摘要
水分含量是影响炸药产品质量的重要指标之一,同时也是出厂必测项目之一,它的多少不但直接关系到炸药质量的高低,而且对整个系统的安全性具有极大的影响。然而,目前炸药生产的自动化程度不高,劳动强度大,水分控制过程中检测严重滞后且波动范围大,比例难以调整。为此,研究如何建立炸药生产线改性硝铵水分控制系统而实现对水分的实时调节,对实现粉状炸药生产过程的全自动化、稳定炸药质量以及提高企业竞争力都具有重要意义。
本文通过对粉状炸药生产工艺研究,分析了对炸药生产线改性硝铵水分的影响因素。首先,在分析炸药生产工艺的基础上,确定了影响炸药生产线改性硝铵水分的主要因素——硝酸铵和水分,利用线性回归方法,提出建立硝酸铵浓度,流速与硝酸铵流量,蒸汽流量的模型,紧接着对BP(Back Propagation)神经网络PID算法进行了研究,并用现场的数据对其进行了验证,得到了很好的仿真效果。通过在线调节神经网络的权值,调节PID参数,控制硝酸铵流量和蒸汽流量,使炸药成品水分含量满足设计质量指标。
开发了基于BP网络PID的炸药生产线改性硝铵水分控制系统,论文描述了系统的结构、功能和系统软件设计。控制系统采用DCS(Distributed Control System)结构,实现了监控画面、报警提示、趋势显示等功能。仿真结果表明控制算法满足炸药产品的水分标准,超调小,调节时间短,提高了炸药的质量和生产效率。
Moisture content is one of important factors that affect the quality of power dynamite. Meanwhile, it is also one of safety items that must be measured before the product is going to the market. Its content influences not only the performance of the dynamite, but also the security of producing system thereof. However, the automation degree's not satisfactory, together with workers'intensive labour in current dynamite product line make the coupling of influent factors exsited in producing process. All factors mentioned above lead to the poor ingredient adjustment and difficult realization of a real-time control of the whole producing process. Therefore, it is significant to research on how to implement real-time control on the ingredient through establishing a moisture control system of features changed ammonium nitrate in dynamite production line so as to realize a steady and high production of power dynamite and enhancing the competitive power of enterprises.
This paper analyzes the major factors which influence moisture of features-changed ammonium nitrate in dynamite production line through the research on the producing process of power dynamite. At first, the product process is analyzed. Flow of ammonium nitrate and steam the main factors that influence the moisture of features changed ammonium nitrate in dynamite production line. Using methods of linear regression, a model that contains speed of liquid, concentration of ammonium nitrate and flow of ammonium nitrate, flow of steam is proposed. And then, we give research to the BP neural network PID and validate it with the data on site to get a reliable simulation result. Through online regulation on the parameters of neural network and PID, making moisture of dynamite products satisfy designed quality index.
A moisture control system of features changed ammonium nitrate in dynamite production line based on the BP neural network PID is developed in this paper. The structure and function of system are presented, in which the technologies such as composition of hardware are introduced in detail. The controlling system adopts the structure of DCS., It realizes the function of monitoring, alarming and trends-displaying. The simulation results indicate that the control algorithm meets the standard of reasonable moisture. Small overshoot is got, and short regulating time is gained. The quality of dynamite and producing efficiency are both improved.
