Multiperiod Metro Timetable Optimization Based on the Complex Network and Dynamic Travel Demand

Compared with the single metro timetable optimization aiming at minimizing passenger transfer waiting time, conflicts occur between the timetable optimization of different lines and directions when the metro forms the network. Previous studies often ignore these conflict issues and the uncertain dynamic passenger characteristic. What is more, few studies consider the consistency between the various periods with the same departure interval in consideration of simplifying the model formulation and solution algorithm. To handle the conflicts, this study establishes an importance model to rank the importance of different lines and directions at transfer stations using complex network theory for fully supporting schedules considering temporal and spatial heterogeneity of passenger demand. The concept of node importance is re-defined for better consistency with the metro schedule and passenger-oriented. Based on the importance model, we propose a multiperiod optimization model for solving the metro timetabling problem by considering dynamic passenger travel demand and the convergence between various intervals within a schedule-based metro network. The stable period optimization model generates timetables in a stable period with the same departure interval, which reduces dynamic passenger transfer waiting time and train traction energy. Given the time heterogeneity, we also gave a penalty factor to show the transfer failure impact of energy on passengers on the last train. A multi-objective optimization algorithm, NSGA-II, is adopted to obtain the Pareto Optimality in a stable period of metro timetable. Then, the transitional period optimization model selects timetabling in Pareto Optimality of disparate parts considering transfer connection coordination and operation stability to form the overall schedule timetabling using data envelopment analysis (DEA). Finally, the metro system in Suzhou City, China is chosen as a case study to illustrate the effectiveness and accuracy of the multiperiod metro timetable optimization method considering node and line importance. The method can theoretically and practically contribute to the metro department for scheduled operation.