基于板形的热连轧精轧机组机座参数优化
|
摘要
机座优化技术的研究大致始于上世纪80年代，至今已有近30年的历史，还是一项较新的技术，从最开始的对于机座、辊系以及机架的优化分析到现在的对工作辊和支承辊辊型进行优化计算，这一技术已经取得了长足的发展，并在实际应用中取得了显著的效果。
国内学者在理论计算方面有着丰富的经验，目前大多采用影响函数法来计算辊系的变形，利用变分法计算金属塑性流动，所得到的辊系变形和出口厚度都与实际十分贴近，而优化部分广泛采用的是Powell优化算法，一般过程是先将目标函数通过罚函数法转换为无约束函数，再利用Powell算法进行优化，由于算法本身的缺陷，优化过程速度较慢且容易陷入局部最小值。
本文在基于前人研究的基础上，对优化部分进行了改进，通过采用最新的AMS-PSO（自适应双群微粒群优化算法）优化算法来提高结果的准确度，使其更接近全局最优值，并与Powell算法进行了多方面对比，从而综合比较二者的优缺点，方便用户的选取。
利用本程序对某850轧机进行优化实践，考察所设计的850轧机是否符合设计要求，同时从刚度、重量等方面对其进行优化，给出参考意见，为用户对850的修改以及今后的设计提供科学的建议。
Study on the optimization technology of roll frame largely began in the last century80's, it has been nearly 30 years up to now, is also a relatively new technology, from thevery beginning study of frame,roll system and rack and to the optimization analysis onwork roll and back roll profile optimization, development of this technology has come along way, and achieved remarkable results in the actual application.
Scholars in China has a wealth of experience in the theoretical calculations, nowmostly using influence function method to calculate the deformation of roll system, usingvariational method to calculate the metal plastic flow, the resulting deformation of rollsystem and export thickness are very close to actual, and optimization part widely used isPowell optimization algorithm, the normal procedure is first to translate the objectivefunction by penalty function method to unconstrained function, Then with Powellalgorithm for optimization, because the defects of the algorithm itself, the optimizationprocess is slow and the result vulnerable fall into a local minimum.
This article is based on the basis study of predecessors and bring some improvmentin the optimization part, through the use of the latest AMS-PSO Optimization algorithm toimprove the accuracy of the results, making it closer to the global optimal value, and makecomprehensive comparison and analysis with Powell algorithm, as a result comprehensivecomparing the advantages and disadvantages of the two, that will be helpful for the user'sselection.
Use this program for a 850 mill optimization practices, verify that 850 mill meets thedesign requirements, and though stiffness, weight and other aspects to optimize it, givesome scientific advice on the 850 mill’s design,and that will be usefull for the 850 mill’smodification and other mill’s design in the furure.
国内学者在理论计算方面有着丰富的经验，目前大多采用影响函数法来计算辊系的变形，利用变分法计算金属塑性流动，所得到的辊系变形和出口厚度都与实际十分贴近，而优化部分广泛采用的是Powell优化算法，一般过程是先将目标函数通过罚函数法转换为无约束函数，再利用Powell算法进行优化，由于算法本身的缺陷，优化过程速度较慢且容易陷入局部最小值。
本文在基于前人研究的基础上，对优化部分进行了改进，通过采用最新的AMS-PSO（自适应双群微粒群优化算法）优化算法来提高结果的准确度，使其更接近全局最优值，并与Powell算法进行了多方面对比，从而综合比较二者的优缺点，方便用户的选取。
利用本程序对某850轧机进行优化实践，考察所设计的850轧机是否符合设计要求，同时从刚度、重量等方面对其进行优化，给出参考意见，为用户对850的修改以及今后的设计提供科学的建议。
Study on the optimization technology of roll frame largely began in the last century80's, it has been nearly 30 years up to now, is also a relatively new technology, from thevery beginning study of frame,roll system and rack and to the optimization analysis onwork roll and back roll profile optimization, development of this technology has come along way, and achieved remarkable results in the actual application.
Scholars in China has a wealth of experience in the theoretical calculations, nowmostly using influence function method to calculate the deformation of roll system, usingvariational method to calculate the metal plastic flow, the resulting deformation of rollsystem and export thickness are very close to actual, and optimization part widely used isPowell optimization algorithm, the normal procedure is first to translate the objectivefunction by penalty function method to unconstrained function, Then with Powellalgorithm for optimization, because the defects of the algorithm itself, the optimizationprocess is slow and the result vulnerable fall into a local minimum.
