Abstract:

The content of timetable synchronization in the last period includes not only a guarantee on network accessibility, but also a reduction on transfer waiting time. Given that the connection relationships among trains are unfixed, binary variables are introduced along with penalty coefficients to provide an accurate representation of the transfer relationship of passengers in network. By adjusting trains running times, station dwell times and headways among trains, a model can construct high-quality timetables of the last trains’hour that minimize transfer waiting times with the distribution characteristics of transfer passengers. Then the CPLEX is used to solve the model. Finally, an empirical analysis of Beijing rail transit network is made to invalidate the validity and accuracy. Results show that: the model can improve the synchronization significantly compared with the current timetable; and it can provide a decision support to network operation in urban rail transit network.

Key words: urban traffic, the last trains’hour, mixed integer linear program, timetable synchronization, network operation

摘要：

在城市轨道交通末班车时段，时刻表协同优化不仅需保障网络可达性，还要减 少乘客等待时间.基于各线路列车间的衔接顺序不确定，通过引入0-1 变量和换乘失败惩 罚系数，来描述末班车时段列车间衔接关系，建立网络换乘模型；根据末班车时段内网络 衔接客流的分布特征，通过调整发车间隔、停站时间及区间运行时间等，构建以总乘客换 乘等待时间最小化为目标的末班车时段时刻表协同优化模型；并利用CPLEX对优化模 型进行求解.最后，以北京市轨道交通网络为案例，验证模型的有效性和准确性.结果表 明，该模型能有效生成末班车时段网络协同时刻表，为网络化运营组织提供决策依据.

关键词: 城市交通, 末班车时段, 混合整数线性规划, 时刻表协同, 网络化运营