Abstract: Based on Cournot behavioral assumption, with the consideration of passenger travel behavior during rush hour, we investigate transit system optimization. The interval of transit vehicle is assumed to be determined by transit companies, while the total number of vehicle is the result of passengers’ equilibrium riding behavior. Considering demand elasticity, arrival early penalty and congestion in carriages, we propose two optimization models of the rush hour transit system on the base of passengers’ equilibrium riding behavior. The first model aims at maximizing the net benefit of the system, gives the optimal value of the demand realized totally and the number of trains required. The second model maximizes every company’s profit in an oligopoly market regime, determines the optimal time interval of dispatching trains. Numerical results are presented for illustrating analytical results. It is found that, the number of transit company could influence system manifestation, which could be cutthroat competition without government regulation, accompanied by the phenomenon of wasting social resources.

Key words: urban traffic, arrival early penalty, congestion in carriages, elastic demand, equilibrium model, cournot behavior assumption

摘要： 基于古诺模型假设,在高峰期乘客交通选择行为的基础上,研究公共交通系统的优化问题. 建立模型时,假设公交车辆的发车间隔时间由公交公司决定,车次总数由乘客的均衡班次选择行为所致. 弹性需求下，考虑乘客的早到成本和车内拥挤成本，在均衡班次选择行为的基础上，建立了高峰期公共交通系统的社会净收益最大化模型，得到了公交系统实现的最优运量和最佳班次数. 在古诺寡头市场条件下，以公司利润最大为目标，确定了公交系统的发车班次间隔. 研究发现,公交系统的公司数目会对系统表现造成影响,如果缺乏政府管制,可能出现恶性竞争,导致社会资源的浪费.

关键词: 城市交通, 早到成本, 车内拥挤成本, 弹性需求, 均衡模型, 古诺假设