提钒炼钢厂物料信息管理与预测系统的研究
|
摘要
炼钢是间歇和连续作业方式相混杂的多工序生产过程,其生产数据的采集与传送实时性强,工序间联系紧密,各道工序产生的生产实时数据多,生产状况千变万化,各基层与生产管理部门、生产指挥系统数据交换频度高。提钒炼钢厂生产工艺流程中的所有物料信息在时间、空间及工艺流程中,构建了企业内部最直接、最基本的信息流。因此为提钒炼钢厂提供准确的物料实时重量信息和消耗预测信息,对于理顺其物流、合理衔接各工序生产、最大程度发挥工序能力、及时调整产品结构、构筑信息化工程、实施精细化生产和精细化管理是十分重要的。
本文以提钒炼钢厂为对象,在研究了该厂复杂的工艺流程和现有生产管理系统的基础上,建立物料重量信息网络化采集管理与消耗预测系统,使达到自动采集与管理各个生产工序的物料重量信息、工艺数据和进行物料消耗预测的目的,方便操作人员的生产管理和调度。
系统的方案设计主要是根据生产工艺流程的特点,采用分片区管理的模式,在各个片区设置相应的片区采集站,采集各自片区的物料重量信息和相关生产参数,使工作人员及时掌握生产数据,方便生产过程跟踪和物料信息的管理。
系统主要研究并实现了物料称重数据传输、采集、管理和物料的消耗预测等功能。通过TCP/IP通信、串口通信、Windows Sockets通信、无线电台传输、射频识别等关键技术实现称重数据的传输,使重量信息准确快速地进入到采集站的计算机中;利用相关软件实现称重数据的自动采集与存储,以及相关数据的查询、统计、报表等功能;利用SQL Server 2000以及数据库维护计划和作业功能,进行生产数据的存储、读取及维护,以保证数据库长期稳定运行和数据的完整可靠,并且采用B/S模式,通过ASP技术实现重量信息的在线查询和统计;利用遗传算法改进神经网络,分析并建立了物料消耗预测模型,以准确实现对下一周期物料需求信息的预报。
该系统的使用,极大地减少了操作人员工作强度和人为因素的影响,提高了物料重量数据的准确性、实时性;实现了全厂工艺历史数据的综合查询与分析,方便了现场技术人员、企业管理人员进行全厂生产过程分析,很好地支持了炼钢厂生产管理和生产计划调度。
Steelmaking is a multi-procedure production which consists of intermittent and continuous work types. In the steelmaking, the real-time of production data acquisition and transmission is strong; the contacts among the procedures are tight; the real-time production data in each procedure are rich; the states of production are protean; the exchange frequency among grass-roots, production management Departments and production command system are high. In Vanadium Refining & Steel Plant, all the materials information is based on the process which builds the most direct and basic information flow of enterprise in time and space. Therefore, it’s very important to supply the accurate materials weight information and consumption forecast information for Vanadium Refining & Steel Plant so as to streamline its logistics, converge each procedure reasonably, bring the maximum capacity of procedure into play, adjust product structure in time, build information projects, implement the sophisticated production and management.
The paper takes the Refining & Steel Plant as its subject. Based on the research of complex production process and the existing management system, the paper regards the building of materials weight information system which is based on network management and forecast to implement automatic acquisition and management of materials weight information materials and process data of each procedure, support materials consumption forecast, and facilitate the operation people’s production management and scheduling as the purpose.
According to the characteristics of production process, the system adopts a sub-section management model, establishes respective collection stations in various production sections, which are used to collect materials weight information and interrelated production parameters in each section, so that the staff can master production data in time and facilitate the production process tracking and materials information management.
The system mainly researches and achieves some functions, including materials weighing data transmission, acquisition, management, and materials consumption forecast. The implementation of materials weighing data transmission is realized through the key technologies of TCP / IP communications, serial communications, Windows Sockets communications, radio transmission, radio frequency identification technology, so the weight information can enter the collecting station computers quickly and accurately; the weighting data can be collected and saved automatically, as well as the implementation of some functions, such as related data query, statistics and report forms, by some relevant software; the production data can be stored, accessed and maintained through the use of SQL Server 2000 database, including its maintenance plans and operations in order to guarantee the long-term stable operation of the database and the integrity and reliability of the data, and in addition the online query and statistics of weight information can be obtained by B/S model and ASP technology; an materials consumption forecast model is established according to the neural network improved by genetic algorithm in order to achieve accurate forecast of materials requirement Information for the next cycle.
