考虑拥挤效应的随机交通悖论分析

发布时间：2018-04-12 19:51

本文选题：交通悖论 + 随机用户平衡　；参考：《哈尔滨工业大学》2017年硕士论文

【摘要】：交通悖论在交通规划与设计中的作用不容忽视。论文选用多元logit模型和多元weibit模型进行随机交通分配,探讨拥挤效应和随机感知误差对随机交通悖论的影响。选取三种不同结构的交通网络对随机交通悖论的产生特征进行分析,分别为:简单的单起讫点对两线网络、单起讫点对多路径网络、多起讫点对多路径网络,获得一些一般化的结论。首先,论文从出行者出行费用的角度分析了在改善或新增一条路段的情况下,随机交通悖论在不同结构的三种交通网络中有何异同。对比无拥挤情况,讨论了拥挤效应及随机感知误差参数对随机交通悖论产生条件的影响。其次,分析随机路径选择模型中的感知误差参数对出行者出行费用的影响,研究信息悖论即增大随机感知误差可能导致网络中出行者平均出行费用降低在三种不同结构交通网络中的产生情况,给出信息悖论发生的需求区间,研究了是否可以通过调整随机感知误差参数使流量分配达到系统最优。最后,论文以logit随机分配模型为例,从出行负效用的角度研究了三种交通网络中随机交通悖论的产生情况,分别在改善或新增一条路段的条件下分析拥挤条件和随机分配模型中的随机感知误差参数对发生悖论的边界条件所产生的影响,对比其与基于出行费用的随机交通悖论特征的差异性。在应用多元logit模型和多元weibit模型讨论随机交通悖论的基础上,本文发现拥挤效应可以显著地影响随机分配悖论的产生,讨论了信息悖论的产生条件,说明了随机用户均衡模型的平衡解可以接近系统最优解,发现在拥挤的交通网络中,基于出行负效用的随机分配悖论可能会产生。本文得到的结果有助于进行科学合理的交通管理与控制,提高道路交通效率。
[Abstract]:The role of traffic paradox in traffic planning and design can not be ignored.In this paper, the multivariate logit model and the multivariate weibit model are used for stochastic traffic assignment, and the effects of congestion effect and random perceived error on the stochastic traffic paradox are discussed.Three kinds of traffic networks with different structures are selected to analyze the characteristics of stochastic traffic paradox. They are: simple single-point to two-line network, single-origin point-to-multipath network, multi-origin point-to-multipath network.Some general conclusions are obtained.Firstly, this paper analyzes the similarities and differences of random traffic paradox in three traffic networks with different structures under the condition of improving or adding a new road section from the point of view of traveller travel cost.Compared with the case of no congestion, the effects of congestion effect and random perceptual error parameters on the production conditions of stochastic traffic paradox are discussed.Secondly, the influence of perceived error parameters in the random path selection model on the travel cost is analyzed.The information paradox, that is, increasing the random perception error, may lead to the decrease of the average travel cost in three different traffic networks, and the demand interval of the information paradox is given.This paper studies whether the optimal flow allocation can be achieved by adjusting the random perceptual error parameters.Finally, taking the logit stochastic assignment model as an example, this paper studies the generation of stochastic traffic paradox in three traffic networks from the perspective of travel negative utility.Under the condition of improving or adding a new section of road, the influence of the random perceptual error parameters in the random allocation model on the boundary condition of paradox is analyzed.The differences between the characteristics of stochastic traffic paradox based on travel cost and that based on travel cost are compared.On the basis of discussing stochastic traffic paradox by using multivariate logit model and multivariate weibit model, this paper finds that congestion effect can significantly affect the generation of stochastic distribution paradox, and discusses the conditions for generating information paradox.It is shown that the equilibrium solution of the stochastic user equilibrium model can approach the optimal solution of the system. It is found that the stochastic assignment paradox based on the travel negative utility may occur in congested traffic networks.The results obtained in this paper are helpful to scientific and reasonable traffic management and control and improve road traffic efficiency.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F503

【参考文献】