Back to home
 

Integrated method for constructing knowledge base system for proactive trouble prevention of nuclear power plant

YANGMing1, ZHANG Zhijian1, YOSHIKAWA Hidekazu2, LINDMorten3,
ITOKyoko4, TAMAYAMA Kiyoshi5, and OKUSA Kyoichi5
 
1. College of Nuclear Science and Technology, Harbin Engineering University, Harbin, Heilongjian, 150001 China  (yangming@hrbeu.edu.cn)
2. College of Nuclear Science and Technology, Harbin Engineering University, Harbin, Heilongjian, 150001 China
  (hidekazuyoshikawa@nike.eonet.ne.jp)
3. Department of Electrical Engineering, Technical University of Denmark, DK-2800 Lyngby Denmark (mli@elektro.dtu.dk)
4. Graduate School of Engineering Science, Osaka University, Toyonaka-shi, Osaka-fu, 580-8531 Japan (ito@cscd.osaka-u.ac.jp)
5. Fast Breeder Reactor Research & Development Center, Japan Atomic Energy Agency, Tsuruga, Japan (tamayama.kiyoshi@jaea.go.jp, okusa.kyoichi@jaea.go.jp) 
 
Abstract: Integrated method for constructing knowledge base (KB) system for proactive trouble prevention of complex machine system such as nuclear power plant has been studied in this paper. The new ideas employed in this study are: (i) Structuralizing trouble KB for trouble prediction and  prevention, (ii) Realizing such KB by using web database, (iii) Modeling plant system by the combination of solid matters and non-solid matters, (iv) Modeling solid matters as object-oriented KB with the combination of structural components and electrical circuits, usage and environmental conditions and knowledge on troubles, and (v) Modeling non-solid matters by a revised Multilevel Flow Model (MFM). The framework is proposed on how to systemize those five different ideas on the basis of graphical modeling method by the revised MFM. A preliminary study is also conducted on applying the proposed method to a fast reactor prototype “Monju”, in order to reduce problems to be improved for the usage of revised MFM for real scale application for nuclear power plant.
Keyword: proactive trouble prevention; knowledge base system; multilevel flow model; Monju
 
      Click here for viewing full text.