基于梯形模糊数期望值的多维偏好群决策模型  

Group decision-making model for multidimensional analysis of preference on trapezoid fuzzy number expected values

作　　者：刘於勋[1,2] 沈轶[1] 谢妞妞[2] 

机构地区：[1]华中科技大学控制科学与工程系,武汉430074 [2]河南工业大学信息科学与工程学院,郑州450001

出　　处：《控制与决策》2009年第9期1377-1379,1384,共4页Control and Decision

基　　金：国家自然科学基金项目(70701013);河南省自然科学基金项目(0611030100)

摘　　要：提出一种基于梯形模糊数距离期望值的多维偏好群决策模型,以解决偏好和属性值均为梯形模糊数的群决策问题.其算法为:首先定义在β截集下主/客观偏好之间的偏差函数,通过构造目标规划模型,求解属性的权重向量;然后集结不同β截集下所有决策者的加权规范化模糊决策矩阵,形成总加权规范化模糊决策矩阵;最后求出各备选方案与模糊理想解的相对贴近度iδ,按大小排序确定最优方案.Group decision-making model for multidimensional analysis of preference on trapezoid fuzzy number distant expected values are proposed. The group decision-making problem are solved when preference and attribute are given by trapezoid fuzzy number. A distortion function between subjective / objective analysis of preference under β-cut is defined. The weighted vector of the attribute by constructing a criterion-programming model. Then the weighted normalization fuzzy decision matrices-of all the decision-makers under differentβ-cut are cengregated to form a total weighted normalization fuzzy decision matrix. Finally, relative closeness δi of each alternative adjustment decision is obtained and sorted by size to determine the optimal program.

关 键 词：梯形模糊数期望值 决策矩阵 TOPSIS方法(逼近理想解的排序)方法 

分 类 号：C934[经济管理—管理学]