论文部分内容阅读
根据内燃机车的故障特点,研究故障诊断流程的二叉树化。将知识库分解为故障现象、故障原因和故障诊断规则3个即相互独立又互为关联的数据库,在知识库管理模块的控制下,分别从故障现象库和故障原因库中选取相应的知识构成故障诊断规则的前件和后件,并以IF……THEN……的形式存储在故障诊断规则库中。采用基于故障匹配率的启发式搜索机制,结合专家的经验与故障发生的概率,通过修改估价函数,实现对较大概率故障的优先搜索,以提高系统的故障诊断效率。目前,以故障诊断二叉树分析法为基础的内燃机车故障诊断专家系统已在南昌铁路局各下属机务段成功装车运用,达到了预期目标。
According to the fault characteristics of diesel locomotive, the binary tree of fault diagnosis process is studied. The knowledge base is decomposed into the symptom, the cause of the fault and the fault diagnosis rules, that is, three independent and interrelated databases. Under the control of the knowledge base management module, the corresponding knowledge is selected from the database of the symptom and the reason of the fault The front and back of the troubleshooting rules are stored in the troubleshooting rule base in the form of IF ... THEN .... Based on the heuristic search mechanism based on the failure matching rate, combined with the expert’s experience and the probability of the failure, the priority search of the higher probability failure can be realized by modifying the valuation function so as to improve the system’s fault diagnosis efficiency. At present, fault diagnosis expert system of diesel locomotive based on the binary tree analysis of fault diagnosis has been successfully loaded and used in the locomotive depots of all subordinates in Nanchang Railway Administration, reaching the expected goal.