Fault diagnosis with enhanced neural network modelling

Authors:
Ding-Li Yu;Thoon-Khin Chang
Affiliations:
Control Systems Research Group, Liverpool John Moores University, UK;Control Systems Research Group, Liverpool John Moores University, UK
Venue:
ISNN'06 Proceedings of the Third international conference on Advances in Neural Networks - Volume Part III
Year:
2006

Citing 2
Cited 0

Fault diagnosis in dynamic systems using analytical and knowledge-based redundancy—a survey and some new results

Automatica (Journal of IFAC)
Fault diagnosis of machines via parameter estimation and knowledge processing: tutorial paper

Automatica (Journal of IFAC) - Special section on fault detection, supervision and safety for technical processes

Quantified Score

Hi-index	0.00

Visualization

Abstract

A neural network (NN) based fault detection and isolation (FDI) approach for unknown non-linear system is proposed to detect both actuator and sensor faults. An enhanced parallel (independent) NN model is trained to represent the process and used to generate residual. A mean-weight strategy is developed to overcome the un-modelled noise and disturbance problem. A signal pre-processor is also developed to convert the quantitative residual to qualitative form and applied to a NN fault classifier to isolate different faults. The developed techniques are demonstrated with a multi-variable non-linear tank process.