City Research Online

Intelligent monitoring of small transients in a complex non-linear system using artificial neutral networks

D'Souza, L. J. (2006). Intelligent monitoring of small transients in a complex non-linear system using artificial neutral networks. (Unpublished Doctoral thesis, City, University of London)


This project uses Artificial Neural Networks (ANNs) to develop a prototype computer based Operator’s Advisory System for the early detection and diagnosis of plant transients.

Three transient monitors were developed two of which are ANN based. Each of the independently developed ANN classifiers was then integrated into a multi-level operator advisory system OAS. The first level of diagnosis provides information to the plant operator of the presence of a major transient. Should a transient be detected a corresponding module provides more detailed information on the size of the transient. To validate the diagnosis two methods are used in the OAS: User confirmation and a comparison with simulated plant data. The diagnosis is reproduced in an independently developed PWR simulator and the plant parameters compared. If in agreement, a high level of confidence was attached to the diagnoses, a poor match would suggest that the transient is not one that the diagnostic module had been trained on. The OAS was evaluated on a wide range of scenarios. The results of the tests were encouraging with the OAS successfully identifying a range of standard transients. However, tests on the robustness of the OAS proved inconclusive.

The project concludes with suggestions for future work.

Publication Type: Thesis (Doctoral)
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Q Science > QP Physiology
Departments: School of Science & Technology > Computer Science > Human Computer Interaction Design
School of Science & Technology > School of Science & Technology Doctoral Theses
Doctoral Theses
[thumbnail of D'Souza thesis 2006 PDF-A.pdf]
Text - Accepted Version
Download (8MB) | Preview


Add to AnyAdd to TwitterAdd to FacebookAdd to LinkedinAdd to PinterestAdd to Email


Downloads per month over past year

View more statistics

Actions (login required)

Admin Login Admin Login