基于Rough Sets-C4.5的故障征兆提取与判别

针对原始信息系统往往存在大量重复样本和冗余属性,从而影响实际故障诊断的精度和速度这一问题,介绍了一种基于粗糙集和决策树C4.5算法相融合的故障诊断模型,用于设备的精确和快速故障诊断.利用粗糙集具有较强的处理不确定和不完备信息的能力,对原始样本集进行离散化及约简处理;同时,利用决策树C4.5算法对约简后的决策表进行快速学习并形成树状故障分类器.以实例介绍了利用该模型进行故障诊断的完整过程.

Fault Diagnosis Based on Rough Sets and C4.5 Decision Tree

It was found that the precision and speed of fault diagnosis is unsatisfied due to large-scale repeating data and redundant attributes in information system (decision table) during practical applications. To solve the problem,a new model based on rough sets and decision tree C4.5 is presented. The theory of rough sets as a new mathematical tool is strong at dealing with incomplete and uncertain information and used to discretize and reduce the initial sample sets,while the C4.5 decision tree is used to learn quickly the reduced decision tables and form a tree classifier. An example is given to show the whole fault diagnosis process of RH-KTB vacuum metallurgical system by use of the new model.