Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Development and assessment of an Al expert system for the monitoring and diagnosis of nuclear power plants

Conference · · Transactions of the American Nuclear Society; (USA)
OSTI ID:5432806
Due to the rapid progress in microcomputer and software development, artificial intelligence (AI) expert systems of practical value can be built into microcomputers. An expert system for nuclear plant surveillance, diagnostics, and prognostics was developed using the Texas Instruments AI shell, Personal Consultant Plus (PC-plus) on an IBM PCAT. This expert system runs in a surveillance mode to find an abnormal operating condition. Once an abnormal behavior is found,it switches to a diagnostics mode to identify the cause of difficult, such as steam generator tube rupture (SGTR) and leak. Then, the prognostics mode can be activated to predict the consequences. For this purpose, the knowledge of experts at Westinghouse for nuclear safety has been collected and processed to construct parameters and rules within the framework of a logic tree. The expert system may be used in an on-line mode via a connection to the plant computer, safety parameter display system, or a plant simulator. In addition to evaluating the diagnosis of an event and providing appropriate information required to generate an event report, this tool can also be used to review the normal recorded plant data daily to assure that no abnormal events have occurred. A limited assessment of the expert system was performed and is presented.
OSTI ID:
5432806
Report Number(s):
CONF-881011--
Conference Information:
Journal Name: Transactions of the American Nuclear Society; (USA) Journal Volume: 57
Country of Publication:
United States
Language:
English