Passenger-oriented Train Rescheduling Under Disruptions For Urban Rail Transit Lines

With the rapid development of urban rail transit system and the continuous increase in passenger demand,various unexpected events,e.g.,signaling failures,occur frequently,which highly affect the operational reliability and safety.Therefore,when a line encounters unexpected disruptions,it is particularly important to make effective rescheduling decisions quickly and arrange the available resources optimally to maximize the operation of the line.At the same time,releasing emergency information in time,dealing with changes in passenger flow and evacuating stranded passengers are also of great significance.This thesis mainly focuses on train rescheduling and passenger flow organization for an urban rail transit line under disruption and presents the following contributions:(1)Based on the typical configuration of an urban rail transit line and complete blockage disruption scenarios,a macroscopic rescheduling model which takes short-turning as main dispatching strategy is established by considering relevant train operation constraints.It worth to note that the short-turning stations are fixed for the current model(one station for each operation direction).In addition,a dynamic passenger flow evaluation model is built on the basis of time-dependent passenger OD matrix,which describes the behaviors of passengers in the line,e.g.,entering station,waiting for trains,boarding and alighting process.An optimization framework which integrates random search and Tabu list is developed to generated different rescheduling measures for disruption.Moreover,the passenger flow evaluation model calculates the passengers’ travel time under corresponding rescheduling measures.Through finite evaluations and analysis of the computational results,the proposed models can provide suggestions for train rescheduling under disruptions.(2)Considering the severe passenger flow accumulation caused by disruptions with long durations,the rescheduling strategy based on fixed short-turning station cannot satisfy the operation requirements due to the limitation of infrastructure capacity,such as short-turning ability and/or platform capacity.Therefore,a train rescheduling model based on multiple short-turning stations is then presented to satisfy the infrastructure constraints.All feasible short-turning station stations are included in the model to provide all possible choices for short-turning train operations during disruption.Meanwhile,in order to effectively alleviate the crowdedness of passengers at stations to ensure passengers’ safety,the passenger control strategy is introduced into the previous passenger flow evaluation model.The numbers of passengers allowed to enter a station is controlled in real-time according to the variations of passengers waiting on the platform.Hence,the decisions on the train rescheduling and the passenger flow control are optimized collaboratively to minimize the impact of disruptions on passengers(including the passengers affected by delayed train and those affected by train services cancellations).(3)In order to transport the large amount of stranded passengers caused by disruption to their destinations as soon as possible,changing stop pattern,changing rolling stock circulation,adding train services,etc.could be adopted when the causes of disruptions are eliminated.Here,we consider the strategy of adding additional train services to quickly transport passengers after disruption.By analyzing how train services connections and rolling circulation plan vary as the number of additional train services are added,an integrated model is established taking the constraints of depot location and rolling stock inventory into account.Multiple train schedules are designed to test the efficiency of passenger evacuation and to verify the validity of the model further.