一类群决策问题的最优分配解法

The optimal assignment method for group decision making of multi-decision-maker

在群决策中存在着多个决策者根据自己的知识、经验和偏好独立选择决策方法 ,对决策方案集作出决策后 ,应综合确定群体决策方案的多人多指标群决策问题 .针对这类决策问题具有决策信息量较少 ,而决策一致性较差的特点 ,分析了总体偏差法、足码法及其相关改进方法的优缺点 ,提出了一种多人综合评价的最优分配解法 .该方法通过建立决策方案集与决策名次之间的关联矩阵 ,并根据关联矩阵对方案集进行分类 ,在此基础上构造最优决策模型 ,利用整数规划中的最优分配解法求解最优模型 ,从而得到方案集的综合排序 .该方法简单实用 ,并且克服了总体偏差法。

After selecting decision method independently according to their knowledge, experience and preference, making decision to decision making alternatives set by multi decision maker, the multi decision maker and multi indexes group decision making problem of how to comprehensive ascertain group decision alternatives set exists in group decision making. To the problem of little decision information and bad decision consistency, after analyzing respective characteristics and limitation of collectivity deviation method, leg mark method and the correlation improved method, an optimal assignment solution of multi decision maker comprehensive evaluation is proposed . By building relative matrixes between decision making alternatives set and decision making order,and sort the alternatives set on relative matrites, the method constructs an optimization decision making model. By getting the optimal solution using integer programming method, we obtain the comprehensive ranking of alternatives set. The method avoids the shortage of collectivity deviation method, leg mark method and the correlation improved method. At last, an applicable computation is illustrated.