Fault diagnosis of non-linear dynamical systems using analytical and soft computing methods

• Autorzy: Korbicz J.
• Języki publikacji: EN

Abstrakty

EN

The paper deals with the problems of robust fault detection using analytical methods (observers and unknown input observers) and soft computing techniques (neural networks, neuro-fuzzy networks and genetic programming). The model-based approach to Fault Detection and Isolation (FDI) is considered. In particular, observers for non-linear Lipschitz systems and extended unknown input observers are discussed. In the case of soft computing techniques, the main objective is to show how to employ the bounded-error approach to determine the uncertainty of the GMDH and neuro-fuzzy networks. It is shown that based on soft computing models uncertainty defined as a confidence range for the model output, adaptive thresholds can be defined. The final part of the paper presents two illustrative examples that confirm the effectiveness of the unknown input observers and the neuro-fuzzy networks approaches.

Słowa kluczowe

EN

fault detection; unknown input observer; dynamical neural networks; neuro-fuzzy system; evolutionary algorithms

• Wydawca: Łukasiewicz Industrial Research Institute for Automation and Measurements PIAP
• Czasopismo: Journal of Automation Mobile Robotics and Intelligent Systems
• Rocznik: 2007
• Tom: Vol. 1, No. 1
• Strony: 7--23
• Opis fizyczny: Bibliogr. 90 poz., rys.

Twórcy

autor

Korbicz J.

• Deputy Rector for Scientific Research and International Cooperationthe and the professor at the Institute of Control and Computation Engineering, University of Zielona Góra,

