Statistical algorithm for automated signature analysis of power spectral density data

Conference · Sat Jan 01 00:00:00 EST 1977

OSTI ID:7079317

Piety, K. R.

A statistical algorithm has been developed and implemented on a minicomputer system for on-line, surveillance applications. Power spectral density (PSD) measurements on process signals are the performance signatures that characterize the ''health'' of the monitored equipment. Statistical methods provide a quantitative basis for automating the detection of anomalous conditions. The surveillance algorithm has been tested on signals from neutron sensors, proximeter probes, and accelerometers to determine its potential for monitoring nuclear reactors and rotating machinery.

• Research Organization: Oak Ridge National Lab., Tenn. (USA)
• DOE Contract Number: W-7405-ENG-26
• OSTI ID: 7079317
• Report Number(s): CONF-770902-2; TRN: 78-000886
• Resource Relation: Conference: Specialists meeting on reactor noise, Gatlinburg, Tennessee, USA, 19 Sep 1977
• Country of Publication: United States
• Language: English

Related Subjects

22 GENERAL STUDIES OF NUCLEAR REACTORS
POWER REACTORS
REACTOR NOISE
DATA ANALYSIS
ALGORITHMS
MONITORING
ON-LINE CONTROL SYSTEMS
ON-LINE MEASUREMENT SYSTEMS
REACTOR CONTROL SYSTEMS
SPECTRAL DENSITY
STATISTICS
CONTROL SYSTEMS
FUNCTIONS
MATHEMATICAL LOGIC
MATHEMATICS
ON-LINE SYSTEMS
REACTORS
SPECTRAL FUNCTIONS
220100* - Nuclear Reactor Technology- Theory & Calculation