SDG故障诊断专家系统及其Visual Prolog实现
|
摘要
故障诊断技术大致分为基于模型的方法和基于历史的方法,其中符号有向图SDG模型是一种深层知识模型,能够揭示复杂系统的变量间内在因果关系及影响,是定性仿真的一个重要分支,是一种重要的故障诊断方法。本文在总结前人有关SDG故障诊断研究的基础上,对基于SDG的故障诊断方法进行了系统而深入的研究,着重研究了SDG建模和推理机制,对复杂系统SDG故障传播规律进行了详细的研究,发展并完善了基于图论的故障诊断方法。
为了克服SDG模型的固有缺陷,提高诊断分辨率和诊断速度,阐述了将SDG方法和专家系统方法相结合的思想,对基于SDG的故障诊断专家系统进行了详细研究,摒弃了传统由SDG模型推导产生式规则的浅层知识库设计方法,提出直接以SDG图模型深层知识作为所开发专家系统的深层知识库,同时将专家的知识和经验作为浅层知识库。最后对该专家系统的推理机算法进行了详细研究并且给出算法。
本文的SDGExpert专家系统是基于Visual Prolog语言实现的,SDGExpert专家系统采用了用户层、中间层、存储层的三层架构设计,完全应用了面向对象程序设计方法,应用了良好的软件设计模式,不仅实现了专家系统应有的功能,还最大程度的实现了SDGExpert专家系统的可复用和可扩展性,以应付以后的进一步开发。
文章最后在多功能仿真平台的液位离心泵仿真系统进行了SDGExpert专家系统案例仿真。结果证明该SDGExpert能够发现系统的故障根源,而且软件的友好的用户界面可以直观的对故障模式进行观察。
The technology of fault diagnosis can be divided into the method based on model and the method based on process history. The SDG model is deep knowledge model, which can express the complex cause and effect relations, and has very large capacity of containing process potential information. The SDG based fault diagnosis method is one of important branch of qualitative simulation, as well as an important fault diagnosis method. Based on the past research of SDG, The thesis systematically and deeply studies the SDG method including the modelling method and reasoning mechanism of SDG, especially on the laws of fault propagation of complex system.
To overcome the intrinsic defectiveness of SDG model and enhance the diagnosis resolution and speed, the thesis presents the method of combining the SDG method and expert system method. The thesis puts forward the method taking the SDG model directly as the deep knowledge base of the expert system, as well as expert knowledge and experience as shallow knowledge base. Finally the reasoning mechanism is studied and the algorithm is given.
In this thesis the SDGExpert software is realized in Visual Prolog language. It employs the three-layer architechture including the user-interface-layer, middle-layer and memory-layer from top to bottom. The software completely applies the object-oriented programming method and nice software design patterns which not only realizes the basic function of the expert system, but also achieves the reusability and expansibility for further development.
Finally, the diagnosis simulation of centrifuge pump system based on SDGExpert is carried out on the MPCE simulation platform. The result proved that the SDGExpert can find out the root fault with friendly user interface showing the fault.
为了克服SDG模型的固有缺陷,提高诊断分辨率和诊断速度,阐述了将SDG方法和专家系统方法相结合的思想,对基于SDG的故障诊断专家系统进行了详细研究,摒弃了传统由SDG模型推导产生式规则的浅层知识库设计方法,提出直接以SDG图模型深层知识作为所开发专家系统的深层知识库,同时将专家的知识和经验作为浅层知识库。最后对该专家系统的推理机算法进行了详细研究并且给出算法。
本文的SDGExpert专家系统是基于Visual Prolog语言实现的,SDGExpert专家系统采用了用户层、中间层、存储层的三层架构设计,完全应用了面向对象程序设计方法,应用了良好的软件设计模式,不仅实现了专家系统应有的功能,还最大程度的实现了SDGExpert专家系统的可复用和可扩展性,以应付以后的进一步开发。
文章最后在多功能仿真平台的液位离心泵仿真系统进行了SDGExpert专家系统案例仿真。结果证明该SDGExpert能够发现系统的故障根源,而且软件的友好的用户界面可以直观的对故障模式进行观察。
The technology of fault diagnosis can be divided into the method based on model and the method based on process history. The SDG model is deep knowledge model, which can express the complex cause and effect relations, and has very large capacity of containing process potential information. The SDG based fault diagnosis method is one of important branch of qualitative simulation, as well as an important fault diagnosis method. Based on the past research of SDG, The thesis systematically and deeply studies the SDG method including the modelling method and reasoning mechanism of SDG, especially on the laws of fault propagation of complex system.
