散货码头集散控制系统的容错控制
|
摘要
集散控制系统是以微处理机为核心，以满足现代化企业生产、相互关系日益复杂的控制对象的要求为前提。从过程综合自动化的角度出发，集中处理信息、集中管理，而控制权分散的一种多级分布式计算机控制系统，目前已经广泛应用于工业生产和控制管理领域。
容错控制是一种提高系统安全性和可靠性的新技术，容错控制以大量的在线故障检测、参数辨识和模型计算为基础，在系统出现故障时，适时修正系统的控制率，使系统继续维持既定的运行目标，具有适应能力强、可靠性高、功能强等特点。
本文以华能南通电厂输煤控制系统工程实例为对象，采用主动性容错的方法构建系统的容错控制律重构模型，解决系统的实时故障诊断和容错控制律重构两大问题。
本文第二章对控制对象根据功能的不同分解为三个相互独立的子系统，运用状态空间理论及矩阵构造的方法对其中的流程控制系统设计了容错控制模型。第三章采用基于径向基神经网络的方法进行容错控制的实时故障诊断，并用Matlab给出了仿真结果。第四章进一步探讨了主动容错在流程控制中的实现问题，即根据故障诊断的结果，实时计算故障函数，在线修改系统的控制律的方法来实现容错控制律重构。第五章采用VC++编制仿真程序，对控制系统进行仿真，作出验证。第六章阐述了如何利用OPC技术，打通与FIX组态软件的缺口，将容错子程序外挂到基于FIX平台的输煤控制系统控制程序中。这种方法结合目前高速微处理器的发展，少量实时计算带有高斯函数的方程，为目前集散控制系统的发展提供了一种带有容错化的思想。
Distributed Control System is not only based mainly on the microprocessor technique, but also the requirement of modern enterprise manufacture and the co-relationship among the control objects with the increasing complexity. From the view of process integrate automation, it can be regarded as one kinds of multi-level distributed computer control system, which has the concentrated way in dealing with information and management, while the scattered solution in control. It has been widely used in industrial production and the fields of control and management.
Fault Tolerant Control is one kind new technique with the aim to enhance the system security and system reliability. FTC is basically depended on a great deal of online detection, parameter identification and model calculation. It can revise the system control rules when the fault occurred and then maintain the system operation according the original goal. And it has the specialty of stronger adaptability, higher reliability and better function etc.
The paper regards the coal-transporting systen of the Huaneng Nantong power plant as the study object, applied the active tolerance to design the reconstructed model of tolerant control rule, to solve the two problems of real-time FDD and CRR.
In the second chapter we divide the system to three independent parts according to the different function of the object, then towards one part of system, flow system, designed fault tolerant model using state space theory and matrix construction. In the third chapter it is applied the RBF neural network to solve the real-time FDD, and the simulation result is given by MatlabS. 3 software. In the fourth chapter the more details of realizing the fault tolerant flow control is discussed, and and the CRR is settled by the conclusion of FDD and the real-time function calculation which achieve the goal of the system control rule on-line reconstruction. In the fifth chapter we complied the simulated program by Visual C++ for the whole theory to draw the conclution. In the sixth chapter we applied the OPC technology to the actual object, then the tolerant program can be linked with the main
II
coal-transporting program based on FIX software by a little modification. The way is associated with the development of high speed CPU so as to real-time compute some Gaussian function. It presents a way with tolerant idea for solving the security and reliability in DCS now.
容错控制是一种提高系统安全性和可靠性的新技术，容错控制以大量的在线故障检测、参数辨识和模型计算为基础，在系统出现故障时，适时修正系统的控制率，使系统继续维持既定的运行目标，具有适应能力强、可靠性高、功能强等特点。
本文以华能南通电厂输煤控制系统工程实例为对象，采用主动性容错的方法构建系统的容错控制律重构模型，解决系统的实时故障诊断和容错控制律重构两大问题。
本文第二章对控制对象根据功能的不同分解为三个相互独立的子系统，运用状态空间理论及矩阵构造的方法对其中的流程控制系统设计了容错控制模型。第三章采用基于径向基神经网络的方法进行容错控制的实时故障诊断，并用Matlab给出了仿真结果。第四章进一步探讨了主动容错在流程控制中的实现问题，即根据故障诊断的结果，实时计算故障函数，在线修改系统的控制律的方法来实现容错控制律重构。第五章采用VC++编制仿真程序，对控制系统进行仿真，作出验证。第六章阐述了如何利用OPC技术，打通与FIX组态软件的缺口，将容错子程序外挂到基于FIX平台的输煤控制系统控制程序中。这种方法结合目前高速微处理器的发展，少量实时计算带有高斯函数的方程，为目前集散控制系统的发展提供了一种带有容错化的思想。
Distributed Control System is not only based mainly on the microprocessor technique, but also the requirement of modern enterprise manufacture and the co-relationship among the control objects with the increasing complexity. From the view of process integrate automation, it can be regarded as one kinds of multi-level distributed computer control system, which has the concentrated way in dealing with information and management, while the scattered solution in control. It has been widely used in industrial production and the fields of control and management.
