基于专家知识库属性重要度的故障诊断方法研究

利用粗糙集理论中重要性度量的概念和专家知识库集成技术，研究了复杂系统故障诊断的一类新方法。首先，在给出多领域专家知识库集成系统定义和复杂系统状态赋值规则的基础上，利用粗糙集理论中的属性重要度的概念建立了专家知识库属性重要度模型；然后，提出了该模型运行机理的算法规则，并结合多领域专家知识库集成技术，构造了一种能对实时动态环境进行监控和故障诊断的网络化诊断系统；最后，仿真试验验证了该方法的有效性和合理性。此方法成功应用于复杂系统故障诊断的实际工程实例表明，这种诊断方法具有良好的更新能力和应用前景。

A New Method of Fault Diagnosis Based on Attibute Sighificance of Multi-field Experts Database

Based on attribute signifieance of rough set and integrated technology of experts database, a new method of fault diagnosis to comlex systems is researched. First of all, the attribute significance model of experts database is presented by combing integrated system of multi-field experts and assignment rule of fault state with attribute significance concept in rough set theory. Secondly, a network model based on multi-field experts database is constructed, which can be used to monitor the dynamic systems in real-time and diagnosis fault. Finally, a simulated example has shown that the technique is effectively and exactly. The presented method has used in the fault diagnosis of complex systems and good results have been achieved, and the theoretical analysis indicats that the presented algorithm has a broad prospect for practical application.