考虑停车换乘的多方式交通网络出行行为分析

发布时间：2018-02-14 13:15

本文关键词： 停车换乘多方式交通网络用户均衡出行选择行为瓶颈模型出行时间不确定 Stackelberg博弈模型　出处：《北京交通大学》2015年博士论文　论文类型：学位论文

【摘要】：在制定交通管控措施以及进行交通评价时,在充分了解和掌握交通需求状况的基础上,还需要充分了解出行者的出行选择行为。在不同交通网络环境中,出行者的出行选择行为有显著的不同,特别是将出行者的停车行为考虑到整个出行过程中时。停车换乘(park-and-ride, PR)作为一种重要的交通需求管理手段,是一种既不影响驾车便利性又可以鼓励出行者使用公共交通的措施,可以转移交通需求,优先发展公共交通,从而提高城市交通资源的利用效率。本文运用系统分析的方法,分别在确定的交通网络环境条件下和随机的交通网络环境条件下,将出行者的停车行为考虑进出行者的出行选择行为中去。在考虑PR的多方式交通网络上研究出行者的出行选择行为,为城市交通和停车设施的规划与管理、交通管控措施的制定与评价提供了理论参考。本论文的主要工作如下： (1)基于用户偏好的多方式交通网络随机用户均衡模型。考虑了出行者对出行方式、换乘站点以及出行路径的选择偏好对出行者出行选择行为的影响,基于Logit离散选择模型,建立了多用户多方式混合交通随机用户均衡的变分不等式模型,并讨论了城市中心区的停车费、地铁或公交的发车频率、总需求以及不同时间价值的出行者所占比例的变化对出行选择偏好以及网络上的流量分布的影响。结果表明,城市中心区的停车费以及公交车的发车频率的变化对道路交通影响显著；不同出行者所占的比例的变化,网络上不同时间价值的出行者的流量分布显著不同。 (2)考虑停车换乘的异质通勤出行者的出行选择。在考虑交通瓶颈的城市快速路和停车换乘方式相互竞争的多方式交通网络上,分析比较了不同用户均衡状态类型存在的必要条件及其稳定性。为研究停车收费对异质出行者出行选择行为的影响,论文构建了社会总成本最小时的最优停车费组合模型,并给出了模型的解析结果。同时还研究了瓶颈处的通行能力以及出行者所占比例的变化对出行者的出行选择行为和社会总成本的影响。 (3)出行时间不确定时的停车费定价和方式分担。在一个有城市快速路和轨道交通线路竞争的随机交通网络上,提出了一个关于停车收费的定价和方式分担的双层规划模型。由于网络结构的特殊性,建立的双层规划模型在用户均衡状态时转化为一个极小化问题,在此基础上,本论文给出了模型的解析结果。同时论文分析比较了平均出行时间、出行时间预算、平均-超额行程时间这三种出行时间选择标准对最优的停车费和均衡状态时交通网络上流量分布情况的影响。作为模型的拓展,本文还探讨了不同参数变化在不同出行时间选择标准下对出行选择的影响,以及道路路段上走行时间的分布对出行者出行选择的影响。结果表明,一方面,不管是否考虑出行时间的不确定性,以及如何对待出行时间的不确定性,提高地铁的服务水平和运营效率都可以有效地诱导出行者的选择行为,调整不同出行方式在交通网络上的流量分布；另一方面,降低交通网络上出行时间分布的方差比仅仅减少出行时间分布的均值对调整网络上的流量分布以及降低社会总成本的效果要更加明显。 (4)停车时间不确定时的网络均衡分析。将交通网络上的停车场分为不同的群集,并假设同一群集的停车场有相同的收费价格和相同的停车时间。在探讨网络上出行时间的不确定时,考虑出行者停车时间的可靠性。本文基于出行时间的可靠性和出行者对出行负效用理解的随机误差,建立了与用户均衡条件等价的变分不等式模型。通过数值算例,分析比较了不同交通总需求、不同置信水平、不同交通管控措施以及停车搜索和进出停车场时间的分布对出行者出行选择行为的影响。结果表明,降低停车时间的分布的方差,可以降低整个出行的出行时间的不确定性,进而影响整个网络上的流量分布；合理地调整城市中心区的停车费或地铁票价,可以有效的调整不同出行方式在交通网络上的流量分布,降低全程开车的分担率,从而缓解城市道路交通拥堵。 (5)简单交通网络上的停车位分配与定价。当政府拥有所有的停车位时,将出行者按照是否在城市中心区有保留车位和时间价值进行分类,为研究政府与私人运营商之间的相互竞争,对考虑道路路段交通瓶颈与否的两种情况,分别建立了三层Stackelberg博弈的停车位分配与定价模型。从考虑道路交通瓶颈的解析结果发现,在终点没有保留车位的出行者由于其时间价值比较低全都会选择收费价格比较低的停车换乘地铁的方式出行。作为模型的拓展,进一步给出了弹性需求条件时的模型构建情况。
[Abstract]:In the development of traffic control measures and traffic evaluation, based on fully understand and grasp the traffic demand situation, also need to fully understand the travel choice behavior. In the different traffic network environment, travel choice behavior is significantly different, especially the parking behavior of traveler considering the entire during the trip. The park and ride (park-and-ride, PR) as an important means of traffic demand management, is a does not affect the driving convenience and can encourage travelers to use public transport measures can transfer traffic demand, the priority development of public transport, so as to improve the utilization efficiency of city traffic resources. This paper applies the method of the system analysis, respectively in the transportation network to determine the environmental conditions and stochastic traffic network environment, the parking behavior of traveler's travel choice for considering import Travel behavior is studied in the multi way traffic network considering PR. It provides a theoretical reference for the planning and management of urban traffic and parking facilities, and the formulation and evaluation of traffic control measures.
(1) multiple user preference model based on stochastic user equilibrium traffic network. Considering the traveler on the way to travel, transfer stations and preference to select travel routes to influence travel choice behavior, Logit discrete choice model based on the established mixed stochastic user equilibrium traffic multi user multi type inequality model. And discuss the city center parking fee, subway or bus departure frequency, the total demand and the different value of time traveler of the proportion of choice preference and the network traffic distribution of travel. The results show that the influence of city center's parking fees and change frequency buses on the road significant changes in the proportion of travelers; the network of different time value of the flow distribution of walkers are significantly different.
