基于知识矩阵推理的小型模块化反应堆并发故障诊断方法研究  

Research on Small Modular Reator Concurrent Fault Diagnosis Method Based on Knowledge Matrix Reasoning

作　　者：彭俏[1] 马杰 刘明辉 绳东杰 Peng Qiao;Ma Jie;Liu Minghui;Sheng Dongjie(School of Nuclear Science and Technology,Naval University of Engineering,Wuhan,430033)

机构地区：[1]海军工程大学核科学技术学院,武汉430033

出　　处：《核动力工程》2024年第3期170-173,共4页Nuclear Power Engineering

基　　金：国防科工局“十三五”核能开发科研项目(2021246010)。

摘　　要：针对小型模块化反应堆(SMR)并发故障诊断问题,提出了一种基于知识矩阵进行知识表达和推理的并发故障诊断方法,该方法给出了一种高效的专家知识表达方式,将传统专家系统复杂的逻辑推理转换为简捷的矩阵运算,提高了故障诊断效率,可满足SMR在线并发故障诊断的要求,并以SMR发生主冷却剂管路破损和控制棒掉棒并发故障为诊断示例证明了本方法的有效性。Aiming at the problem of concurrent fault diagnosis of small modular reactor(SMR),a concurrent fault diagnosis method based on knowledge matrix for knowledge expression and reasoning is proposed.This method gives an efficient expert knowledge expression,which transforms the complex logical reasoning of traditional expert system into simple matrix operation,improves the efficiency of fault diagnosis,and can meet the requirements of online concurrent fault diagnosis of SMR.The effectiveness of this method is proved by the diagnosis examples of main coolant pipeline damage and control rod failure in SMR.

关 键 词：小型模块化反应堆(SMR) 并发故障诊断 知识矩阵推理 

分 类 号：TL364.1[核科学技术—核技术及应用] TP18[自动化与计算机技术—控制理论与控制工程]