基于系统直觉投票排序的ANP加权矩阵新构造方法  

A New Approach to Constructin the Cluster-weighted Matrix of ANP Based on Voting Rank by Using System Intuition

作　　者：孙永河[1] 段万春[1] 谢晖[1] 李春好[2] 柳文炜[1] 

机构地区：[1]昆明理工大学管理与经济学院,云南昆明650093 [2]吉林大学管理学院,吉林长春130025

出　　处：《系统工程》2015年第5期114-120,共7页Systems Engineering

基　　金：国家自然科学基金资助项目(71261013;71263031;71371083);教育部新世纪优秀人才支持计划项目(NCET-09-0968);教育部人文社会科学研究青年基金资助项目(10YJC630218);云南省科技计划项目(2010ZC060);云南省哲学社会科学创新团队支持项目(2014cx05)

摘　　要：在网络分析法(ANP)的应用中,构造加权矩阵是其中的一个技术关键。然而,目前关于ANP加权矩阵的传统构造方法存在着因素集之间的两两比较机理模糊不清问题,并且基于决策试行与实验室(DEMATEL)的构造方法有忽视系统因素自身强度信息的内在不足。为克服上述缺陷,本文提出了一种基于系统直觉投票排序的ANP加权矩阵构造新方法,该方法既从专家遴选、情境分析方面完善了现有的投票决策程序,而且采用系统直觉思维使专家对复杂问题的判断更为可行。实例验证结果表明,新方法是科学可行的,有着较强的实际应用可操作性。The analytic network process(ANP)is an effective decision approach to dealing with complex systems issues.In recent years,it has been widely applied to project selection of information systems,product planning and strategic decision,etc.In ANP procedures,how to construct cluster-weighted matrix(CWM)is an important technologic core.However,nowadays the traditional constructing method for CWM has the drawback that the comparison mechanism among clusters is unclear.Furthermore,the CWM construction method based on the decision making trial and evaluation laboratory(DEMATEL)ignores the strength of system factors,so the method is also problematic.To overcome the above disadvantages,we suggest a new constructing approach for CWM based on voting rank by using system intuition in this paper.In the new approach,expert selection and scenario analysis are both considered to improve exiting voting procedures.Also,the system intuition judgment way adopted in this paper is more suitable to solve the complex issues.Finally,the new approach is validated to be reasonable and would be well applied to solve the real world problems.

关 键 词：复杂系统 网络分析法 加权矩阵 系统直觉 投票排序 数据包络分析 

分 类 号：N941.4[自然科学总论—系统科学] O151.21[理学—数学]