Narx model in rotating machinery diagnostics

• Autorzy: Bednarz J. , Barszcz T. , Uhl T.

Warianty tytułu

PL

Zastosowanie modelu Narx w diagnostyce maszyn wirnikowych

• Języki publikacji: EN

Abstrakty

EN

Rotating machines are often described using linear methods with acceptable accuracy. Some malfunctions, however, are of non-linear nature. The most common examples of those malfunctions are loose bearings and rotor rubs. Accurate detection and identification of such malfunctions requires more accurate methods. One of such methods can be NARX - nonlinear systems identification. This method is based on neural networks approach and is especially efficient in modeling and diagnostics of nonlinear systems. Application of this method leads to shorter and less costly tuning of the model to the object, which is the key requirement when practical application of a method is concerned. The paper presents how NARX can be applied for modeling of rotating machinery malfunctions. Idea of the diagnostic algorithm based on such modeling is presented. The proposed algorithm was verified during research on a specialized test rig, which can generate vibration signals. The paper also presents results of an application of the NARX method for data collected at a wind turbine.

PL

Maszyny wirnikowe są często opisywane przy użyciu metod liniowych z zadowalającą dokładnością. Niektóre awarie mają jednak charakter nieliniowy; jako przykład można wymienić m.in. takie usterki jak luźne łożyska i przytarcia wirnika. Precyzyjne wykrywanie i identyfikacja ww. usterek wymaga bardziej dokładnych metod. Jedną z takich metod może by NARX - model umożliwiający nieliniową identyfikację systemów. Metoda ta opiera się na sieciach neuronowych i jest szczególnie skuteczna w zakresie modelowania i diagnostyki układów nieliniowych. Zastosowanie tej metody prowadzi do krótszego i mniej kosztownego dostrajania modelu do obiektu, co jest kluczowym wymogiem w przypadku praktycznego zastosowania tej metody w diagnostyce. W artykule przedstawiono, jak modele NARX mogą być stosowane do modelowania awarii maszyn wirnikowych, a także ideę algorytmu diagnostycznego z wykorzystaniem takiego modelowania. Proponowany algorytm został zweryfikowany podczas badań na specjalnym stanowisku badawczym, które umożliwią generowanie sygnałów drganiowych oraz wprowadzanie symulowanych uszkodzeń. W artykule przedstawiono również wyniki zastosowania metody NARX dla danych zarejestrowanych w elektrowni wiatrowej.

Słowa kluczowe

EN

NARX model; rotating machinery diagnostics; wind-turbine damage detection

PL

model NARX; diagnostyka maszyn wirnikowych; wykrywanie uszkodzeń turbin wiatrowych

• Wydawca: AGH University of Science and Technology Press
• Czasopismo: Mechanics and Control
• Rocznik: 2011
• Tom: Vol. 30, no. 2
• Strony: 41--52
• Opis fizyczny: Bibliogr. 33 poz., rys., wykr.

Twórcy

autor

Bednarz J.

autor

Barszcz T.

autor

Uhl T.

• AGH University of Science and Technology, Faculty of Mechanical Engineering and Robotics, Department of Robotics and Mechatronics, al. A. Mickiewicza 30, 30-059 Krakow, Poland,

