论文信息 - A Stochastic Approach to Dynamic System Modeling With Application to Gas Turbines

A Stochastic Approach to Dynamic System Modeling With Application to Gas Turbines

Dynamic system mathematical models are a cornerstone for model based fault diagnostic techniques. However, achieving a robust mathematical model, based only on analytical derivations, cannot always be a realistic task especially for dynamic systems possessing physical parameter uncertainties and modeling errors. One attractive alternative is the huge capacity of stochastic methods to model non-deterministic highly complicated dynamic systems. In this paper, a general sensor-based modeling strategy is presented using stochastic techniques with application to a stationary gas turbine. The turbine sensors are monitored and recorded in real time to implement a signal modeling technique based on time series principles and turbine operation states. A subsequent recognition strategy, based on the turbine operation states and the Markov chain approach, is implemented to estimate the different signals. To experimentally validate the developed technique, input and output sensor data for a 4.5 MW stationary gas turbine has been recorded and analyzed during the critical start up phase. The proposed modeling strategy offers a considerable capability to recreate an unobserved signal from the observation of the other signals.Copyright © 2005 by ASME

John R. Wagner | Hanny Bassily

[1] Hasan Al-Nashash,et al. EEG signal modeling using adaptive Markov process amplitude , 2004, IEEE Transactions on Biomedical Engineering.

[2] Peter J. Fleming,et al. System identification strategies applied to aircraft gas turbine engines , 2000 .

[3] Björn Olsson,et al. A hybrid method for protein sequence modeling with improved accuracy , 2001, Inf. Sci..

[4] Lawrence R. Rabiner,et al. A tutorial on hidden Markov models and selected applications in speech recognition , 1989, Proc. IEEE.

[5] Chiman Kwan,et al. A novel approach to fault diagnostics and prognostics , 2003, 2003 IEEE International Conference on Robotics and Automation (Cat. No.03CH37422).

[6] Yongyong He,et al. Hidden Markov model-based fault diagnostics method in speed-up and speed-down process for rotating machinery , 2005 .

[7] Jong Min Lee,et al. Diagnosis of mechanical fault signals using continuous hidden Markov model , 2004 .

[8] G. G. Kulikov,et al. Application of Markov Chains to Identification of Turbine Engine Dynamic Models , 2002 .

[9] Paul M. Frank,et al. Issues of Fault Diagnosis for Dynamic Systems , 2010, Springer London.

[10] Paul D. Gader,et al. Generalized hidden Markov models. II. Application to handwritten word recognition , 2000, IEEE Trans. Fuzzy Syst..

[11] Yangsheng Xu,et al. Hidden Markov model-based learning controller , 1994, Proceedings of 1994 9th IEEE International Symposium on Intelligent Control.