Bibliografia

• [1] Abonyi J., Babuška R. and Szeifert F., "Modified Gath-Geva fuzzy clustering for identification of Takagi-Sugeno fuzzy models", IEEE Trans. Systems, Man, and Cybernetics, 2002,Part B, vol. 32, no. 5, pp. 612-621.
• [2] Atkinson A.C. and Donev A.N.,Optimum Experimental Designs, New York: Oxford University Press, 1992.
• [3] Babuška R., Fuzzy Modeling for Control, London: Kluwer Academic Publisher, 1998.
• [4] Blanke M., Kinnaert M., Lunze J. and Staroswiecki M.,Diagnosis and Fault-Tolerant Control, New York: Springer-Verlag, 2003.
• [5] Boutayeb M. and Aubry D., "A strong tracking extended Kalman observer for nonlinear discrete-time systems",IEEE Trans. Automatic Control, 1999, vol. 44, no. 8, pp.1550-1556.
• [6] Calado J.M.F., Sa da Costa J.M.G., Bartys M. and Korbicz J., "FDI approach to the DAMADICS benchmark problem basedon qualitative reasoning coupled with fuzzy neural networks",Control Engineering Practice , 2006, vol. 14, no.6, pp. 685-698.
• [7] Chen J. and Patton R.J.Robust Model Based Fault Diagnosis for Dynamic Systems,London: Kluwer Academic Publishers,1999.
• [8] Chen J-Q., Xi Y-G. and Zhang Z-J., "A clustering algorithmfor fuzzy model identification", Fuzzy Sets and Systems,1998, vol. 98, pp. 319-329.
• [9] DAMADICS (2002):Website of the RTN DAMADICS: Development and Application of Methods for Actutator Diagnosis in Industrial Control Systems.At: http://diag.mchtr.pw.edu.pl/damadics
• [10] Delebecque F., Nikoukah R. and Rubio Scola H., "Test signal design for failure detection: A linear programming approach" - Int. J. Appl. Math. and Comput. Sci., 2003, vol.13, no. 4, pp. 515-526
• [11] Esparcia-Alcazar A.I. (1998): Genetic Programming for Adaptive Digital Signal Processing, Ph.D. dissertation,Glasgow University.
• [12] Frank P.M., (1990):“Fault diagnosis in dynamic systems using analytical and knowledge-based redudancy - a survey and some new results” Automatica, 1990, vol. 26, pp. 459-474.
• [13] Frank P.M. and Köppen-Seliger B., "New developments using AI in fault diagnosis",Artificial Intelligence, 1997,vol. 10, no. 1, pp. 314.
• [14] Frank P.M., Schreier G. and Garcia E.A.,Nonlinear observers for fault detection and isolation, New Directions in Nonlinear Observer Design (Nijmeijer H. and T. Fossen, (Eds.), Berlin: Springer-Verlag, 1999.
• [15] Gertler J.,Fault Detection and Diagnosis in Engineering Systems, New York: Marcel Dekker, 1998.
• [16] Goldberg D.E.,Genetic Algorithms in Search, Optimization and Machine Learning, Massachusetts: Addison-Wesley,1989.
• [17] Gupta M.M., Jin L. and Homma N.,Static and Dynamic Neural Networks. From Fundamentals to Advanced Theory, New Jersey: Wiley, 2003.
• [18] Haykin S., Neural Networks. A Comprehensive Foundation, New Jersey: Prentice-Hall, 1999.
• [19] Hui S.and Zak S.H., "Observer design for systems with unknown input", Int. J. Appl. Math. and Comput. Sci., 2005,vol. 15, no. 4, pp. 431-446.
• [20] Isermann R.,Fault Diagnosis Systems. An Introduction from Fault Detection to Fault Tolerance, New York: Springer-Verlag, 2006.
• [21] Ivakhnenko A.G. and Mueller J.A., "Self-organizing of nets of active neurons”, System Analysis Modelling Simulation, vol. 20, pp. 93-106.
• [22] Janczak A., "Identification of Nonlinear Systems Using Neural Networks and Polyno-mial Models. A Block-oriented Approach",Lecture Notes in Control and Information Sciences, 2005, vol. 310, Berlin: Springer-Verlag.
• [23] Karpenko M., Sepehri N. and Scuse D., "Diagnosis of process valve actuator faults using multilayer neural network",Control Engineering Practice, 2003, vol. 11, pp. 1289-1299.
• [24] Köppen-Seliger B. and Frank P.M., "Fuzzy logic and neural networks in fault detection", In: Fusion of Neural Networks,Fuzzy Sets, and Genetic Algorithms(L. Jain and N. Martin,Eds.), New York: CRC Press, 1999, pp. 169-209.
• [25] Korbicz J., "Fault detection using analytical and soft computing methods",Bulletin of the Polish Academy of Sciences: Technical Sciences, 2006, vol. 54, no. 1, pp.75-88.
