陈可心

随着社会经济的快速发展、生活水平的日益提高，人们对出行质量的要求也越来越高，小汽车出行需求的迅速增长使城市道路基础设施的容量捉襟见肘，城市道路交通拥挤问题愈发严峻。为缓解交通拥挤，新加坡、特隆赫姆、墨尔本等城市首先尝试了拥挤收费策略并卓有成效，此后拥挤收费在更多的城市得到实施并成为一种有效的交通拥挤管理手段。另一方面，机制设计理论所要做的，正是设计一个机制，使经济参与者即使在自利行为的驱使下做出决策，也能实现机制设计者所设想的社会目标，这种理念恰恰与拥挤收费策略不谋而合。因此，本文尝试将机制设计理论引入到拥挤收费策略当中，以求设计一种拥挤收费方案，使出行者在追求自身出行利益的同时，自发地实现交通管理者想要达到的社会利益最大化的目标。

本文在梳理了已有的拥挤收费理论研究与实践经验、机制设计理论的基本概念的基础上，将机制设计理论中的激励机制理念引入到拥挤收费策略中，建立了带有激励机制的多用户单方式以及多用户多方式的拥挤收费模型，结合SQP算法、遗传算法、内点算法求解最佳拥挤费额与相应的流量分配。

之后，本文构建了停车激励、公交票价激励两种具体的激励机制，并分别引入到多用户单方式的收费模型与多用户多方式的收费模型上。设计了简单路网，以数值计算分析了基于停车激励机制、公交票价激励机制的拥挤收费策略效果，发现其确实能够影响出行者的路径及方式选择。与无收费策略、实施传统收费策略情况下的策略效果进行对比，发现基于机制设计理论的拥挤收费策略可以平衡路段流量、获得更低的路网总行程时间，并带来一定的社会经济利益，具有一定的实施价值。在此基础上，本文对停车激励、公交票价激励等形式的激励机制提出了操作方式的建议，并给出了基于不同激励机制的拥挤收费策略的实现方式及相应的配合管理措施。

此后，本文研究了拥挤收费的公众可接受程度，利用多元方差法及回归分析法对问卷调查数据进行分析，得到了五个显著影响策略可接受性的因素以及它们之间的关系。最后，针对基于机制设计理论的拥挤收费策略的收入再分配问题给出建议。

关键词：拥挤收费策略，机制设计理论，激励机制，模型，管理

     With the rapid development of social economy and increasingly improved living standards, people have raised higher requirements of travel quality, and demand for travelling by car keeps growing, which brings a shortage of urban road infrastructure and a more crowded traffic condition. For the purpose of relieving traffic congestion, cities like Singapore, Trondheim, and Melbourne implemented the strategy of congestion pricing which was proved highly effective in these cities. Since then, congestion pricing has been carried out in more cities as an effective measure of urban traffic congestion management. On the other hand, what the mechanism design theory tries to do is to design a mechanism which can achieve the designer’s goal even when the economy participants are making decisions driven by their self-interest. This idea exactly coincides with the congestion pricing strategy. In consequence, this thesis tries to introduce mechanism design theory into congestion pricing strategy, in order to design a pricing scheme which will naturally obtain the maximum road network benefits when the travelers are pursuing their private profits.

     After the literature review on congestion pricing theory and its practical experience and the introduction on mechanism design theory, this thesis introduces the idea of “incentive mechanism” from mechanism design theory into congestion pricing strategy, and establishes two pricing models based on incentive mechanism. One of them is a multi-user and single-mode bi-level programming model, and the other one is a multi-user and multi-mode pricing model. SQP algorithm, Genetic algorithm and interior point algorithm are used to solve these models.

     Then, the thesis establishes two different incentive mechanisms. Through the calculation on an assumed road network, the effects of different congestion pricing strategies (congestion pricing based on two different incentive mechanisms, congestion pricing without incentive mechanism, and no pricing strategy at all) are compared. It is proved that congestion pricing strategies based on incentive mechanisms can truly influence the travelers’ choices, balance the traffic flow on each road link, relieve traffic pressure and bring some social economic benefits. On this basis, the thesis offers realizing forms of three different incentive mechanisms and specifies how congestion pricing based on these incentive mechanisms will be carried out.

     Afterwards, the thesis conducts a research on the acceptability of congestion pricing with the data collected through a questionnaire survey. Through a multi-factor analysis of variance, it is found that five variables can make significant impacts on the acceptability, and a binomial logistic model is established and calibrated to determine their relationship. Lastly, this thesis gives proposals for managing the income and expenses of congestion pricing together with incentive mechanisms.

Key Words: congestion pricing strategy, mechanism design theory, incentive mechanism, model, management