基于时间和资源约束的多Agent双边动态协商模型  

A Dynamic Bilateral Negotiation Model between Multi-agent under Time and Resource Constraints

作　　者：王灿[1,2] 梁永全[1] 

机构地区：[1]山东科技大学,山东青岛266590 [2]山东商业职业技术学院,山东济南250103

出　　处：《山东农业大学学报（自然科学版）》2014年第4期605-609,共5页Journal of Shandong Agricultural University：Natural Science Edition

基　　金：山东省自然科学基金(ZR2012FM003;ZR2013FM023);青岛市科技发展计划项目(KJZD-13-29-JCH);高等学校博士学科点专项科研基金(20133718110014)

摘　　要：双边交易Agent协商中因时间、资源的约束和信息的不完整,常会导致协商失败和较差的协商收益。本文提出一种基于时间和资源双重约束下的多Agent动态协商模型,使用出价函数和出价曲线来设定Agent的策略,采用特定协议来描述协商资源对心理状态的影响。同时,定义了折中、保守和急躁三种协商策略以及心理状态对策略的动态改变规则。通过实验模拟发现"多买方、单卖方"条件下,采用提出的协商策略和动态协商算法能够提高协商效率、降低成交价格和提升协商满意度。The bilateral bargaining agents with incomplete information will often fall into failure or cause poor utility ofnegotiation under time and resource constrains. We present a dynamic negotiation model between multi-agent under time andresource constraints by using offer function and offer curve to design agent’s strategy and adopting special protocol todescribe the influence of negotiating resource on mental states. As well as we also define three negotiating strategies namedcompromised、conservative and impatient strategies and rules of the strategy’s changes with the transformation of mentalstates. Experiment shows that the proposed negotiating strategy and algorithm can improve the efficiency of negotiation,reduce the transaction price and enhance the satisfaction between buyers and seller under the “many buyers-to-one seller”model.

关 键 词：动态协商 协商策略 协商算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]