To overcome the intrinsic defectiveness of SDG model and enhance the diagnosis resolution and speed, the thesis presents the method of combining the SDG method and expert system method. The thesis puts forward the method taking the SDG model directly as the deep knowledge base of the expert system, as well as expert knowledge and experience as shallow knowledge base. Finally the reasoning mechanism is studied and the algorithm is given.
In this thesis the SDGExpert software is realized in Visual Prolog language. It employs the three-layer architechture including the user-interface-layer, middle-layer and memory-layer from top to bottom. The software completely applies the object-oriented programming method and nice software design patterns which not only realizes the basic function of the expert system, but also achieves the reusability and expansibility for further development.
Finally, the diagnosis simulation of centrifuge pump system based on SDGExpert is carried out on the MPCE simulation platform. The result proved that the SDGExpert can find out the root fault with friendly user interface showing the fault.
引文
[1]叶银忠,潘日芳,蒋慰孙.动态系统的故障检测与诊断方法(综述)[J].信息与控制,1985,15(6):27-34
[2]周东华,孙优贤.控制系统的故障检测与诊断技术.北京:清华大学出版社,1994
[3]Lapp S A,Powers G J.Computer-aided synthesis of fault trees[J].IEEE Transactions on Reliability,1977,26(2):2-12
[4]Iri M,Aoki K,O'Shima E,et al.An algorithm for diagnosis of system failures in the chemical process[J].Computer and Chemical Engineering,1979,3(1-4):489-493
[5]Umeda T,Kuriyama T.A graphical approach to cause and effect analysis of chemical processing system[J].Chemical Engineering Science,1980,35:2379-2388
[6]Shiozaki J,Matsuyama H,Tano K,O'Shima E.Fault diagnosis of chemical processes by the use of signed,directed graphs.Extension to five-range patterns of abnormality[J].International Chemical Engineering,1985,25(4):651-659
[7]Tsuge Y,Shiozaki J,Matsuyama H,et al.Fesibility study of a fault diagnosis system for chemical plants[J].International Chemical Engineering,1985,25(4):660-667
[8]Kramer M A,Palowitch B L.A rule-based approach to fault diagnosis using the signed directed graph[J].AIChE Journal,1987,33(7):1067-1078
[9]Shiozaki J,Shibata B,Matsuyama H,et al.Fault diagnosis of chemical processes utilizing signed directed graphs-improvement by using temporal information[J].IEEE Transactions on Industrial Electronics,1989,36(4):469-474
[10]Chang C C,Yu C C.On-line fault diagnosis using the signed directed graph[J].Industrial and Engineering Chemistry Research,1990,29(7):1290-1299
[11]Yu C C,Lee C C.Fault diagnosis based on qualitative/quantitative process knowledge[J].AIChE Journal,1991,37(4):617-628
[12]Wang X Z,Yang S A,Yang S H,et al.The application of fuzzy qualitative simulation in safety and operability assessment of process plants[J].Computers Chem.Engng,1996,20:671-676
