钢铁企业板坯QM合同匹配与负荷分配问题的研究
|
摘要
随着经济一体化进程的不断加快,钢铁企业作为国家经济命脉上的重要环节,在国家经济体系中扮演着重要的角色。目前国际原料供应日趋紧张,致使许多钢铁企业产品成本居高不下,利润空间越来越小。然而,钢铁企业难以再通过传统的手段来降低产品的成本。因此,如何利用科学的管理思想来降低钢铁企业的生产成本,成为近年来钢铁企业提高其产品市场竞争力关注的焦点。
本文以降低生产成本为目的,对钢铁企业热轧板坯QM合同问题和热连轧精轧负荷分配问题进行了研究,具体内容概括如下。
(1)热轧板坯QM合同匹配问题。
热轧QM合同与普通的期货合同不同,它代表一种make-to-stock的生产方式。与常见的无委托板坯匹配问题相比,热轧板坯QM合同匹配需要预测在库无委托板坯的期货配成率、考虑控制库存水平和设计虚拟合同。本文对钢铁企业热轧板坯QM合同匹配问题进行了细致的分析,提出了无委托板坯期货配成率计算方法及无委托板坯库存计算方法,并根据热轧板坯QM合同匹配问题的特点,分别建立了无委托板坯确定产线模型和热轧QM合同聚类优化模型。本文所建立的模型均未在公开发表的文献中见到过。并且,对于热轧QM合同聚类优化问题,本文首次提出了一个基于启发式的多项式时间最优化算法。最后,本文基于所提出的模型与算法,开发了热轧板坯QM合同匹配自动优化决策支持系统,为钢铁企业的热轧板坯QM合同匹配过程提供了科学、便捷、可靠的决策支持。
(2)热连轧精轧负荷分配问题。
钢铁企业热连轧精轧负荷分配问题是钢铁企业为了适应现代市场向低成本市场转变,积极响应国家“建设资源节约型社会”的号召而做的生产过程中精轧机组能耗(负荷)分配方法的研究。本文基于钢铁企业热连轧精轧机组负荷分配问题的目标、生产实际约束和实际生产中出现的问题,在经典轧制理论的基础上,提出了热连轧精轧机组负荷分配问题的优化模型。针对钢铁企业热连轧精轧机组负荷分配问题的特点,本文采用差分进化算法进行求解,并提出了一个线性变换的解码规则和一个基于启发式的修复策略。本文在标准差分进化算法的框架上,通过引入自适应交义概率递增策略和带扰动的局部搜索策略,提出了自适应差分下降算法用来改善标准差分进化算法的求解效果。最后,本文将所采用的智能优化算法与基于Know-how知识的人工负荷分配算法进行了比较,验证了算法的有效性。
Along with the progress of economic integration, steel enterprises, as one of the most essential parts for the economic development in our country, paly an important role in the national economic system. Nowadays, the increasing high product costs and low profit for steel enterprises are attributed to the rising price of the intenational raw materials. Steel enterprises, however, can no longer reduce the production costs via traditional methods. Therefore, a scientific and advanced method to regulate the production costs has been welcomed by steel enterprices during recent years.
The main content for this dissertation, for the purpose of reducing the production costs in steel enterprises, are QM-order slabs matching problem and load allocation problem in hot strip finish rolling.
(1) QM-order slabs matching problem.
Being different from the common orders, QM-orders represent the marke-to-stock mode of production. Comparing to the open-order slabs matching problem, QM-order slabs matching problem have its distict characteristcs. At first, it requires the sales probability of all slabs in storage. Moreover, it needs to consider the storage level factors. At last, it asks to design the virtual contracts. After minutely studied the QM-order slabs matching problem, this dissertation proposes a method to calculate the sales probility of open-order slabs and another method to reckon the storage level for open-order slabs. This dissertaition also establishes a model to choose the hot-rolling plant for open-order slabs and another model to cluster these QM orders. Futhermore, this dissertation puts forward a heuristic-based optimization algorithm to cluster QM orders, which belongs to polynomial time algorithm. At last, based on the models and algorithm mentioned above, this dissertation developes a QM-order slabs matching decision support system, from which steel enterprises are able to get credibile and expedient decision supports.
(2) Load allocation problem in hot strip finishing rolling.
