用户名: 密码: 验证码:
多变复杂锌精矿湿法炼锌信息系统
详细信息    本馆镜像全文|  推荐本文 |  |   获取CNKI官网全文
摘要
湿法炼锌是一个流程较长的冶金过程而且工艺比较成熟,包括锌精矿的焙烧、浸出、净化、电解、阴极锌的熔铸等过程。论文首先叙述了锌精矿种类,沸腾炉的焙烧过程,常规浸出各种影响因素及电解沉积锌的经济技术指标,为后续章节信息系统的开发奠定了基础。
     本文主要研究锌精矿的焙烧、浸出、电解三个过程,首先建立了锌精矿的配矿信息系统。该系统能提供精确的配矿和符合生产条件下配矿成本最低化,系统中优化了十种矿样的45种组合,而且全部数据输入和输出都使用数据库操作。然后根据锌精矿沸腾炉硫态化焙烧原理建立了热力学模型的信息系统,根据物料平衡和热平衡建立方程组来确定焙烧矿以及烟气的成分。由于焙烧过程的时间变化性,该热力学模型无法全部描述的内在机制等,使得结果存在着一定的偏差,但此模型能够较好地拟合焙烧过程的主要趋势。最后,在锌常规浸出和过电解沉积锌过程中设计了BP神经网络来预测浸出过程中的浸出率、浸出渣率、浸出液上清率、新液合格率、渣含水以及电解过程中的电流效率等因素。网络采用了近30组的训练样本,样本数据范围大,网络的训练误差精度可以达到10~(-5)。新的嫁接BP神经网络预测适应性较广、精度较高。可以实现离线预测,并且为在线操作提供了参数指标。
     湿法炼锌信息系统的程序采用了Visual Basic 6.0和Matlab两种语言混合编写,系统的数据库采用Microsoft Access创建和维护。Visual Basic 6.0编程语言简单实用,可视化功能强大,具有严密的封装性,而且还提供许多ActiveX控件;Matlab编程语言不仅有较强的矩阵运算功能和绘图能力,而且带有12个功能强大的工具箱;Microsoft Access编写的数据库具有随时对数据进行修改和补充。程序运用Matlab解方程组的功能和神经网络工具箱,建立沸腾炉焙烧物料平衡信息系统和锌常规浸出和锌电积神经网络预测信息系统两大系统,再把解方程过程中BP神经网络可视化接到VB的封装体系中,所涉及到的BP神经网络函数都以脚本文件的形式存在,这样既减少程序的复杂性又提高了程序的运行效率,
Zinc hydrometallurgy is a complicated and sophisticated process, which includes roasting of zinc concentrate, leaching, purifying, electrolysis and casting of zinc cathode. The zinc concentrate, the principle of fluidized bed furnace roasting, the factors influencing the general leaching and the economic and technological indexes in the zinc electrolysis process are indicated in this thesis, which lay a foundation for the information systems.
    The roasting, leaching and electrolysis of ZnSO4 are mainly studied in the thesis. The information system of mixing of Zinc concentrate is found first, which offers accurate mixing and costs the least and all the data needed can be input with database operating and ten kinds of ores are
    
    
    optimized by the way of forty-five combinations. And then thermodynamic model is established in the process of fluidized roasting. So the components of roasting ores and gas can be determined through material balance and heat equilibrium. Because of the in fluctuation of the roasting time, this model can't describe the inner mechanism of the process and brings in certain deviations of calculation, but it can describe the roasting tendency. Finally, BP neural network is designed in the process of leaching and electrolysis to predict the leaching rate, the residue ratio and so forth and the factors such as electric current efficiency in the electrolyzing. Thirty groups of training data are used in the BP neural network and their range is wide. The training error goal can reach 10~5. So the prediction of the new designing BP neural network is reasoned with industrial process and its precision is high. It not only forecasts off-line but also provides the index of parameter for the direction of industrial process
    .
    This information system uses visual basic 6.0 and matlab 6.1 to programme and Microsoft Access to build and modify database. The visual basic 6.0 is simple and practical, is easy to realize visualization, is convenient to package and offer a lot of ActiveX widgets. The matlab 6.1 is not only convenient to operate the matrix and to draw a graph, but also has twelve toolboxes. The database of Microsoft Access may be revised and supplemented at any moment. Two information systems of fluidized bed furnace roasting through material balance and heat equilibrium and BP neural network of electrolysis of ZnSO4 are built by matlab 6.1 of neural network toolbox and the function of dealing with equation. Those processes are displayed by visualization in the system of visual basic 6.0. The involved function of BP neural network exist the function of script, which decrease the complexity of the program and improve the running efficiency of the program.
引文
[1]冯桂林主编,有色金属矿产资源的开发及加工技术,昆明:云南科技出版社,2000,5
    [2]魏昶 编著,湿法炼锌理论与技术,昆明:昆明理工大学,2002,8
    [3]丁朝模,有色金属进展,长沙:中南大学出版社,1995
    [4]徐鑫坤,魏(永口),锌冶金学,昆明:云南科技出版社,1996
    [5]曹异生,铅锌工业现状及发展前景的预测,冶金经济分析,1995(7):19
    [6]刘志宏,国内外锌冶炼技术的现状及发展动向,有色金属,2000(1):23~25
    [7]李治国、杨如中等,湿法炼锌的电解节能,有色矿冶,1998(1):42~44
    [8]张乐如,我国炼锌能耗与余热利用,有色金属(冶炼部分),1996(1):6~10
    [9]吴绍清,影响电锌含铅偏高的因素分析,湿法冶金,2001(3):136~138
    [10]吕少祥、戴曦,降低锌电积直流电耗生产实践,有色金属(冶炼部分),2001(6):13~15
    [11]雷桂萍,刘中华,锌工业近十年的统计与发展趋势分析,上海有色金属,2001(4):30~32
    [12]徐采栋、林蓉等,锌冶金物理化学,上海:上海科技出版社,1979,7
    [13]陈维东 主编,国外有色冶金工厂(铅锌下册),冶金部长沙矿冶研究院印刷,1988,7
    [14]孙倬、张驾等主编,重有色金属冶炼设计手册(铅锌铋卷),北京:冶金工业出版社
    [15]王春伟、吴坚,铸铁熔炼计算机辅助配料系统软件,铸造,2002(5):313~315
    [16]E.M.Beale.Cycling in the Dual SimplexAlgorithm[J].NavalResearch Logistics Quarterly. 1953,(2):269-276.
    [17]G.G.Bland.New Finite Pivoting Rules of the Simplex Method[J].Math.of Operation Research, 1977. (2):103-107
    [18]何蔼平,汁算机在冶金中的应用,云南科技出版社
    [19]许璐等.一类退化的线性规划模型构造.肇庆学院学报.第23卷第2期.2002年4月:13-15
    [20]林立军、程斌等编著,Visual Basic 6.0数据库开发指南,西安:西安电子科
    
