一种航空发动机连续数据的粗糙集离散化方法

摘要

Aiming at the problem of traditional data discretization method based rough set, a novel method of data discretization, based on rough set, is put forward, which is based on discretization knowledge of attributes and information entropy. Aeroengine spectrometric oil analysis data is used to verify the method. The result proves that the method can maintain the compatibility relation of decision table and get discrete intervals with actual meanings.%针对传统粗糙集属性离散化方法存在离散化区间会破坏数据实际含义的问题,结合连续属性的领域知识,提出了一种基于信息熵的粗糙集属性离散化方法.利用航空发动机的滑油光谱分析数据对其进行了验证,结果表明其能够在保持决策表相容关系的前提下,获得具有较好实际意义的离散化区间.