After the system goes into effect, it greatly reduces the working strength of operational staff and the impact of human factors, increases the accuracy and real-time of materials weight data. It also realizes the query and analysis of historical data in the entire plant, facilitates the locale technical personnel and production management personnel to analyze the production process, which greatly supports production management, production planning and scheduling for steel plant.
本文以提钒炼钢厂为对象,在研究了该厂复杂的工艺流程和现有生产管理系统的基础上,建立物料重量信息网络化采集管理与消耗预测系统,使达到自动采集与管理各个生产工序的物料重量信息、工艺数据和进行物料消耗预测的目的,方便操作人员的生产管理和调度。
系统的方案设计主要是根据生产工艺流程的特点,采用分片区管理的模式,在各个片区设置相应的片区采集站,采集各自片区的物料重量信息和相关生产参数,使工作人员及时掌握生产数据,方便生产过程跟踪和物料信息的管理。
系统主要研究并实现了物料称重数据传输、采集、管理和物料的消耗预测等功能。通过TCP/IP通信、串口通信、Windows Sockets通信、无线电台传输、射频识别等关键技术实现称重数据的传输,使重量信息准确快速地进入到采集站的计算机中;利用相关软件实现称重数据的自动采集与存储,以及相关数据的查询、统计、报表等功能;利用SQL Server 2000以及数据库维护计划和作业功能,进行生产数据的存储、读取及维护,以保证数据库长期稳定运行和数据的完整可靠,并且采用B/S模式,通过ASP技术实现重量信息的在线查询和统计;利用遗传算法改进神经网络,分析并建立了物料消耗预测模型,以准确实现对下一周期物料需求信息的预报。
该系统的使用,极大地减少了操作人员工作强度和人为因素的影响,提高了物料重量数据的准确性、实时性;实现了全厂工艺历史数据的综合查询与分析,方便了现场技术人员、企业管理人员进行全厂生产过程分析,很好地支持了炼钢厂生产管理和生产计划调度。
Steelmaking is a multi-procedure production which consists of intermittent and continuous work types. In the steelmaking, the real-time of production data acquisition and transmission is strong; the contacts among the procedures are tight; the real-time production data in each procedure are rich; the states of production are protean; the exchange frequency among grass-roots, production management Departments and production command system are high. In Vanadium Refining & Steel Plant, all the materials information is based on the process which builds the most direct and basic information flow of enterprise in time and space. Therefore, it’s very important to supply the accurate materials weight information and consumption forecast information for Vanadium Refining & Steel Plant so as to streamline its logistics, converge each procedure reasonably, bring the maximum capacity of procedure into play, adjust product structure in time, build information projects, implement the sophisticated production and management.
The paper takes the Refining & Steel Plant as its subject. Based on the research of complex production process and the existing management system, the paper regards the building of materials weight information system which is based on network management and forecast to implement automatic acquisition and management of materials weight information materials and process data of each procedure, support materials consumption forecast, and facilitate the operation people’s production management and scheduling as the purpose.
According to the characteristics of production process, the system adopts a sub-section management model, establishes respective collection stations in various production sections, which are used to collect materials weight information and interrelated production parameters in each section, so that the staff can master production data in time and facilitate the production process tracking and materials information management.
The system mainly researches and achieves some functions, including materials weighing data transmission, acquisition, management, and materials consumption forecast. The implementation of materials weighing data transmission is realized through the key technologies of TCP / IP communications, serial communications, Windows Sockets communications, radio transmission, radio frequency identification technology, so the weight information can enter the collecting station computers quickly and accurately; the weighting data can be collected and saved automatically, as well as the implementation of some functions, such as related data query, statistics and report forms, by some relevant software; the production data can be stored, accessed and maintained through the use of SQL Server 2000 database, including its maintenance plans and operations in order to guarantee the long-term stable operation of the database and the integrity and reliability of the data, and in addition the online query and statistics of weight information can be obtained by B/S model and ASP technology; an materials consumption forecast model is established according to the neural network improved by genetic algorithm in order to achieve accurate forecast of materials requirement Information for the next cycle.
