System diagnostics using qualitative analysis and component functional classification
A method for detecting and identifying faulty component candidates during off-normal operations of nuclear power plants involves the qualitative analysis of macroscopic imbalances in the conservation equations of mass, energy and momentum in thermal-hydraulic control volumes associated with one or more plant components and the functional classification of components. The qualitative analysis of mass and energy is performed through the associated equations of state, while imbalances in momentum are obtained by tracking mass flow rates which are incorporated into a first knowledge base. The plant components are functionally classified, according to their type, as sources or sinks of mass, energy and momentum, depending upon which of the three balance equations is most strongly affected by a faulty component which is incorporated into a second knowledge base. Information describing the connections among the components of the system forms a third knowledge base. The method is particularly adapted for use in a diagnostic expert system to detect and identify faulty component candidates in the presence of component failures and is not limited to use in a nuclear power plant, but may be used with virtually any type of thermal-hydraulic operating system. 5 figures.
- DOE Contract Number:
- W-31109-ENG-38
- Assignee:
- Univ. of Chicago, IL (United States)
- Patent Number(s):
- US 5265035; A
- Application Number:
- PPN: US 7-885132
- OSTI ID:
- 5377429
- Resource Relation:
- Patent File Date: 18 May 1992
- Country of Publication:
- United States
- Language:
- English
Similar Records
A first-principles generic methodology for representing the knowledge base of a process diagnostic expert system
Systematic construction of qualitative physics-based rules for process diagnostics