列车运行监控记录装置故障诊断系统的研究

摘要

In this paper, we summary the advantages and disadvantages of the fault diagnosis system of train running supervisory and control arrangement based on fault tree analysis(FTA), and present a solution on the basis of the theory of bidirectional associative memory(BAM). The system structure, functions and diagnostic flow are introduced briefly, and the modeling, matrix and the rule of inference are described. Finally, we point out the merits and main problems of the fault diagnosis system.%概括分析了基于故障树分析的列车运行监控记录装置故障诊断系统的特点及存在的问题，并在此基础上，提出了采用双向联想记忆神经网络的故障诊断系统方案。简要介绍了双向联想记忆神经网络故障诊断系统的系统结构、软件系统功能及诊断流程，并以主机无显示故障为例，详细描述了故障诊断模型、M矩阵的建立、推理过程、解释机制的构建原理。最后还概括分析了该系统的优越性，并点出了系统在技术实现中存在的主要问题。