After the system goes into effect, it greatly reduces the working strength of operational staff and the impact of human factors, increases the accuracy and real-time of materials weight data. It also realizes the query and analysis of historical data in the entire plant, facilitates the locale technical personnel and production management personnel to analyze the production process, which greatly supports production management, production planning and scheduling for steel plant.
引文
[1]马竹语.冶金工业自动化[M].北京:机械工业出版社, 2007.
[2]孙彦广.支撑钢铁工业可持续发展的冶金自动化技术[J].冶金自动化, 2006, 2:1-4.
[3]漆永新.我国钢铁工业信息化与自动化发展评述[J].冶金管理, 2005, 4:46-48.
[4]孙彦广.冶金自动化技术现状和发展趋势[J].冶金自动化, 2004, 1:1-5.
[5]朱继民.新型工业化道路及其在首钢的实践[J].管理世界, 2006, 1:2-13.
[6]周一峰. SAP软件在钢铁企业ERP中的应用[J].上海钢研, 2006, 2:60-62.
[7]杜贵霞.刍议炼钢生产成本控制[N].冶金财会, 2007, 4:36.
[8]罗艳萍,段黎明,卢涛,王延伦.炼钢企业计算机辅助质量控制系统[J].重庆大学学报, 2003, 10:56-59.
[9]陈志讯,彭开香.炼钢—连铸过程计算机控制系统更新与改造[J].冶金设备, 2007, 5: 42-45.
[10]李成龙.攀钢炼钢过程计算机系统应用分析[J].冶金自动化, 2003(增刊):54-56
[11]陈莉.攀钢2号板坯连铸过程计算机系统介绍[J].控制与检测, 2004, 3:31-34.
[12]丛培勇,范玉青,刘忠献. ERP与MES集成关系研究[J].制造业自动化, 2007, 8:1-5.
[13]易向军,王雪晗,赵维军,章乐乐. MES在炼钢—中厚板生产计划信息管理中的应用[J].工业控制计算机, 2003, 9:1-3.
[14] SU Dong-ping, CHEN Wen-ming, LUO Shou-zhang, et al. Adjustment of planning and scheduling for steelmaking and continuous casting with man2machine interactive[J]. Metallurgical Industry Automation, 2005, 3:17-20.
[15] P. P. MOHANTY. An agent-oriented approach to resolve the production planning complexities for a modern steel manufacturing system[J]. International Journal Advance Manufacturing Technology, 2004, 24:199– 205.
[16] B.S.Sahay, K.B.C.Saxema, Ashish Kumar. World-Class Manufacturing and Information Age Competition[J]. Industrial Management, 2001, 5:134-146.
[17] Harjunkoski, Grossmann. A decomposition approach for the scheduling of a steel plant production[J]. Computers and Chemical Engineering, 2001.6:147-160
[18] Lixin Tang, Jiyin Liu, Aiying Rong, Zihou Yang. A mathematical programming model for scheduling steelmaking-continuous casting production[J]. European Journal of Operational Research, 2000, 120(2):423 - 435.
[19]何腊梅等.攀钢炼钢生产物流仿真分析[J].重庆大学学报, 2004年, 4:57-61.
[20]陈东奎,田乃媛.八钢炼钢厂计算机辅助生产管理系统数据库的设计[J].冶金自动化,2003, 6:10-13.
[21]王家军,赵建东,冯彦民.炼钢厂生产信息采集系统[J].计量管理, 2005, 11:48-50
[22]吴卫祖,陈谋文,孙永林.计算机网络技术基础[M].北京:清华大学出版社, 2006.
[23]黄燕,邹仁明,吴平.计算机网络教程[M].北京:人民邮电出版社, 2004.
[24]杨鑫,黄宏光,何玮华.基于RS-232的无线数据传输系统的实现[J].工艺与技术, 2006, 5:40-44.
