基于有限理性的个体出行路径选择进化博弈  

Evolutionary game mode about individual travel route choice based on bounded rationality

作　　者：刘建美[1] 马寿峰[1] 

机构地区：[1]天津大学系统工程研究所,天津300072

出　　处：《控制与决策》2009年第10期1450-1454,1462,共6页Control and Decision

基　　金：国家自然科学基金项目(70671073);国家863计划项目(2006AA11Z210)

摘　　要：首先,从出行者的信息集和理性层次的差异两个方面给出两种改进的"最优反应动态"模型,并将其应用于一个算例,得到了与利用复制者动态方程所求得的进化稳定策略一致的结果;然后,为验证结论的普遍性,将出行者和路径数量扩充,构造了相应的算法,并将此算法及其复制者动态方程应用于一个规模较大的路网;最后,将不同机制下的实验结果进行了比较,并对其实际意义进行了分析.Firstly, the paper gives the original best-response dynamics travel model, and two improved models from two aspects of strategy adjustment and difference among the travelers＇ rationality levels under the assumption that all travelers are regarded as hounded rational. Then, a small example is given and solved by three different learning mechanisms with consistent learning results obtained by the continuous replicator dynamic equations. In order to confirm the universality of the conclusion, the number of travelers and routes are extended and a relevant algorithm is constructed. And the algorithm and continuous replieator dynamic equation are applied to one large example. Finally, the results under different mechanisms are compared, and the practical meaning is analyzed.

关 键 词：个体出行 有限理性 最优反应动态 复制者动态 进化博弈 

分 类 号：U491[交通运输工程—交通运输规划与管理]