Fault Tolerant Control is one kind new technique with the aim to enhance the system security and system reliability. FTC is basically depended on a great deal of online detection, parameter identification and model calculation. It can revise the system control rules when the fault occurred and then maintain the system operation according the original goal. And it has the specialty of stronger adaptability, higher reliability and better function etc.
The paper regards the coal-transporting systen of the Huaneng Nantong power plant as the study object, applied the active tolerance to design the reconstructed model of tolerant control rule, to solve the two problems of real-time FDD and CRR.
In the second chapter we divide the system to three independent parts according to the different function of the object, then towards one part of system, flow system, designed fault tolerant model using state space theory and matrix construction. In the third chapter it is applied the RBF neural network to solve the real-time FDD, and the simulation result is given by MatlabS. 3 software. In the fourth chapter the more details of realizing the fault tolerant flow control is discussed, and and the CRR is settled by the conclusion of FDD and the real-time function calculation which achieve the goal of the system control rule on-line reconstruction. In the fifth chapter we complied the simulated program by Visual C++ for the whole theory to draw the conclution. In the sixth chapter we applied the OPC technology to the actual object, then the tolerant program can be linked with the main
II
coal-transporting program based on FIX software by a little modification. The way is associated with the development of high speed CPU so as to real-time compute some Gaussian function. It presents a way with tolerant idea for solving the security and reliability in DCS now.
引文
[1]王常力 集散型控制系统的设计与应用 清华大学出版社1993年6月
[2]Astrom K J,Intelligent control.In:Proc.Of 1st European Control Conference.Grenoble,1991.2328-2329
[3]黄苏南，邵慧鹤，分散容错控制的最优设计，《智能控制与智能自动化》，北京：科学出版社，1993，1499-1503
[4]Yu D, Shields O N. billinear fault detection filter.Int.J.Control,1997,68(3):417-430
[5]郑叶飞 基于模型的集散系统智能化容错控制 硕士学位论文2000．5
[6]阎平凡，张长水 人工神经网络与模拟进化计算 2000年11月 信息技术丛书 清华大学出版社
[7]张志华，郑南宁，郑海兵 径向基函数(RBF)神经网络的一种极大熵学习算法，《计算机学报》第24卷 第5期Vol.24 No.5 2001.5
[8]闻新，张洪钺，周露 控制系统故障诊断和容错控制 1998．2电气自动化新技术丛书 机械工业出版社
[9]Intellution Introduction to iFix 2.1软件简介
[10]Sunil Eianayar V.T.,Yung C.Shin State estimation of contiunous-time radial basis function networks Automatica 36(2000) 399-407
[11]Marcantonio Catelani,Ada Fort Fault diagnosis of electronic analog circuits using a radial basis function network classifier Measurement 28(2000) 147-158
[12]蒋斌 颜钢峰 基于RBF神经网络的模糊自适应控制《电气自动化》第1期2001年
[13]杜耀武 蒋伟康 基于人工神经网络的大型机械系统的剩余寿命预报模型 《机械强度》Vol 19 No．1 March 1997
[14]顾伟 褚建新 基于故障统计模型的可修系统维修周期预测法 《机械强度》Vol．22 NO．1 March 2000
[15]张金玉等 基于RBF的范例检索方法的研究 《机械工程学报》Vol．36 No．10 Oct．2000
[16]郑士荣等 基于模糊径向基函数神经网络的航空煤油干点估计 《机电工程》Vol．17 No 3 2000
[17]解学军 张大雷 基于径向基函数网络的非线性离散时间系统的自适应控制 《自动化学报》Vol．26 No．3 May2000
[18]付扬 基于SIMULINK矢量控制系统的仿真 《工业仪表与自动化装置》 第6期2000年
[19]周鸣崎 徐军 控制系统故障诊断 《计算机自动测量与控制》 8(3)2000．
[20]张定会 故障诊断专家系统实时数据库获取现场实时数据的方法 《自动化仪表》第20卷第2期Vol．20 No．2 1999
[21]曹晓红 杨衍崎 基于DCS的C／S模式应用《自动化仪表》第20卷第2期Vol．20 No．2 1999
[22]卓敏 黄道 DCS与微机局域网的集成 《自动化仪表》 Vol．20 No．2 1999