[25]史建民,黄有方,嘉红霞. RS-232串口设备远程通讯功能的实现[J].起重运输机械, 2003, 6:30-32.
[26]李长林,高洁. Visual C++串口通信技术与典型实例[M].北京:清华大学出版社, 2006.
[27]郎锐,孙方. Visual C++网络通信程序开发基础及实例解析[M].北京:机械工业出版社,2006.
[28]张建斌.基于带内扩频技术的无线电台数字通信的设计与实现[J].电子科技, 2005, 12:5-8.
[29]吴永祥,曹珍贯,吴攀.基于无线电台的风井综合参数安全监控系统[J].煤矿安全, 2006, 12:37-38.
[30]李彩红.无线射频识别(RFID)技术及其应用[J].广东技术师范学院学报, 2006, 6:26-28.
[31] Finkenzeller K. RFID Handbook: Fundamentals and Applications in Contactless Smart Cards and Identification[M]. Second Edition. West Sussex Land: John Wiley and Sons, 2003.
[32]龙怀冰,李白桦,郭智博. SQL Server 2000实用开发教程[M].北京:人民邮电出版社,2006.
[33]李涛,刘凯奎,王永皎. Visual C++ SQL Server数据库开发与实例[M].北京:清华大学出版社, 2006.
[34]陈万米,何智伟. Microsoft SQL Server数据库安全备份和恢复策略[J].微型电脑应用, 2006, 1:55-56.
[35]任泰明.基于B/S结构的软件开发技术[M].西安:西安电子科技大学出版社, 2006.
[36]张秋颖.基于ASP、SQL Server技术的门户网站设计与实现[J].计算机与现代化, 2006, 12:125-128.
[37] Martin T.Hagan, Howard B.Demuth, Mark Beale. Neural Network Design[M].北京:机械工出版社, 2002.
[38]陈光华.人工神经网络在证券价格预测中的应用[J].计算机仿真, 2007, 10:244-248.
[39]玄光男,程润伟.遗传算法与工程优化[M].北京:清华大学出版社, 2004.
[40]李敏强等.遗传算法的基本理论与应用[M].北京:科学出版社,2002.
[41]杨宏韬,张德江,李秀兰,王秀英.遗传神经网络能耗预测模型在钢铁企业中的应用[J].长春工业大学学报(自然科学版), 2007, 28(增刊): 186-189.
[42]梁勤欧.基于遗传克隆选择算法优化BP神经网络的地理信息预测研究[J].测绘通报, 2007, 10:29-31.
[43] Paul S. Heckerling, Ben S. Gerber, Thomas G. Tape, Robert S. Wigton. Use of genetic algorithms for neural networks to predict community-acquired pneumonia[J]. Artificial Intelligence in Medicine, 2004, 30:71–84.
[44] Ying-Pin Chang, Chinyao Low. Optimization of a passive harmonic filter based on the neural-genetic algorithm with fuzzy logic for a steel manufacturing plant[J]. Expert Systems with Applications, 2008, 34:2059–2070.
[45] Michael X. Cohen, Jennifer Young, Jong-Min Baek, Christopher Kessler, Charan Ranganath. Individual differences in extraversion and dopamine genetics predict neural reward responses[J]. Cognitive Brain Research, 2005, 25:851– 861.
[46] Lin Xin-Bo, Zhang Zhi-Liang, Ruan Xue-Yu. Prediction of material flow stress in warm forging with BP neural network[J]. Journal of Shanghai Jiaotong University, 2002, 36(4): 459-462.
[47] Sadeghi, B H M. BP-neural network predictor model for plastic injection molding process[J]. Journal of Materials Processing Technology, 2001,103(3):411-416.
[48]张琳.基于GA-BP混合算法的转炉终点优化控制模型[D].重庆大学. 2004, 5.
[49]田慧欣.基于遗传算法和BP网络的LF炉终点温度预报[D].东北大学. 2005, 2.
[50]雷英杰,张善文,李续武,周创明. MATLAB遗传算法工具箱及应用[M].西安:西安电子科技大学, 2005.
