基于小生境遗传算法的自动谈判模型  

Automatic Negotiation Model Based on Niche Genetic Algorithm

作　　者：陈培友[1,2] 李义玲[2] 

机构地区：[1]哈尔滨工业大学管理学院,哈尔滨150027 [2]黑龙江科技学院经济管理学院,哈尔滨150027

出　　处：《计算机工程》2011年第1期194-196,200,共4页Computer Engineering

基　　金：黑龙江省博士后基金资助项目(LBH-Z05129);黑龙江省研究生创新科研基金资助项目(YJSCX2009-071HLJ)

摘　　要：为削弱自动谈判模型中谈判Agent对偏好信息汇报的主观度,引入直觉模糊区间数的多准则算法,得到各方案的区间直觉模糊集,并结合不完全确定的权系数信息建立非线性规划模型,得出谈判Agent对各属性的谈判权重。为克服遗传算法早熟现象,运用小生镜遗传算法对谈判空间进行全局快速寻优,提高收敛性。仿真计算表明该模型是有效的,而且精度高、数值稳定性好,能快速客观地辅助谈判Agent达成一致协议。A multi-criteria ranking method about interval-valued intuitionistic fuzzy sets is introduced into the automatic negotiation model to weaken the subjectivity degree of preference information report from negotiation Agent and the interval intuitionist fuzzy sets of alternatives are obtained. Considering the incomplete certain information on weights, a nonlinear programming model is developed and the weight of various attribute is obtained. In order to overcome the precocious phenomenon on Genetic Algorithm（GA）, the Niche Genetic Algorithm（NGA） is proposed to search optimization and improve astringency. Simulation test shows the model is feasible and effective, the precision is high and the value-stability is good, which can help negotiation Agent to reach fast a mutually beneficial agreement objectively.

关 键 词：自动谈判 议价区间 区间直觉模糊数 小生境遗传算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]