[13]Tarifa E E,Scenna N J.Fault diagnosis,Direct Graphs,and Fuzzy Logic[J].Computer and Chemical Engineering,1997,21(Suppl):649-654
[14]Vaidhyanathan R,Venkatasubramanian V.Experience with an expert system for automated HAZOP analysis[J].Computers Chem.Engng,1996,20:1589-1594
[15]Vedam H,Venkatasubramanian V.Signed digraph based multiple fault diagnosis[J].Computer and Chemical Engineering,1997,21(6):655-660
[16]Mylaraswamy D,Venkatasubramanian V.A hybrid framework for large scale process fault diagnosis[J].Computer and Chemical Engineering,1997,21(9):935-940
[17]Dash S,Venkatasubramanian V.Challenges in the industrial applications of fault diagnosis systems[J].Computer and Chemical Engineering,2000,24(2-7):785-791
[18]Venkatasubramanian V,Zhao J,Viswanathan S.Intelligent system for HAZOP analysis of complex process plants[J].Computer and Chemical Engineering,2000,24(9/10):2291-2302
[19]黄卫东,王克昌.基于定性和定量关系的液体火箭发动机故障诊断[J].航空动力学报,1996,11(3):281-284,331
[20]黄卫东,王克昌.基于定性模型的液体火箭发动机故障诊断[J].推进技术,1996,17(4):5-9
[21]黄卫东,王克昌.基于深层知识规则的液体火箭发动机故障诊断[J].宇航学报,1997,18(4):61-65
[22]李安峰,夏涛,张贝克,等.化工过程SDG建模方法[J].系统仿真学报,2003,15(10):1364-1368
[23]吴重光,夏涛,张贝克.基于符号定向图(SDG)深层知识模型的定性仿真[J].系统仿真学报,2003,15(10):1351-1355
[24]牟善军,姜春明,吴重光.SDG方法与过程安全分析的关系[J].系统仿真学报,2003,15(10):1381-1384
[25]杜殿林,张光红,吴重光.基于知识故障诊断系统所用的深层知识及SDG方法[J].化工自动化及仪表,2005,32(4):8-10
[26]李安峰,夏涛,张贝克,等.基于SDG的计算机辅助危险与可操作性分析[J].系统仿真学报,2003,15(10):1394-1397
[27]张钊谦.基于符号定向图模型的故障树自动建树研究[D].北京:北京化工大学,2005
[28]王强,吴重光,张贝克,等.基于SDG的传感器分布优化设计[J].江南大学学报(自然科学版),2006,5(4):391-393
[29]王强,吴重光,张贝克,等.基于SDG故障诊断的传感器分布优化设计[J].计算机仿真,2006,23(7):308-312
[30]王强.基于SDG的控制系统有效性方法研究[D].北京:北京化工大学,2006
[31]张贝克,夏涛,吴重光.集成化SDG建模、推理与信息处理软件平台[J].系统仿真学报,2003,15(10):1360-1363
[32]张贝克,夏涛,吴重光.SDG标准测试图及推理引擎论证[J]I系统仿真学报,2003,15(10):1369-1373
[33]夏涛,张贝克,吴重光.石油化工SDG故障诊断系统仿真实验系统[J].系统仿真学报,2003,15(10):1377-1380
[34]刘玥,张贝克,吴重光.基于模糊推理的SDG故障诊断新方法[J].计算机应用,2005,25(11):2661-2664
[35]张玎,吴重光.基于符号定向图的计算机辅助自动建模研究[J].计算机工程与设计,2005,26(4):882-884
[36]刘玥,张贝克,吴重光.基于模糊推理的SDG故障诊断新方法[J].计算机应用,2005,25(11):2661-2664
[37]刘列,张贝克,吴重光.基于模糊推理的计算机辅助SDG-HAZOP新方法[J].计算机工程与应用,2006,11:201-204
[38]杨帆,萧德云.SDG建模及其应用的进展[J].控制理论与应用,2005,22(5):767-774
[39]杨帆,萧德云.基于SDG的复杂系统故障传播规律分析[J].高技术通讯,2005,15(10):33-36
[40]杨帆,萧德云.大规模复杂系统的定性SDG建模方法[J].化工自动化及仪表,2005,32(5):8-11
[41]刘敏华.基于SDG模型的故障诊断及应用研究[D].北京:清华大学,2005
[42]刘敏华,萧德云.基于SDG模型和模糊融合的故障诊断方法[J].控制工程,2006,13(1):10-14
[43]刘敏华,萧德云.基于趋势分析和SDG模型的故障诊断[J].控制理论与应用,2006,23(2):306-310
[44]杨帆,萧德云.概率SDG模型及故障分析推理方法[J].控制与决策.2006,21(5):487-491,496
[45]杨帆,萧德云.结构残差在基于SDG故障分离中的应用[J].控制工程,2007,14(3):320-324
[46]曹文亮,王兵树,马永光,等.基于SDG深层知识模型的电站热力系统故障诊断方法[J].华北电力大学学报,2005,32(5):79-83
[47]曹文亮,王兵树,马良玉,等.主元统计法与符号有向图模型相结合的故障诊断方法[J].动力工程,2005,25(6):870-875
[48]曹文亮,王兵树,马良玉,等.基于SDG推理规则和定量分析的故障诊断技术[J].热能动力工程,2005,20(6):619-623
