摘要
R-fuzzy集以粗糙集的形式给出了隶属函数,按照隶属度与描述符的相关程度将其进行分类.完全符合描述符的隶属度划分到下近似集,与描述符有关的隶属度划分到上近似集.如果能够得到上近似集中隶属度的重要性,将拓展R-fuzzy的应用领域.进一步讲,如果能够引入一种方法对R-fuzzy上近似集中的隶属度重要性进行量化,将可实现对隶属度重要性的量化排序,实现更高的分辨力.本文提出的优势测度概念可以很好的实现这个要求.首先,给出了优势测度理论框架,证明了优势测度与1型模糊集的等价性,接着,论证了与R-fuzzy集的一致性,指出了优势测度模糊集本质就是R-fuzzy粗糙隶属集的验证器.最后,通过人类视觉感知实验及优势测度的可视化,研究了不同类别群体共识与个识对确定R-fuzzy隶属度测度的影响,分析了R-fuzzy粗糙近似隶属度集的优势测度方法对于人类群体感知辨识的优势.
R-fuzzy membership function is given in the form of rough set,and classification is made according to the membership degree and the relevance of the descriptor to it.The membership values fitting perfectly with the descriptor are divided into the lower approximation set,while membership values related to the descriptor into upper approximation set.If you can introduce a kind of method on R-fuzzy approximation set to quantify the importance of membership values,then quantitative ordering of the membership degrees can be achieved,as such to achieve higher resolution to discriminate the differences among these membership values.This requirement can be realized well by the advantage measure concept proposed in this paper.Advantage measure theory is put forward firstly,and then the equivalence of the advantage measure to the type-1 fuzzy set is proved,and moreover the consistency of the concept with the R-fuzzy set is demonstrated.In essence,advantage measure fuzzy set serves as the validator for the R-fuzzy rough membership set.Finally,through the human visual perception experiment and visualization of advantage measures,the influence of group consensus and individual perception on the advantage measure of the R-fuzzy set is investigated followed by analysis of the advantages of the advantage measure over human cognitive identification.
作者
李守军
马小平
杨春雨
LI Shoujun;MA Xiaoping;YANG Chunyu(School of Mechanical and Electrical Engineering,Suqian College,Suqian 223800,China;School of Information and Control Engineering,China University of Mining and Technology,Xuzhou 221116,China)
出处
《系统工程理论与实践》
EI
CSSCI
CSCD
北大核心
2019年第6期1602-1609,共8页
Systems Engineering-Theory & Practice
基金
国家自然科学基金(61374043,61603392)
江苏高校品牌专业建设工程项目(PPZY2015C252)~~