(2) consider the heterogeneity of the park and ride commuters travel choice. In considering the multi way traffic network competing traffic bottleneck of the city road and parking transfer methods, analysis and comparison of different types of necessary condition for user equilibrium and its stability. The research of parking charge on heterogeneous travelers' choice of this paper, constructs the optimal combination model of parking fees total social cost minimum, and the analytical results of the model are given. We also studied the bottleneck of the traffic capacity and the traveler proportion change of travel choice behavior and social effects of the total cost.
(3) share the travel time of the parking fee pricing way. And in a city expressway and rail transit line competition stochastic traffic network, puts forward a bi level programming model of sharing a parking fee pricing and the way. Because of the particularity of the network structure, the bi level planning model is established for the transformation a minimization problem in user equilibrium, on this basis, this paper gives the analytical results of the model. At the same time, this paper compares the average travel time, travel time budget, the average excess travel time of the three travel time selection criteria of the optimal parking fees and equilibrium traffic flow on the network distribution effect. As the model expansion, this paper also discussed the different parameters selection criteria in different travel time effect on travel choice, and the road section travel time. The distribution of impact on travel choice. The results show that, on the one hand, regardless of whether or not to consider the travel time uncertainty, and how to deal with the travel time uncertainty, choice behavior of enhancing the service level and the operating efficiency can effectively induce the travelers, adjust the flow distribution of different travel modes in the traffic network; on the other hand, reduce the variance of traffic network travel time distribution than just reduce the average travel time distribution of flow rate adjustment on the network distribution and reduce the social total cost effect to be more obvious.
(4) analysis of the network equilibrium parking time is uncertain. The parking lot traffic on the network is divided into different clusters, and assuming the parking lot the same cluster have the same price and the same parking time. On the uncertain network travel time reliability, considering the travelers stop time in this paper, the reliability. The travel time and traveler travel disutility of random error based on understanding, established with the user equilibrium condition of equivalent variational inequality model. By numerical examples, analysis and comparison of different traffic demand, different levels, different traffic control measures and parking search and import time distribution on the parking lot the influence of travel choice behavior. The results show that the reduced variance distribution of parking time, can reduce the travel time uncertainty, thereby affecting the entire network traffic Reasonable distribution of parking fees or subway fares in urban central area can effectively adjust the flow distribution of different travel modes on traffic network and reduce the sharing rate of whole driving, thus easing the traffic jams of urban roads.
(5) parking space allocation and pricing of simple traffic on the network. When the government owns all the parking spaces, the traveler according to whether in the city center have reserved parking spaces and the time value of classification, to compete with each other on the government and private operators, to consider two cases of road traffic bottlenecks or not the parking space allocation and pricing model of three layer Stackelberg game were established. Results from analysis of road traffic bottleneck in consideration, no parking reservation end point of travelers due to transfer to the subway stop and the time value is relatively low will all choose to charge a lower price of the travel mode. As a model of development, further given. Construction of elastic demand model of the situation.

【学位授予单位】：北京交通大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：U491

【参考文献】