考虑混合评价信息和指标间关联关系的设计方案择优  

Evaluation of Schemes Considering Hybrid Evaluation Information and the Indicator Association Relationships

作　　者：张永政[1] 叶春明[1] 耿秀丽[1] ZHANG Yongzheng;YE Chunming;GENG Xiuli(Business School,University of Shanghai for Science of Technology,Shanghai 200093,China)

机构地区：[1]上海理工大学管理学院,上海200093

出　　处：《工业工程与管理》2023年第1期34-41,共8页Industrial Engineering and Management

基　　金：国家自然科学基金资助项目(71840003);上海市一流学科资助项目(S1201YLXK);上海市高原学科资助项目(GYXK1201);上海理工大学科技发展资助项目(16KJFZ028)。

摘　　要：考虑设计方案评价信息的模糊性和不确定性,将指标类型拓展为精确数、区间数、三角模糊数、犹豫模糊数和概率语言术语集5种类型的混合评价信息,针对指标特性、类型的不同,分别采用不同类型的评价信息。当评价信息为混合信息时,考虑到指标间的相互影响关系更加复杂,采用基于模糊测度的广义Shapley函数分析指标间相互影响关系并确定指标权重。针对将混合信息转为同一种信息会导致计算复杂和信息缺失的问题,采用Mo-RVIKOR方法对设计方案进行综合评估并择优。最后以某企业特色播种机设计方案择优为例,对所提方法的有效性和可行性进行了验证。In consideration of the fuzziness and uncertainty of evaluation information of the design schemes,indicator types were added to five types of mixed evaluation information:real numbers,interval numbers,triangular fuzzy numbers,hesitant fuzzy numbers and probabilistic linguistic term sets. According to the different characteristics and types of indicators,different types of evaluation information were obtained. When the evaluation information was mixed information,generalized Shapley function based on fuzzy measures was used to analyze the more complex interaction relationship between indicators and determine the weight of indicators. The hybrid information was converted into the same information. It would lead to computational complexity and information loss. Mo-RVIKOR was proposed to rank the concepts to solve the problem. Finally,a case study of evaluating a company’s featured seeder schemes was given to verify the effectiveness and feasibility of the proposed method.

关 键 词：方案选择 混合型多属性决策 广义Shapley函数 Mo-RVIKOR 

分 类 号：TH122[机械工程—机械设计及理论] N94[自然科学总论—系统科学]