Modeling Strategic Route Choice and Real-Time Information Impacts in Stochastic and Time-Dependent Networks
Modeling Strategic Route Choice and Real-Time Information Impacts in Stochastic and Time-Dependent Networks
复制标题
DOI：
10.1109/tits.2012.2187197
复制
发表时间：
2012-09-01
影响因子：
8.5
通讯作者：
Gao, Song
中科院分区：
文献类型：
--
作者：
Gao, Song
本文建立了一个总体框架来研究广义均衡条件下随机时间相关交通网络中实时信息对用户路由决策和系统成本的影响。假设用户做出策略路由决策，将用户当前状态（包括节点、时间和信息）映射到下一个节点的决策的规则被定义为路由策略。该定义考虑到了各种各样的信息可访问性情况，因此排除了通常简化的假设，例如没有信息或完整信息。用户的选择集包含路由策略而不是简单路径。给出了用户均衡的定点问题公式，并设计了连续平均启发式方法。计算测试是在假设的网络中进行的，其中随机事件是随机性的来源。比较了具有不同信息可访问性情况的三种模型得出的系统成本。战略路线选择导致均衡时预期旅行时间更短。作为副产品，获得了较小的行程时间方差。实时信息的价值是事件概率的递增函数。然而，它并不是信息市场渗透率的单调函数，这表明在设计旅行者信息系统或路线引导系统时，需要仔细选择信息渗透率以实现利益最大化。
This paper establishes a general framework to study the impacts of real-time information on the users' routing decisions and the system cost in a stochastic time-dependent traffic network under a generalized equilibrium condition. Users are assumed to make strategic routing decisions, and the rule that maps a user's current state, including node, time, and information, to a decision on the next node to take, is defined as a routing policy. This definition allows for a wide variety of information accessibility situations, thus excluding the usually simplified assumptions, such as either no information or full information. A user's choice set contains routing policies rather than simple paths. A fixed-point problem formulation of the user equilibrium is given, and a method of successive average heuristic is designed. Computational tests are carried out in a hypothetical network, where random incidents are the source of stochasticity. System costs derived from three models with different information accessibility situations are compared. The strategic route choices lead to shorter expected travel times at equilibrium. Smaller travel time variances are obtained as a byproduct. The value of real-time information is an increasing function of the incident probability. However, it is not a monotonic function of the market penetration of information, which suggests that in designing a traveler information system or route guidance system, the information penetration needs to be chosen carefully to maximize benefits.