This article is based on the basis study of predecessors and bring some improvmentin the optimization part, through the use of the latest AMS-PSO Optimization algorithm toimprove the accuracy of the results, making it closer to the global optimal value, and makecomprehensive comparison and analysis with Powell algorithm, as a result comprehensivecomparing the advantages and disadvantages of the two, that will be helpful for the user'sselection.
Use this program for a 850 mill optimization practices, verify that 850 mill meets thedesign requirements, and though stiffness, weight and other aspects to optimize it, givesome scientific advice on the 850 mill’s design,and that will be usefull for the 850 mill’smodification and other mill’s design in the furure.
引文
[1]闻邦椿.产品设计理论与方法的发展趋向及产品的现代设计[J].机电工程,2011,28(3):256-257.
[2]王钢.热连轧带钢轧机辊系优化设计[J].重型机械，1985,(1):1-7.
[3]邹家祥.轧钢机械[M].北京:冶金工业出版社.2007:97-104.
[4]王琍,郭希学. HC轧机辊系优化设计[J].太原重型机械学院学报,1996,17(1):6-9.
[5]王伟.板带冷轧机机型选择与辊系参数优化理论及其应用[D].秦皇岛:燕山大学机械设计及理论学科工学博士论文,1999:24-32.
[6]田勇,胡贤磊,王昭东,等. 2500 mm中厚板精轧机支承辊辊型优化设计[J].钢铁研究学院,2006,18(10):23-24.
[7]戚向东,连家创. 1780热连轧机精轧机组辊型优化技术的设计及应用[J].钢铁,2010,(10):68.
[8]陈华昶,吴首民.辊型优化技术的发展及应用[J].冶金设备,2009,(2):66-69.
[9] N. Chakraborti, B. Siva Kumara, V. Satish Babu a, S. Moitra a, A. Mukhopadhyay b. Optimizingsurface profiles during hot rolling: A geneticalgorithms based multi-objective optimization[J].Computational Materials Science,2006(37):159-165.
[10] Yushi Miyake,Ikuo Yarita, Keiichi Hamada. Development of Hot Rolling Technology forImproving Strip Profile and Flatness[J]. Kawasaki Steel Technical Report,1985(12):1-14.
[11] Lars Nolle , Alun Armstrong , Adrian Hopgood ,Andrew Ware. Optimum Work Roll ProfileSelection in the Hot Rolling of Wide Steel Strip using Computational Intelligence[J]. LectureNotes in Computer Science,1999(1625):435-457.
[12] Lars Nolle. SASS applied to optimum work roll profile selection in the hot rolling of wide steel[J].Knowledge-Based Systems,2007(20):203-208.
[13] Arif.S. Malik, Ramana.V. Grandhi. A computational method to predict strip profile in rollingmills[J]. Journal of materials processing technology,2008(206):263-274.
[14] Sudipta Sikdar , Sabita Kumari. Neural network model of the profile of hot-rolled strip[J]. AdvManuf Technol,2009(42):450-462.
[15] Guo,Remn-Min,Schunk, Jerry H. Determination of optimal work roll crown for a hot strip mill[J].Iron and Steel Engineer,1989,66(8):52-60.
[16]耿玉磊,张翔.多目标优化的求解方法与发展[J].机电技术,2004,(z1):105-107.
[17]甘敏,彭辉,王勇.多目标优化与自适应惩罚的混合约束优化进化算法[J].控制与决策,2010,25(3):34-37.
[18]马小姝.传统多目标优化方法和多目标遗传算法的比较综述[J].电气传动自动化,2010,(3):47-48.
[19]刘真.差分进化算法在多目标优化中的应用研究[D].青岛:中国石油大学(华东)计算机科学与技术学科工学硕士论文,2010:37-43.
[20]崔逊学.基于多目标优化的进化算法研究[D].合肥:中国科学技术大学模式识别与智能系统学科工学博士论文,2001:15-17.
[21]王宏旭.四辊冷轧带材中张应力横向分布的求解[J].钢铁研究学报,1996,(1):14-15.