• [26] Korbicz J. (Ed.), "Soft Computing in Control and Fault Diagnosis", Int. J. Appl. Math. and Comput. Sci.,2006a, vol. 16, no. 1, pp. 7-114
• [27] Korbicz J., "Computer designing of diagnostic systems",In: Engineering of Machine Diagnostics(Zółtowski B. and Cz. Cempel, Eds.), Radom: Instytut Technologii Eksploatacji (Institute for Sustainable Technologies in Radom,Poland), 2004, pp. 453-474, (in Polish).
• [28] Korbicz J. and Bidyuk P.I.,State and Parameter Estimation.Digital and Optimal Filtering, Applications, Technical University of Zielona Góra Press, 1993.
• [29] Korbicz J. and Kus J. (1999): "Dynamic GMDH neural networks and their application in fault detection systems",Proc. European Control Conference, ECC, Karlshure,Germany, 31 August3 Sept., CD ROM.
• [30] Korbicz J. and Mrugalski M. (2007): "Confidence estimation of GMDH neural networks and its application in fault detection systems", Int. J. System Science, (accepted).
• [31] Korbicz J. and Kowal M., "Neuro-Fuzzy Systems in Processes Diagnostics", In: Fuzzy Sets and their Applications(Chojcan J. and Leski J., Eds.), Gliwice, Poland: Wydawnictwo Politechniki Slaskiej (Silesian University of Technology Press), 2001, pp. 333-379, (in Polish).
• [32] Korbicz J., Koscielny J.M., Kowalczuk Z. and Cholewa W.(Eds.),Fault Diagnosis. Models, Artificial Intelligence,Applications, Berlin: Springer-Verlag, 2004.
• [33] Korbicz J., Koscielny J.M., Kowalczuk Z. and Cholewa W.(Eds.),Processes Diagnosis. Models, Methods of Artificial Intelligence, Applications, Warsaw: Wydawnictwo Naukowo Techniczne, WNT, 2002, (in Polish).
• [34] Korbicz J., Obuchowicz A. and Ucinski D.,Artificial Neural Networks. Fundations and Applications, Warsaw: Akademicka Oficyna Wydawnicza EXIT, 1994, (in Polish).
• [35] Korbicz J., Patan K. and Obuchowicz A., "Neural network fault detection system for dynamic processes",Bulletin of the Polish Academy of Sciences. Technical Sciences, 2001, vol. 49, no. 2, pp. 301-321.
• [36] Korbicz J., Patan K. and Obuchowicz A., "Dynamic neural networks for process modelling in fault detection and isolation systems",Int. J. Appl. Math. and Comput. Sci. ,1999, vol. 9, no. 3, pp. 510-546.
• [37] Korbicz J., Witczak M. and Puig V., "LMI-based strategies for designing observers and unknown input observers for non-linear discrete-time systems",Bulletin of the Polish Academy of Sciences, Technical Sciences, 2007, vol. 55, no. 1, pp. 31-42.
• [38] Kościelny J.M.,Diagnostics of Automatic IndustrialProcesses. Warsaw: Akademicka Oficyna Wydawnicza EXIT,2001, (in Polish).
• [39] Kowal M., "Optimization of Neuro-fuzy Structures in Technical Diagnostics",Lecture Notes in Control and Computer Science, 2005, vol. 9, University of Zielona Góra Press.
• [40] Kowal M. and Korbicz J., "Robust fault detection using neuro-fuzzy models", Przegląd Elektrotechniczny, 2006, no.1, pp. 32-36.
• [41] Kowal M. and Korbicz J., "Robust fault detection under fuzzy-neural model uncertainty",Pomiary, Automatyka,Kontrola, 2005a, no. 9, pp. 93-95, (in Polish).
• [42] Kowal M. and Korbicz J., "Neuro-fuzzy structures in FDI system", Proc. 16-th IFAC World Congress, 2005b, Prague,Czech Republic, CD-ROM.
• [43] Kowal M. and Korbicz J., "Self-organizing Takagi-Sugeno fuzzy model in a fault detection system",Proc. 6-th Nat.Conf. Diagnostics of Industrial Processes 2003 DPP'03, Władysławowo, Poland, pp. 253-258, (in Polish).
• [44] Kowal M. and Korbicz J., Fault detection using neuro-fuzzy networks, Proc. 14-th Polish Conf. Automatic Control, KKA,Zielona Góra, Poland, 2002a, pp. 595-600, (in Polish).
• [45] Kowal M. and Korbicz J., "Faul detection using Takagi-Sugeno neuro-fuzzy networks and ARX models",Proc. Nat.Conf. Fuzzy Systems, 2002b , Cracow, Poland, pp. 198-205.
• [46] Kowal M., Korbicz J., Mendes M.J.G.C. and Calado J.M.F.,"Fault detection using neuro-fuzzy networks",Systems Science, 2002, vol. 28, no. 1, pp. 45-57.