[49]曹文亮,王兵树,马良玉,等.基于改进SDG的电站热力系统故障诊断方法研究[J].中国电机工程学报,2005,25(23):124-128
[50]曹文亮,王兵树,马进,等.基于符号有向图节点定量化模型的故障诊断方法[J].中国电机工程学报,2006,26(7):67-71
[51]曹文亮.基于符号有向图的热力系统故障诊断方法研究[D].北京:华北电力大学,2006
[52]Maurya M R,Rengaswamy R,Venkatasubramanian V.A signed directed graph-based systematic framework for steady-state malfunction diagnosis inside control loops[J].Chemical Engineering Science,2006,61:1790-1810
[53]雷英杰,刑清华,孙金萍,等.Visual Prolog智能集成开发环境评述[J].空军工程大学学报(自然科学版),2002,3(5):39-43
[54]雷英杰,刑清华,王涛,等.人工智能(AI)程序设计(面向对象语言)[M].北京:清华大学出版社,2005
[55]GUIDE用户指南[EB/OL].http://www.ad.siemens.com.cn/sce/promotions/2OO6compete/download/mpce.pdf,2006.8
[2]周东华,孙优贤.控制系统的故障检测与诊断技术.北京:清华大学出版社,1994
[3]Lapp S A,Powers G J.Computer-aided synthesis of fault trees[J].IEEE Transactions on Reliability,1977,26(2):2-12
[4]Iri M,Aoki K,O'Shima E,et al.An algorithm for diagnosis of system failures in the chemical process[J].Computer and Chemical Engineering,1979,3(1-4):489-493
[5]Umeda T,Kuriyama T.A graphical approach to cause and effect analysis of chemical processing system[J].Chemical Engineering Science,1980,35:2379-2388
[6]Shiozaki J,Matsuyama H,Tano K,O'Shima E.Fault diagnosis of chemical processes by the use of signed,directed graphs.Extension to five-range patterns of abnormality[J].International Chemical Engineering,1985,25(4):651-659
[7]Tsuge Y,Shiozaki J,Matsuyama H,et al.Fesibility study of a fault diagnosis system for chemical plants[J].International Chemical Engineering,1985,25(4):660-667
[8]Kramer M A,Palowitch B L.A rule-based approach to fault diagnosis using the signed directed graph[J].AIChE Journal,1987,33(7):1067-1078
[9]Shiozaki J,Shibata B,Matsuyama H,et al.Fault diagnosis of chemical processes utilizing signed directed graphs-improvement by using temporal information[J].IEEE Transactions on Industrial Electronics,1989,36(4):469-474
[10]Chang C C,Yu C C.On-line fault diagnosis using the signed directed graph[J].Industrial and Engineering Chemistry Research,1990,29(7):1290-1299
[11]Yu C C,Lee C C.Fault diagnosis based on qualitative/quantitative process knowledge[J].AIChE Journal,1991,37(4):617-628
[12]Wang X Z,Yang S A,Yang S H,et al.The application of fuzzy qualitative simulation in safety and operability assessment of process plants[J].Computers Chem.Engng,1996,20:671-676
[13]Tarifa E E,Scenna N J.Fault diagnosis,Direct Graphs,and Fuzzy Logic[J].Computer and Chemical Engineering,1997,21(Suppl):649-654
[14]Vaidhyanathan R,Venkatasubramanian V.Experience with an expert system for automated HAZOP analysis[J].Computers Chem.Engng,1996,20:1589-1594
[15]Vedam H,Venkatasubramanian V.Signed digraph based multiple fault diagnosis[J].Computer and Chemical Engineering,1997,21(6):655-660
[16]Mylaraswamy D,Venkatasubramanian V.A hybrid framework for large scale process fault diagnosis[J].Computer and Chemical Engineering,1997,21(9):935-940
