变精度粗集模型在故障诊断专家系统中的应用

Application of VPRS model in fault diagnosis expert system

将变精度粗集模型应用于故障诊断专家系统的知识获取和知识库的更新;利用变精度知识约简和正则条件熵进行故障特征选择,实现最简诊断知识的获取,以建立专家系统知识库;利用属性的相对依赖性实现知识库的维护与更新.滚动轴承故障诊断的实验结果表明,该方法有效地克服了传统规则式故障诊断专家系统的不足,实现了故障诊断功能.

A variable precision rough set model approach was proposed for knowledge acquisition and update of fault diagnosis expert system. Knowledge reduction with variable accuracy β, and regular conditional entropy were utilized for the knowledge acquisition to establish the knowledge base. The maintenance and update of the base were realized by the relative dependency. Simulation of rolling bearings fault diagnosis illustrates that the approach can compensate for the defects of rulebased ES and carry out the fault diagnosis effectively.