Feedforward Neural Networks Based Input-Output Models for RailwayCarriage System Identification

Authors:
Tommy W. S. Chow;Oulian Shuai
Affiliations:
Department of Electronic Engineering, City University of Hong Kong, Hong Kong. E-mail: eetchow@cityu.edu.hk;Department of Electronic Engineering, City University of Hong Kong, Hong Kong. E-mail: eetchow@cityu.edu.hk
Venue:
Neural Processing Letters
Year:
1997

Citing 6
Cited 1

Fault diagnosis in dynamic systems: theory and application

Fault diagnosis in dynamic systems: theory and application
Consistent order selection for noncausal autoregressive models via higher-order statistics

Automatica (Journal of IFAC)
Arbitrary nonlinearity is sufficient to represent all functions by neural networks: a theorem

Neural Networks
Identification using feedforward networks

Neural Computation
A least third-order cumulants objective function

Neural Processing Letters
Control of nonlinear dynamical systems using neural networks. II. Observability, identification, and control

IEEE Transactions on Neural Networks

Formal approach to railway applications

Formal methods and hybrid real-time systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

Depending on the representationability of neural networks for a nonlinear process,this paper develops procedures to locate the inputspace dimensions of feedforward neural networks forgeneral practical nonlinear systems identificationwhen only the outputs are accessible observations. Thesize of the input space is directly related to theinvolved system order. Hence based on our developedmodel, we are able to determine whether the system isfaulty or not by monitoring the output error change.The methods are demonstrated in vibration signals thatwere measured by an accelerometer mounted on atraction centre of a railway carriage running at about60 mph along the Hong Kong railway.