[51]葛哲学,孙志强.神经网络理论与MATLAB R2007实现[M].北京:电子工业出版社, 2007.
[52]凌振华.钢铁企业能源管理信息系统的研究[D].中南大学. 2005, 5.
[53]张晶.基于MATCOM的VC++与MATLAB混合编程[J].自动化枝术与应用, 2007, 26(5):59-61.
[2]孙彦广.支撑钢铁工业可持续发展的冶金自动化技术[J].冶金自动化, 2006, 2:1-4.
[3]漆永新.我国钢铁工业信息化与自动化发展评述[J].冶金管理, 2005, 4:46-48.
[4]孙彦广.冶金自动化技术现状和发展趋势[J].冶金自动化, 2004, 1:1-5.
[5]朱继民.新型工业化道路及其在首钢的实践[J].管理世界, 2006, 1:2-13.
[6]周一峰. SAP软件在钢铁企业ERP中的应用[J].上海钢研, 2006, 2:60-62.
[7]杜贵霞.刍议炼钢生产成本控制[N].冶金财会, 2007, 4:36.
[8]罗艳萍,段黎明,卢涛,王延伦.炼钢企业计算机辅助质量控制系统[J].重庆大学学报, 2003, 10:56-59.
[9]陈志讯,彭开香.炼钢—连铸过程计算机控制系统更新与改造[J].冶金设备, 2007, 5: 42-45.
[10]李成龙.攀钢炼钢过程计算机系统应用分析[J].冶金自动化, 2003(增刊):54-56
[11]陈莉.攀钢2号板坯连铸过程计算机系统介绍[J].控制与检测, 2004, 3:31-34.
[12]丛培勇,范玉青,刘忠献. ERP与MES集成关系研究[J].制造业自动化, 2007, 8:1-5.
[13]易向军,王雪晗,赵维军,章乐乐. MES在炼钢—中厚板生产计划信息管理中的应用[J].工业控制计算机, 2003, 9:1-3.
[14] SU Dong-ping, CHEN Wen-ming, LUO Shou-zhang, et al. Adjustment of planning and scheduling for steelmaking and continuous casting with man2machine interactive[J]. Metallurgical Industry Automation, 2005, 3:17-20.
[15] P. P. MOHANTY. An agent-oriented approach to resolve the production planning complexities for a modern steel manufacturing system[J]. International Journal Advance Manufacturing Technology, 2004, 24:199– 205.
[16] B.S.Sahay, K.B.C.Saxema, Ashish Kumar. World-Class Manufacturing and Information Age Competition[J]. Industrial Management, 2001, 5:134-146.
[17] Harjunkoski, Grossmann. A decomposition approach for the scheduling of a steel plant production[J]. Computers and Chemical Engineering, 2001.6:147-160
[18] Lixin Tang, Jiyin Liu, Aiying Rong, Zihou Yang. A mathematical programming model for scheduling steelmaking-continuous casting production[J]. European Journal of Operational Research, 2000, 120(2):423 - 435.
[19]何腊梅等.攀钢炼钢生产物流仿真分析[J].重庆大学学报, 2004年, 4:57-61.
[20]陈东奎,田乃媛.八钢炼钢厂计算机辅助生产管理系统数据库的设计[J].冶金自动化,2003, 6:10-13.
[21]王家军,赵建东,冯彦民.炼钢厂生产信息采集系统[J].计量管理, 2005, 11:48-50
[22]吴卫祖,陈谋文,孙永林.计算机网络技术基础[M].北京:清华大学出版社, 2006.
[23]黄燕,邹仁明,吴平.计算机网络教程[M].北京:人民邮电出版社, 2004.
[24]杨鑫,黄宏光,何玮华.基于RS-232的无线数据传输系统的实现[J].工艺与技术, 2006, 5:40-44.
[25]史建民,黄有方,嘉红霞. RS-232串口设备远程通讯功能的实现[J].起重运输机械, 2003, 6:30-32.
[26]李长林,高洁. Visual C++串口通信技术与典型实例[M].北京:清华大学出版社, 2006.
