摘要
针对带有决策者期望的风险型多属性决策问题,提出一种结合直觉模糊熵和直觉模糊相似度的属性权重优化方法,并应用前景理论和证据推理处理决策数据。首先,将不同类型的决策数据规范化后统一描述为直觉模糊数;其次,以期望向量为参考点,运用前景理论求得前景决策矩阵;再次,计算属性的直觉模糊熵和直觉模糊相似度,结合期望值对属性权重进行优化,在运用概率函数计算出不同自然状态的概率后,先后运用两次证据推理方法融合前景决策信息,并利用记分函数对方案进行排序;最后,通过算例对比验证了该方法的可行性和优越性。
For the problem of risky multi-attribute decision making with decision-maker’s expectation,an attribute weight optimization method combining intuitionistic fuzzy entropy and intuitionistic fuzzy similarity is proposed,and prospect theory and evidential reasoning is applied to process decision data.Firstly,different types of decision data are standardized and then described as intuitionistic fuzzy numbers.Secondly,the prospect decision matrix is obtained by using the prospect theory with the expectation vector as a reference point.Thirdly,the weight of attributes are optimized by combining with the fuzzy entropy,fuzzy similarity and expected value.Furthermore,the evidential reasoning method is used twice to fuse the information of prospect decision after using the probability function to calculate the probability of different natural states,and then the score function is used to rank the scheme.Finally,the feasibility and superiority of the method are verified by comparison of examples.
作者
夏梦颐
王应明
XIA Mengyi;WANG Yingming(College of Economics and Management,Fuzhou University,Fuzhou 350108,China)
出处
《重庆理工大学学报（自然科学）》
CAS
北大核心
2020年第11期254-262,共9页
Journal of Chongqing University of Technology：Natural Science
基金
国家自然科学基金资助项目(61773123)。
关键词
直觉模糊数
前景理论
证据推理
风险型多属性决策
intuitionistic fuzzy number
prospect theory
evidential reasoning
risky multi-attribute decision making
