信息不完全确定的动态随机多属性决策方法  

Dynamic stochastic multiple attribute decision making method with incomplete certain information

作　　者：汪新凡[1] 杨小娟[1] 

机构地区：[1]湖南工业大学理学院,株洲412008

出　　处：《系统工程理论与实践》2010年第2期332-338,共7页Systems Engineering-Theory & Practice

基　　金：湖南省教育厅科学研究项目(07C232)

摘　　要：针对属性权重信息不完全确定、属性值为正态分布随机变量且数据信息来自于不同时期的动态随机多属性决策问题,给出了正态分布数的运算法则,定义了正态分布数加权算术平均(NDNWAA)算子和动态正态分布数加权算术平均(DNDNWAA)算子,进而提出了一种信息不完全确定的动态随机多属性决策方法.该方法利用DNDNWAA算子和NDNWAA算子对正态分布属性值进行集成;利用正态分布属性值的方差和属性权重的随机性,通过建立优化模型确定最优属性权重;利用正态分布3σ原则、区间数比较的可能度公式和互补判断矩阵的排序公式对决策方案进行排序和择优.最后,实例分析表明了该方法的可行性和有效性.For dynamic stochastic multiple attribute decision making problems, in which the information on attribute weights is incomplete certain, the attribute values are normal distribution stochastic variables and the argument information are given at different numbers are given, the normal distribution number periods, some operational laws of normal distribution weighted arithmetic averaging （NDNWAA） operator and the dynamic normal distribution number weighted arithmetic averaging （DNDNWAA） operator are proposed, and then an approach for solving dynamic stochastic multiple attribute decision making with incomplete certain information is developed. In this method, normal distribution attribute values are aggregated by the DNDNWAA operator and the NDNWAA operator, some optimal models are constructed to determine the optimal attribute weights by using the variance of normal distribution attribute values and the randomicity of attribute weights, ranking of alternatives is performed by using 3σ principle of normal distribution, possibility degree formula for comparing two interval numbers and formula for priority of complementary judgement matrix. Finally, an example shows the feasibility and effectiveness of this method.

关 键 词：动态随机多属性决策 集成算子 正态分布 信息熵 

分 类 号：N945.25[自然科学总论—系统科学]