层次分析法中高阶平均随机一致性指标(RI)的计算  

Caculation on High-ranked R I of Analytic Hierarchy Process

作　　者：洪志国[1] 李焱[2] 范植华[1] 王勇[1] 

机构地区：[1]中国科学院软件所GSL实验室,北京100080 [2]中国科学技术大学计算机系,合肥230027

出　　处：《计算机工程与应用》2002年第12期45-47,150,共4页Computer Engineering and Applications

基　　金：中国科学院知识创新工程国防军工方向性重大项目<大型数字对象应用环境及其并行模拟>资助

摘　　要：利用层次分析法分析和解决问题时,要对通过两两比较判断出的矩阵一致性进行检验犤1犦。高阶平均随机一致性指标的值一般无法直接通过查表而得,这一难点阻碍着层次分析法大面积的推广应用犤2犦。文章在深刻剖析层次分析法的基础上,给出根据平均随机一致性指标的定义计算高阶平均随机一致性指标值的算法,并且基于windows环境在delph6.0下予以程序实现。该算法已成功运用于中国科学院知识创新工程某智能决策系统中。To analyze and solve problems with Analytic Hierarchy Process needs a check on the consistency of a ma-trix coming from a two-two comparison .Usually people can hardly get the value of High-Ranked R.I.immediately through consulting related tables,and as a result,it hinders a mass of applications of Analytic Hierarchy Process .On the basis of a thorough analysis of Analytic Hierarchy Process,the algorithm to calculate out the value of High-Ranked R I according to the definition of R I is offered in the paper.The program of the algorithm under windows using delphi 6.0is provided as well.The algorithm has been successfully applied to a project of the intelligent decision system in the knowledge innovation of Chinese Academy of Sciences.

关 键 词：高阶随机判断矩阵 平均随机一致性指标 乘幂法 层次分析法 

分 类 号：O225[理学—运筹学与控制论]