基于出行时间可靠性的交通配流问题  

Travel time reliability-based traffic assignment problem

作　　者：邵虎[1,2] 林兴强[2] 孟强[3] 谭美琳[2] 

机构地区：[1]中国矿业大学理学院数学系,徐州221116 [2]香港理工大学土木及结构工程学系 [3]新加坡国立大学土木工程系,新加坡117576

出　　处：《管理科学学报》2009年第5期27-35,共9页Journal of Management Sciences in China

基　　金：Research Committee of Hong Kong Polytechnic University(87PV);中国矿业大学青年科研基金项目(2007A028);徐州市科技项目(XM08C095).

摘　　要：提出一类由需求随机性所导致的基于出行时间可靠性的交通配流问题.由于每一天交通需求的随机变化,出行者的出行时间不是确定的,而是随机变量.假设出行者在过去经验的基础上能够得知出行时间的随机分布,提出一类新准则去刻画出行者在出行时间不确定情况下的路径选择行为.这种准则可以表示为一种以路径流量为变量的变分不等式模型.对于这类新的模型,给出了解的存在性证明,并且引入一个启发式的算法去求解该问题.数值算例展示了模型在应用上的特性和算法的有效性.This paper presents a travel time reliability-based traffic assignment model in view of the daily demand variation. Due to the stochastic travel demand, path travel times are not constant, which can be viewed as random variables. Assuming that the travelers are able to learn the variation of the path travel time based on past experiences, this paper proposes a travel time reliability-based user equilibrium （TRUE） principle to characterize travelers＇ path choice behavior under uncertainty in travel times raised from demand variation. This principle can be formulated as an equivalent path-based variation inequality （VI） problem. It is rigorously proved that there is at least one solution for the VI problem. A heuristic solution algorithm is employed for solving the VI problem. Numerical examples are used to illustrate the applications of the proposed model and the solution algorithm.

关 键 词：需求不确定性 出行时间可靠性 交通配流问题 变分不等式 

分 类 号：U491[交通运输工程—交通运输规划与管理]