Feedforward Neural Networks Based Input-Output Models for Railway Carriage System Identification

Depending on the representation ability of neural networks for a nonlinear process, this paper develops procedures to locate the input space dimensions of feedforward neural networks for general practical nonlinear systems identification when only the outputs are accessible observations. The size of the input space is directly related to the involved system order. Hence based on our developed model, we are able to determine whether the system is faulty or not by monitoring the output error change. The methods are demonstrated in vibration signals that were measured by an accelerometer mounted on a traction centre of a railway carriage running at about 60 mph along the Hong Kong railway.

[1]  P. Vega,et al.  A neural networks based approach for fault detection and diagnosis: application to a real process , 1995, Proceedings of International Conference on Control Applications.

[2]  David M. Himmelblau,et al.  FAULT DIAGNOSIS IN COMPLEX CHEMICAL PLANTS USING ARTIFICIAL NEURAL NETWORKS , 1991 .

[3]  A. J. Morris,et al.  Artificial neural networks in process engineering , 1991 .

[4]  Kumpati S. Narendra,et al.  Identification Using Feedforward Networks , 1995, Neural Computation.

[5]  A. Bahgat,et al.  Fault detection and identification using a hierarchical neural network-based system , 1993, Proceedings of IECON '93 - 19th Annual Conference of IEEE Industrial Electronics.

[6]  P. J. Dalianis,et al.  Fault diagnosis in complex systems using artificial neural networks , 1994, 1994 Proceedings of IEEE International Conference on Control and Applications.

[7]  Rolf Isermann,et al.  Process fault detection based on modeling and estimation methods - A survey , 1984, Autom..

[8]  Paul M. Frank,et al.  Fault diagnosis in dynamic systems: theory and application , 1989 .

[9]  Lee A. Feldkamp,et al.  Neurocontrol of nonlinear dynamical systems with Kalman filter trained recurrent networks , 1994, IEEE Trans. Neural Networks.

[10]  Jitendra K. Tugnait,et al.  Consistent order selection for noncausal autoregressive models via higher-order statistics , 1990, Autom..

[11]  Vladik Kreinovich,et al.  Arbitrary nonlinearity is sufficient to represent all functions by neural networks: A theorem , 1991, Neural Networks.

[12]  K S Narendra,et al.  Control of nonlinear dynamical systems using neural networks. II. Observability, identification, and control , 1996, IEEE Trans. Neural Networks.

[13]  Kumpati S. Narendra,et al.  Identification and control of dynamical systems using neural networks , 1990, IEEE Trans. Neural Networks.

[14]  Tommy W. S. Chow,et al.  A least third-order cumulants objective function , 1996, Neural Processing Letters.

[15]  Jie Chen,et al.  Fault diagnosis in nonlinear dynamic systems via neural networks , 1994 .

[16]  T. L. Ruchti,et al.  Kalman based artificial neural network training algorithms for nonlinear system identification , 1993, Proceedings of 8th IEEE International Symposium on Intelligent Control.

[17]  Timo Sorsa,et al.  Neural networks in process fault diagnosis , 1991, IEEE Trans. Syst. Man Cybern..

[18]  S. Fararooy,et al.  On-line condition monitoring of railway equipment using neural networks , 1995 .

[19]  Chrysostomos L. Nikias,et al.  Higher-order spectral analysis , 1993, Proceedings of the 15th Annual International Conference of the IEEE Engineering in Medicine and Biology Societ.

[20]  E. Zafiriou,et al.  Use of neural networks for sensor failure detection in a control system , 1990, IEEE Control Systems Magazine.

[21]  C. L. Nikias,et al.  Higher-order spectra analysis : a nonlinear signal processing framework , 1993 .