基于动态交通信息的路径选择行为的实证研究

The potential of ATIS (Advanced Traveler Information System) for alleviating traffic congestion has been of interest to researchers and practitioners since 1990s. It is well known that ATIS can help drivers make better choices of modes, departure times and routes, and therefore reduce travel times and vehicle miles travelled (VMT) on the network. The potential of ATIS is highly dependent on travelers' responses to provided information. Trip characteristics, drivers' socioeconomic characteristics and perceptions of the relevance and reliability of traffic information were found to be important factors affecting drivers' route choice behavior. However, there is little evidence of how drivers perceive the information with different patterns of travel time information, and how the perception affects drivers' behavior. In addition, the effect of travel information in alleviating traffic congestion has been heavily investigated; however, the net increase in drivers' utility under the presence of travel information has been neglected. A further research in this field is necessary..This study aims to examine drivers' route choice behavior in response to real-time travel time information. The study begins with a literature review of the previous research since 1990s, and this is followed by a design of Stated Preference (SP) survey which is conducted in Dalian city. In the survey, the travel information is provided with three patterns: single-point travel time information, minimum- maximum travel time information and multiple-point travel time information. A set of route choice modes incorporating drivers' perceptions to the travel time information is estimated based the survey data. The study also examines the changes among different drivers' group by exploring market segmentation method. Finally a utility-based approach is proposed to access drivers' benefits from travel time information provision.

先进的交通信息服务系统通过提供实时的交通路况信息，帮助机动车驾驶员合理选择交通出行行为，改变出行方式、出发时间和路径，从时间和空间两方面分流出行需求，对于缓解交通拥挤具有显著作用。基于交通信息的交通出行行为研究是近年来国内外的研究热点之一。. 本项目在国内外既有研究成果的基础上，设计交通意愿调查问卷、收集驾驶员路径选择行为数据，分析比较在不同的动态交通信息下（单点出行时间信息、多点出行时间信息、最小最大出行时间信息），驾驶员对于交通信息的感知度，建立以感知出行时间为主要变量的路径选择模型，并应用该模型研究各类驾驶员群体出行行为的变化。本项目还提出基于期望最大效用计算交通信息的经济效益的方法。. 本项目提出的基于感知出行时间的建模方法以及基于期望最大效用计算交通信息的经济效益的方法，具有重要的科学研究意义；研究成果可为设计有效的交通信息发布方式及内容提供参考依据。

随着我国国民经济的持续快速发展，城市机动车保有量急剧攀升，交通运输供需不平衡的矛盾日益尖锐。目前，我国大中城市道路严重堵塞，机动车污染日益严重，道路交通安全形势依然严峻，城市交通问题已经影响了城市社会经济的正常发展。.提供实时的交通路况信息可以帮助机动车驾驶员合理选择交通出行行为，改变出行方式、出发时间和路径，从时间和空间两方面分流出行需求，进而有助于缓解交通拥挤。.本项目在国内外既有研究成果的基础上，设计交通意愿调查问卷、收集驾驶员路径选择行为数据，分析比较在不同的动态交通信息下，驾驶员对于交通信息的感知度，建立以感知出行时间为主要变量的路径选择模型，并应用该模型研究各类驾驶员群体出行行为的变化，还提出基于期望最大效用计算交通信息经济效益的方法。 .（1）本项目提出基于驾驶员感知出行时间的建模方法，引入驾驶员本身的感知出行时间等变量，研究ATIS（先进交通信息系统）准确性对驾驶员的符合描述性信息和路径选择行为的影响。利用二元logit模型，表明ATIS的精度不仅影响驾驶员的合规性，也影响每条选择路径的可靠性，同时发现在决策过程中对待风险的态度同样重要。.（2）从交通信息的内容和类别入手，综述了交通信息对出行者行为的影响，包括出行者在出发时间、出行方式、目的地选择和出行路径等方面。.（3）建立基于期望最大效用计算交通信息经济效益的方法。对3个案例进行调查，用累积前景理论（CPT）更精确地描述驾驶员的路径选择行为。结果表明，Prelec的概率加权函数更适合驾驶员的路径选择行为；驾驶员对收益的敏感性降低，对损失的敏感性降低。