[22]刘立文.冷轧板形控制理论的发展[J].钢铁研究学院,1997,(6):22-23.
[23]刘洋.板带热连轧轧制力及其设定的研究[D].洛阳:河南科技大学材料加工工程学科工学硕士论文,2007:21-27.
[24]李长生,梅瑞斌,刘相华.刚塑性有限元求解板材轧制过程的奇异点问题[J].鞍钢技术,2010,(5):1-6.
[25]刘相华,白光润.万能孔型轧制H型件的刚塑性有限元分析[J].力学与实践,1987,(S1):81-83.
[26]李俊洪.板带轧机板形计算理论及其在宽带轧制中应用研究[D].秦皇岛:燕山大学机械设计及理论学科工学博士论文,2003:78-85.
[27]刘德义.三维弹塑性摩擦接触多级边界元法和四辊轧机轧制模拟[D].秦皇岛:燕山大学机械设计及理论学科工学博士论文,2003:55-59.
[28]李新星.非对称板材轧制压力分布的研究[D].武汉:武汉科技大学机械电子工程学科工学硕士论文,2010:27-31.
[29]王英睿,刘宏民.基于流面条元法的热带钢连轧仿真分析[J].塑性工程学报,2003,10(3):53-55.
[30]童云春.辊颈受均布载荷的四辊轧机辊系变形探讨[D].武汉:武汉科技大学机械电子工程学科工学硕士论文,2008:41-43.
[31]李新文.热连轧机板形控制系统分析研究[D].沈阳:东北大学材料加工工程学科工学硕士论文,2009:29-34.
[32]连家创,刘宏民.板厚板形控制[M].北京:兵器工业出版社.1998:32-45.
[33]丁开荣.板形标准曲线的理论计算方法研究[D].秦皇岛:燕山大学机械设计及理论学科工学博士论文,2006:63-67.
[34]王淑华.宽带钢热连轧板形控制参数优化[D].唐山:河北理工大学材料加工工程学科工学硕士论文,2007:26-33.
[35]戚向东.宝钢新建板带轧机轧制规程及机型选择的研究[D],秦皇岛:燕山大学机械设计及理论学科工学博士论文,2002:58-67.
[36]宋司兵.闭式机架优化设计研究初步[J].一重技术,2007,(1):21-23.
[37] Guo Defu,Shao Jian,He Anrui,Yang Quan,Liao Zhi,Wang Shende. Multi-objective and satisfactoryoptimization design of backup roll chamfer parameters for hot strip mills[J]. MechanicAutomation and Control Engineering,2010:5467-5471.
[38] Wei Guang Cheng,Cao Jian Guo, Zhang Jie,Hao Jian Wei. Optimization and application of CVCwork roll contour on 2250 hot strip mills[J]. Zhongnan Daxue Xuebao,2007,38(5):937-942.
[39] R. Nandan, R. Rai, R. Jayakanth, S. Moitra, and N. Chakraborti. Regulating Crown and FlatnessDuring Hot Rolling: A Multiobjective Optimization Study Using Genetic Algorithms[J]. Materialsand Manufacturing Processes,2005(20):459-478.
[40]黄庆学,肖宏,孙斌煜.轧钢机械设计[M].北京:冶金工业出版社,2007:32-41.
[41]陈庆军,康永林,王黎明,等.中厚板粗轧机轧辊强度的有限元模拟[J].计算机辅助工程,2006,15(9)：441-442.
[42]杜凤山,黄华贵,许志强.大型非均质轧辊辊间接触应力分布规律的研究[J].工程力学,2006,23(7):176-177.
[43]张玲娟.中厚板轧机工作辊疲劳断裂测算算法研究[J].河北冶金,2009,(3):7-9.
[44]刘国志,苗晨,杜翼辰.一个与Powell搜索相结合的混合免疫进化算法[J].长春理工大学学报,2010,(2):121-124.
[45]李志,陈年生,郭小珊,等.粒子群算法及其改进技术研究[J].湖北师范学院学报,2011,31(2):104-105.
[46]孙慧.基于PSO的组合评价算法研究[D].天津:天津大学工程管理学科工学硕士论文,2008:34-39.
[47]王伯成,施锦丹,王凯.粒子群优化算法的研究现状与发展概述[J].电视技术,2008,48(5):7-10.
