Development of diagnosis and maintenance support system for nuclear power plants with flexible inference function and knowledge base edition support function

Autor: Eiji Seki, Makoto Fujii, Morioka Toshihiko, Ichiro Tai
Rok vydání: 1988
Předmět:
Zdroj: Journal of the Atomic Energy Society of Japan / Atomic Energy Society of Japan. 30:1110-1118
ISSN: 0004-7120
DOI: 10.3327/jaesj.30.1110
Popis: For the reliable and efficient diagnosis and inspection work of the nuclear power plant equipments, “Diagnosis and Maintenance Support System” has been developed. This system has functions to assist operators or engineers to observe and evaluate equipment conditions based on the experts' knowledge. These functions are carried out through dialogue between the system and users.This system has two subsystems: diagnosis subsystem and knowledge base edition support subsystem.To achieve the functions of diagnosis subsystem, a new method of knowledge processing for equipment diagnosis is adopted. This method is based on the concept of “Cause Generation and Checking”. Knowledge for diagnosis is represented with modularized production rules. And each rule module consists of four different type rules with hierarchical structure. With this approach, the system is equipped with sufficient performance not only in diagnosis function but also in flexible man-machine interface.Knowledge base edition support subsystem (Graphical Rule Editor) is provided for this system. This editor has functions to display and edit the contents of knowledge base with tree structures through the graphic display. With these functions, the efficiency of constructing expert system is highly improved.By applying this system to the maintenance support of neutron monitoring system, it is proved that this system has satisfactory performance as a diagnosis and maintenance support system.
Databáze: OpenAIRE