Profit-oriented Line Planning For High-speed Railway Networks

After more than ten years of vigorous construction and rapid development,the scale of high-speed railway network now ranks the first in the world.It has contributed to promote economic and social development,support regional coordinated development strategies and assist the "Belt and Road" economic development strategy.With the constant improvement of high-speed railway network,the unceasingly expansion and the increasing associability of the network scale,the passenger demand is also increasing year by year.The high-quality development of economy has improved people’s living standard.Passengers have put forward diversified and personalized demands for passenger transportation services.In railway transportation,passengers and railway operators are interdependent and games with each other.Railway operators formulate the operation plan according to the passenger demand.Passengers choose the transportation services based on their travel preferences,which forms differentiated transportation resources utilization.Railway operators adjust the operation plan according to the utilization of transportation resources.Railway operators hope to improve the operational income as much as possible to meet the passenger travel demand,while passengers expect to travel quickly and conveniently with lower travel cost.How to optimize passenger transportation services to better meet diversified passenger demand and adjust supply side structure of railway transport so as to improve railway operational income as become the focus of the current and future development strategies of railway operators.With the continuous improvement of the “eight vertical and eight horizontal” high-speed railway network,the compilation of operation plan is no longer satisfied with single line optimization but needs to coordinate the profit and passenger demand for network operation.The operational profit of the railway used in this thesis is the net profit determined by ticket income and operational cost.The ticket income depends on high-speed railway network passenger demand and ticket price.This study optimizes network line planning from the aspects of different dimension of passenger demand and dynamic pricing so as to ensure passenger satisfaction,improve operational profit and make good use of transportation resources.The main research contents are as follows:(1)Based on the network passenger demand,the optimization of profit-oriented line planning for high-speed railway network is studied.A mixed integer nonlinear programming model is established by transforming the generalized travel cost of passengers into constraint and aiming at maximizing operational profit.Heuristic passenger assignment method is adopted to ensure passenger travel utility: the train service network is constructed based on network line plan,the generalized travel cost shortest path is searched for each passenger OD pair and the passenger is assigned according to multi-commodity network flow method.The framework of adaptive simulated annealing algorithm is built and the neighborhood search method based on alternative line plan set is designed,such as reducing line by alternative line,extending line by alternative line,increasing stopping,reducing stopping,adding line by alternative line and removing line,which are used to adjust network line plan.The initial network line plan is constructed by combining the passenger volume and long-distance passenger priority.Based on the initial network line plan,the adjustment network line plan and heuristic passenger assignment are iterated.In the instance,compared with the initial network line plan,the operational profit obtained by the optimized network line plan is increased by 44.70%,the average capacity rate of trains is increased by 18.37% and the difference of train capacity utilization rate between trains is small,which improves the matching degree of railway transportation supply and passenger demand.In addition,a greedy heuristic algorithm is designed to compare with the adaptive simulated annealing algorithm.The results show that the adaptive simulated annealing algorithm has high solving quality and fast convergence rate,which verifies the superiority of the proposed model and algorithm.(2)Considering the characteristics of passenger demand in hourly period,the mixed integer nonlinear programming model is extended to optimize the profit-oriented network line planning.The hourly train service network is constructed based on the hourly network line plan.Network passengers should follow the principles when choosing the feasible path of traveling: passengers can only take the trains if the operation hour period is later than their planned time and the feasible transfer train of the transfer passengers should be later than the hourly period when the passenger arrives at the transfer station.First of all,according to the passenger demand of hourly period,the initial network line plan is constructed and the feasibility check is carried out,that is,passengers of each hourly period should have feasible path.If there is no feasible travel path,the hourly period line plan based on that OD pair is added to the network line plan.Then,using neighborhood search method based on the alternative line plan set to optimize the initial hourly period network line plan.The diversity of the hourly period network line plan has been improved in the aspects of line plan operational hourly period,stopping pattern and train travel paths.In the calculation example,the optimized hourly period network line plan can improve the operational profit by 88.79%.Compared with the network line plan not considering the hourly period passenger demand,the network line plan optimization considering the hourly period passenger demand can obtain higher operational profit.By designing the passenger assignment based on physical network,the passenger volume on each section can be obtained.Comparing with the transportation capacity on each section provided by optimized hourly operated network line plan,the fluctuation trend of the transportation supply and passenger demand is similar and the matching degree is high.(3)Dynamic pricing has been widely applied in air transportation and has been achieved in small scale in railway transportation.Considering the demand fluctuation of multi-level passengers with the same departure date in multi-stage of the pre-sale period,firstly,based on the passenger demand model,the multistage and multi-level network line planning model is established.Secondly,a dynamic pricing method driven by train capacity utilization is constructed,which is one of the bases to determine the passenger demand sharing ratio of high-speed railway and the choice of passenger travel path.A dynamic pricing submodel of multi-stage and multi-level high-speed railway network is established.Thirdly,using the improved simulated annealing algorithm,three neighborhood search methods and evaluation functions are designed to iteratively optimize the network line plan and determine the dynamic ticket prices of multi-stage and multi-level passenger travel paths.The instance experiment results show that the constructed model can induce different types of passengers to choose different travel paths through dynamic ticket prices and different types of passengers to choose trains with different prices between the same origin and destination stations.Finally,this study compares the changes of passenger demand and operational profits of highspeed railway network line plan under different initial ticket discount rates: with the increase of initial ticket discount rates,the passenger demand of high-speed railway network gradually decreases,while the operational profit gradually increases.It is suggested to adopt the initial ticket discount rate of 0.95.The model considers the travel demand of passengers with different price sensitivity and time sensitivity and takes into account the efficient travel demands of time-sensitive passengers and the economical travel demands of price-sensitive passengers.