[27]郎锐,孙方. Visual C++网络通信程序开发基础及实例解析[M].北京:机械工业出版社,2006.
[28]张建斌.基于带内扩频技术的无线电台数字通信的设计与实现[J].电子科技, 2005, 12:5-8.
[29]吴永祥,曹珍贯,吴攀.基于无线电台的风井综合参数安全监控系统[J].煤矿安全, 2006, 12:37-38.
[30]李彩红.无线射频识别(RFID)技术及其应用[J].广东技术师范学院学报, 2006, 6:26-28.
[31] Finkenzeller K. RFID Handbook: Fundamentals and Applications in Contactless Smart Cards and Identification[M]. Second Edition. West Sussex Land: John Wiley and Sons, 2003.
[32]龙怀冰,李白桦,郭智博. SQL Server 2000实用开发教程[M].北京:人民邮电出版社,2006.
[33]李涛,刘凯奎,王永皎. Visual C++ SQL Server数据库开发与实例[M].北京:清华大学出版社, 2006.
[34]陈万米,何智伟. Microsoft SQL Server数据库安全备份和恢复策略[J].微型电脑应用, 2006, 1:55-56.
[35]任泰明.基于B/S结构的软件开发技术[M].西安:西安电子科技大学出版社, 2006.
[36]张秋颖.基于ASP、SQL Server技术的门户网站设计与实现[J].计算机与现代化, 2006, 12:125-128.
[37] Martin T.Hagan, Howard B.Demuth, Mark Beale. Neural Network Design[M].北京:机械工出版社, 2002.
[38]陈光华.人工神经网络在证券价格预测中的应用[J].计算机仿真, 2007, 10:244-248.
[39]玄光男,程润伟.遗传算法与工程优化[M].北京:清华大学出版社, 2004.
[40]李敏强等.遗传算法的基本理论与应用[M].北京:科学出版社,2002.
[41]杨宏韬,张德江,李秀兰,王秀英.遗传神经网络能耗预测模型在钢铁企业中的应用[J].长春工业大学学报(自然科学版), 2007, 28(增刊): 186-189.
[42]梁勤欧.基于遗传克隆选择算法优化BP神经网络的地理信息预测研究[J].测绘通报, 2007, 10:29-31.
[43] Paul S. Heckerling, Ben S. Gerber, Thomas G. Tape, Robert S. Wigton. Use of genetic algorithms for neural networks to predict community-acquired pneumonia[J]. Artificial Intelligence in Medicine, 2004, 30:71–84.
[44] Ying-Pin Chang, Chinyao Low. Optimization of a passive harmonic filter based on the neural-genetic algorithm with fuzzy logic for a steel manufacturing plant[J]. Expert Systems with Applications, 2008, 34:2059–2070.
[45] Michael X. Cohen, Jennifer Young, Jong-Min Baek, Christopher Kessler, Charan Ranganath. Individual differences in extraversion and dopamine genetics predict neural reward responses[J]. Cognitive Brain Research, 2005, 25:851– 861.
[46] Lin Xin-Bo, Zhang Zhi-Liang, Ruan Xue-Yu. Prediction of material flow stress in warm forging with BP neural network[J]. Journal of Shanghai Jiaotong University, 2002, 36(4): 459-462.
[47] Sadeghi, B H M. BP-neural network predictor model for plastic injection molding process[J]. Journal of Materials Processing Technology, 2001,103(3):411-416.
[48]张琳.基于GA-BP混合算法的转炉终点优化控制模型[D].重庆大学. 2004, 5.
[49]田慧欣.基于遗传算法和BP网络的LF炉终点温度预报[D].东北大学. 2005, 2.
[50]雷英杰,张善文,李续武,周创明. MATLAB遗传算法工具箱及应用[M].西安:西安电子科技大学, 2005.
[51]葛哲学,孙志强.神经网络理论与MATLAB R2007实现[M].北京:电子工业出版社, 2007.
[52]凌振华.钢铁企业能源管理信息系统的研究[D].中南大学. 2005, 5.
[53]张晶.基于MATCOM的VC++与MATLAB混合编程[J].自动化枝术与应用, 2007, 26(5):59-61.