• [47] Kowalczuk Z. and Białaszewski (2006): "Nichingmechanisms in evolutionary computation",Int. J. Appl. Math. and Comput. Sci, vol. 16, no. 1, pp. 59-84.
• [48] Koza J.R.,Genetic Programming: On the Programming of Computers by Means of Natural Selection, Cambridge: The MIT Press, 1992.
• [49] Leith D.J. and Leithead W.E., "Analytic framework for blended multiple model systems using local models",Int.J. Control, 1999, vol. 72, no. 7/8, pp. 605-619.
• [50] Mendes M.J.G.C., Kowal M., Korbicz J. and Sa da Costa J.M.G., "Neuro-fuzzy structures in FDI system",Proc. 15-th IFAC Triennial World Congress, 2002, Barcelona, Spain,CD-ROM.
• [51] Michalewicz Z.,Genetic Algorithms + Data Structures =Evolution Programs, Berlin: Springer-Verlag, 1996.
• [52] Milanese M., Norton J., Piet-Lahanier H. and Walter E.Bounding Approaches to System Identification,Bounding Approaches to System Identification, New York:Plenum Press, 1996.
• [53] Mrugalski M., Neural Network Based Modelling of Non-linear Systems in Fault Detection Schemes, Ph.D. thesis, 2004,Faculty of Electrical Engineering, Computer Science and Telecommunications, University of Zielona Góra, (in Polish).
• [54] Mrugalski M.,Neural Network Based Modelling of Nonlinear Systems in Fault Detection Schemes, Ph.D. thesis, 2003,University of Zielona Góra, Faculty of Electrical Engineering,Computer Sience and Telecommunications, (in Polish).
• [55] Mrugalski M. and Korbicz J., "Application of the MLP neural network to the robust fault detection",Proc. 6-th IFAC Symp. Fault Detection Supervision and Safety of Technical Processes, SAFEPROCESS, 2006, Beijing, China, CD-ROM.
• [56] Mrugalski M. and Korbicz J. (2005):Robust fault detection via GMDH neural networks, Proc. 16-th IFAC World Congress Prague, Czech Republic, CD-ROM.
• [57] Mueller J.E. and Lemke F., Self-organising Data Mining,Hamburg: Libri, 2000.
• [58] Nelles O.,Nonlinear System Identification. From Classical Approaches to Neural Networks and Fuzzy Models, Berlin: Springer-Verlag, 2001.
• [59] Norgard M., Ravn O., Poulsen N. and Hansen L., Networks for Modelling and Control of Dynamic Systems, London: Springer-Verlag, 2000.
• [60] Obuchowicz A.,Evolutionary Algorithms for Global Optimization and Dynamic System Diagnosis, Zielona Góra: Lubuskie Towarzystwo Naukowe, 2003.
• [61] Osowski S.,Artificial Neural Networks for Information Processing, Warsaw: Oficyna Wydawnicza Politechniki Warszawskiej (Warsaw University of Technology Press),2006, (in Polish).
• [62] Patan K., "Stability analysis and the stabilization of a class of discrete-time dynamic neural networks",IEEE Trans.Neural Networks, 2007, vol. 18, (accepted).
• [63] Patan K., Korbicz J. and Prętki P., "Global stability conditions of locally recurrent neural networks",Lecture Notes in Computer Science, 2005, vol. 3697, pp. 191-196.
• [64] Patan K. and Parisini T., "Identification of neural dynamic models for fault detection and isolation: The case of a real sugar evaporation process", J. Process Control, 2005, vol.15, pp. 67-79.
• [65] Patton R.J. and Korbicz J. (Eds.), "Advances in Computational Intelligence for Fault Diagnosis Systems",Int. J. Appl. Math. and Comput. Sci., Special Issue, 1999,vol. 9, no. 3, pp. 471-735.
• [66] Patton R.J., Frank P.M. and Clark R.N.,Issues of Fault Diagnosis for Dynamic Systems, Berlin: Springer-Verlag,2000.
• [67] Patton R.J., Korbicz J. and Lesecq S. (Eds.), "A Benchmark Study of Fault Diagnosis for an Industrial Actuator",Control Engineering Practice, 2006, vol. 14, no. 6, pp. 575-717.
• [68] Patton R.J., Korbicz J., Witczak M. and Uppal F., "Combined computational intelligence and analytical methods in fault diagnosis", In: Intelligent Control Systems using Computational Intelligence Techniques (Rauno A.E. (Ed.), London: The IEE Press, 2005, pp. 349-392.
• [69] Piegat A. (2003): Modelling and Fuzzy Systems, Warsaw: Akademicka Oficyna Wydawnicza EXIT, (in Polish).