The load allocation problem in hot strip finish rolling comes from the idear to adapt to the recent trends in modern market, which grow toward to low-cost product. This idear is also a positive respsiveness to the national exhortation to build an economic society. After carefully analysed the charactics and the difficulties encounted in hot strip finishing rolling, this dissertation designs an optimization model to solve this problem. This model is based on the classic rolling theory, and reflects the typical characertstics of load allocation problem in hot strip finish rolling, such as its objective, subjects, and so on. In order to surmount some difficulties of this problem, such as the characteristics of high-order non-linear, continuous variables and the involute subjects, this dissertation uses differential evolution algorithm to solve this problem. This dissertation also proposes a decode-rule, which based on linear transformation, and a heuristic repair strategy, which based on linear transformation, to make differential evolution algorithm being suitable for this load allocation problem. Moreover, this dissertation designs an intelligent optimization algorithm based on differential evolution algorithm, adaptive differential descent algorithm, to obtain an effective computational result. This new algotithm adds an adaptive and increasing crossover probability strategy and a local search with disturbance strategy on the typical differential evolution algorithm. At last, this dissertation verifies the efficienty of those algorithms, by comparing these algorithms with the artificial algorithm based on the Know-how knowledge.
本文以降低生产成本为目的,对钢铁企业热轧板坯QM合同问题和热连轧精轧负荷分配问题进行了研究,具体内容概括如下。
(1)热轧板坯QM合同匹配问题。
热轧QM合同与普通的期货合同不同,它代表一种make-to-stock的生产方式。与常见的无委托板坯匹配问题相比,热轧板坯QM合同匹配需要预测在库无委托板坯的期货配成率、考虑控制库存水平和设计虚拟合同。本文对钢铁企业热轧板坯QM合同匹配问题进行了细致的分析,提出了无委托板坯期货配成率计算方法及无委托板坯库存计算方法,并根据热轧板坯QM合同匹配问题的特点,分别建立了无委托板坯确定产线模型和热轧QM合同聚类优化模型。本文所建立的模型均未在公开发表的文献中见到过。并且,对于热轧QM合同聚类优化问题,本文首次提出了一个基于启发式的多项式时间最优化算法。最后,本文基于所提出的模型与算法,开发了热轧板坯QM合同匹配自动优化决策支持系统,为钢铁企业的热轧板坯QM合同匹配过程提供了科学、便捷、可靠的决策支持。
(2)热连轧精轧负荷分配问题。
钢铁企业热连轧精轧负荷分配问题是钢铁企业为了适应现代市场向低成本市场转变,积极响应国家“建设资源节约型社会”的号召而做的生产过程中精轧机组能耗(负荷)分配方法的研究。本文基于钢铁企业热连轧精轧机组负荷分配问题的目标、生产实际约束和实际生产中出现的问题,在经典轧制理论的基础上,提出了热连轧精轧机组负荷分配问题的优化模型。针对钢铁企业热连轧精轧机组负荷分配问题的特点,本文采用差分进化算法进行求解,并提出了一个线性变换的解码规则和一个基于启发式的修复策略。本文在标准差分进化算法的框架上,通过引入自适应交义概率递增策略和带扰动的局部搜索策略,提出了自适应差分下降算法用来改善标准差分进化算法的求解效果。最后,本文将所采用的智能优化算法与基于Know-how知识的人工负荷分配算法进行了比较,验证了算法的有效性。
Along with the progress of economic integration, steel enterprises, as one of the most essential parts for the economic development in our country, paly an important role in the national economic system. Nowadays, the increasing high product costs and low profit for steel enterprises are attributed to the rising price of the intenational raw materials. Steel enterprises, however, can no longer reduce the production costs via traditional methods. Therefore, a scientific and advanced method to regulate the production costs has been welcomed by steel enterprices during recent years.
The main content for this dissertation, for the purpose of reducing the production costs in steel enterprises, are QM-order slabs matching problem and load allocation problem in hot strip finish rolling.
(1) QM-order slabs matching problem.
Being different from the common orders, QM-orders represent the marke-to-stock mode of production. Comparing to the open-order slabs matching problem, QM-order slabs matching problem have its distict characteristcs. At first, it requires the sales probability of all slabs in storage. Moreover, it needs to consider the storage level factors. At last, it asks to design the virtual contracts. After minutely studied the QM-order slabs matching problem, this dissertation proposes a method to calculate the sales probility of open-order slabs and another method to reckon the storage level for open-order slabs. This dissertaition also establishes a model to choose the hot-rolling plant for open-order slabs and another model to cluster these QM orders. Futhermore, this dissertation puts forward a heuristic-based optimization algorithm to cluster QM orders, which belongs to polynomial time algorithm. At last, based on the models and algorithm mentioned above, this dissertation developes a QM-order slabs matching decision support system, from which steel enterprises are able to get credibile and expedient decision supports.
