Data Mining for Monitoring Loose Parts in Nuclear Power Plants

Abstract

Monitoring the mechanical impact of a loose (detached or drifting) part in the reactor coolant system of a nuclear power plant is one of the essential functions for operation and maintenance of the plant.Large data tables are generated during this monitoring process. This data can be “mined ” to reveal latent patterns of interest to operation and maintenance. Rough set theory has been applied successfully to data mining. It can be used in the nuclear power industry and elsewhere to identify classes in datasets, finding dependencies in relations and discovering rules which are hidden in databases. This paper can be considered as one of a series, the earlier ones being summarized in Guan & Bell (2000a). These methods can be used to understand and control aspects of the causes and effects of loose parts in nuclear power plants. So in this paper we illustrate the use of our data mining methods by means of a running example using Envelope Rising Time data ERT on monitoring loose parts in nuclear power plants.