,1998,03)：101-4.
[58]易武,李力争.模糊控制器的一种加权平均算法[J].控制与决策,1992,7(5)：382-5.