一种基于粗糙-模糊集集成模型的决策分析方法  

A decision analysis method based on rough-fuzzy sets integration model

作　　者：何亚群[1] 胡寿松[1] 

机构地区：[1]南京航空航天大学自动化学院

出　　处：《控制与决策》2004年第3期315-318,共4页Control and Decision

基　　金：国家自然科学基金重点资助项目(60234010);航空科学基金资助项目(02E52025).

摘　　要：针对信息系统为连续属性的情况,提出一种将粗糙集与模糊集相结合来获取决策规则的方法.这种基于粗糙-模糊集集成模型求取决策规则的方法通过一个模糊隶属函数将连续属性值表示成模糊值,从而避免了连续属性的离散化问题.同时给出了连续属性值转换成模糊值的表示形式,提出了模糊相似关系和模糊相似类的概念,给出了粗糙-模糊近似空间的下、上近似及其性质以及模糊相似关系下属性约简的方法.最后以自修复飞行控制系统的效能评估为例,给出了自修复效能评估的决策规则.A method of decision rules acquisition is proposed for continuous attributes decision system by combining rough sets and fuzzy sets. The method transforms the continuous attributes value into the fuzzy value with fuzzy membership functions, and avoids the discretization of continuous attributes. The fuzzy value expression is given, and the concepts of fuzzy similarity relation and fuzzy similarity classes is proposed. Lower and upper approximations of rough-fuzzy approximation space and their properties are given.A method of system reduction is also presented. Finally, as an example of effectiveness evaluation of the self-repairing flight control system, the decision rules of self-repairing effectiveness evaluation are given.

关 键 词：粗糙集 模糊集 集成模型 连续属性 效能评估 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]