电子商务中基于遗传算法的多Agent自动协商  

Genetic Algorithm Needed for Automated Negotiations among Multi-agents in Electronic Commerce Environment

作　　者：朱冬梅[1] 盛建瓴[1] 杜杏菁[1] 

机构地区：[1]华北科技学院计算机系,河北三河065201

出　　处：《煤炭技术》2011年第4期215-216,共2页Coal Technology

摘　　要：讨论了现有的多Agent协商技术,提出了一个基于遗传算法的多边多议题自动协商模型,并且给出了相应算法,该算法在评价函数、种群编码策略、交叉和变异操作等方面都对传统遗传算法进行了改进。实验表明:该算法能够迅速求出协商解,且Agent能保持较高的效用,具有一定的通用性。In this paper,current techniques of negotiation are discussed,a new many-to-many multi-issue automated negotiation model is proposed and the associated negotiation algorithms is provided.The algori-thm adopted real coding,adjusted fitness function,improved cross and variation operation.The model can solve problems rapidly and keep the utility of agents in a high level.

关 键 词：多AGENT 自动协商 遗传算法 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]