基于路径运行时间可靠度的随机系统最优拥挤收费模型

应用交通网络平衡模型和边际成本收费理论相结合的方法,研究了运行时间可靠度下的随机系统最优拥挤收费问题,建立了运行时间可靠度及内生ATIS市场渗透率条件下随机系统最优交通拥挤收费模型.分析了基于运行时间可靠度下的随机系统最优拥挤收费对用户出行行为的影响.发现了与确定性网络用户平衡流中的情形类似,对于考虑运行时间可靠度下的随机交通网络,边际成本收费理论仍然适用,即采用边际社会成本流函数代替单位路段成本流函数,可以使随机网络随机用户平衡流变为随机网络随机系统最优流.算例分析结果表明:在传统的拥挤收费模型中,拥挤收费仅与路径(路段)运行时间和路径(路段)流量有关.现实中,在确定他们的出行路线时,用户往往还会考虑网络运行时间可靠度因素,而不仅仅是路径运行时间或成本.用户对于运行时间可靠度的置信度要求越高,传统的拥挤收费执行效果越不理想.因此,现实生活中传统的拥挤收费不一定能使网络效益达到最优或缓解交通拥挤.

Path travel time reliability-based stochastic system optimum congestion pricing model

The reliability-based stochastic system optimum congestion pricing was studied. The approach employed in our study was a combined application of the traffic network equilibrium modeling and the Marginal Cost Pricing (MCP) theory. A reliability-based stochastic system optimum congestion pricing model with respect to the stochasticity of network travel times under Advanced Traveler Information Systems (ATIS) with endogenous market penetration was established. The influence of the reliability-based stochastic system optimum congestion pricing on traveler's travel behaviour was analyzed. It is pointed that analogous to the situation in the deterministic case, the principle of marginal cost pricing is still applicable in a travel time reliability-based stochastic traffic network. That is, the Stochastic Network-Stochastic User Equilibrium (SN-SUE) flow pattern can be driven to the Stochastic Network-Stochastic System Optimum (SN-SSO) flow pattern by using the marginal cost-flow functions in place of the link cost-flow functions. Analysis results indicate that in traditional congestion pricing models, congestion pricing is only related to route (link) travel time and route (link) traffic flow. However, in real life, travelers often choose their travel routes based on the network travel time reliability, not only on route travel times or costs. The higher the travel time reliability the traveler is concerned with, the worse the actual effect of the traditional marginal cost pricing. Therefore, the traditional congestion pricing may neither drive the traffic network toward the system optimum nor reduce total network travel cost.