The Research On Optimization For Urban Rail Transit Train Schedule Based On Time-space Distribution Of Passenger Flow

With the continuous expansion of population size and the rapid increase in the number of car ownership,most cities face problems such as traffic congestion and serious traffic pollution.As a kind of transportation mode with independent road rights,urban rail transit can solve these problems effectively.The train schedule determines the time and space trajectory of the train and thus is the base of urban rail transit operation organization.Therefore,the optimization of time schedule will directly affect the transport capacity and service level of the urban rail transit.Under the premise of ensuring operational safety,the optimization of train timetable should combine the actual passenger flow rules,optimize resource allocation and improve passenger service levels.To provide facility in the network operation of urban rail transit,this study establishes a multi-period based optimization model of urban rail transit timetable based on the spatial and temporal imbalance distribution of passenger flow in urban rail transit network and designs an efficient algorithm based on dynamic passenger flow.This study adopts Beijing urban rail transit network for case study analysis.The results show that the validity of the model and the efficiency of the algorithm are verified.The main research contents of this study are as follows:(1)Using urban rail transit data from the Automatic Fair Collection System(AFC),this study conducts statistical analysis for AFC data and provides an in-depth study for the spatial and temporal distribution imbalance characteristics of the network passenger flow,which is served as data foundation for multi-period based optimization model of urban rail transit timetable.(2)Based on the characteristics of urban rail transit,passenger travel decisions can be divided into pre-trip decisions and in-trip decisions.The pre-trip decision indicates that the passenger determines the travel route according to his OD pair(origin and destination)before traveling according to the personal experience.Specfically,the travel route decision process can be divided into path search and path selection:pre-trip path selection is conducted used the two-layer network modeling and the path selection model is adopted to match the actual travel path for under the constraints of action travel time of passengers.(3)The in-trip decision indicates that the passenger determines the travel train according to the network state which he senses in real-time traveling during the trip.Considering the dynamic interaction process between passengers and trains,the train dynamic simulation traffic assignment algorithm is proposed to realize the distribution of passenger flow in the network,calculate the average waiting time of passengers,and evaluate the performance of the scheduled timetable.(4)A multi-period based optimization model of urban rail transit timetable based on dynamic passenger flow is established and the model adopts the operating period of each line and the departure interval of each operating period as the decision variables.A corresponding algorithm is proposed to combine the genetic algorithm framework and train dynamic simulation traffic assignment algorithm together in order to minimize the comprehensive cost of passengers and operating companies.Finally,this study adopts Beijing urban rail transit network as case study to optimize the timetable schedule,the results show that the optimized timetable can effectively reduce the waiting time of passengers and the operating cost of the operating unit,which proves the validity and feasibility of the model.