| In recent years,the rapid development of China’s high-speed railway has changed people’s traditional travel mode and created a new time-space pattern.At the same time,the time-space imbalance of passenger flow also brings great challenges to the operation of high-speed railway.Because most of China’s EMUs adopt fixed formation and cyclic operation according to the EMU circulation,this unbalanced passenger flow structure will cause the full load rate of some EMUs with small passenger flow to be very low,resulting in the waste of transport resources of EMUs.At present,some railway bureaus in China have tried the EMU combination and split at the station,which can better adapt to the time-space imbalance of passenger flow by flexibly changing the formation of EMUs.This mode is significant for reducing the waste of transport resources of EMUs and improving the utilization efficiency of EMUs.Under this background,this paper focuses on the operation mode of EMU combination and split at the station and discusses the optimization method of EMU circulation plan under different problem boundaries.The specific research contents are as follows:(1)This paper analyzes and defines the new characteristics of EMU circulation plan under the mode of EMU combination and split at the station.Firstly,this paper summarizes the current situation of EMU combination and split at the station at home and abroad,and analyzes the new characteristics of EMU circulation plan under this operation mode based on the operation characteristics of high-speed railway in China.Secondly,for the operation mode of EMU combination and split at the station,the characteristics of EMU circulation plan under the condition of definite train formation are analyzed and defined.Then,facing the imbalance of passenger flow,the characteristics of the integrated optimization of EMU circulation plan and train formation under the condition of indefinite train formation are analyzed and defined.(2)Based on the definite train formation,the optimization method of EMU circulation plan under the mode of EMU combination and split at the station is studied.In view of the influence of different formation types on EMU circulation plan under the mode of EMU combination and split at the station,this paper uses the connection network to describe the EMU circulation plan and describe the EMU combination and split process,and establishes the optimization models based on arc and path respectively.The optimization model based on arc can be solved directly by solver Gurobi.For the optimization model based on path,this paper designs a hybrid algorithm of Lagrange relaxation and column generation.Through the calculation of different scale cases,it is found that the two models can effectively solve the problem.The model and solution method based on arc are suitable for small-scale problems.While the model and algorithm based on path can obtain a satisfactory upper bound feasible solution in a short time,and have obvious advantages in solving large-scale problems.(3)Based on the indefinite train formation,the integrated optimization method of EMU circulation plan and train formation for passenger flow imbalance is studied.Considering that the train formation is indefinite in this problem,and there is a complex connection relationship between different trains due to the EMU combination and split at the station,this paper uses the space-time-state network to describe the adaptability of different formation types to passenger flow and the connection relationship between trains.Then,the optimization model based on path is established,and a hybrid algorithm based on Lagrange relaxation and column generation is designed.Then,this paper uses the hybrid algorithm to solve the cases,and studies the influence of passenger flow factor,the mode of EMU combination and split at the station and model parameters on the integrated optimization.The cases show that the model and algorithm can adapt to and effectively solve the integrated optimization problem of EMU circulation plan and train formation.There are 40 figures,35 tables and 51 references. |