Operation estimation on multiple public transport timetables integrated with vehicle scheduling in practice

Even-headway and even-load timetables have been in practice for the past 50 years. The former stipulates constant departures to cultivate users' habits accordingly, whereas the latter is oriented from demand regulations aiming to further reduce waiting times. However, striking the balance between operation reliability (derived from even-headway timetabling) and an on-demand response to passenger fluctuation (referring to even-load timetabling) requires addressing major challenges presented by peak or alternatively off-peak demands. Our work addresses this imbalance by comparable estimation. The focused problem involves timetabling, vehicle scheduling, fleet size, and operation reliability based on an identical modeling framework simultaneously involving the three models. Nonetheless, their compatibility warrants a unified measure estimation. Hence, a mixed integer linear programming model is built. Finally, multiple timetabling performance comparisons are observed by the Auckland public transport system yielding sensitivity analysis. An identical modeling framework involving the four widespread timetables in practice.Trader-off between the operation reliability and accommodating the passenger demand fluctuation.A mixed integer linear programming model with cost-based benchmark.A neighborhood search algorithm associated with a-priori valid cuts is employed to facilitate an efficient solution.Investigating four operation schemes, a balance is identified between timetabling, vehicle scheduling, fleet size, and operation reliability.