(2) Load allocation problem in hot strip finishing rolling.
The load allocation problem in hot strip finish rolling comes from the idear to adapt to the recent trends in modern market, which grow toward to low-cost product. This idear is also a positive respsiveness to the national exhortation to build an economic society. After carefully analysed the charactics and the difficulties encounted in hot strip finishing rolling, this dissertation designs an optimization model to solve this problem. This model is based on the classic rolling theory, and reflects the typical characertstics of load allocation problem in hot strip finish rolling, such as its objective, subjects, and so on. In order to surmount some difficulties of this problem, such as the characteristics of high-order non-linear, continuous variables and the involute subjects, this dissertation uses differential evolution algorithm to solve this problem. This dissertation also proposes a decode-rule, which based on linear transformation, and a heuristic repair strategy, which based on linear transformation, to make differential evolution algorithm being suitable for this load allocation problem. Moreover, this dissertation designs an intelligent optimization algorithm based on differential evolution algorithm, adaptive differential descent algorithm, to obtain an effective computational result. This new algotithm adds an adaptive and increasing crossover probability strategy and a local search with disturbance strategy on the typical differential evolution algorithm. At last, this dissertation verifies the efficienty of those algorithms, by comparing these algorithms with the artificial algorithm based on the Know-how knowledge.
引文
[1]F.J. Vasko, F.E. Wolf, K.L. Stott,1989. A set covering approach to metallurgical grade assignment [J]. European Journal of Operational Research,38:27-34.
[2]F.Q. Yang, T.L. Sun, C.H. Zhang,2009. An efficient hybrid data clustering method based on K-harmonic means and Particle Swarm Optimization [J]. Expert Systems with Applications,36(6):9847-9852.
[3]H. Manuel, I. Joaquin, M. Idel, G.-A. Juan, V.R. Jose,2009. Identification of surgical practice patterns using evolutionary cluster analysis [J]. Mathematical and Computer Modelling, In Press.
[4]Y.-K. Hu, Y.P. Hu.,2007. Global optimization in clustering using hyperbolic cross points [J]. Pattern Recognition,40(6):1722-1733.
[5]A. Ortega, S. Marco, T. sundic, J. Samitier,2000. New pattern recognition systems designed for electronic noses [J]. Sensors and Actuators B:Chemical,69(3):302-307.
[6]叶潇,胡建华,王清心,2008.对多路数组聚集方法的研究及其优化[J],武汉理工大学学报(信息与管理工程版),30(1)：93-96.
[7]罗彪,郑金华,2008.多目标进化算法中基于动态聚集距离的分布性保持策略[J],计算机应用研究,25(10)：2934-2938.
[8]钱开余,孙发国,2006.数据仓库的聚集优化[J],信息系统,5：39-41.
[9]梁国平,1980.关于轧机的最佳负荷分配问题[J],钢铁,15(1)：42-48.
[11]刘战英,1996.轧制变形规程优化设计[M],北京：冶金工业出版社：182-186.
[12]白埃民,周和敏,2001.综合等负荷函数的逆向算法[J],钢铁,36(1)：38-41.
[13]王国栋,刘相华,2000.金属轧制过程人工智能优化[M].北京：冶金工业出版社21-412
[14]王萍,邹美娟,李胜祗,黄贞益,2001.热连轧带钢负荷分配法的进展[J],安徽工业大学学报,18(4)：295-299.
[15]R. Storn, K. Price,1995. Differential evolution-A simple and efficient adaptive scheme for global optimization over continuous spaces [J]. Technical Report, International Computer Science Institute, (8):22-25.
[16]K. E. Parsopoulos, D. K. Tasoulis, N. G. Pavlidis,2004. V. P. Plagianakos, M. N. Vrahatis. Vector evaluated differential evolution for multiobjective optimization [A]. IEEE Congress on Evolutionary Computation, Portland,1:204-211.
[17]B. V. Babu, M. M. L. Jehan,2003. Differential evolution for multi-objective optimation [A]. IEEE Congress on Evolution Computation, Canberra,4:2696-2703.
[18]G. Onwubolu, D. Davendra,2006. Scheduling flow shops using differential evolution algorithm [J]. European Journal of Operational Research,171(2):674-692.
[19]S. Ay din, H. Temeltas,2003. Time optimal trajectory planning for mobile robots by differential evolution algorithm and neural networks [A].2003 IEEE International Symposium on Inderstrial Electronics,1:352-357.
