基于效用图的依赖关系与协商策略研究  

Research on dependence and negotiation strategies based on utility graphs

作　　者：柴玉梅[1] 谷明[1] 王黎明[1] 

机构地区：[1]郑州大学信息工程学院,河南郑州450001

出　　处：《计算机工程与设计》2010年第3期586-590,629,共6页Computer Engineering and Design

摘　　要：依赖关系一直是多议题协商的重点和难点。在基于依赖关系的多议题协商背景下,对依赖关系和依赖度进行了严格的定义和度量。在协商过程中,协商Agent强化了自身的协商策略,并使用多目标遗传协商算法对协商进行优化。实验结果表明,在该协商背景下依赖关系的定义以及依赖强度的度量是合理的,卖方Agent使用多目标遗传协商算法和各种协商策略是可行的,并取得了较好的协商效果,且使协商结果达到了pareto最优。The dependencies among multi-issues negotiation is the key and difficult items.Under the background of multi-issues negotiation based the dependencies,a strict definition of the dependencies and measurement of dependencies degree are given;during the procedure of negotiation,the negotiation agents strengthen their strategies and use the multi-objective genetic negotiation algorithms （MOGNA）to optimize the negotiation.Experimental results show that,under this background the definition of the dependencies and measurement of dependencies degree are reasonable,negotiation agents using the multi-objective genetic negotiation algorithms（MO- GNA）and strengthening their strategies is viable,and the negotiation result reaches a pareto-optimality.

关 键 词：多议题协商 依赖关系 依赖度 多目标遗传协商算法 PARETO最优 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]