引文
[1]张亚明,杨国富.影响膨化硝铵炸药性能的因素分析[J].爆破器材,2002,10,31(5)：19-21
[2]肖国,刘厚平,郭子庭.浅析美日中工业炸药的发展近况[J].爆破器材,2003,6：164-170
[3]陆明,吕春绪.种粉状铵梯油炸药配方的分析与评价[J].含能材料,2001,6：134-136
[4]周向群.试论乳化炸药连续生产技术的发展[J].爆破器材,2004,8：1-3
[5]刘厚平,肖国,郭子庭.粉状硝铵炸药的连续自动化生产及其生产本质安全性问题[J].矿冶工程,2003,6：134-135
[6]吕春绪.膨化硝铵炸药[M].北京：兵器工业出版社,2001
[7]丁炜,魏孔平.可编程控制器在工业控制中的应用[M].北京：化学工业出版社.2004
[8]Wang Yu-jia, Zhang Ming-jun.Study of fuzzy neural networks model for system condition monitoring of AUV [J]. Journal of marine science and application,2002,1(2):42-45
[9]段仁君.乳化炸药连续生产自动控制系统[J].矿业研究与开发,2005,3(15)：60-63
[10]Yu H J, Chen C Z, Zhang S. Intelligent diagnosis based on neural network[M]. Bei-jing, China, Metallurgy Industry Press,2002
[11]陈之功,吴龙祥,潘涛,等.自控技术在乳化炸药连续化生产中的应用[J]. 金属矿山,2005,11：43-45
[12]童进秋,吴燕.乳化炸药生产线集散控制系统的研究和应用[J].煤矿爆破,2005,20-24
[13]贺建军,喻寿益,桂卫华.基于微机监控的乳化炸药全自动连续生产线[J].中南工业大学学报,2000,30：422-425
[14]Johnell Gonzales-Lujan, Robert J. Burnside, George H. Tompkins. A simulation study of high power detonator production transition[C]. Proceedings of the Winter Simulation Conference,2002
[15]朱凌云,蒋峰,何端阳.乳化炸药连续化生产线计算机控制系统[J].现代电子技术,2003,12(3)：68-70
[16]肖慧荣,陈学岗,乐淑萍.基于两级闭环控制的型砂水分在线测控系统[J].仪表技术与传感器,2006,1
[17]贺成柱.基于DSP的烧结混合料中子水分控制系统[J].机械研究与应用,2006,4：56-65
[18]韦尔特,王红英,宋立霞.我国膨化颗粒饲料水分控制研究技术发展[J].饲料工业,2008,29：88-90
[19]唐淑娟,李新杰.工业炸药水分测定的标准化问题[J].分析与检测,2004,9：25-30
[20]刘向东,常江,张崇会,等.自适应模糊PID控制方法在温度控制中的研究[J].佳木斯大学学报,2007,5：18-25
[21]张葛祥,唐钟.基于IGA的多层神经网络BP算法及在系统辨识中的应用[J].西南工学院学报,2002,17(2)：9-13
[22]刘永红.神经网络理论的发展与前沿问题[J].信息与控制,2003,28(1)：31-46
[23]王旭,王宏,王文辉.人工神经元网络原理与应用[M].沈阳：东北大学出版社,2000
[24]吕砚山,赵正琦.BP神经网络的优化及应用研究[J].北京化工大学学报,2001,(1)：67-69
[25]姚一波,王纪亮.提高BP网络训练速度的研究[J].信息技术,2002,(1)：4-6
[26]FANG Fang, WEI Le, SUN Wanyun. Simulation of sensor fault diagnoses based on the neural network observer group [J]. Computer Simulation,2001,18 (4):24-28
[27]He Yigang, Li Yun, Liu Meirong. BP neural network approach to module fault diagnosis for large analog circuit, proc[C]. IEEE TONCON'2004, Chiang Mai, Thailand,2004,11
[28]A. Alessandri. Fault diagnosis for nonlinear systems using a bank of neural estimators[J]. Computers in Industry,2003(52):271-28
[29]Martin T H, Howard B D, Mark H B神经网络设计[M].北京：机械工业出版社,2004
[30]蒋宗礼.人工神经网络导论[M].高等教育出版社,2001
[31]Yu H J, Chen C Z,Zhang S. Intelligent diagnosis based on neural network[M]. Bei-jing, China,Metallurgy Industry Press,2002
[32]朱旭东,戴文战,朗燕峰.基于神经网络的方法在故障诊断中的应用[J].机电工程,2003,20(5)：75-78
[33]王宏伦,吕庆风,佟明安.模糊逻辑系统的GA+BP学习算法[J].控制理论 与应用,2001,18(1)：17-20
[34]Qingfeng Zeng,Jiakui Zu,Litong Zhang.Designing Expert System with Artificial Neural Networks for in Situ Toughened Si3N4[J]. Materials and Design,2004, 23 (3):287-290