[48]章万国.粒子群优化算法[J].计算机应用研究,2003,(12):52-53.
[49]刘春晓.基于SOM和PSO的聚类算法研究[D].成都:西南交通大学计算机应用技术学科工学硕士论文,2009:54-62.
[50]何志华.基于非线性规划的微粒群算法研究[J].计算机仿真,2010,27(9):212-214.
[51]潘章明.自适应双群微粒优化算法[J].计算机应用与软件,2010,27(7):239-241.
[2]王钢.热连轧带钢轧机辊系优化设计[J].重型机械，1985,(1):1-7.
[3]邹家祥.轧钢机械[M].北京:冶金工业出版社.2007:97-104.
[4]王琍,郭希学. HC轧机辊系优化设计[J].太原重型机械学院学报,1996,17(1):6-9.
[5]王伟.板带冷轧机机型选择与辊系参数优化理论及其应用[D].秦皇岛:燕山大学机械设计及理论学科工学博士论文,1999:24-32.
[6]田勇,胡贤磊,王昭东,等. 2500 mm中厚板精轧机支承辊辊型优化设计[J].钢铁研究学院,2006,18(10):23-24.
[7]戚向东,连家创. 1780热连轧机精轧机组辊型优化技术的设计及应用[J].钢铁,2010,(10):68.
[8]陈华昶,吴首民.辊型优化技术的发展及应用[J].冶金设备,2009,(2):66-69.
[9] N. Chakraborti, B. Siva Kumara, V. Satish Babu a, S. Moitra a, A. Mukhopadhyay b. Optimizingsurface profiles during hot rolling: A geneticalgorithms based multi-objective optimization[J].Computational Materials Science,2006(37):159-165.
[10] Yushi Miyake,Ikuo Yarita, Keiichi Hamada. Development of Hot Rolling Technology forImproving Strip Profile and Flatness[J]. Kawasaki Steel Technical Report,1985(12):1-14.
[11] Lars Nolle , Alun Armstrong , Adrian Hopgood ,Andrew Ware. Optimum Work Roll ProfileSelection in the Hot Rolling of Wide Steel Strip using Computational Intelligence[J]. LectureNotes in Computer Science,1999(1625):435-457.
[12] Lars Nolle. SASS applied to optimum work roll profile selection in the hot rolling of wide steel[J].Knowledge-Based Systems,2007(20):203-208.
[13] Arif.S. Malik, Ramana.V. Grandhi. A computational method to predict strip profile in rollingmills[J]. Journal of materials processing technology,2008(206):263-274.
[14] Sudipta Sikdar , Sabita Kumari. Neural network model of the profile of hot-rolled strip[J]. AdvManuf Technol,2009(42):450-462.
[15] Guo,Remn-Min,Schunk, Jerry H. Determination of optimal work roll crown for a hot strip mill[J].Iron and Steel Engineer,1989,66(8):52-60.
[16]耿玉磊,张翔.多目标优化的求解方法与发展[J].机电技术,2004,(z1):105-107.
[17]甘敏,彭辉,王勇.多目标优化与自适应惩罚的混合约束优化进化算法[J].控制与决策,2010,25(3):34-37.
[18]马小姝.传统多目标优化方法和多目标遗传算法的比较综述[J].电气传动自动化,2010,(3):47-48.
[19]刘真.差分进化算法在多目标优化中的应用研究[D].青岛:中国石油大学(华东)计算机科学与技术学科工学硕士论文,2010:37-43.
[20]崔逊学.基于多目标优化的进化算法研究[D].合肥:中国科学技术大学模式识别与智能系统学科工学博士论文,2001:15-17.
[21]王宏旭.四辊冷轧带材中张应力横向分布的求解[J].钢铁研究学报,1996,(1):14-15.
[22]刘立文.冷轧板形控制理论的发展[J].钢铁研究学院,1997,(6):22-23.
[23]刘洋.板带热连轧轧制力及其设定的研究[D].洛阳:河南科技大学材料加工工程学科工学硕士论文,2007:21-27.
[24]李长生,梅瑞斌,刘相华.刚塑性有限元求解板材轧制过程的奇异点问题[J].鞍钢技术,2010,(5):1-6.
