Research On Collaborative Optimization Of Rail Transit Timetable

Under the multi-line operation,passenger flow demand is growing rapidly,the imbalance between supply and demand of urban transit is being aggravated constantly.Strengthening passenger transport organization,especially the optimization and adjustment the timetable of whole lines,is the most effective means to alleviate the contradiction between supply and demand.However,the optimization and adjustment of current timetable faces many problems,such as the difficulty of network state perception,the low degree of refinement,and the untimely feedback of optimization effect,it is difficult to achieve collaborative optimization of network level timetable.Based on the research related to this field.First,this paper analyzes the structural characteristics of rail transit network to construct the network model and proposes a path search algorithm based on depth first search method to obtain all effective paths between each OD pairs.Then,combining passengers’ swipe card data and train timetable data,this paper designs a schedule matching model which can obtain all possible travel time-chains of passengers.According to the time difference of getting off and leaving station in the time chain,the most likely travel path of the rest passengers is determined by referring to the passengers who only match one time-chain,which provides accurate passenger flow input for the collaborative optimization of timetables.Second,the interactive model of rail transit passenger flow is constructed.The rail transit system is regarded as a discrete interactive system composed of stations,trains and passengers.The travel process of passengers in rail transit is analyzed to design the interactive events.Then,the events are triggered according to time sequence and logical sequence,so as to complete the deduction process.This model can quickly reconstruct the travel time and space trajectory of passengers under different timetable schemes and provide support for the calculation of passenger travel cost in the timetable optimization model.Then,according to the characteristics of rail transit passenger flow,a two-stage optimization strategy is proposed.In the process of center-lines schedule optimization,taking the departure time of each train on each line at the starting station as the decision variable,the collaborative optimization model is constructed considering the constraints of train capacity and the actual operation of long and short routes.After the completion of center-lines optimization,considering the connection between the suburban line and the main line,the suburban line schedule is adjusted to minimize the connection time.In the aspect of solving algorithm design,considering the potential parallel computing properties and good search ability of genetic algorithm,a genetic algorithm based on parallel computing is designed to solve the model,and to obtain the optimal schedule adjustment scheme under current passenger flow mode.The evaluation index of timetable optimization is constructed,and the optimization effect of the model is evaluated from the aspects of line network,line,station and transfer connection.Finally,Nanjing Metro is taken as an example to verify the effectiveness of the model and the algorithm.The results show that the average waiting time of passengers on the central line is reduced from 235 s to 219 s,and the transfer connection of suburban lines is improved,and the transfer waiting time is slightly reduced.