Aiming at the problems of low efficiency and easy errors in the modeling of fault trees for unplanned shutdown and reactor trip of nuclear power units, this paper proposes a hierarchical automatic generation method based on graph database. By fusing equipment configuration information and fault mode libraries, and combining the semantic understanding and logical reasoning capabilities of large language models, the fully automatic modeling of fault trees from top events to equipment levels is realized. Taking the low-low water level signal of steam generator (SG) in M310 unit as an example, this method successfully constructs a fault tree including system level, equipment level and support equipment level, which improves the efficiency greatly compared with traditional manual modeling, and provides an efficient tool for unit reliability analysis and risk assessment.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Automatic Generation Method of Fault Tree for Unplanned Shutdown and Reactor Trip of Nuclear Power Units

  • Bin Cheng,
  • Xiaohu Yang

摘要

Aiming at the problems of low efficiency and easy errors in the modeling of fault trees for unplanned shutdown and reactor trip of nuclear power units, this paper proposes a hierarchical automatic generation method based on graph database. By fusing equipment configuration information and fault mode libraries, and combining the semantic understanding and logical reasoning capabilities of large language models, the fully automatic modeling of fault trees from top events to equipment levels is realized. Taking the low-low water level signal of steam generator (SG) in M310 unit as an example, this method successfully constructs a fault tree including system level, equipment level and support equipment level, which improves the efficiency greatly compared with traditional manual modeling, and provides an efficient tool for unit reliability analysis and risk assessment.