[23]张立明 人工神经网络的模型及其应用 复旦大学出版社 1994年7月
[24]王永骥 涂健 神经元网络控制 机械工业出版社电气自动化新技术丛书 1998年2月
[25]周东华 孙优贤 控制系统的故障检测与诊断技术 清华大学出版社 1994年9月
[26]胡寿松 主编 自动控制原理 国防工业出版社 1984年6月
[27]Piggio T, Girosi F.A theory of networks for approximation and learning.A I Memo No 1140,Artificial Intelligence Laboratory. Massachusetts Institute of Technology.Cambridge Mass.1989
[28]Monica Bianchini,etal.Learning without local minima in radial basis function networks.[J] IEEE Trans on Neural Network.1995.6930:749～755
[29]Moody J,Darken C.East learning in networks of locally-turned processing units.[J] Neural Computation.1989,(1):281～294
[30]Bruce A,etc.Cooperative-competitive genetic evolution of radial basis function centers and widths for time series prediction.[J] IEEE Transactions on Networks july 1996.Vol.7.No.4,868～880
[31]曹先彬、高隽、王熙法．基于生态竞争模型的遗传强化学习．[J]软件学报1999．6．Vol、10，No．6，658～661
[32]周佩玲 陶小丽 傅忠谦 彭虎 王新跃 改进RBF神经网络用于降雨量预测 小型微型计算机系统 2001 Vol．22 No．2 P．244-246
[33]Ye Y Z, Zhang j M,Liu Y J.OPC technology and its application in marine
engine simulation system.In:Proceedings of the 4th international marine electri c technology conference (IMECE'2000),Nov.2000.Shanghai.146～151
[34]Pentland A P Automatic extraction of deformable part models J International Journal of Computer Vision 1990 4
[35]Lowe D G Adaptive radial basis function nonlinearities,and the problem of generalization C In 1st IEE Conference on Artificial Neural Networks.London.UK 1989
[36]Zemel R S Hinton G E Learning population codes by minimum description length J Neural Computation 1995 7 3 549-564
[37]Rissanen J Universal coding,information, prediction and estimation J IEEE Transactions on Information Theory 1984 30 629-636
[38]Haykin S Neural network-a comprehensive foundation M Published by:Macmillan College Publishing Company,New York,USA 1994
[39]DO'Leary Enterprise knowledge management J IEEE Computer 1998 31 3 54-61
[40]Dimitris Karagiannis (ed ACM SIGOIS Bulletin Special issue: business process reengineering J ACM Special Interest Group on Office Information Systems 1995
[41]Jana Roehler Planning from second principles-a logic-based approach R Research Report RR-94-13 DFKI-German Research Center for Artificial.Intelligence 1994
[42]MErdmann The data warehouse as a means to support knowledge management C In Workshop of Knowledge-Based System for Knowledge Management in Enterprise in KI97 Sept 1997
[43]AAamodt MNygard Different roles and mutual dependencies of data,information,and knowledge-An AI Perspective on Their Integration,J In Data & Knowledge Engineering 16() Elsevier North-Holland 191222
[2]Astrom K J,Intelligent control.In:Proc.Of 1st European Control Conference.Grenoble,1991.2328-2329
[3]黄苏南，邵慧鹤，分散容错控制的最优设计，《智能控制与智能自动化》，北京：科学出版社，1993，1499-1503
[4]Yu D, Shields O N. billinear fault detection filter.Int.J.Control,1997,68(3):417-430
[5]郑叶飞 基于模型的集散系统智能化容错控制 硕士学位论文2000．5
[6]阎平凡，张长水 人工神经网络与模拟进化计算 2000年11月 信息技术丛书 清华大学出版社
[7]张志华，郑南宁，郑海兵 径向基函数(RBF)神经网络的一种极大熵学习算法，《计算机学报》第24卷 第5期Vol.24 No.5 2001.5
[8]闻新，张洪钺，周露 控制系统故障诊断和容错控制 1998．2电气自动化新技术丛书 机械工业出版社
[9]Intellution Introduction to iFix 2.1软件简介
[10]Sunil Eianayar V.T.,Yung C.Shin State estimation of contiunous-time radial basis function networks Automatica 36(2000) 399-407