[17]Dash S,Venkatasubramanian V.Challenges in the industrial applications of fault diagnosis systems[J].Computer and Chemical Engineering,2000,24(2-7):785-791
[18]Venkatasubramanian V,Zhao J,Viswanathan S.Intelligent system for HAZOP analysis of complex process plants[J].Computer and Chemical Engineering,2000,24(9/10):2291-2302
[19]黄卫东,王克昌.基于定性和定量关系的液体火箭发动机故障诊断[J].航空动力学报,1996,11(3):281-284,331
[20]黄卫东,王克昌.基于定性模型的液体火箭发动机故障诊断[J].推进技术,1996,17(4):5-9
[21]黄卫东,王克昌.基于深层知识规则的液体火箭发动机故障诊断[J].宇航学报,1997,18(4):61-65
[22]李安峰,夏涛,张贝克,等.化工过程SDG建模方法[J].系统仿真学报,2003,15(10):1364-1368
[23]吴重光,夏涛,张贝克.基于符号定向图(SDG)深层知识模型的定性仿真[J].系统仿真学报,2003,15(10):1351-1355
[24]牟善军,姜春明,吴重光.SDG方法与过程安全分析的关系[J].系统仿真学报,2003,15(10):1381-1384
[25]杜殿林,张光红,吴重光.基于知识故障诊断系统所用的深层知识及SDG方法[J].化工自动化及仪表,2005,32(4):8-10
[26]李安峰,夏涛,张贝克,等.基于SDG的计算机辅助危险与可操作性分析[J].系统仿真学报,2003,15(10):1394-1397
[27]张钊谦.基于符号定向图模型的故障树自动建树研究[D].北京:北京化工大学,2005
[28]王强,吴重光,张贝克,等.基于SDG的传感器分布优化设计[J].江南大学学报(自然科学版),2006,5(4):391-393
[29]王强,吴重光,张贝克,等.基于SDG故障诊断的传感器分布优化设计[J].计算机仿真,2006,23(7):308-312
[30]王强.基于SDG的控制系统有效性方法研究[D].北京:北京化工大学,2006
[31]张贝克,夏涛,吴重光.集成化SDG建模、推理与信息处理软件平台[J].系统仿真学报,2003,15(10):1360-1363
[32]张贝克,夏涛,吴重光.SDG标准测试图及推理引擎论证[J]I系统仿真学报,2003,15(10):1369-1373
[33]夏涛,张贝克,吴重光.石油化工SDG故障诊断系统仿真实验系统[J].系统仿真学报,2003,15(10):1377-1380
[34]刘玥,张贝克,吴重光.基于模糊推理的SDG故障诊断新方法[J].计算机应用,2005,25(11):2661-2664
[35]张玎,吴重光.基于符号定向图的计算机辅助自动建模研究[J].计算机工程与设计,2005,26(4):882-884
[36]刘玥,张贝克,吴重光.基于模糊推理的SDG故障诊断新方法[J].计算机应用,2005,25(11):2661-2664
[37]刘列,张贝克,吴重光.基于模糊推理的计算机辅助SDG-HAZOP新方法[J].计算机工程与应用,2006,11:201-204
[38]杨帆,萧德云.SDG建模及其应用的进展[J].控制理论与应用,2005,22(5):767-774
[39]杨帆,萧德云.基于SDG的复杂系统故障传播规律分析[J].高技术通讯,2005,15(10):33-36
[40]杨帆,萧德云.大规模复杂系统的定性SDG建模方法[J].化工自动化及仪表,2005,32(5):8-11
[41]刘敏华.基于SDG模型的故障诊断及应用研究[D].北京:清华大学,2005
[42]刘敏华,萧德云.基于SDG模型和模糊融合的故障诊断方法[J].控制工程,2006,13(1):10-14
[43]刘敏华,萧德云.基于趋势分析和SDG模型的故障诊断[J].控制理论与应用,2006,23(2):306-310
[44]杨帆,萧德云.概率SDG模型及故障分析推理方法[J].控制与决策.2006,21(5):487-491,496
[45]杨帆,萧德云.结构残差在基于SDG故障分离中的应用[J].控制工程,2007,14(3):320-324
[46]曹文亮,王兵树,马永光,等.基于SDG深层知识模型的电站热力系统故障诊断方法[J].华北电力大学学报,2005,32(5):79-83
[47]曹文亮,王兵树,马良玉,等.主元统计法与符号有向图模型相结合的故障诊断方法[J].动力工程,2005,25(6):870-875
[48]曹文亮,王兵树,马良玉,等.基于SDG推理规则和定量分析的故障诊断技术[J].热能动力工程,2005,20(6):619-623
[49]曹文亮,王兵树,马良玉,等.基于改进SDG的电站热力系统故障诊断方法研究[J].中国电机工程学报,2005,25(23):124-128
[50]曹文亮,王兵树,马进,等.基于符号有向图节点定量化模型的故障诊断方法[J].中国电机工程学报,2006,26(7):67-71
[51]曹文亮.基于符号有向图的热力系统故障诊断方法研究[D].北京:华北电力大学,2006
[52]Maurya M R,Rengaswamy R,Venkatasubramanian V.A signed directed graph-based systematic framework for steady-state malfunction diagnosis inside control loops[J].Chemical Engineering Science,2006,61:1790-1810
[53]雷英杰,刑清华,孙金萍,等.Visual Prolog智能集成开发环境评述[J].空军工程大学学报(自然科学版),2002,3(5):39-43
[54]雷英杰,刑清华,王涛,等.人工智能(AI)程序设计(面向对象语言)[M].北京:清华大学出版社,2005
[55]GUIDE用户指南[EB/OL].http://www.ad.siemens.com.cn/sce/promotions/2OO6compete/download/mpce.pdf,2006.8