Urban Railway Transit Timetable Optimization Research Based On Passengers' Boarding Decision Behaviors

Recently,urban railway transportation,which is the main urban public transportation in alleviating the urban congestion,plays the fundamental role in promoting the urban sustainable development of green traffic.Urban railway transit timetable can be characterized as the urban transportation product which connects the urban railway managerial department with passengers closely,thus the rationality in making urban railway transit timetable affects the passengers' travel efficiency and satisfaction deeply;due to different passengers' demands in the urban railway system and complicated passengers' behaviors,in the dissertation,we choose the passengers' boarding decision behaviors to analyze the relationship between passengers' boarding decision behaviors and urban railway transit timetable adjustment.Obviously,this research has list of advantages in terms of satisfying passengers' daily demands,improving daily urban railway operation management and process of making urban railway transit timetable.On the basis of literature and AFC data recorded in the automatic ticket checking system,the passengers' boarding decision model is formulated after taking full consideration of passengers' path decisions and passengers' walking flow in the station,and moreover,the algorithm of figuring out the posterior stationary distribution is designed.In addition,the urban railway transit timetable optimization model based on passengers' boarding decision behaviors and its heuristic algorithm are constructed,with taking the headway,dwelling time and train servicing as the constraints.Several works and conclusions are listed as follows.(1)Enlighten by the works of literature around the world,several preparations are conducted to provide theoretical basis for data collection and timetable optimization,including analyzing the relationship among passengers' transferring behaviors,passengers' boarding decision behaviors and timetable optimization,investigating the AFC data characteristics,summarizing the process of making urban railway transit timetable.(2)The decisions of passengers' path choices would be instrumental for obtaining the number of passengers in the single line,and thus,in the network of metro system,the shortest path finding based on BFS is constituted.After aggregating transferring passengers into the specific line,the passengers' behaviors in the station can be uniform as follows: entering the station by swiping card,walking to the platform,waiting at the platform,boarding into the train,alighting from the train,walking out of the station,leaving the station by swiping card.With the inclusion of the passengers' walking flow and the approximate distribution of passengers' walking time in the station,the passengers' boarding decision model is utilized,and the MCMC method is elaborately designed to figure out the probability of the passengers' boarding decision behaviors.(3)Motivated by passengers' boarding decision behaviors,the formulation of passengers' average waiting time is generalized.In order to decrease the passengers' average waiting time and increase the probability of the passengers' boarding decision behaviors,with these concerns,the bi-level programming model is constructed,and in particular,the upper-level is formulated by adjusting headway and dwelling time to optimize the timetable,and additionally,the lower-level is constituted according to the passengers' boarding decision model.On the basis of the actual passengers' data in the Line one of the Being Metro System,results generated from the mixed SAGA-MCMC algorithm is shown to provide better performances in terms of probability of passengers' boarding behaviors,passengers' average waiting time,the number of passengers' boarding behaviors in one time,social and economic benefits compared to the practical timetable.The research and results would provide scientific references for the operation management in the metro system.