考虑参考依赖偏好的交通拥挤收费机制优化与福利分析研究

In recent years, road congestion problem in cities has brought the unprecedented challenges to the traffic management.Congested road-use pricing as an effective means for both managing road traffic demand and raising additional revenue for road construction and maintenance or subsidizing the public transport has been extensively studied by both transportation researchers and economists.An incontestable fact is that there is a great need for the development of efficient road-use pricing models in a general way. Under this background, the research will take the travelers' reference-dependent congnitive preferences into the traditional congesion pricing model, and then design and optimize the congestion pricing schemes.Firstly, the static congestion pricing models with reference-dependent preferences are constructed to analyze the estimation bias of ignoring the congnitive preferences in several different cases.Secondly, the dynamic congestion pricing models with reference-dependent preferences are further designed, to be specific, the dynamic congestion pricing levels for different time in a day and the day-to-day dynamic congestion pricing strategies are optimized, respectively. Thirdly, the welfare effect for congestion pricing refunding schemes are investigated. Finally, the behavioral experiments are conducted to support the previous theoretical observations. In a word,the research wishes to help correct the bias both in pricing levels and public acceptance due to the absence of cognitive preferences, and provide some valuable suggestions to traffic management sector.

近年来,城市道路拥堵问题给交通管理带来了前所未有的挑战。拥挤道路使用收费不仅能够有效地管理交通需求，所得的额外收入还可用于道路建设和维护、补贴公共交通等，正在世界范围内被检验和完善。理论和实践都迫切需要研究更为"感性"的一般拥挤收费模型。在此背景下，本项目以交通拥挤收费作为研究对象，在传统的拥挤收费模型中添加出行者参考依赖的认知偏好，首先构建具有参考依赖偏好的静态拥挤收费模型，从多个角度研究忽略认知偏好对最优收费水平造成的估计偏差；然后，设计基于参考依赖偏好的动态收费模型，分别对一日内不同时间段的动态拥挤收费水平和日常（Day-to-day）动态收费策略进行优化；其次,对拥挤收费返还计划的福利效果进行分析；最后，借助行为实验为前面的理论模型提供进一步的现实依据。本项目的研究结果旨在帮助纠正因忽略认知偏好导致的拥挤收费定价和认识偏差，为交通管理部门提供有价值的参考意见。

本项目在传统的交通网络均衡模型中添加出行者参考依赖的认知偏好，从理论建模和实证调查两个方面阐述了参考依赖的认知偏好对出行行为和交通政策实施效果的影响。首先研究了参考依赖特性对收费道路利润和社会福利的影响，其次设计了动态拥挤收费策略，分析了两参考点模式下的日常出行流量均衡演化，然后考察了参考依赖特性对出行方式和出行路径选择结果的影响，讨论了不同的道路优化管理策略对个体感知价值提高的作用，并通过调查问卷数据验证了参考依赖特性对个体选择行为的影响，最后设计了集可交易通行权和鼓励合乘于一体的组合新政，分析了该政策对各类用户和整个交通系统的影响。.本项目的研究结果有助于加深对个体通勤行为的理解，引导交通需求在路网上合理分布，为交通管理部门提供优化管理的参考意见。到目前为止，已在国内外重要刊物和学术会议上共发表标注论文12篇，其中SCI/SSCI收录论文2篇， EI收录论文8篇，CSSCI 收录论文1篇，其他国际期刊（SCI收录期刊的子刊）论文1篇，指导2名硕士研究生顺利毕业，目前在读硕士研究生4名。