• [70] Rodrigues M., Theilliol D., Aberkane S. and Sauter D., "Fault tolerant control design for polytopic LPV systems", Int. J. Appl. Math. and Comput. Sci., 2007, vol. 17, no. 1, pp. 27-37.
• [71] Rutkowska D.,Neuro-Fuzzy Architectures and Hybrid Learning, Heidelberg, New York: Physica-Verlag, Springer-Verlag Company, 2002.
• [72] Rutkowski L.,Methods and Techniques of Artificial Intelligence, Warsaw: Wydawnictwo Naukowe PWN, 2005,(in Polish).
• [73] Tadeusiewicz R., Neural Networks, Warsaw: Akademicka Oficyna Wydawnicza, 1993, (in Polish).
• [74] Thau F.E., "Observing the state of non-linear dynamic systems", Int. J. Control , 1973, vol. 17, no. 3, pp, 471-479.
• [75] Uciński D.,Optimal Measurement Methods for Distributed Parameter System Identification, Boca Raton: CRC Press,2005.
• [76] Uppal F., Patton R.J. and Witczak M., "A neuro-fuzzy multiple-model observer approach to robust fault diagnosis based on the DAMADICS benchmark problem",Control Engineering Practice, 2006, vol. 14, no. 6, pp. 699-717.
• [77] Walter E. and Pronzato L.,Identification of Parametric Models from Experimental Data, London: Springer, 1997.
• [78] Wang S.H., Davison E.J. and Dorato P., "Observing the states of systems with unmeasurable disturbances",IEEE Trans. Automatic Control , 1975, vol. 20, no. 5, pp. 716-717.
• [79] Witczak M., "Modelling and Estimation Strategies for Fault Diagnosis of Non-linear Systems",Lecture Notes in Control and Information Sciences, 2007, no. 354, Berlin, Heidelberg:Springer.
• [80] Witczak M., "Advances in model-based fault diagnosis with evolutionary algorithms and neural networks",Int. J. Appl.Math. and Comput. Sci., 2006a, vol. 16, no. 1, pp. 85-99.
• [81] Witczak M., "Toward the training of feed-forward neural networks with the D-optimum input sequence",IEEE Trans.Neural Networks, 2006b, vol. 17, no. 2, pp. 357-373.
• [82] Witczak M., "Identification and Fault Detection of Nonlinear Dynamic Systems", Serie: Lecture Notes in Control and Computer Science, Technical University of Zielona Góra Press, 2003, vol. 1.
• [83] Witczak M. and Korbicz J., "Design of observers for Lipschitz non-linear discrete-time systems", Proc. 14-th IFAC Symp. System Identification, SYSID, Newcastle,Australia, 2006, pp. 985-990, CD-ROM.
• [84] Witczak M. and Korbicz J., "Observers and genetic programming in the identification and fault diagnosis of non-linear dynamic systems", In: Fault Diagnosis. Models,Artificial Intelligence, Applications (Korbicz J., Kościelny J.M., Kowalczuk Z. and Cholewa W., (Eds.), Berlin: Springer-Verlag, 2004, pp. 457-509.
• [85] Witczak M., Korbicz J., Mrugalski M. and Patton R.J.,"A GMDH neural network-based approach to robust fault diagnosis: application to the DAMADICS benchmark problem", Control Engineering Practice, 2006a, vol. 14,no. 6, pp. 671-683.
• [86] Witczak M., Korbicz J. and Puig V., "An LMI approach to designing observers and unknown input observers for nonlinear systems",Proc. 6-th IFAC Symp. Fault Detection Supervision and Safety of Technical Processes, SAFEPROCESS,2006b, Beijing, China, CD-ROM.
• [87] Witczak M., Obuchowicz A. and Korbicz J., "Genetic programming based approaches to identification and fault diagnosis of non-linear dynamic systems", Int. J. Control,2002, vol. 75, no. 13, pp. 1012-1031.
• [88] Witczak M., Obuchowicz A. and Korbicz J., "Design of nonlinear state observers using genetic programming",Proc.3-rd Nat. Conf. Evolutionary Algorithms and GlobalOptimization , 1999, Potok Złoty, Poland, pp. 345-352.
• [89] Witczak M., Prętki P., Korbicz J. and Puig V., "Design of an extended unknown input observer",Proc. Int. Workshop Advanced Control Diagnosis, ACD, 2006, Nancy, France,CD-ROM.
• [90] Zolghadri A., Henry D. and Monsion M., "Design of nonlinear observers for fault diagnosis. A case study", Control Engineering Practice, 1996, vol. 4, no. 11, pp. 1535-1544. .