[20]N. Chakraborti, A. Kumar,2003. The optimal scheduling of a reversing strip mill: Studies using multipopulation genetic algorithms and differential evolution [J]. Materials and Manufacturing Processes,18(3):433-445.
[21]付晓刚,俞金寿,2009.基于混沌迁徙策略的多种群差分进化算法[J],华东理工大学学报(自然科学版),35(2)：308-312.
[22]黄福令,高慧敏,2009.基于文化算法和改进差分算法的混合算法[J],计算机应用,29(5)：1264-1269.
[23]李昕,颜学峰,2009.基于群体信息挖掘的协同差分算法及应用[J],化工进展,28(5)：778-783.
[24]谭跃,谭冠政,2009.具有混沌局部搜索策略的差分进化全局优化算法[J],计算机工程与应用,45(14)：15-17.
[25]邓泽喜,刘晓冀,2008.差分进化算法的交叉概率因子递增策略研究[J],计算机工程与应用,44(27)：33-36.
[26]周艳平,顾幸生,2007.差分进化算法研究进展[J],化工自动化及仪表,34(3)：1-5.
[27]孙一康,2007.带钢热连轧的模型与控制[M],北京：冶金工业出版社：38-49.
[28]祝孔林,单旭沂,2007.精轧负荷分配自动实现技术及应用[J],钢铁,42(1)：45-49.
[29]薛展,2007.精轧轧制规程的建模与优化[D],上海：上海交通大学.
[30]单旭沂,1998.一热轧和二热轧机组精轧数模剖析[J],宝钢技术,1：33-37.
[31]吕程,赵启林,林立忠,刘相华,王国栋,2001.热连轧精轧机组负荷分配的优化算法[J],钢铁研究学报,13(1)：26-29.
[32]赵适宜,2005.基于改进遗传算法的热连轧精轧机组负荷分配优化研究[D],沈阳：东北大学.
[33]张志勇,2005.宝钢热轧生产计划编制方法的研究与应用[D],大连：大连理工大学.
[34]J.M. YANG, H.J. CHE, F.P. Dou, T. ZHOU,2008. Genetic algorithm-based optimization used in rolling schedule [J], Journal of iron and steel research, international,15(2): 18-22
[35]朱涛,吕敬东,向阳,康明,常军,2002.热轧板凸度控制研究[J],钢铁钒钛,23(2)：23-28.
[36]胡浩平,伍清河,费庆,卢春宁,2006.热轧中板扎变形规程的优化设计[J],北京理工大学学报,26(7)：569-573.
[37]C.N. Patel,2002. Differential evolution A method of global optimization[D], USA:The University of Texas at Arlington.
[38]吴亮红,王耀南,袁小芳,周少武,2006.自适应二次变异差分进化算法[J],控制与决策,40(1)：53-56.
[39]S. Sayah, K. Zehar,2008. Modified differential evolution algorithm for optimal power flow with non-smooth cost functions [J]. Energy Conversion and Management,49(11): 3036-3042.
[2]F.Q. Yang, T.L. Sun, C.H. Zhang,2009. An efficient hybrid data clustering method based on K-harmonic means and Particle Swarm Optimization [J]. Expert Systems with Applications,36(6):9847-9852.
[3]H. Manuel, I. Joaquin, M. Idel, G.-A. Juan, V.R. Jose,2009. Identification of surgical practice patterns using evolutionary cluster analysis [J]. Mathematical and Computer Modelling, In Press.
[4]Y.-K. Hu, Y.P. Hu.,2007. Global optimization in clustering using hyperbolic cross points [J]. Pattern Recognition,40(6):1722-1733.
[5]A. Ortega, S. Marco, T. sundic, J. Samitier,2000. New pattern recognition systems designed for electronic noses [J]. Sensors and Actuators B:Chemical,69(3):302-307.
[6]叶潇,胡建华,王清心,2008.对多路数组聚集方法的研究及其优化[J],武汉理工大学学报(信息与管理工程版),30(1)：93-96.
[7]罗彪,郑金华,2008.多目标进化算法中基于动态聚集距离的分布性保持策略[J],计算机应用研究,25(10)：2934-2938.
[8]钱开余,孙发国,2006.数据仓库的聚集优化[J],信息系统,5：39-41.
[9]梁国平,1980.关于轧机的最佳负荷分配问题[J],钢铁,15(1)：42-48.
[11]刘战英,1996.轧制变形规程优化设计[M],北京：冶金工业出版社：182-186.