[35]田华,李相林.一种基于混合学习算法的模糊神经网络控制[M].哈尔滨工业大学学报,2002,7(1)
[36]安艳秋,陈举华,张洪才.基于进化神经网络的齿轮可靠性预测[J].山东大学学报(工学版),2002,32(3)：227-231
[37]Pan Xu-wei. Knowledge modeling techniques for knowledge management[J]. Computer Integrated Manufacturing Systems,2003,9(7):517-521
[38]He Yigang, Li Yun, Liu Meirong. BP neural network approach to module fault diagnosis for large analog circuit, proc[C]. IEEE TONCON,2004, Chiang Mai, Thailand,2004,11
[39]Li Rongzheng. The development of the expert control technology and the intelligent control theory[J]. Process Automation Instrumentation,2004:1-7
[40]Meng J E, Liao J, Lin J Y. Fuzzy neural networks based quality prediction system for sintering process[J]. IEEE Transactions on Fuzzy Systems.2000, 8(3):314-324
[41]庞中华,崔红.系统辨识与自适应控制MATLAB仿真[M].北京：北京航空航天大学出版社,2009
[42]胡守仁.神经网络应用技术[M].国防科技大学出版社,2005,1
[43]Li X F, Liu T B. Neural network forecast model of the temperature of cement rotary kiln[J]. Microcomputer Information,2009,12(l):277-284
[44]魏海坤,徐嗣鑫,宋文忠.神经网络泛化理论和泛化方法[J].自动化学报,2001,27(6)：806-814
[45]汪学军,唐焕文.前馈神经网络泛化性能的系统分析[J].系统工程理论与实践,2000,(8)：36-40
[46]Zhan J. Zhang Q. Application of Neural Networks Optimized by Genetic Algorithm in Forecasting Electric Field Aging Technic[C]. International Conference on Multi-Media and Information Technology,2008,30(1)
[47]陈小前,罗世彬,王振国,等.BP神经网络应用中的前后处理过程研究[J].系统工程理论与实践,2001,(1)：65-70
[48]阎平.人工神经网络的容量、学习与计算复杂性[J].电子学报,2005,1：23-67
[49]Yang M, Wang Y. Prediction of coal and gas outburst intensity with Incorporate Genetic Algorithm Based Back Propagation Neural Network[J]. Journal of Chongqing University,2010,12(3):32-37
[50]FANG Fang, WEI Le, SUN Wanyun. Simulation of sensor fault diagnoses based on the neural network observer group[J]. Computer Simulation,2001,18 (4): 24-28
[51]Xue Z T, Li Z. Control system design of cement rotary kiln based on PID neuron networks [C]. International Conference on Test and Measurement,2009, 12(1):21-24
[52]李海民,吴成柯.基于BP网络的遗传算法[J].模式识别与人工智能,2004,12(2)：223-227
[53]贾毅,陆鑫达,佘雷.基于遗传算法的动态负载平衡研究[J].计算机工程与应用,2003,(12)：84-86
[54]王学梅,王义和.模拟退火与遗传算法的结合[J].计算机学报,2005,20(4)：381-384
[55]Liu,Peiqi. Knowledge representation of extended production rule[J]. Hsi-An Journal of Xi'an Jiaotong University,2004,38(6):587-590
[56]张华军,赵金.基于遗传算法和神经网络预测的再励学习[J].计算机工程,2010,36(1)：91-96
[57]李敏强,徐博艺,寇纪松.遗传算法与神经网络的结合[J].系统工程理论与实践,2001,2：65-69
[58]王洪燕,杨敬安,蒋培.进化设计人工神经网络[J].计算机科学,2000.27(6)：43-45
[59]Rudolph G. Convergence analysis of canonical genetic algorithm[J]. IEEE Trans,2004,5(1):96-101
[60]Simon Haykin.神经网络原理[M].北京：机械工业出版社,2004
