Research On The Optimization Model For Rail Passenger Transport Network Revenue Management

Revenue management(RM)has been an important approach of modern management science and operational research.Chinese High-speed railway has formed the large-scale network with four-vertical and four-horizontal lines across the mainland benefit from the extensive infrastructure construction by the Government.However,the existing methods of operation management for High-speed raiway are not adjusted and promoted to suit for the actual operating situation,leading to the restriction of the service improvement.The application of RM to the rail passenger transport is necessary for increasing the operating revenue and improving the service to the passengers.Based on the literature review of line planning and revenue management for railway operation,a optimization model for large-scale rail passenger transport operation is proposed in this paper.In addition,we introduce a column generation algorithm and fast heuristic algorithms.Combine the optimization of line planning with the dynamic revenue management,to solve the railway network operation problem which is larger and more complicated.The objective is maximizing the overall operational revenue.Numerical results of randomly generated data are presented.Our model and algorithms can solve large-scale revenue management problem in a short time.Currently,Chinese high-speed raiway tickets are sold at full price without any discount.While the high-speed raiway capacity is increasing fast in China,selling discounted tickets is an inevitable trend.In order to research the price discounting and seats control,we extend the model to build the optimization model for rail passenger transport network revenue management,which could take consideration the multi-level seats,multi-level dicounts,and the transfer of passenger flow between the different ticket prices and between different time periods.Through the numerical trial analysis,we found the most rational discount strategy.At the same time,the model could also propose the reserved seats for passengers with different demands as a reference for seats control.XPRESS software is adopted for the system development and model implementation.The numerical results of randomly generated data verify the logic of both models.The second model can solve the optimization problem for large-scale rail passenger transport network revenue management validly.The trial results show that the column generation algorithm and fast heuristic algorithms proposed in this paper can decrease the scale of model,reduce the difficulty of solution and satisfy the actual optimization need,under the premise of not reducing the solution accuracy.