    技出版社,1999,9
    [21]胡斌,国际铅锌精矿计价方法及如何规避风险,有色金属,2000(7):44~47
    [22]杜清枝、杨继舜主编,物理化学,重庆:重庆大学出版社,1997
    [23]谭春娥:昆明理工大学硕士学位论文,2001
    [24]黄波:昆明理工大学硕士学位论文,1998
    [25]H.A.因G.H.盖格 著,李光,赵宇译,冶金过程物料平衡计算手册,冶金工业出版社,1983(5)
    [26]叶大伦、胡建华编著,实用无机物热力学数据手册,冶金工业出版社 2001
    [27]郑君杰、黄峰,在VB6.0中调用和操作Matlab,微机发展,2002(4):15~17
    [28]蒙以正,Matlab5.X应用与技巧[M],北京:科技出版社,1999
    [29]MIDEVA. MATCOM & Visual MATCOM USER'S GuideV4.5relcase[M]. Mathv Tools ltd,1999
    [30]B Muller Reinhaudt.Neural Network: An Introduction.New York: Springer-Verlag, 1990
    [31]施鸿堡,神经网络及其应用,西安电子科技大学出版社,1993
    [32]沈清、胡德文、时春,神经网络应用技术,国防科技大学出版社,1993
    [33]胡瑞敏,BP网络的全集成运行开发环境,微计算机应用,1996,17(2):1~5
    [34]陆宁编著,Matlab语言即学即会,北京:机械工业出版社,2000,8
    [35]王学辉、张明辉等编著,Matlab6.1最新应用详解,北京:中国水利水电出版社,2002,1
    [36]黄文梅、杨勇等编著,系统分析与仿真(Matlab语言及应用),长沙:国防科技大学出版社,1999,8
    [37]李兰友等编著,Visual Basic绘图与图象处理,人民邮电出版社,1999,9
    [38]闻新,周露等编著,Matlab神经网络应用设计,北京:科学出版社,2000,9
    [39]白鹏、莫卫东等编著,Visual Basic 6.0高级编程技巧,西安:西安交通大学出版社,1999,12
    [40]Francesco Balena著、翔实翻译组译,Visual Basic编程技术大全,北京;机械工业出版社,2000,6
    [41]Evangelos Petroutsos Kevin Hough著、邱仲潘等译,Visual Basic5.0高级开发指南,北京:电子工业出版社,1998,9
    [42]G.Cybenko.Continuous value neural networks with two hidden layers are sufficient.Math.,Contr.,Sign.& Sys.,Vol.2,pp303-314,1989
    
    
    [43]K.Hornik,M.Stinchcombe & H.White.Universal approximation of an unknown mapping and its derivatives using multiplayer feedforward networks.Nerual Networks.Vol.3,pp551-560.1990
    [44]邱东强,神经网络控制的现状与展望,自动化与仪器仪表,2001(5):1~7
    [45]Kosko B.Neural Network and Fuzzy Systems:A Dynamical System Approach to Machine Intlligence.NJ:Prentice+Hall, 1992:299-338
    [46]从爽编著,面向Matlab工具箱的神经网络理论与应用,中国科学技术大学出版社,1998年
    [47]邝孔武,王晓敏编著,信息系统分析与设计,北京:清华大学出版社,1999,6
    [48]Buckley, J.J and Hayashi,Y..Fuzzy Neural Networks: a survey.Fuzzy Sets and System,1994,66(1):1-13
    [49]黄家英、葛一楠等,误差反向传播学习算法及B -P网络,自动比与仪器仪表,1997(3):46~52
    [50]Buckley, J.J and Hayashi,Y..Numerical relationships between neural networks,continuous function,and fuzzy systems,Fuzzy Sets and systems, 1993,60(1):1-8
    [51]徐建林,基于人工神经网络的球铁缺陷预报系统,冶金自动化,200(4):11~14
    [52]Funahashi.K.I:On the Appoximate Realization of Continuous Mapping by Neural Networks 1989.1
    [53]Robert H.N., Thory of the Back-Progation Neural Network. Pro.VCNN.Washing D.C 1989.1

© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号

地址:北京市海淀区学院路29号 邮编:100083

电话:办公室:(+86 10)66554848;文献借阅、咨询服务、科技查新:66554700