Model-based fault detection using hierarchical artificial neural network
In this paper, a two-stage approach integrating a neural network dynamic estimator and a neural network fault classifier is proposed to overcome the problem of malfunction in sensors. The process estimator is designed to predict the dynamic behaviour of the normal or fault-free operating process eve...
| Main Authors: | , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English |
| Published: |
2001
|
| Subjects: | |
| Online Access: | http://eprints.utm.my/967/ http://eprints.utm.my/967/1/RSCE2001.pdf |
| Summary: | In this paper, a two-stage approach integrating a neural network dynamic estimator and a neural network fault classifier is proposed to overcome the problem of malfunction in sensors. The process estimator is designed to predict the dynamic behaviour of the normal or fault-free operating process even in the presence of sensor failures. The difference between this estimated “normal� values and the actual process measurements, termed the residuals are fed to the classifier for fault detection purposes. The classifier then identifies the source of faults. The scheme was implemented under dynamic operating conditions of the Tennessee Eastman challenge process and was successful in detecting various sensor faults introduced within the system. |
|---|