[25]刘相华,白光润.万能孔型轧制H型件的刚塑性有限元分析[J].力学与实践,1987,(S1):81-83.
[26]李俊洪.板带轧机板形计算理论及其在宽带轧制中应用研究[D].秦皇岛:燕山大学机械设计及理论学科工学博士论文,2003:78-85.
[27]刘德义.三维弹塑性摩擦接触多级边界元法和四辊轧机轧制模拟[D].秦皇岛:燕山大学机械设计及理论学科工学博士论文,2003:55-59.
[28]李新星.非对称板材轧制压力分布的研究[D].武汉:武汉科技大学机械电子工程学科工学硕士论文,2010:27-31.
[29]王英睿,刘宏民.基于流面条元法的热带钢连轧仿真分析[J].塑性工程学报,2003,10(3):53-55.
[30]童云春.辊颈受均布载荷的四辊轧机辊系变形探讨[D].武汉:武汉科技大学机械电子工程学科工学硕士论文,2008:41-43.
[31]李新文.热连轧机板形控制系统分析研究[D].沈阳:东北大学材料加工工程学科工学硕士论文,2009:29-34.
[32]连家创,刘宏民.板厚板形控制[M].北京:兵器工业出版社.1998:32-45.
[33]丁开荣.板形标准曲线的理论计算方法研究[D].秦皇岛:燕山大学机械设计及理论学科工学博士论文,2006:63-67.
[34]王淑华.宽带钢热连轧板形控制参数优化[D].唐山:河北理工大学材料加工工程学科工学硕士论文,2007:26-33.
[35]戚向东.宝钢新建板带轧机轧制规程及机型选择的研究[D],秦皇岛:燕山大学机械设计及理论学科工学博士论文,2002:58-67.
[36]宋司兵.闭式机架优化设计研究初步[J].一重技术,2007,(1):21-23.
[37] Guo Defu,Shao Jian,He Anrui,Yang Quan,Liao Zhi,Wang Shende. Multi-objective and satisfactoryoptimization design of backup roll chamfer parameters for hot strip mills[J]. MechanicAutomation and Control Engineering,2010:5467-5471.
[38] Wei Guang Cheng,Cao Jian Guo, Zhang Jie,Hao Jian Wei. Optimization and application of CVCwork roll contour on 2250 hot strip mills[J]. Zhongnan Daxue Xuebao,2007,38(5):937-942.
[39] R. Nandan, R. Rai, R. Jayakanth, S. Moitra, and N. Chakraborti. Regulating Crown and FlatnessDuring Hot Rolling: A Multiobjective Optimization Study Using Genetic Algorithms[J]. Materialsand Manufacturing Processes,2005(20):459-478.
[40]黄庆学,肖宏,孙斌煜.轧钢机械设计[M].北京:冶金工业出版社,2007:32-41.
[41]陈庆军,康永林,王黎明,等.中厚板粗轧机轧辊强度的有限元模拟[J].计算机辅助工程,2006,15(9)：441-442.
[42]杜凤山,黄华贵,许志强.大型非均质轧辊辊间接触应力分布规律的研究[J].工程力学,2006,23(7):176-177.
[43]张玲娟.中厚板轧机工作辊疲劳断裂测算算法研究[J].河北冶金,2009,(3):7-9.
[44]刘国志,苗晨,杜翼辰.一个与Powell搜索相结合的混合免疫进化算法[J].长春理工大学学报,2010,(2):121-124.
[45]李志,陈年生,郭小珊,等.粒子群算法及其改进技术研究[J].湖北师范学院学报,2011,31(2):104-105.
[46]孙慧.基于PSO的组合评价算法研究[D].天津:天津大学工程管理学科工学硕士论文,2008:34-39.
[47]王伯成,施锦丹,王凯.粒子群优化算法的研究现状与发展概述[J].电视技术,2008,48(5):7-10.
[48]章万国.粒子群优化算法[J].计算机应用研究,2003,(12):52-53.
[49]刘春晓.基于SOM和PSO的聚类算法研究[D].成都:西南交通大学计算机应用技术学科工学硕士论文,2009:54-62.
[50]何志华.基于非线性规划的微粒群算法研究[J].计算机仿真,2010,27(9):212-214.
[51]潘章明.自适应双群微粒优化算法[J].计算机应用与软件,2010,27(7):239-241.