[11]Marcantonio Catelani,Ada Fort Fault diagnosis of electronic analog circuits using a radial basis function network classifier Measurement 28(2000) 147-158
[12]蒋斌 颜钢峰 基于RBF神经网络的模糊自适应控制《电气自动化》第1期2001年
[13]杜耀武 蒋伟康 基于人工神经网络的大型机械系统的剩余寿命预报模型 《机械强度》Vol 19 No．1 March 1997
[14]顾伟 褚建新 基于故障统计模型的可修系统维修周期预测法 《机械强度》Vol．22 NO．1 March 2000
[15]张金玉等 基于RBF的范例检索方法的研究 《机械工程学报》Vol．36 No．10 Oct．2000
[16]郑士荣等 基于模糊径向基函数神经网络的航空煤油干点估计 《机电工程》Vol．17 No 3 2000
[17]解学军 张大雷 基于径向基函数网络的非线性离散时间系统的自适应控制 《自动化学报》Vol．26 No．3 May2000
[18]付扬 基于SIMULINK矢量控制系统的仿真 《工业仪表与自动化装置》 第6期2000年
[19]周鸣崎 徐军 控制系统故障诊断 《计算机自动测量与控制》 8(3)2000．
[20]张定会 故障诊断专家系统实时数据库获取现场实时数据的方法 《自动化仪表》第20卷第2期Vol．20 No．2 1999
[21]曹晓红 杨衍崎 基于DCS的C／S模式应用《自动化仪表》第20卷第2期Vol．20 No．2 1999
[22]卓敏 黄道 DCS与微机局域网的集成 《自动化仪表》 Vol．20 No．2 1999
[23]张立明 人工神经网络的模型及其应用 复旦大学出版社 1994年7月
[24]王永骥 涂健 神经元网络控制 机械工业出版社电气自动化新技术丛书 1998年2月
[25]周东华 孙优贤 控制系统的故障检测与诊断技术 清华大学出版社 1994年9月
[26]胡寿松 主编 自动控制原理 国防工业出版社 1984年6月
[27]Piggio T, Girosi F.A theory of networks for approximation and learning.A I Memo No 1140,Artificial Intelligence Laboratory. Massachusetts Institute of Technology.Cambridge Mass.1989
[28]Monica Bianchini,etal.Learning without local minima in radial basis function networks.[J] IEEE Trans on Neural Network.1995.6930:749～755
[29]Moody J,Darken C.East learning in networks of locally-turned processing units.[J] Neural Computation.1989,(1):281～294
[30]Bruce A,etc.Cooperative-competitive genetic evolution of radial basis function centers and widths for time series prediction.[J] IEEE Transactions on Networks july 1996.Vol.7.No.4,868～880
[31]曹先彬、高隽、王熙法．基于生态竞争模型的遗传强化学习．[J]软件学报1999．6．Vol、10，No．6，658～661
[32]周佩玲 陶小丽 傅忠谦 彭虎 王新跃 改进RBF神经网络用于降雨量预测 小型微型计算机系统 2001 Vol．22 No．2 P．244-246
[33]Ye Y Z, Zhang j M,Liu Y J.OPC technology and its application in marine
engine simulation system.In:Proceedings of the 4th international marine electri c technology conference (IMECE'2000),Nov.2000.Shanghai.146～151
[34]Pentland A P Automatic extraction of deformable part models J International Journal of Computer Vision 1990 4
[35]Lowe D G Adaptive radial basis function nonlinearities,and the problem of generalization C In 1st IEE Conference on Artificial Neural Networks.London.UK 1989
[36]Zemel R S Hinton G E Learning population codes by minimum description length J Neural Computation 1995 7 3 549-564
[37]Rissanen J Universal coding,information, prediction and estimation J IEEE Transactions on Information Theory 1984 30 629-636
[38]Haykin S Neural network-a comprehensive foundation M Published by:Macmillan College Publishing Company,New York,USA 1994
[39]DO'Leary Enterprise knowledge management J IEEE Computer 1998 31 3 54-61
[40]Dimitris Karagiannis (ed ACM SIGOIS Bulletin Special issue: business process reengineering J ACM Special Interest Group on Office Information Systems 1995
[41]Jana Roehler Planning from second principles-a logic-based approach R Research Report RR-94-13 DFKI-German Research Center for Artificial.Intelligence 1994
[42]MErdmann The data warehouse as a means to support knowledge management C In Workshop of Knowledge-Based System for Knowledge Management in Enterprise in KI97 Sept 1997
[43]AAamodt MNygard Different roles and mutual dependencies of data,information,and knowledge-An AI Perspective on Their Integration,J In Data & Knowledge Engineering 16() Elsevier North-Holland 191222