The Vehicle Current Calculating And A Robust Approach Of Making A Period Plan In The Railway Freight Industry

According to the statistical data, Chinese railway uses only six percents of the whole world's network to meet the need of twenty-four percents of the whole world's transport task. Because of the rare research and onerous transportation task, it is meaningful to enhance the efficiency and the profit of the whole industry from different view. In a macro context, Vehicle Current Calculating plays an important role in the strategic level. Vehicle Current Calculating is the precondition of making transportation plan, making scheduling plan in marshalling yard and adjusting the vehicle current. However, Vehicle Current Calculating is usually conducted by human and lack of transparency, completeness and instantaneity. This leads to traffic congestion. In a micro context, marshalling yards, where inbound trains are disassembled into different groups of cars and then new groups are reassemble to form outbound trains, play an important role in improving dwell time of each train and profit of the whole railway transportation industry. In most marshalling yards the period plan-which means the scheduling plan in a period of in a marshalling yard-is made artificially. In this way, the period plan is usually not optimal and leads to the backlog of cars and an extra cost to store these cars.Based on these requirements, this paper studied the Vehicle Current Calculating problem and how to make a period plan under uncertain information of inbound trains. Projections for traffic problems, according to various existing data on the integration scheduling platform of the China Railway Corporation, the author designed a real-time Vehicle Current Calculating system. The system aimed at forecasting the distribution of vehicle current in a future time. Because data from the integration scheduling platform is various and in large scale, data Structure was designed before the system had been established. Based on the existing API of pathways calculation, the author proposed an average-speed-based algorithm to conduct the Vehicle Current Calculating. Finally, the automatic processing logic was designed so that each sub-module of the system was able to execute in the background following a certain order, while the result is shown on the on the integration scheduling platform.For the period plan making problem, the author propose a robust optimization approach to solve this scheduling problem because deviation of arrival time and related parameters are known only within certain bounds. Additionally, probability distributions governing these parameters often cannot be estimated. Based on former work, first, a standard MIP model is proposed to describe four operations in a normal Chinese marshalling yard. After a series of pre-processing is conducted, the robust counterparts of the nominal MIP model are presented. Then a CPLEX-based algorithm is proposed to solve this model. Finally, the performance is tested with a real case introduced from South Shijiazhuang station. In this way the robustness of the model is test and the solution efficiency is improved by adjusting the parameters of the CPLEX.