一种基于FCM聚类的粗糙集连续数据故障诊断方法

摘要

There is the problem that rough set theory can not process the quantitative data in the equipment test data. One FCM method is given to discretize the quantitative data, and the center value of the each cluster is calculated in order to discretize. And then, the attribution reduction algorithm and rules extract methods are applied to get the last diagnosis rules, which is applied to some aero radio equipment fault diagnosis to verify its validity.%针对粗糙集无法对航空电子装备测试数据中的连续数据进行处理的问题,提出一种基于FCM聚类的离散化方法,通过计算各聚类中心值进行离散化处理,并应用粗糙集属性约简和规则提取方法,获得最终诊断规则,并在某型机载电台故障诊断中验证其有效性.