Research On Coordinated Scheduling Of Integrated Public Transport System

Convenient transfer is a key factor of public transport service quality that influences residents’ choices about trip modes.High ratio of direct transit routes cannot completely solve the inconvenience caused by transfers.Providing passengers a more reliable,continuous,convenient public transport service with schedule synchronization has already become a fundamental consensus.In effect,with the development of information technology,public transport system is ready to implement coordinated scheduling.The research on timetable design and optimization is therefore of great significance to intelligent coordinated scheduling in practice.It is a common case that transferring between lines with an overlapped sub-route where multiple alternative transfer stops inevitably induces various ridership distributions along the common sub-route.The paper therefore develops a disaggregate transfer optimization method(DTOM)in an effort to minimize transfer waiting time by searching for the optimal offset times of scheduled terminal departure times,considering heterogeneous choice inertia and random travel time.Simulation-based comparison and sensitivity analyses are carried out to explore the effects on the minimum transfer waiting time of heterogeneous choice inertia in different scenarios as well as the potential benefits of the DTOM in various situations.This paper also builds a bus network scheduling model to reduce total transfer waiting time(including the average term and average absolute deviation term)in bus network by searching for the optimal offset times of scheduled terminal departure times,considering bus random travel time.In this model,the offset times for different trips during the time horizon of the study should be identical.In other words,the headways between two consecutive trips should not be adjusted after schedule synchronization.Moreover,a novel calculation method of transfer waiting time is incorporated into the model.Due to bus operation randomness and large scale of practical bus network,the genetic algorithm coupled with Monte Carlo method is proposed to solve the model.This paper develops a novel time control point strategy coupled with schedule coordination for solving a multi-objective schedule design problem:improving schedule adherence performance and reducing bus-and-train transfer waiting time.In order to obtain the optimal slack time scheme,a robust mixed integer nonlinear programming(MINLP)model incorporating travel time randomness,drivers’ recover efforts and passengers’ transfer demand has been developed.The MINLP model is equivalently transformed into a robust mixed integer linear programming(MILP)model and then be approximated by a regular MILP model based on Monte Carlo simulation method.Therefore,the optimal scheduled arrival time at each time control point can be exactly obtained by the branch and bound method.The paper develops a coordinated schedule design approach for feeder bus service given the schedules of rail transit service,allowing for transferring passenger queuing phenomena.The transfer waiting time calculation method considering the limit of bus capacity is first proposed.Then a MINLP model is built to minimize transfer waiting time(i.e.,passengers’ cost)and operator’s cost.The decision variables include first terminal departure time and headways within certain time period.The genetic algorithm incorporating enumeration method is developed to solve the MINLP model.This paper proposes an optimal train schedule coordination problem for last train service in an urban rail transit system by offsetting and perturbing the original train schedules to reduce transfer failures across different lines,considering effects of schedule adjustments.The proposed problem is then formulated as a MINLP model incorporating elaborate train schedule coordination strategies.The MINLP model is equivalently transformed into a MILP model that can be exactly solved by the branch and bound method.This study also proposes the tangible bus schedule coordination strategies for solving the last transfer problem to ensure that at least one of the last k(l)buses serving a target bus line l has a smooth connection with the corresponding last train service.Accordingly,a MINLP model incorporating the enumerated bus schedule coordination strategies was developed to minimize the number of transfer failures from bus service to train service.The MINLP model is transformed into an equivalent MILP model that can be exactly solved by the branch and bound method.