考虑有限理性的交通流Day-to-day动态演化研究

发布时间：2018-03-06 13:15

本文选题：累积前景理论　切入点：诱导信息　出处：《南京大学》2014年硕士论文　论文类型：学位论文

【摘要】：城市交通网络中,日常路径选择是一个参与者众多的动态学习与博弈过程,交通个体不断获得新的路网信息并更新路网认识,进而调整出行路线,因此,具有重复性的日常出行是动态的学习与更新过程,其宏观表现为网络交通流的动态演化。本文首先分析了日常城市交通出行的动态学习与更新的行为与心理过程,考虑出行者择路过程中的有限理性和参考点依赖,建立了基于累积前景理论的day-to-day动态交通系统,并研究不同精确程度的诱导信息和不同学习机制下出行者日复一日的动态路径选择以及相应的交通系统演化规律,最后,针对具体出行情境,在考虑决策规则异质性的交通流day-to-d ay动态分配方面做了初步探讨。本文具体研究内容和研究结果如下：首先,根据出行者个体出行行为决策特征,建立基于累积前景理论的出行决策行为框架,并将行为决策科学的相关研究成果引入动态交通出行行为机理研究,建立了诱导信息下基于累积前景理论的day-to-day动态交通出行概念模型。随后,考虑出行者的学习与信息更新过程,在路段通行能力退化和交通诱导信息下建立了参考点调整变化的day-to-day动态交通模型,研究所建立的动态交通系统能否演化到收敛以及演化过程中的网络流量分布；探讨诱导信息的精确程度以及出行者学习更新机制的不同对网络交通流演化规律的影响。最后,从具体出行情境出发,考虑由出行者个体特征以及外部环境引起的决策行为差异,分别建立了使用与不使用交通诱导信息系统的两类出行者的决策规则,研究出行者决策规则异质性对交通流分布形态与动态演化的影响,以期通过相应的显性社会特征和个体特征对出行者进行分类,有针对性地制定更加合理的信息诱导策略对出行者进行分类引导。
[Abstract]:In urban transportation network, daily path selection is a dynamic learning and game process with many participants. Traffic individuals constantly acquire new road network information and update road network knowledge, and then adjust travel routes. Daily trip with repeatability is a dynamic learning and updating process, and its macroscopic manifestation is the dynamic evolution of network traffic flow. Firstly, this paper analyzes the dynamic learning and updating behavior and psychological process of daily urban traffic travel. Considering the limited rationality and reference point dependence in the traveler's road selection process, a day-to-day dynamic traffic system based on cumulative foreground theory is established. The paper also studies the dynamic path selection and the corresponding rules of transportation system evolution under different degree of guidance information and different learning mechanisms. Finally, according to the specific travel situation, The dynamic assignment of day-to-d ay in traffic flow considering the heterogeneity of decision rules is discussed. The specific contents and results are as follows: firstly, according to the characteristics of individual travel behavior decision, The framework of trip decision behavior based on cumulative foreground theory is established, and the related research results of behavioral decision science are introduced into the study of dynamic traffic travel behavior mechanism. A day-to-day dynamic traffic travel conceptual model based on cumulative foreground theory under induced information is established. Then, the learning and information updating process of travelers is considered. The day-to-day dynamic traffic model with reference point adjustment change is established under the condition of road capacity degradation and traffic guidance information. Can the dynamic traffic system evolve to convergence and network flow distribution in the evolution process are established. This paper discusses the accuracy of induced information and the influence of different learning and updating mechanisms on the evolution of network traffic flow. Finally, starting from the specific travel situation, Considering the difference of decision behavior caused by the individual characteristics of the traveler and the external environment, the decision rules of the two types of travelers using and without the traffic guidance information system are established, respectively. This paper studies the influence of the heterogeneity of traveler's decision rules on the traffic flow distribution and dynamic evolution, in order to classify the traveler through the corresponding dominant social characteristics and individual characteristics. A more reasonable information guidance strategy should be developed to guide the travelers.
【学位授予单位】：南京大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491

【相似文献】