基于参考点的大群体信息融合方法  

Information Fusion Approach of Large Group Based on Reference Point

作　　者：江文奇[1] 丁健美 

机构地区：[1]南京理工大学经济管理学院,江苏南京210094

出　　处：《系统工程》2013年第11期8-12,共5页Systems Engineering

基　　金：国家自然科学基金资助项目(71271116);教育部人文社科基金资助项目(09JD630522)

摘　　要：大群体评价信息融合中,评价成员数量多,权重难以有效界定。首先,借助于前景理论中的参考点思想,分析了期望值、正理想点和负理想点作为参考点的缺陷,以组合参考点和这三种参考点差异最小为目标函数,建立优化模型;其次,运用成员评价信息与参考点之间的距离来表示成员与群体意见的差异来确定权重,并给出了大群体信息融合步骤;最后,用实际算例说明了方法的有效性。The number of group is large, and weight vector of evaluators is very difficult to be designed in the information fusion process of large groups. Firstly, this paper analyzes the defects of expected value and ideal point and negative point by means of the reference point of prospect theory. An optimization model is proposed, in which the objective function is minimum distance between combination reference point and the above three reference points. Secondly, the difference between each memher and decision group is depicted as the distance between their evaluation information and reference point, which can be used to determine weights. The steps of group information fusion are given. Finally, a practical calculation example is used to illustrates the effectiveness of the approach.

关 键 词：大群体决策 信息集结 参考点 权重 

分 类 号：N945[自然科学总论—系统科学]