[12]白埃民,周和敏,2001.综合等负荷函数的逆向算法[J],钢铁,36(1)：38-41.
[13]王国栋,刘相华,2000.金属轧制过程人工智能优化[M].北京：冶金工业出版社21-412
[14]王萍,邹美娟,李胜祗,黄贞益,2001.热连轧带钢负荷分配法的进展[J],安徽工业大学学报,18(4)：295-299.
[15]R. Storn, K. Price,1995. Differential evolution-A simple and efficient adaptive scheme for global optimization over continuous spaces [J]. Technical Report, International Computer Science Institute, (8):22-25.
[16]K. E. Parsopoulos, D. K. Tasoulis, N. G. Pavlidis,2004. V. P. Plagianakos, M. N. Vrahatis. Vector evaluated differential evolution for multiobjective optimization [A]. IEEE Congress on Evolutionary Computation, Portland,1:204-211.
[17]B. V. Babu, M. M. L. Jehan,2003. Differential evolution for multi-objective optimation [A]. IEEE Congress on Evolution Computation, Canberra,4:2696-2703.
[18]G. Onwubolu, D. Davendra,2006. Scheduling flow shops using differential evolution algorithm [J]. European Journal of Operational Research,171(2):674-692.
[19]S. Ay din, H. Temeltas,2003. Time optimal trajectory planning for mobile robots by differential evolution algorithm and neural networks [A].2003 IEEE International Symposium on Inderstrial Electronics,1:352-357.
[20]N. Chakraborti, A. Kumar,2003. The optimal scheduling of a reversing strip mill: Studies using multipopulation genetic algorithms and differential evolution [J]. Materials and Manufacturing Processes,18(3):433-445.
[21]付晓刚,俞金寿,2009.基于混沌迁徙策略的多种群差分进化算法[J],华东理工大学学报(自然科学版),35(2)：308-312.
[22]黄福令,高慧敏,2009.基于文化算法和改进差分算法的混合算法[J],计算机应用,29(5)：1264-1269.
[23]李昕,颜学峰,2009.基于群体信息挖掘的协同差分算法及应用[J],化工进展,28(5)：778-783.
[24]谭跃,谭冠政,2009.具有混沌局部搜索策略的差分进化全局优化算法[J],计算机工程与应用,45(14)：15-17.
[25]邓泽喜,刘晓冀,2008.差分进化算法的交叉概率因子递增策略研究[J],计算机工程与应用,44(27)：33-36.
[26]周艳平,顾幸生,2007.差分进化算法研究进展[J],化工自动化及仪表,34(3)：1-5.
[27]孙一康,2007.带钢热连轧的模型与控制[M],北京：冶金工业出版社：38-49.
[28]祝孔林,单旭沂,2007.精轧负荷分配自动实现技术及应用[J],钢铁,42(1)：45-49.
[29]薛展,2007.精轧轧制规程的建模与优化[D],上海：上海交通大学.
[30]单旭沂,1998.一热轧和二热轧机组精轧数模剖析[J],宝钢技术,1：33-37.
[31]吕程,赵启林,林立忠,刘相华,王国栋,2001.热连轧精轧机组负荷分配的优化算法[J],钢铁研究学报,13(1)：26-29.
[32]赵适宜,2005.基于改进遗传算法的热连轧精轧机组负荷分配优化研究[D],沈阳：东北大学.
[33]张志勇,2005.宝钢热轧生产计划编制方法的研究与应用[D],大连：大连理工大学.
[34]J.M. YANG, H.J. CHE, F.P. Dou, T. ZHOU,2008. Genetic algorithm-based optimization used in rolling schedule [J], Journal of iron and steel research, international,15(2): 18-22
[35]朱涛,吕敬东,向阳,康明,常军,2002.热轧板凸度控制研究[J],钢铁钒钛,23(2)：23-28.
[36]胡浩平,伍清河,费庆,卢春宁,2006.热轧中板扎变形规程的优化设计[J],北京理工大学学报,26(7)：569-573.
[37]C.N. Patel,2002. Differential evolution A method of global optimization[D], USA:The University of Texas at Arlington.
[38]吴亮红,王耀南,袁小芳,周少武,2006.自适应二次变异差分进化算法[J],控制与决策,40(1)：53-56.
[39]S. Sayah, K. Zehar,2008. Modified differential evolution algorithm for optimal power flow with non-smooth cost functions [J]. Energy Conversion and Management,49(11): 3036-3042.