Bibliografia

• Barszcz T., Czop P., Uhl T. 2004, System identification and its limitations relating to the diagnosis of rotating machinery faults. Proc. of 10th IEEE International Conference Methods and Models in Automation and Robotics, Międzyzdroje, Polska.
• Barszcz T., Uhl T., Bednarz J. 2006, Nonlinear system identification in diagnostics of rotating machinery. Proc. of the 7th IFToMM-Conference on Rotor Dynamics, Vienna, Austria.
• Barszcz T. 2005, Nonlinear system identification for diagnostic of turbine control system. Proc. of 11th IEEE International Conference Methods and Models in Automation and Robotics, Międzyzdroje, Polska.
• Basso M., Giarre L., Groppi S., Zappa G. 2005, NARX model of an industrial power plant gas turbine. IEEE Transaction on control systems technology, vol. 13, No. 4.
• Bednarz J., Barszcz T, Uhl T. 2007, Rotating machinery diagnostics based on NARX models. CAMES: Computer Assisted Mechanics and Engineering Sciences, vol. 14, No. 4, pp. 557-567.
• Bednarz J., Barszcz T, Uhl T 2006, Application of OMAX method and NARX models for rotating machinery diagnostics. Pomiary Automatyka Kontrola, No. 5, pp. 23-26.
• Chen S., Bilings SA. 1989, Representations of non-linear systems: The NARMAX model. International Journal of Control, vol. 49, pp. 1013-1032.
• Eisenmann R.C. 1997, Machinery Malfunction Diagnosis and Correction. Hewlett Packard Professional Books.
• Goldman P., Muszyńska, A. 1994, Chaotic Behavior of Rotor/Stator Systems with Rubs. Journal of Engineering for Gas Turbines and Power, vol. 116, pp. 692-701.
• Grybos R. 1994, Rotating machinery Dynamics. Wydawnictwo Naukowe PWN, Warsaw.
• Jing X.J., Lang Z.Q., Billings S.A. 2007, New bound characteristics of NARX model in the frequency domain. International Journal of Control, vol. 80, pp. 140-149.
• Kerschen G., Worden K., Vakakis A.F., Golinval J.C. 2006, Past, present and future of nonlinear system identification in structural dynamics. Mechanical Systems and Signal Processing, vol. 20, pp. 505-592.
• Kicinski J. 2005, Rotordynamics. Wydawnictwo IMP PAN, Gdansk.
• Kicinski J. 2004, Non-linear vibrations as a new diagnostic tool - crack detection example. Diagnostics, vol. 30, pp. 249-256.
• Korbicz J., Obuchowicz A., Ucinski D. 1994, Artificial neural networks. Akademicka Oficyna Wydawnicza, Warsaw.
• Koscielny J.M. 2001, Diagnosis of automated processes. Wydawnic-two EXIT, Warszawa.
• Kraemer E. 1993, Dynamics of Rotors and Foundations. Springer-Verlag.
• Leonards I.J., Billings S.A. 1985, Input-output parametric model for non-linear systems: Part I: deterministic non-linear systems. International Journal of Control, vol. 41, No. 2, pp. 303-328.
• Leonards I.J., Billings S.A. 1985, Input-output parametric model for non-linear systems: Part II: stochastic non-linear systems. International Journal of Control, vol. 41, No. 2, pp. 329-344.
• Lorenz E.N. 1963, Deterministic nonperiodic flow. Journal of Atmospheric Sciences, vol. 20.
• Muszyńska A. 2005, Rotordynamics. Taylor & Francis Group, USA.
• Nalinaksh S.V., Satishkumar D. 2001, Artificial neural network design for fault identification in a rotor-bearing system. Mechanism and Machine Theory, vol. 36, pp. 157-175.
• Narendra K.S., Parthasarathy K. 1990, Identification and control of dynamical systems using neural networks. IEEE Transactions on Neural Networks, vol. 1, No. 1, pp. 4-27.
• Norgaard M., Ravn O., Poulsen N.K., Hansen L.K. 2000, Neural Networks for Modeling and Control of Dynamic Systems. Springer, Londyn.
• Norgaard M. 2000, Neural Network Based System Identification Tool-box. Tech. Report. 00-E-891. Department of Automation, Technical University of Denmark.
• Oberholster A.J., Heyns P.S. 2006, On-line fan blade damage detection using neural network. Mechanical Systems and Signal Pro-cessing, vol. 20, pp. 78-93.
• Orlowski Z. 2001, Diagnostics in the life of steam turbines. WNT, Warsaw.
• Peng Z.K., Lang Z.Q., Billings S.A. 2007, Crack detection using nonlinear output frequency response function. Journal of Sound and Vibrations, vol. 301, pp. 777-788.
• Piroddi L., Spinelli W. 2003, An identification algorithm for polynomial NARX models based on simulation error minimization. International Journal of Control, vol. 76, No. 17, pp. 1767-1781.
• Sanz J., Perera R., Huerta C. 2007, Fault diagnosis of rotating machinery based on auto-associative neural networks and wavelet trans-forms. Journal of Sound and Vibration, vol. 302, pp. 981-999.
• Smit W.G., Heyns P.S. 2002, Fan blade damage detection using on-line vibration monitoring. R and D Journal, vol. 18, No. 3, pp. 77-90.
• Stoisser CM., Audebert S. 2008, A comprehensive theoretical nume-rical and experimental approach for crack detection in power plant rotating machinery. Mechanical Systems and Signal Pro-cessing, vol. 22, pp. 818-844.
• Thouverez F., Jezequel L. 1996, Identification of NARMAX models on a modal base. Journal of Sound and Vibration, vol. 189, No. 2, pp. 193-213.