Title :
Method of Multiple Fault Isolation in Large Scale Systems
Author :
Jan Maciej Koscielny;Micha ł Bartys;Micha ł Syfert
Author_Institution :
Institute of Automatic Control and Robotics, Warsaw University of Technology, Warsaw, Poland
Abstract :
In this paper, problems of the online diagnostics of multiple faults in large scale systems are presented. Possibilities of recognition of multiple faults based on the knowledge of signatures of the single faults are discussed. The distinction between simultaneous faults and the series of faults is introduced and defined. The necessary condition for diagnostic reasoning assuming an occurrence of multiple faults is formulated. This paper will show that multiple faults appearing simultaneously or within short time intervals are correctly isolated under the assumption of single faults, if the subsets of faults suitable for isolation are disjunctive. A novel, practical and easy to implement algorithm designed for reasoning multiple faults is presented. An industrial example of the online diagnostics of a steam-water line of the power boiler is given.
Keywords :
"Large-scale systems","Boilers","Cognition","Fault detection","Cooling","Neural networks","Knowledge engineering"
Journal_Title :
IEEE Transactions on Control Systems Technology
DOI :
10.1109/TCST.2011.2162587