基于故障树和产生式规则的故障诊断专家系统设计

Design of fault diagnosis expert system based on fault tree and production rules

  • 摘要:
    目的 为充分运用核动力装置的运行管理经验辅助核动力操纵人员进行故障诊断,设计一种船用核动力装置故障诊断专家系统。
    方法 首先,根据故障树与产生式规则之间的逻辑一致性,提出一种将故障树知识转化为产生式规则的方法;然后,对采用正、反向混合推理方法的专家系统知识库和推理机进行优化设计,并依据故障树最小割集和重要度分析结果设计正向推理策略以简化推理流程;最后,根据人工对故障状态判断的思路设计状态监测模块,实时采集关键设备参数以转化为专家系统可识别的设备信息。
    结果 结果显示,采用所提方法可解决专家系统知识获取困难的问题,能在保证推理准确度的前提下提升推理效率,实现了专家系统的在线故障诊断功能。
    结论 研究表明采用所提方法可提升专家系统获取知识的能力和推理效率,对保障核动力装置的运行管理安全具有重要意义。

     

    Abstract:
    Objective To fully utilize the experience of nuclear power plant operation and management to assist nuclear power operators in fault diagnosis, a marine nuclear power plant fault diagnosis expert system is designed.
    Method First, based on the logical consistency between fault trees and production rules, a method is proposed to transform fault tree knowledge into production rules. The knowledge base and inference machine of the expert system are then optimized by using a mixed forward and backward inference method, and a forward inference strategy is designed to simplify the inference process based on the minimum cut set and importance analysis results of the fault tree. Finally, based on the idea of manually judging the fault status, a status monitoring module is designed to collect key equipment parameters in real time and convert them into equipment information that can be recognized by expert systems.
    Results The results show that the proposed method solves the problem of difficult knowledge acquisition in expert systems and improves inference efficiency while ensuring inference accuracy, thereby achieving the online fault diagnosis function of expert systems.
    Conclusion Using the proposed method can enhance the knowledge acquisition ability and inference efficiency of expert systems, which is of great significance for ensuring the operational management safety of nuclear power plants.

     

/

返回文章
返回