敏捷供应链中的谈判决策系统研究

《敏捷供应链中的谈判决策系统研究》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、糊评价之后，会选用与自己上一次建议接近同时又使自己利益最大的建议，这样既能保证自身的利益不会过分减少，也能使对方的利益增大。如此反复进行，直到双方达成协议，这时谈判的结果接近或达到pareto最优，并保证了社会利益的最大化，同时实现了谈判各方的合作“双赢／多赢，’o(5)在谈判决策过程分析和谈判决策算法分析的基础上，设计了基于黑板模式的谈判Agent和基于MAS的敏捷供应链谈判决策系统的体系结构．在此基础上采用JADE、MySQL等开源软件设计并实现了一个敏捷供应链谈判决策原型系统。具体应用实例的测试表明：该系统界面友好，实用性强，能实现敏捷供

2、应链谈判决策的多次交互。(6)为了使该谈判系统具有更好的智能性，采用模糊神经网络方法构建了谈判决策的学习算法模型，谈判Agent可以利用两阶段混合学习算法在谈判中不断优化谈判策略，从而缩短谈判进程，在更短的时间内获得更好的谈判结果。关键词：敏捷供应链谈判决策MAS模糊综合评价学习算法ABSTRACTAgilesupplychaincameforthwithagilemanufacturinganddynamicalliance．Agilesupplychainshouldberestructuredandadjustedrapidlyalong

3、withthereconfigurationordisorganizationofthedynamicalliance，whichshouldmeetthedevelopmentalrequirementofglobalmarketcompetitionandindividualuser．Sincethepatterenterprisesofagiledynamicalliancearethedistributed，self-interestedandautonomousentities，Negotiationdecisionsystemofa

4、gilesupplychainbecamethehotresearchtopicsbecauseofthedistributed，flexible，complicatedanddynamicprocessofnegotiationdecisionandestablishmentofsymbiosisintheagilesupplychain．Multi-agentsystem(MAS)hasthegoodadvantagesinmodeling,coordinatingandcontrollingtheagilesupplychain,whic

5、himprovestheagilesupplychainmanagementsystemaloLThe—glltAgentnegotiationmechanismshouldbeintroducedtothedynamicalliance，whichcanCometoanavailableagreementquickly,restructuretheagilesupplychainadaptively,andimprovethewholeperformanceeffectively．Themaincontentsofthisdissertati

6、oncanbespecifiedasfollows：(1)ThenegotiationdecisionprocessisanalyzedcompletelyusingNashEquilibriumandsuchrelatedgametheory,andthenthesymbiosisofagilesupplychainisestablishedwithnegotiation．Throughnegotiation，thepartnerenterprisesofdynamicallianceconstituteanagilesupplychain，

7、whichcallnotonlyreduce"BullwhipEffect”，butalsobringthepositivestructureeffect．Addingthescaleeffect,itcanmakethesystememergesomenewcharacterscontinuously，whichcallimprovethepayoffandcompetenceoftheagilesupplychain．(2)Amulti-agent,multi—attribute,multi·issueandmulti—levelcompr

8、ehensiveevaluationmodelbasedonfurylogicispresented．Whendiscussinghowtoselec