摘要： 介绍基于多Agent的谈判对手选择方法，利用灰色理论对信息不确定情况下的谈判对手属性进行模糊区间处理，从知识库中获取灰色区间值信息，在此基础上构造关系-利益双因素模型。通过计算理想对手评估向量，设置白化权函数，对多属性的谈判对手进行综合灰色聚类分析，给出具体算例进行验证。该方法能够完善不确知信息条件下的谈判对手选择，提高谈判对环境信息的适应性。

关键词: 自动谈判, 对手选择, 多Agent, 灰色聚类, 双因素模型

Abstract: In the opponent selection process of Multi-Agent-based negotiation, this paper uses the gray theory to deal with the fuzzy interval attributes of negotiation opponents under uncertainty information condition. It gets the gray interval value information from knowledge base and constructs a two-factor model of relationship and profit, calculates the ideal opponent evaluation vector, sets the bleaching weight function, analyzes multi-attribute gray cluster of negotiation opponents, and gives a specific example for illustration. This method perfects the opponent selection method under uncertain information, and improves the environmental adaptability of negotiations.

Key words: automated negotiation, opponent selection, Multi-Agent, grey clustering, two-factor model

中图分类号: 

• TP391