Variable Selection and Sensor Fusion
in Automatic Hierarchical Fault Monitoring of Large Scale Systems
Abstract
In this paper, the problem of human overload in monitoring the many sensors
of a complex industrial plant is addressed. A hierarchical automated fault
monitoring and analysis system for high-level decision making is presented.
The monitoring system operates in parallel with the traditional channels.
It has no effect other than being able to display its findings to the human
operator, and to point out potential problems and their perceived causes. The
prospects as well as difficulties in realizing such a monitoring system are
analyzed by discussing a prototypical implementation of such a system on a
sophisticated quantitative large-scale model of a nuclear power plant.
Interested in reading the
full paper?
(13 pages, 594,878 bytes, pdf)
Homepage
Last modified: January 19, 2006 -- © François Cellier